ࡱ> n$dMwP \=.]PNG  IHDR XsRGBIDATx^]`߻K(P,9X8Xذg{DQAE v{64( XTQP 5 - ɥ_ٹl6{ۮ$|po~3;ff5dZ:DFEGF{&JgtZD|h}y|(į" " " @@ Q! tk[k{mLLTD'Stdd0_-|cYn\Qd2 #" " " J=| _9R~vuU7U}FLdDmd6Fh"P^ Gha<D8h9؞u!" "Ťt)ad֗L/KB#VhCu)QwЦ$袹 ɪ,T[ןtjo4)"%\&2Lod 0^ws4C!"pf!uY" "" q]b4؉Cxn_0䳉7j;.ґv[&d5%%5nKxvОصPq9bB,Hu9⯞XL]s#vV}0EdR xyWS_"[1P5" aGƺD" ".α rNP_e&re ^.pzv{jꍽ7q@CTP~zxBÆ_4:W{vj VBw)q ]nFW8* =&.`V3ntTոjYR[nﲟ}e;)&xxl4@(8aO@D@DC`ljhjw0+!uB`Y'@<]n%eO|T$wW(.u= \Gt#q^ Xp#!7k܄wohoMm~(_r#T˜K=լs}k"]LbR$΁@D@6G͋D,Fj:\`Lp|q o)8_>7w_ Qj KGu[${"`\ؾ2o, ' |%a`aȃ3~捸hÝnrCO~ʎ,{fyP=eS( Mǀ_>M/"htDÀ NĀ @8WBOtHL=b8S!)Z~fbk=,-h6'p .+H%\N2(%Y " "Ќ)^I"^ve,e{MfipdaQ%枸oN/|ϖNY J4:%:C, cv,^b3Ps#>l璕[(rž./at/ʹXtATՒNqٱ ?jL嗎tO0CPL7\uHMXmWa,GpJkC]!U}z_4%KX( " "pF @LGgFY')} p3kld3dz{|:gs{.zׯ\y%d _Te>fRN=-K~&Xzv(ts ~צ.d wFoݗ2y}cG K/|U$|$igF2NAs~'!S nD*I$;#/ԷNk kDKv_"d[TU6 x9E ioKK{~\[A-_wүX8LsϽjAgIK|$4T=NXR/|g|<`\}iztQG(e@ˋy;M{11u$sY Hqage+z i} HeD@$Xs @+ p'+]C =gx.L>Y0 9dw{ (twK'?ƙ'JrU K?vg4Z ^5NW^?eV$9Y@BX[1jH4/3  Dؽih 58=@L/kHw(.P_dRO'/\rB^\cfyf8 _!N(2 Ԁs߲~mc`ײ|0E1bH沂wي^*[?DՀ2" Ra{vf"p Y> lTa :ͣz8'xD'}]s )!dSD,^dgpӛ ^;*l!Ӧ\즉м)а(*O'wrm~ TEccݔLr(xǒ';YD@k YTBS_>`DI{*߶x H?%=AfN#ܼ&aq: L]zU2IxEoI)Kb$>@݇˭NC._1Р\P[n-rEmDMKd)a)E1cF껄 %5oRFt &"" ". BNnn>}sݿx f)18̲e4ӣS\+)2ċ f;qg Qeҗ/C2b?gNkv{@]`Ź}k/[w Vm!k`-G>8]SUׁv |B;)O&!WvtA1C;Q #Wv< 4o*HU\uT֗\YC좎*f X!z<JJqMQjl&ñaOh;J%B`d/>C =WAbKS(R UVn0';BtHrғ%Hc*F沂K"OU"*qF1D@D@6CnqɖWMo9肃~gBa)%bD|LgvdVvn.ZF _@cb7g6ޣ AÏ3_0q2iKN&saa-׏)t<~sgi>Oaxc%!p.=̢tT-DHSEAR wL:pl6y*S42%T3!" "d2_ x<}>.!%~XSd+B sI\j@嬛?7Kcuz",ql=@^~q|*Ԁ( дk= ކa "gW7N]&:ɮ%ioeuiY4 ^zlδuVBBT}رr?5X&2Jɾ2P0w*ffS`=>ǜC=@#"_PiFd#ǥp,Lh&`>Md5mn& M7HψzFYip[\{wb@Q " Y>;BFo$0>eH{-0B#^QD &d"#>:d-981 uF/ &fd4l ~a1|͢=nߟ#ڦX ǁq=]FD@D@D t6RJ8YK"J^ IvyWyi5%e5Llt߱9jcKJ$BNWзA,6z;orD#'=ZBxL^mu@Z҃qpx{jCD@D@ -ߞ& "۵S6&ЧbgEsëp~S{Bp,YG5< c=/fy&1n[I<>LvF܋P@D@D@DmY*aͧ} $Qþ'r_w"Gx "p&" ~3.h3"c4jwf4dgxT qQz kir$\n:" " " ه@m]\VGu4HP qUj3,ud\9J,ቪ%.KU i= YӁЄ%x*8aqW%%&* WȠAo9BȠj~h`;LG*"j[qdP/*xE~\mQ@D8E"" " "=Hs,@D@D@D SE@4iL<,\KUWRN<艉(6 :&:BfZqc˪JJLHUF)ҝ,;0d v=2/AVl7QJOHUzښ"  q!~gpF#F,N|XDvfq 0\YUX pn0"Hne&2͗B8=@D@D@D@\<4b.D@D@D@B5jdԗJ"x s r iWu.c@ڢLBSIͭ#2Vvҥ2MaI{ks/" "< rG*'jA;!h+{Or=^Av _XtS~Lip(d168?%TseBW~;/Ѝ4L,xDžG' 4e2)JމcA%E9P@D@H 'o@[P^W:.j^hsPgk¯ok~yjsZ•z>a#mv|W"mVbAK77(ʴ \PS[O ՘2" "#bu8՜s R9\vQt(>& ĂpMs@vMa[,\hB /8{iH(~`~)1hg#M{CsE%@msR/8B/#*\ԻEw ha33?&ќO =so !ʊf`y4pb")Nb"iM)\$DS!i/r`r1 T)!ZeW0 |\":Å3AD@3? G\)|pNqӏe<|h)gN38; AkQuʂGe<#reU NrD8SaIPX`lGE9[V'>Â-IډCp7l)x@E겪j+صt""-6_nˑ!0#yҁelj\Fe$ZqdP{#bH‹r"H!J%7:=@D@D@D@KW{Zjld%UJyWK T%wA~Ӥ*@\u V`PT]VURbB2h!Bʍ+jl)6_",CVU`62@ڬ eUWPYG5?#JC:W*uK-( @=%_M?'Phn͕tm)@.:(` $.$JFRBD@ݶxci" " " $x,@D@D@Dm@ܶxci" " " `%2Lz$s"  86#YEDE`Pϗ;cC1OsaCAiM8S}bNxBi34^kw/| T^#|)IkC>06P` !>hܞ>|cCu`W＝Э{RpGQ>_XSSSy%=笩5VBĉʊڣG^};u əyfk$͏WU8DCfOog^ߓ/Iu)Q k,6Y>8;oXW2H*[0[Q&X O /2-owrJKnTK nLBmCڙI{|r^=Bz}zp|{tk86}x]侹DC~`ߑ&ٳ9^TDyɚտ6F/Hv jbo[. ͼPUpHC|i~[+ erE^/zL>Z Wٺ.F:~9{07/Cw^ .$Bqpէ,\mʀoǍ۳{ǡTv'?ujԇOgYĪƀ;:|nL#M;/UP颕_q>qLjsjd㳳sNVd򆈳m>0mOsKύ%1=Θs2oK۲j'ߌ4xX`ɳ fIOO~c[C]nwZסvȃ+mjr2G,ּKÖ04A]8ׅ//|P韫w6B']B Mw,[>`opxV3{lQ?u\ԉc|1iic-^ z1W}!1yF-9F4P'ET^-Vد!]Q \pVRJ?E ,f- I;S']__H97+To+ptvۅɝ}C>xLxݿ]kj EZ[ tlO [\}9۳D Ϡ VlEmi;I&ЏT ?{jvŁ/e%81kpg],qH` ;!p_qnV);xpPZ*7HTt԰acd/6 [0L8`o8b#٩78-[K ֋+p,.N蜪5!,_|j2'&&&^/c[8ڶlqun*Zټ<|铥k:ۅ[=_}볟hӠ'3mY |` FmUIᯢѱtnUjd2'37Lo03:̼mVsHk1g;3+n 4޵zmM>`q9,pŪߗ\/;sH} Am4nqY^X cJ?wRh8QZW>>[_8 ĔD{v+?.p)ri ׉xkkD8Ԃ>kV.6.ϱ[r[+DCRCIi O'/lN'MV$K`tz步e=6Ԛ(456҃k|LQyJpl+& &qce5ks7&ե;iXGYoun޹u?~J@.-4U(`LQ@I*(o)%Kn>홠o'dsFTETeX@f)կ8JD+,5 ύQ ;bU*"*C?+Z*EF*2 ~=} R?u-}|G.ʷ廕tJJ˶Xa,tÀ,nk3& +?"~iedfA}A4xb = c'Շ+k>r9RJe)Kc-m|&!y{=;0}~"x8FXa֭z-<˿YoNhD*ߓA=R)TK2GN8x1`$ryU\{C+ćeA5&*̀`cw]oޥ˿g<5w^5f,-9Wouia(sN#2!$\Gd.+T nGF@=1\,͑ Ԗu K?b]V+_}[mMO>rb)?EBrolJ_3*`hRƿЩ^& f,v]peEػe웥FI-2&vuN:1}J80.xD&i{ ዧ^|q$gp k/XI\X,{N t đi&i&CvNZkr&9lɇ2ۏ2 qv5p6ޘ> 8[om*d4K)=Wm0!{M{vNup&9+䬱²feJŶ<`cA&-=Qe:tS3\P];&u!6z ; ܾ] Lvy=]mLHWcYo,Y6sgs6幸^Cǡ/?[,0`Zt_wi(=ogמh6^3)LZ.zn 'Y9{P:\\`IХ^;zd铥G3l  zC1AHRʥ<J#L=ێ4 T""S  :92`[izi]y /K&5;\&Ӕ4^oXKfCHr9ȷݾ~KQaSBtC9(CEPM#H9p&{%ݦc`]ɴy0LgtI΃eʥW\>l"gH|CLs3T~)5c@jM# R߫JgsgK`02O)ΕVL\>a5D} "6Z RRЕD: l_0ѯ=WQ='%&僔mR)2^dq_>1 H/]I&\y]:twN CNV4Ps띯^55>zb4_S~xHT!E…găԴ'"RdAq:%,X-^d=x`]dd}<~R:Rp@XF gelkqZwCv&m<Ӟ@'xBSha;.kDS@wCibbW~'P񸎌i#u.d1.FޭIش:y\(P$(Ƃ@,}`-`,%z :[Ʀ7WdqDen0˫7UX@6LW1!0˜cR"Y"irJ-=)=kdyrA'qXHAID!f6z`K6$FT? G ɴj_ksƒ<auxCx1&80jy}:ɀ!F~QaN"RG&WDoa{Rz9xsygr6T7>L1[MwL+BCa0$YC6A~ #Z˥|qqȤTÈ޵K%R%2 _JxW8W n GZ) ާb YN=j?\Eh+գԷ]x!Us/Q.yf9 ̸V>45٥\fR~/#GQ/-8]K^|mk#m4 871VGF\qjr)"S!nbSl˵i-MwL2klk -F I]K_Y%5M"`s%JqA?&x8-Mfs 9:.En̬0K^_df,V5k,2xp)&DdٰJj̛`w/xXUq넡+[#>sǛvZvmջniy퓐ҧKn6e_' U^ܺyNv_\0e06d۲J #MV"!qx1Lzɉq-l++74Cj"2IQp#~%?'@C|L蝘`j ,RDgd'f-Do5X,RG͐g[䁊=U-9d*SKrC ɠb:q$lGHU g@??O?Ӈ-92 5 -(1i һ:<0a9ƒ}"0iU6U {7r&*' Kd? T敲Y*E=Η#$0P>7^{T@⭣b{-.}٪̗>|+zwyR=f(P`yhwOs|);S̙>?sf1=;kNvVf xoɝm<-3 Ƃ@孬t_v v{M uU}nnJnGlDžew|یGaPrptWqG *OV(6%x].' J[PXp}tn^KSR&7(wOMzLO`8'cjfF _=v]OB0۪JNL&Ɯg&m(PW+Gz,a l `=<ˬ:䈈 q 1`*<]j́Yy:ܕ\Xv-6&^o^$f.r  PpS vа_sbgFAGSmEG1 3oͼi_OU/_&AX]*$G H& 54]7{BBڔ4L_<J[mYm_)4N(.58{5;#koWvԹ iX|)YBI^~ ,UFCէL(|^Ͻ` bћ&?` abD3py%r ĉRm֑!*0.# DuԵ>;cG&pJݞ+ܞ(MV=!#]U7Av5p dBm0A6^C -~ZyU}MSW6f)Q=݀}/cWRG*zJyk[NiSDP1.\.ru R` 塁H.'x7)ޏQARe噿o)mܲ &SaQS i2m) -w~ @U؄rzh]pKdagP8o,'˘anzFVFF&qͽ&U {#jRkI5Wd k8yV)_6 u~Mo>7-ZL^`[&hEU1x 3Ov׃3Q2WWHz׿3w{z}dgJ8 SSF4 EeT J|鉑P{/ %~P)%nzuɖcɖf,*[F_e1i,5uiůU"pB/f 袵0%/Of=&))=<)uwQ3[FvĆe2mw͐v0\"=~H%meLFM:`ss8sj陳si43g|1s+G){)psZmUVxS^ΐW$>ljJf@c:XY^{.}^hP'l'$YS6ګ+h'ުEFhz{^bG+{_,-Xv_H:8i^ zU*f[&d4$ZݖF|8)NF^X}>4\>a]aԪ=>6ߐၶ@r)7S u@Ӡ(z@$e)fRD 2:[& fOL6ޔf00zxM>2e#wK{= $ w0Òc AVsvgutL- o* /u̘Tugc` 7l-ߑ5@HO6%`p`xO\;3Sj)c.?/F@R4QǮv9=?ąױe>@dXg+wgR, 0)ZӋG9gy!lG^l^PAɭ9V8~ɕB?A>'_&QO*z[D~4"')F8Bi#~=?N ) n j HF=  ิz=!Tp FZ:o%Jp6m4}V[EE=r;`3??4E,؟/#h C#^x_hoPiYU1UUo5.[V 5x{hzeULjZE+ VHra:whY< |O!فYGםPE"s.}@F^P8ZbdvzC7&{%a+/Xu1&0?[}aV8?FT}ZM PO t]xcǻٰ%~2:zCbi ,|.5zZwu>[%ʮ89Kߞ{ӯ>/꜌1\8 Vܿ\h?:b9m$\ /O82Aߣ~D2vTsXס/K`4ATÓFұNMY*E8ZB7GJl{2AxUieK{ezm8 |n'w:U!m\t.}nӯ75&Iv~S5wd6o=eg@"ԗ}9W<=/ kH:5|E{eo,`o -/ Fô29pЅw7: pVY%Ղ90=IdXpgzDm[6vfQ#ΕA g !,OK#b9<)`LV5_m5QppaUǭյU 5 LUɉUntK :vѥjbHM}zE~{Khyؿaexlko}䉣̡؂-z תpAH4$%~YPiwRp=R۱} zfFH b=w/9afzWo\װ-ބ0 ^/jW.٩I +.akɘBzF%)ZaQӆkbyGjDX{W K\8 3+ /T&\}Hf\dܓ #nP~ 6כKfط̐#'~? rvPGH"R}oŝNQٸHlT漢<]#FaJ߮c<+.Miu+?ԱØ1auщ쥫9;c6&unNXtrK 1>-/ _pd0YWҔagAm"~0 䶇T};!VSl&=0T 4`gNܯ‚gL1*˻∿ |Os^?] u@a8Ź?^ 2%/?ff]:.2c덤ݹ#e)0q=}*o }ns :aS/&_5lz~xԩSϛr_v9 bx㑨C'%xRa {|U4W3Qۑ=ڲnK0fFs {c#nj߁<:M`̉ʊ]\'11}F]4hp^} JP@䉒=` )8N /k[GKUߌwЯe^O#=@J+I]!U_"T-a-D(ڶ՝:<Ɂڦ(/e;vs8CTW?_{2ref =atSU{d7ScOKJ˻}yzmᒩ>͸%1<5;^7q3N7…9-x kWR|о{z%;pMTcWg(0̏R ^JM!=)y E~j8={cRVʻ WGVjkGo@O o7PlY~_3W%AnG'LH)m?uR%ο;m`[m=%n\C[Y4{+^`ꏏW@P` pcG=" mc]/r5_h(Z; xYDkRWP)y6VZpe}n-4_h!BEL]96PL+~GD@3'nлdO(kn EE^}ٲS_%s'oΙ("(S\@z"}̚qzc郡+A " ,G:^{X ޞaGBD@%l3kUb" " *p<6&pBE'|/TY !"P9 ӂ}O!dXeD@D@Z}!`TrU@_T$I(B'6cIsJPRtgOTE.q_|TĥXH3go@T(J&,5ĄTe5AerZʩ+Y-G lde%Տ#eGȨYeV6 vZJ`VD@S@ 3f"D#ؓS_' d@/]x~ѵsDSn9 )A@qؼ09n-͎T-4Oۺa)G b~{ =- Ag"a+h9HCB\=!fhSҊܭ)EXqؼnTx6Ǿ{c$FM$mM2W g.a&:^1@򈸉xRu,@A#" =% S[S(6Uhcw*ءg~ƜgQǁâo5D+Esbzt,ROVػفCyXN@X9|5d6q3]>~IaQ7~]#ӫ LEy6}qmNG)IϮ=DTkj n>'g}뒵|']փ $'`̤go}`F f6.]OIP_􂊊bWŌ2שA&D#D@7{ 5t4JoU݆5{67K,U-Ir:KVx>oW&n?罭 oPkJy$8@%-lp0TZNAjss_Pc \%5ezW3%6J)88~P'׾|~Ɵߐ37 0B,3-0Wp0ضscnpYv6L6JA%j|',*- "qKVժ@pHOrdQ$oI Vi-Pf .ooa1?ri]%Aե@q'wE/"+Vg!*ʮ^ rׯzՏ:lt!l U~|. qC#%(*ogʣRЈ!u3-T#.,d0y m:&)2XOQI%{sϏͽEcl{`mtRP{ҽՒ{U^2 9æ$0o|eWu4^7v toGfR !I$BcCQjԉߟ)6Ze">h&pO<Jq?hJА T'83h\R(Hp*w$S E%aUWe-n6JNi󊦽Y4i.O:4h22DƆO􇣟~{42'`0e9RIs7^c*+bOOY2BAzc_`א2|aƧ{A"5/q$mx}xCZBzbS[|`~fO:uM&2)q%5'M*53:=3M@̬lyo\}ON#9g%)$FrNfrvf2S%YXj̤؞[XK)4z f(C<$A|O{1݅A3LM92,J+W_As7ߟ:lr!So6#,aK-U%\\vNQE8m! L6Ua*|a K'%W (셕=*%&R3L Yo|cZjTZ=C-Cq↧ĆGz0PEW,Y@,L NymHx~Q.A`*Yno 7l.%eF&%>)! ;쇯Vo74Z,̟2{2Y\?FB-VӥQhLXMxtlVUg_㩙Z×1v tҩPo1SS z ? sLzb]'Բۀ?ɆAdFóU*4ɐk앖w%Ԓ~D P'Ro Tyl'FQPSrRr8̹E ˇ(5n eg eu=T]K&;ghD~)U2JDm|T,؃0RM6]~$E@Ol [)'aղKtŶamg^۳턑^)73W_} p$UڙE)Iِ"FMęM[3i )))NrM`q_MS^K'%Xi]Yk' ǯ"ҰQLOLZjUqD B-|M`Dr"H|3|&/Z ; K*ǂ_ $2.&HNGhN^oo$eV/`PUdi$TޒCbc$nVd-~a '+3GJ~䀈昫y4#3دjK T\) ̵2d20"sMCUv[3e|"Gv6MAd]ؚ9+)^Ԍ JݯW$OtaP(̪Dc$1 7+F&EM$nZ OIHJJXdž$&&W.֧ &<6$f] R͊K[EWWrs P/~X^m4,-!=P$^K~\5\gul|h}I%3!{u>h;vCe傡 ?-YIctAa6xWX"co{ w"-xۛ7{m@/  6ڒ[HLj"x'Ϲgym$wFq@~j>nG"S)"d̹s9/OcAX\k [rj AUY⁘b4H~fڀ*-EjHDpJ#T s6C\|d\loTZp \pVӃ3gf!fAfg+ٵ/ykEد ^J%RMop7M'7|ІZUz [ Jop"Ygl6W'0x}!>+7'8`D/EDBYc\#&~jU)`[p^yVs{JRK~"~-fsVi@-yxb!]|c&w aQ&\4 0}^p K")A$x_T^A?K*"ei#8 !Zˀbɛce38t3,?1*Wo_h@)MQ'/_lϷk.Qe"?RC$oy.)q[}>L(fߖr?uuR ^>Y2vdYRD]?m ac%vݗC59췒ub Os7nܞd/LF$du oR1z9B ;/ RVJuZv/?0f2Zw*N큯Aeϗ.U/^YK|-Â&$_j,Tb;7)nN S6e^uBlqqn.p`v(   /bbYI=מِ74(ޱSr[p hvk*t:?prc7љ䳁/@a6a-PO(0zhG W>H J.8@~]ҕ.=:)Ph4`Caxr$zd[NIEŽ+˵D D'B )Kϊ-(ߘ^]~ Vlj7Y,Y@Hl۬7RyJѺ^rSso(&:Ƒ2KML AxNl_1Stdip+& nTV^[SZA~ MFyb?prHo&A Pa*y 7p~#5Y[y0{W qHt* ZZ]rލ)ؽPg^mIWX7<0{E*8l6}s~ھWgKJd2zKW<`ŒFN`g[~Y97lr@'ptBaK?mHi`GLo +>^}ѣrÊK§N4jX5݆X`k88f[]p[A-@h>%H*> nEladSɪ@m4T]%Z`a~PFR—>#'s~-!FqN zu@n5v[off*ܲi3%2vܱ~!Sh}:-l_}}EK9v[Ƒ p`*8}جi$ w)d`gXKl6-%~P WMjvU[ 1n5WzᬵĖuG[iLW'n%`L`Yj*.e۱*K?ETwZt.hepSi n~(JDm\ *RJi>~uiO-ڦRjG-/ cm۳ޙTb bjvF m_YLf|C6GdJw aÖNE&V~oy{ (Bq|h{.;psoYX%4XtA2)\]6:ӛIUv{76 L83eq6U+j;P+Z@{F@EYJS\\ZX-y4T]R˥V&إJ~Lj.t pqb,0v$դDYKYPR|LMY+h#Rv"74Kd=9-ԗ__f>d0b!Ӽa8xtx|j&Z;:Ȣ_dYg[O!a?8~!d/dg[MA4:iȄYd길&Wƥʿ׾;z.3}\d_n2b* +iZLTC p,9TCP N@r ?пddI14l81@XE(n(o&n[*UJI@# 0^_QZ%? ,5ڴuSa 魒玪% 0+X񢳱`-e7g8o؄ǂcFCA]„%o2~ {*uI= fL.b2Z)پeyI@ )=~hZH(ӂPHQhHxq'۸3 nGiW\qi; p ЁY}+3Dƌn1[>,^F',Jlƥ?֧?-wƫ+6Wo"hW]Kf&Bb[N 6LF-j}[@@櫍,ɗ?/-9,4!d`W i2p;Zޯ6CNS1;**b 7 t"ӫUvˀjpmo ٸDAD}Pi6 j/Ep_X~EF909lܰ|~LMU-7$6ݢooxm]m:yG %gʦ]Zv "7lcc *mL0uvc)̂wm6cR2¶w{nðʤ/>7PLz`Ǿ_h׀+Tj;PQh4O-" |~ChNa)24Lؿq!&\29~As`p>JRdpt]3^EpK=M@S9o=W]ZW0u>U\Vuo{ټN1!䃏 , kͩU߀w 3ό53c&=yN5A¿%+&d/h]c?0\ƽ̆ͬ U;$*Xw*M[{JIt 9K6ǕR%$Qc_U"*;-CJZMEj,޶I{\Ȼ=_@%<|+ ޻HO%ia|ai al$I~[\X\XRho'%$ :S=_Wov/ؖVQڣ_Ql,l<2(I`ATD!cC\Nm(fUn߮W0͌9 j%: |`4-ӧ$o>^S9-,~bh_3^8y?5*֤6k >LwZ.-m l-PĨ|+&6={66h#iFa|/*8*zvˍbu͟ qz)ʨ_D-/rjX/ =JM=ԁZ@B787Dq="gJeęb*ډA65c?EoD  @+!8l~-|exz.3"f\U7sű" ] e+v} x68B هⰙΝ[+/@8 cCBQ"p!L+k+j:E}!"W%;_^zHY8fDlE@q}-3=p͠{z.3"f \k7 @C۴/o^2N}4K*wOhwaD!fؼp6N4 0 A@dI h PֹcDhF cߵy2/|/DD@'æRfy+Zu(%b^D,F@aD@D@D@@;"[D@D@D@DlE 3>[h"@CGh|DH<&`𩉈Lj-xi/xrH#q*8OeX6W-PqoLьf0$iL# .v*ѽ#q,#" "@7_TUy "wzk:vic1qpcbٻ>8p@C" @a-pZ bܧ .j0ꘚjs`ťu8D@D@N'w1PEK)"IHuMǰ`p;+=q" !0Ut_M:md<9|.b)cmy(g|Ӱ$D@D8}j3FϘM ӇkD5] 4FEj&*2*Fx7Rx0dHS+$z |4)x \xBho ܚ)D°`0Y.!Yӳ=XeLݷ$$ 1؂o X4" " g $?~jwiV#-a4ai4k"vfaFW;N#l b!ڙ"kIg/jFD@DA0uWG x_Z c>tLboFS:";?NiSxB8L8Rh=$ $P`vje)ok2F! Ӵx]$H֫jFDm$7mmoO./h JJ!m tđ0gP3fI/{xa3{ vzT@D (ws?]ty/tS7J6&fzu В7x4b4.^41H#Y6kD$m5t!#4^&4x- toνp7,0H/4^r[a &ng5B`!d;`/: ;0vx{\'OuyoeÇf UJV[U1U}?{]^H^vn{넝F{Cmɯ~2sY)&@mvF{9[Nb" nAfPcA@*tb適r4Xz)QGwil N^0R[弾֫h"Ж nsf.pꪁgkUu^xu`L!&`샭\n=n;zwUUU'OY~8!!*yQ,gF襉Vkl^q:)d/_xù "F'jlUUǣ:~ML}UXS0ȶM'qqmY6 y$҆BY\۵Mie[pEpHʼnbڨQ0?'u‡8|r_[iJѣ H[JZG#O$S= RLn!ҎC&FV`fW1ڣe$D;)I7i=_b?Y\rUC]9>=boha,  @ 0YXDG,0Zح- (%3c'jsCUBՕMt4X 5Qj#4`S=#U 62X줝h߆|6(WqW_5*J(`QB%܀*< [%qq}LC%g_ h%ZD d"(/][wH\ 01"R ٨mТL w~p#aь,{֤g9-G~/]Kr߹3r+n'Հ aEAr`t |zg`쭷{kZ[Qpu.XA7ᣪBT筫V{ju4x0'zj6wU d5Dc?!0r>Ϛ26cFT"u(e@(uSԥ Rƫf }B?3hƨD~Jl i/AP:ܚFd1 ]p`E|Z M[կ,8>wރeyHm_֢CPr3,\ = my 4ު Էkֲp%^[-KЫՀ<'t }bΉ帟^8?㤔_O c'u=SW筫պ=6OMǿ7W5x[ܠtK rQO NL.İkFl6::hi9Cw?9O<+@M/싒,ު؅W[v.'5>ˋc䎜I+h^APBmى ȵd7| p6htAg?K25^1uv\ht,`0Y KA4xͶZ*Q*B%M.j:ķ!`} &ƽIi 7=Xt Q~Nik(iY2[E(ZDNw6(yZRv`hE K `ďNHVU1bF)d2rsXA8"p&"l~5Zͫ'w`_S V5/VP-+U|QW0kaj*PS:v~-=ieu& ? mB l^ztHz?tڿDvsn| M;=qқTCwd9"oln]1g@:ʚ2u(TPSNUi\adsعĵŏ.,=!8(M_~ .0^`0F@ Ny#_] 1jD4e rJg끲~8R$y7tڳ[WD XN4XcQ4 `^Ҵ8p6`vDh?g /W|{"rH>`@t6r2yʭGO|b_]^ؗ_] G12e>~FGLvmw8/pw72fΟؾ=cUc#jʨ>V][~Q 'p]r{bAM"#_Ob0?^pܫHXJ&N ` !~zxRL~ >uH<^sqJ^O|a cK]x?z8+T۲~a_SpJFAo W#7 -J. :Dm.h۾`dAJ3)Dh RU%Nٕ՗ȯ`ףPPߠKQ~i)8 )r?7C8İTl}Ǻ  `g7ߠz߻`S&ō 5FV VxC|灭 DΓnuGx}4G-ɽ:ʊ&O2)` sEↅDHqHᄋ{I3Ξ {E̾,2#" QV>g(kS]`NY:wyvr'"ZCsSH*SZ.`XG>dUxK ZSZĚ=KI8 o,vA6%g0Hr-oT308ɻN, E%]?"5H2x#4[4BjP2JP )T{=~O4>A#&m5xQ#НLl~ӆaVL ZIi Z-ˢFq۲DAiتm*p5"Z\н]MF=jL ZFsp{2T:D<6"m_MŌyב7lLh2Ɍh6͸gbOLQC2צB^W}$¾AQ V⍯S ^s^]PwŕuvT qUw9"fёn_ yJZDq$;kÀ^o|V4/{d|I6}Qdm^C|NI~׶}7QZ,(H-^{p(\]Ǜ zUCAbin9}5u}{kj#%v.^]gw&|\Ej]tR|(@\5`\=';r{A⾯{1YLv2)?rq ;v<]R^~ hiwg÷fe%_ݹcTu^$m?$;}9~ /SYzcWVώY]/<#v Q]>w" یW9e_3T.; x^L~DCjI 5PXFgϐ XxQIoc↦ w7* A }PHԛCoin Jǯ  [KuęʾT ^>O$o,r1.aR IV'JH2_O^ɤZ+kNxRC&Bpa_z" n=Sa v3<9wk ^PLBF0- -z"4c޶@z{0.ˠ{s\xfpx/]hL~v:g@671;kI7kd _nRr_Jn&{m-?vAD$.N[~^OҼؒLp-)T#< Sctܴͦ[xaSv`aS#" "##4K%;w-lni*I^_=Gͣ;f9yQdsD^e^㜽ͻe]v?|Z|hsErƍ [wuFh4}a4Q##jwyPHOQS^Ot.pQaOǩ6 ]\"goyGsk?okSqO$8fR`I-̷7 UeHnUXW3gO8DZƮl~5{ hx뽽/*,#qH! 㺃$WCNT@W_毬C3r`ٱK_EϙeƮXvIo 4/kaҝ(-Wv-L &%pVط `#hn'p3}j8Yru!d?BpH'5'F? 4I 52a_Rc6 " l ܅y>a/gjFrPaͷ]׾w5G4p% ,TA ^r%SulE4DzKK/ 2k`hFǾ@D@D@u Ÿ^W1'ĝSvEFs]_>Z:n(7ÒM`%V4Xcwyygs]p酙1QQ^&Z% ӕ>bÁFØorLFBmidnol!7wKM~3Y_ Ӊǐ BXkyDjuI((ߪ6ėm'3THphR ѐbiDՄsO;ΕM<"ě 2HkA~Z0D͠YTLPM*FЩ7$^tM6ca"@rˍGN8'\;j~sՠwʨ\@ޞL6!IPKއ+[N/P\scɌ͝>3vsnKAF4HXA'$' + k`C,pصew.}g+<,I ٿ )Nn " @G`}bƏy"\)q͐M*#cNğo #O_ᨋ)uCTD\*ipq{^(m jD5nzrD4g`9ٹU'rr}UUW#w&ttetC!؁y7{mM={+ ːû`j]xcprzK$ߺΪ\GpfoSa "  d}{C~;}80Y'04 _؜Ѳa=/Rv؛ʈ"nrה;DF ^ӵh# #" M no}#Z; %8n#I^M/ǩE JgWq 4ؾ3 x_@62d~f@z]61 "p&#Ln`lo]|X?'uN bzxl9{$EAvIe/qİb``xEXM Vڌ-b9}̡J<\"TȉK fP8H MXUũ+^~W {@׏lOӤȠAo972𬃩)eogxg #" 1o]I`@D@D@D@zMDD@D@D  ~ " " ".@.+ " " "Ц͗B@D@D@D@GM 0޴ھX"" " " @;GM Ӯ\ƀrM5E)§ Gѓ[G eDKceڛX_D@D8{P&!V!(H gO7uQ͔Bl3SXN=Wv!tw_<8~NÔ˂q\}tz[~BC(Y&J)8<ԨyxkϛZ2[$jeD@D8P&bd4\in8+@ \"ٚZ~\8p|oHCHX 2m |kl5&2giD@DhE p t  sia쥁GB-<@=ir+:p(AZnW;z}9]WA'| T!pڥ].BWW5D W4͈|\S) }ˈX!VV4;'sI$v IkJ"! aN8~I4O .;i+Q.Q" " m@`kS(:h(.|i .U|~~Q ,ZWG/|B_9 E|/CP*/ _CRCrE?%8+$P +Bpn 7?r]~ɦq`r,W[ls 'BHʖ8%SA\~?I80O HMGk!\ GD@D8%)*&2-@ p ,?AŁu~e{AL))aAgnXR^ [4 \Flif 4-2~@[NLܪfrD@D@VBM 5U<i%PZU-HWN})Mpօs85/U4XQl <|Vf`" " #Ц`J=:theS\hp88gʞϭah5,)`3LC _C4z.mE͈" " mJU9]KC`k``,jƖtz ^E90 >0n3M|fХ]4,6.OwpKi¥-_n t[zb`X$i,ai)%VS.rA It/ 0/zfD@D@0"nlJ7u,ca;c{wti>R@LM$I(B'6cIsJPRtgOTE.q_|TĥXH3go@T(J&,5ĄTe5Aer#Z)jɇŃnxԕ1L}|Fp#'BD@3;gOuDTd jɡDYW,.ۀ `~`e`ŻQl4IW6{W:ZJT: ?rsI2o)Qqp T`6H"|/E kX`d'qVC[`',ٙUBwCԺ_$28BD@$m-u L .1`J r߄etܰU!"N fliqp~#HEEH}IHU{^֢@V)G<CۊK#S@UA?uh|]" " A&g֮X_D@D@D@p!ЦVp " " " mJ`l诓_-SHPHONmƖ\̡J<8\"TȉK fP8H MXj=+BQuYUI w9k&SF/SWrwu0ueu=S_){2,1*AF4^dL\B^'Bu{=2f9c4@9=C!Uк;_!>EUH_D܍J:1" >TM\\( g)gG@Bl< "+ڃk!BD8k 5Ċ gǙg3Z "&ڔ{j@ ^!>3 <;뫈DDBM 0k/"@A`ݯJAkoxzX+W7?a8BD @̎ " Bq9r9HƭNM`" >CD@D@DH$AD@D@D8PE鋅Ou" " " 0iDD@D@D@\T3zh9" " " "@@=@D@`Ie<'{7i /C/ 3YRD@ w?c㖇gϕ801B}J|Av$ƏQ]VGS= =#իEID@5H?I?=>>Ѭ5W4{@/]xgL]S+=ND@@jODGF5r=^Hg)1ˈ] 1jnc^u:+J @3܎EY,E\:"*2S-41H`P4l'"m#8a@D .Mu7jMqu{|D7G4Ǫp3u ^&::KGaqV:kju48HpuθhoX.xԄ "pf#o-D@G>i뮬m㮝ɘhfh}v"%1rHrF >' D ;7/r]!*қ%L@DpL}F"0G[< T۾-&*TĻߗˏO]-qjh%8:pz**N8쎎q{=n7YESn:8:t膝^!8k)gmJT촞 T0qʳ$`۸lUgS=4{,s7k-|dىm[׿nG56[VjF+E . H8&R'KCq-\mʀoǍ۳{ǡB`Jz]?qb;Npl1+bـX՘`_64ZrXT @ѧ ,_tU(OF윓6!,eۢO>L?ùҮ<ks#4qI} `ϣ3 k3) 7#= XlFla+_UsQ qLj$PRN wIK-ߠ'Eȋ&צKgOtBH쫊P.D/\^8=Ea'\(̷A.OOxmpz8n 02\{w" #DF#"ȋ4 IY}pp`sj\.?fySC$*:jذ1 l-Lk`7:oZ r0u G| 6ܲp`anp8N^tNUU՚M_rfY-DEm[8{Ӻi\l`>5{_c{wY믾4#QO40OgLu0ւ"uc򢉢z|步bOg a:qO=i:܊J qu (^Pk`L7?a'Pܤ0  P" "  8t׊~sD2"NT춌,0/ȿAzxMC5\|VGJesgkZ])=;0*kI8!R`e][둷X 23"u142`4Z7,DEm[ٻ̘/y/>ukGs_|8ss02o?&z@g*b"䇭76M~;vOU`)L~t?f#]oܽkgI|599k%VsJ•ymy&s胰Yіk.T^#R3v[VQ }ȭ@ed hbԝsvI ;2TovMD@D@dhShkkFS-'i&L~엱YMij0:}<(Vd _`SlP^J0R:%|&Iշz^!xG* WX8+U j93i)C=Y-]E[Gh񩷎L]J94:4ȅ6ަ7iii O8vm"M4`K,kv՛]erT)`hbx# .u S,( Z(,Be^ܴPNLzl_:}9'gԬ7gg̜a+e4<-A$ Ƃe ]JgNk65Uu)}?M;'fT6ccp4{ACAd;hX8Y- HBȳ($9R@X4Q\nT[{_qeNO{ ā3ŀԦÎ,n*/*SWSfYQ"  "_ڂ;\(Ó#.[-OReg2ۭӡmRe噿o)mܲTdJ5ѐ6hD>)L[Jn4W >[`w*YJl9f9v9&VukX" >y `7>1w^,s3222zW`v V-pf0eI5Wd k8yV)_^Icɳ2_s䛏 ve۪5b7Yk6F}vYi;2p&*.wuն2ej } (_}rRU rNC6p6J ;2jl"@|P" @ΟK>;(7eW(3ȷTDl)m,sEؠ9/bɴdYYV[ɡ~Jj`Fۻ#V514^c(^chT DX(b9pξcfwggvvo?μ{yo;H<u">njL%Y;!miw^M׎>n}yzz5 v)q,,WTY@%ӚgD=6ZQxgqKVo#1 :7N:3SNqcGΪLt#Ɍؐ15t⷏OAńjMwmxrR Y bse͘ɘu*vLS8ZXZٯ|K V[~X]GQ;Rn=멮n8kAR_uW{뛢?K7o}:iZ=fо_zuR/LYە>ɴ%l}/#aEYBI3evW|/{3׎">Ji=<ޚr;1H@J@(Οndпj9M|:v bNm;X_'ΦuV.[u 0ĵAUz`XZ1tKJ]v:>}w{ZõwuV3~pFX`tP;kAM3WA38٘>|At{{hVվUb3n4#ݠCյ|ڼ>MVV^~oaJT9Fuur@Yy ݋Eu`\ͤ1j혻G3zZJW 1eʌY~+);~E“.ﰔܔSXXJxo݊| NXWng=@ $$JwxL*ɔf215'LVcKO5Tq=[;e)cL$Ij6ZEV<('K;E6jPޚjGdG&fR_ Ⱥ1c_u~?4,kF)Ϛ1f3=3n3E, ]W[=ʌ}dv}.IAT eAU\W[7v^7oM6Y{Bx-nPF:Kp Vwz"2Ïur3APu}=|?k5 5[u]D@J<罻~Y%iYdԩi5}Jn7[}w]sw֥brUyy"~ wZlϽ~|JwvmQiVJc|}7p;'ن|G;gν~~:k^Kk3l׾mj4/=cg#Z`K~lyHqmU{N Yua15u[qŚ37_ z{oH;k vDgr΃Oh-&!S˞{t UkU={ߎ_]eC;خ!U(y(S 5wa~`o-0ξy>`nN)tf\3MdNs YC@.oYpNY,RfDn;s t^beK56,gzͲfY?"SG2$>b34A"p;w|ׯsSV⊭Ve-}*χn%ZTT3qv7;&LϿ>Y?ޜH,b³wO߲*m=uȟsnfk(^y7M\#O{#WcWS[2MY^c;yß먋'g-Nh'[R[Ӕndop{:yK@a%BT K׳ĿDK/H}̢|wx~QVwc]k0a)45θWdKObE)uZ+jYxogv7 \lۃ/ 2"3cK#-ut%B@w6^-/Z:oђUzVZP_@ `pM%Y\ _;)^pv :)l Y'tRDT K\jm_X! ~Y EOz?:vM8p{n:x0n}8[];S7/@ ǯrOoIwv_{}V^ڦIzG!&P{NOڔ#ocOmE ZԾiem2J/ln@@w((lA&[謿FpPb2HnfMC/fd̅V|Hѷ<@Rz Wx"zyw FS}[q]]r%oV‘>ivR3cƌ;Go|6Q} kOE?C~{뤯2/{?*,^ %ZjQCVEߖ|R)|ddWY Mm}f|k}iv[.\8k֬9lY>s9W_}u0̥K*QNzS^uU@ @<9v􄉓Ov[1"1"X?i7_37a+8}JCCnm 6_kNQdMkWo}{_LQ'_{.hҤI֓HS6-Asx't>%H|)C>Y= s$!!~tx@%!Hk#/me,KZwt}iL4qW_}_nvwwռosv^pzOx.}[rW\Kɾ6prw*GZBzRuVU༆;-'K?7KV^izLKY0˯c5~P:ꫯfkh_-yiϙ3gɒ%_#4?\YRg>*K%@uz/ܰneY)4״^%ڵkϞ=[[qkΝy]_hҨ:+ ɦJnK_|ŷ~kE 2( [k<[e˖-X@uUی:BFNt4u:`[\ձTkUA:)7͡cmm aZTW_ݛ:#'kR"4R ̊O{ us[Ufϰ@r)e$ 276x.Ϗwl}礍7{ߌN斵=1'>wϘ>t+}BlC꾰^[NuIKYW>6LNb>?O^ם =ICodW^{^-OLo>p֥Vx{yF3|?µf6۰ؼxG6Wg8,ѧby5/{KkIj:i쫕tЋ/OuofnI 67o)v>f\:c5LYZü=P11_~_KwM<ױ\" W^yN3FFN |cY3@R<¢Ioywc9FY&O,l_8po߅-/RCT!xȓ\;ՑΟX@-RI2 zjJ'wUn 3oFmqYgocգuuVK`Ahׁmz['L]2;-i;Q[g*Zϊ wb2Wo3r]9d$ӊ=yr{pod~^ɖ'*ѿ{w4?.=푣|m_=OyJWug'fԱqmӥyBl#t?~[[n/"7)SuQWJ 'plS1W@BNOB;`>|Dڟggy7ЪRzC ;R W^J%4YǍwUVQ;~ivT4yy3g6V^-eg8E4WԼ;l e VU U;3GKʬWqVjJa?^fEO$Eb׿um=ӪnKJof:6裏?j^pEN%yI'W-XW\AÆ }uRg:/ߤERE޶w*2}?kyDI%Ibٔ3]'NiӜq793Un"("x/"h$\ ++Vv)] C-ɔk:i^'OKY>gk.c筳n1p׍3ԛ,eM%HXnmr`!~%gs{_'\U5fêm)*7?v?`_fMMuݬ o3zA]]ߵ9+82[nzda:,Kw߱}YsO-OTn+nL`UիWwAeLF_}Kw)'N¬"ӣҌ=9͖#F9tgw$TT}L7p+$D%V6MIt4 oLirif{=mV/"K3bISUYX*WT;)@uS]RK[ġtiڸNF$&U%fUG+5[ns{,&jKQIJНŢRyjP]Q뮻4hn(6ͺ uFlWzQVW) m@@(NA#Xoܣ5nA3ۤI2p*Mnq6lޔjc?VڷY]F`ו=vKۼ4LyI3l3nw:]UU  RJRu;+;`JCWT׼y Km]]Y\YUUU3 WBW U\C^w;놏ƌy3ڸ#1};N8sCvg;6ly@ZzuO=~GfvWQ߽^dCS3ߝҷ2%ΜRWtIh)wDwt4'_}<+~p|BY$=#8.(Zh *R*F**IͨTғc{%}&| ӫ#qk}HI:/#vDTt%lš[_ֵzL!ľM5xwxyhC'fK Jr垜^^zJqye[']n /*ܭҫ^b*QL]J /jk%1U\xv$VWe٫:lQRLXay3rXX/NRv+㛉3 %d[juB\i,d332cdўBs7Qg* ۔F75{ E3vx3 e6F4!*50[/j6cSWmjǘFߢ >tȯ~ J X~.=vg3Cs3DP_}諗:ooFÎc@G6'ezg^!3rpPČZJv % ֞m|I*ݨRbRV묳Tͨ$Uh~_6>HcS]$:RV餂Rqpe;n:Vv)j8WTU+GPXiT^]9Z,w]Rv7:`{|XꦷrZNRUL7ձ^[t:쪯4'k_%;pdox $^>)Ge; U1!iB[W|9d̳k͒:s-N4,]`%BЋ盆eމ]YQILZa;XjL[i%הkꭷ*%t2JeYZFL TX(i]ѓw'V?)դ0Rq'Q*B~53Vu~СZoIt%tI%Pu^n5Uu5.aU9U2>VeDS%D•o4@Z՘] j ~+f|JP IՁɬ詿b:FhUɴj29 ߔ@7ʸm/ 59EVrWíjhBݶ)H V[gQi.RT!+g@ @>w@_hg-U75nO&l{xK5 vb1-vؠ»{ǥ{kf7j:iw]Iİ5UzC'j$}jUaCںi=TSjkOϞ^.S9)Fr`UJ_uAEq*Z h! :3fȐa5x=7lXof/^9=Ot -Tg~&?lwovO NGi(*&ĉ][|ӼSed-8WN thO4ܝ{4`鱏^0/iJ݇٫_$5}?hgY,t\I5 QvKh&j(⽊iԫU_WQ> `4XQveA5sU A^VъM+RZJM_zlJ5Վְ[2Kjj&EU#]ҚRRh:(Vl_ZIRP4WKCi&Z&}%%䛬)h) z{ےZZK:ozPI%S*TapB%\._ydRߤoeJ$Yq=H+KYgt3rJcV(Iu*٩V:̦u,RnEqyףd͵g̕dV4C̐*3ʌyvJ4<@2xLueq{}Y%xrLc.bXimKJrL3V_'z9Ȫקi nM.U|ypf漵 ~w}cuғ,cfim$3o3WcH|4W1]a}bJ_#DHrh]wsg*)9'Ak4 )|''.)bTMֲ;{|VI*<('9;nk]Uc﷟PAɒRkBmKR4ruUzY`)J*Hjܵ5JђNJKJF=ʒ馛*xmJՊS,i%V,T2jUJ9c1%d9&uAUU'rFǚmQ&"U2TJ Jc׾뭷 jR{ZJ֠f܋UYZ\J*Tc]g죭t^W6xcwyaO/{ˢs"e2SVfo@8wVG}mN[ْu ƙv|lh˅ {@pe¨Ow+snw?1 XqźS7vyYKnhj-[6< >W!LWݶ;SzOm%o{jB}Zl5|G̪{䐾;n~|M(Z]jGuuIXq_/{c1޶1ѫ~xNXGsFUJ9}+G r ɦ3sfъ>kP9RgWJnpȔ>H<7C@tBr2Ka ԯ"G_>pVN[hO8z71wo>K_[ͪU2ύ7F8W!7[QBNRRSHbFJΔ~jSR$0Z%$Aު_ҫ6U 3;*X:J>qTIARd;?V֤|ז9:/(\]R:/M͞YaǾ?A(R2ʒI6˧AQ SĔݻ`JQY*]> _F21{gAEImuyw&SF{F׶\ ,#/Tzկ.IZ կ,v}/_QrC'sj\:_˃@@* ;83y-sZ *w  i'egi7k{J|ėiŠi-U7e`tր]j~k V}{MxjVkku&U6ՓEpST#o5ac\=jĘ}կ5m8g̶Ѽ߾ǝqgiںq޻^`m5HY wsf}3$+땇{(vz#=h{6Z$lWN$2qlkٳWG1mҜy0kfp@/wj*0!;ם{Ki&nDN>]Q>hP 03H$ Em@l!b^5 ۶KO)NkhTJCK+$4}@Cڧ_%#w#cϴ,-+X3ﭴrl|jiR:/UO~mjb|[CÂпVOuQz혱_ukfn$-*^',vUoٿj2ͨuu:ugMX Vѫ1NwW{c?;aA枷ݷo{i^KN͞yo7;a Ğ{f޶9 q{̯3^ٽԯډ脝6'˿:^QvWwԯ=VOLʁa`f& LNT:V8m' #Uݳ.+ݥmҤ8Β!1=yI@)|op7vmxrK(7_t=9gAd\4l^ʓ[.n䇺 O.+^e}c{uU:K%{v׫/h3̹))xxd!fc\:mMM!6pȊgBʚ>+[K]зh… ?_J$K՛N;ԯ_KzADd:[{OAjkegM_k{|*/%t/TOAH sU._zQ oMo=} 466<`{׳. >'GqKvu[[suO<+Az҉MFWҷAۥ=[ +K:S}s+N'4kOʫEu0i}~ v m./z)ԩ4~M:u5WL裏쓓bXFm83g*;Ȣ[;oބvrHR?g^ܒ+&.ܶF-)ELYY6t[N: |Dv5b_3%q8OX|IDAT֑|@A('M pѧ~mx$JO~Wu'RG5Xvv)͉f͛9l;8gu滩;N2;a‡3e6'|an\0)nwtSƙ"o?ϒ{gWMM$K.Dykk&dؓY;gϞ}y,X{۷?UW]#7o޲eƎ W\{ԨQW[m5W?T4c+Gxsung򗿼W:y*KxwL2E2[vuWU嗝%ǯ7Χ6xcVeO:$=Z'8X0>?o=T+'/慤Y :pip5)Dag ~bdrG~_]]]ve%_L0/\kv8a|~7w0z4-3or)0 ZLzFb7ΦfLuO6?;3MX8 zܸq+ 鯚0U{%[ͼ樻b 1&_T0_a|yd!f_MU]G_el>l>'#n `v2o?j^aC$'-~cZ̫;Ut5^thY{ӻii5`Ot|^?_$$Oi?!eK+k \0k|!3fmA n7v^SknX+Љgv,mNoKцm̖Ysc#%_>ܴriY/Af}x:[emo1o o~)3ROKkskϕy/O6ql=ؘB" iUc%MfxN1^sĉ A-, FJK)+)PBnʥA JIZ_@ N;Mc0棏>zKJ*.IjXdz&afUמwm ~{U$5B[bX?$zeSgEk}G4ZEb"oNrO2U:n5~YъT ASנzIQS:7ȾENMb@C ;{ LRXmCկЖԚBݻw_HYǷqs͇9Rvts&7kH9K4/bҮnvXaf 'd)y9gZr/rض"bz.;I:*(&JJ_EJ5RZStYo4!C?J=J: ¹J[3٬U)C9DgD >Xmd(,k*[:ގ{F^o2 Z*I* /K꼂ؒPo:+v/*?`/ik @S%׮Hup(L?ΞfO漖5ѨWJYΥ;Ï6I꿯2=g˝AcOw1N7o~}Ygmf%Dj;.pBZj Wn1W=Lysz:ixWg:uߙg.qGByզ*(LOgaqEhZ*٭*m%VU~%x*Ѽ@ b O]vեSN9EWIGFzA+ dk 4R]T4&Zr؆Uuҕz>fZVAZ ֱfkXѣ4[N9i:c$]֬|nE,Y"U VpXJymlcȬjƯVOijjRۦTZӆJq=v#^C@𹷼-6j6@G@Cg%H4*UcYAU70{fEf0Yik[5#̋w8Sg{47coLs̴ẉg#͔SW鯙XϚfg7y,ktVP'K2=fuͶ;oUVRsfqi7TYl\[2O%%ؤ֤u]UnJI+UZXSg5jW`i eIb$AU.K+B+(Q +?Zv4!Vs dlŹvO$@UמԳlX35ȰļR1HNylF;ܫ2^{F%\l4]_誖s||c=IX5Rq"Gm)Pk*<.z*eOZ@[K[{ G絵{V!`pvy[ +o}UNسW2&MA'A,^G:O ?an +~$ 7[.=5956bv;̅Cc)37_b. Ak]cnrk'GΠ_se3͚zsVK@NJ =R+;7H{6\.%3)#m^g|O`y/xJ̟, /?.>3}95pQ7_g,7:Ze%.Ax)[0γRK@O`ޏߵ4%NۼV$R{LҊLOo&V+Rijjm__O4{=#w\s[TAU;5l4ӌgmʛ̹h.rᝐN*ȩ9wpJ++ K'EO{^@OWv=X$@ED `5O~*@YY[ԶL8ajS_>v&-֒/f:-XR0v\|y}s裍f~6~Cqԙgpp^F=+([e69a8JXEV]N5-Scp ]'@]w[ KK3m52Yf0$2&_"5cȚfFխcJT5myh\#UV٧Ke.*ַj}ڧ[>]uon7+F,欳ūke?|:" ~MmՔQicONƜ{s|yif=F4h}>($23 @@@jşD4fKK^e~3uf\Ynkn<Hvf @2pUY+?SA-Srݼ%1υ4o%6644A?--=9mj4ٳY#rn/PZ;<4Oj~AlX;pz՜H==-Q֝ҥ-)ÿ^iTZ=T2 @ ߎv̴qȪzϑ퐙 ɼE S6T\\.ڢ%,-ц&) ,IUeMY iP=q cNZM`MfND$J Y @T [2Gl }R? ")%Yd,.oeyopt3n㖘c=n+9 JF[گ:9{Ż!ݜ_8r^iՓoVx[Nz6@{#~aB2X"Y9@D . Cô? 2VV1XWd౬Y/͗FL9;unm М-νЬuF O]U ܖVSͲ_';6BÂ[MĮl-+D?_}_޽{g]O?}I''fُmf)U{i;ЁT<$>}t0f;EPEE$.B@n_xkv=33ߣ3M~;ַQYz?̊)">gkAÞ5W!_ujH_UHXi:|mK.>{R\R17߼Xhڵk~zjU"}xmΜ9$yWy睷mj7[X:ctd P҈NeyzS6M?v3YL 33 [H7{fRr HA f|/#:fnWP#߀t+h3ֹʑ#ιuc`Me}~_d}YT:7… ]Z>gku{[GYRn >b:+?뮻+$>=zH^, @@G `-; V[o2Us'{v^Sd"eL׽\ $ {KT%6Y ;B &[6 VdoС+5?Ϝ9Snj7(:|iӦ]w3UDzuOW\1k,/1]zǿ_J 效˖-+˪g':.X@O>c1cF_+x)Iz@Y j#9sy:wީ)z׫]v[oddP^G ^@:y#9~y+ Ma_(zc["ā*뮃JoѢEҖ=jO<[n &HSY_JGKΞ={:)u 7H8u]:?\$\Iֿ/*^SoQ K,%VI4id=iR U )i=QwIʁN8>=?*ۻ[eU((zW**\ ?^/B]UT:kMU+eW+5eAv,YAW9mܸq5^|E+3jkV4l.]TwHuU'o$2zpRwԲ:믿^Y _L)RފH?ˬSU?:")vŧ'!@Y$}Ue_VL{V,c(RZs9$UTDNz#F-}ݥH5{t JR'?>VIIe$BڔpRvԼYd%u^H),HHo'z6%PJ2D ve-pF'wuW5Cؾ\yVVV[n[{u@+z[oB*ZU՞WDZ38Co5UF6g4p@e7xjj[oU|TdV|n N>묳dvw>#ߪ(pګ^^{H? @Ș@ɱ'L|Ұ]/onVq{uͅs?|^ޒ7]xOO?]UG?5@XV]7խܔi)hGFˢZ,mZx{y~?S~dfS;,{\pR><|}YK&ix!X+`m& x4JbIlRWꫯի"K˿D;c% .IWKIX*QaL;;ЁX~xRUmGWrƂ5j?7(p%fnd5% GuBF+xKv4 AH \7)4b\,Y/P"coYpNY,Rfpnn=)@b[dq]c~q讷?,j>~7`:{;SK2snw?EHcV\./kP`u,ћ]׻gr{L}U{g=/#||$oZZ[ZMQȉߚhcT5G 7t `* RQč6H"YbRVJīԲ"W,QQPUO"g-`$$+l裏ʲ!\$--HܓU-t^[Հ۳bjh0hܻIJ \RsUk7(2 4<=Imv Z@(@pVQ8ݽy `n4G45homj61XTXܴ2ҧW6^颰-TɕRtNlK8)$kKL;ZAȓ^5+؎BiԱDKImֈk%֤Y[:Ms9GIڗ0:UTS#TbWzEJWP1bQX> -wOvvMU)|Eb'4Y[nzV_u@!б,(kȲj>E$WUJ#mudVnT\t0l0ݏFlYZ>*K/Cݰfi TVDm=X+Ԭz qA|VuޡTMF>zvE.հ-YW\G_p_{j_u'-`ʗATn~M]}n~ׁt[ON"n^HiZjA5~%S52vmUpUIJ2X~5WeFy!+%-']5zVW5WvS-+;*U2Tq^3u^cw5qW5WONk2Ɉ 0Ivas[lGMeV%̴9+ jXUTG)m,WUX*rCL9R8e- :IJE$5_ZUcŅHCFnFjTȦET V }C92(˺kxj@hyg vfJJM7K[mҕ]ET}vgR";^T8yƊ66h 3y;NmZh۲zmjoiiѳ Aim|?r) D@J{;SK2l^yYŊ/ؔ؟1M%wWI3NCF *]˥s5>7go5 ͦ47S1O=;Vنd=@1%@h 0h.CNjf[6}$]6@K~kU;vxoDE#hkTtpE0.e C?f"6 @(@Ǟb+:" "p_C6/q7u2mU*f}+{o|w.r3P)I@ H C7_9ϲc۶/E@5Vy/illr[Vd[ۻ{S4/1H VY8ll-@m8rO!йg\X%†7Jijg߭{_reGmr1^~f}Ӌԃ7XwG;^7䫃w{jka  #zt1:/yw>ytr:J @ p}5+Łua߬`d%bŝP=@2#Ό 䆀5uߙcf~ݥO e×4LZ?v|ug4f} @@Q@u< |_3=MfE扳̛ۢ3g|AmXb;zy ͗MKSCC @@{ Ou:)fL7}{ʴYr wr¤ɱ=x.~X33Cs拷̛Ϙ8>,Qo?4]lffMs4Kכ9ߙ0oN~flym8 +"Όz6@(Q,'hlSA>KIY 2 m u $LSTqp>g9aXOEsa񚛚0ĜDݞׯ↌FkR1 @(;gtdXɘHHy-;=Yp<o^?Ok5Zqn]eml{?1l@ĆTA@I#)2$JCXN@[lтX+LYs{ kU3qsujSz֥􈭄Q6"yE$.B@n)@-j|bٱּf&s^Mxº 's Dqr7oUAC +b˲^u2n،Ř>8 a&Ҹ`UjHO,ЯnE…N-f<33oPس̌ͤS96ͷ5f<݇0v̷S?6ت1V:c W m.I>v褫mׄ]ZNi4>)n'\ [x-Ⱥ8wȫ8X'|M;=!c&'t;xg| T_F g jxCz"_`|^m>|̟os4ؙ{f̩ϴ uZm,e^T 0Z:zn>^s/e9^$TI>V؋ay&]ntC.rHK$N0`aƳ # 8I@-T@ on*UM`.5}ga9^wm:qsU7k=6yOE@[@~E^܇ p u^λ.0 aM-OoVЄJa&l0ɸ0n#͢Eڑ檑f"lNsni-3gN:̞}ys3sߘ^4'aTt7oN0'j>t曞ݻA.>~ PPm͑F-~oA5i9pr1p8A'"2Hp }YPg9\36gi'S:v@od o6WM7̍0iWRjSO5+G3/lF2b><^{SN1jwn̈{.ov2k;!b6@ b?ִmu|rO FCSpٳ>hUbhxЫn,d}my (gJIydwi-ӂR\N,N&;|孈I~&,WSoo]ŒϕN2zևW*II" +LnKsh czr#޽콷yIӧs,uu4>k@ @ɱ'L|Ұ]&_`O9p ,ُ>^2n>ЦYf\D}~{ /-q?d酬Uaz$?.ݜoke:2rwR./kS~dfS;,:ܶ\`㷙S+Bvgˢs҃W|tTEѵ>ʖ,klXo5eS͜3/j  QΆ=l:c9]#w}u ƍ{Y*m}@̛WL+ d@ {K] [i*y>6|ZEknZsTD^Xu :U@1hYiV1y na#. ؟uk!>Ytj/zCM! /Nx@ @$n'@C @@q@G;% @ @) `)abgՙb$n%yz=9-i8eCL`J,dΒ@q2ii(- "JTS\ϝd@j'Ta'Etӧj̮![LNi´ku:.ßMLgA3`D@ PB NLXtB'n%CL=yizӫjz٫GϾݻڭW׮+*UTt^V%{H$ @Rp1+67qMײ3d<+IIjhqb{+A p 퇪S"%a6J°jғmu3x? ?Jn9S,ҷ3Lm+byC|>Rrؗb؞=NfIhu%d^gLbmCS"OSw4pғЈk0GW]݋7٧/|O L^e}Tүj}֧}U'J_|:p~R @r~@NW!+䄹MB7!m3kŧ|2Ҟ&H' ʇ#'KgM^ w"{Tp^{qHmĐZZ8YNlZ[Z[Z"-M9mj4EZZLsTHsggba @pb)H&)۔ٳBgFpP:xhfQ\7\g)+eڴ'o0Zv,ٳllnmljmh4Kmh2K" 1e,^,)!@OgiF Ӥ)}N.@)qN|Һ֨v)^9{Ż!͵K"5 @!2a ;Uw 4k!aϾxo H`Z͠6 ϔ(͚d=A%I8v xڧrf=.jrĈHYD2G_W9;K  @%Ip_Yi`a=l}ձ7Bo* o2Û+ {;I2az[\=4\ @%P8K]vurty{dT.ϝC;%+Tx&uw>;ܹvu*ںo @ x]_k>.塓#:K;\_[>[}dz;G>?||]vOx_lo5 @F ;)^3ng5fVGfŁ,:Lomlm^n-K{S~*ܿhױOMّ6e3k{ @%m35ϝ\x>d 9G(]7H@ &P4ѸF}QdAxO6'kgPv|6 .%Y_A֥'M>߳H`C,᫖05r(e' u@n=9LKC xCµׇޛ7 i-$K&oPyxtO&LɲX㫋a|罉&I6. ܷɐW֦ttWW6cT-YHˠSNd\k+iT_vX @x>L&~98a-7]' hdBku)@ʦI?ΌB!MV{ %K\}2B @@{K1Ї3"8oɬ.)+;*KI LsIӞf"/ @rD#dNQ%i!wH_0ܝLzY}3E@ \@f Nnzw&dB!l6AeIfM,Yw;^Sg}S _,1m筋/KJqpJ`p?u>fpBNco( YB A @ nqJ_,qG}/x^;"{)az Z#^\'c^;{Q Z M#髚kdHSxBd} iO CXQ5&d-XT\_)zuo( \@ PR`s]A;'D@7q|zM,>k&sҫ!Lf3A^':NV0Lد2O֚ ,ǣ:/F2?) wK&K |K,dKfӵm誀 @H!]U; N2J)7̿3! MctzK @H=Z7GV 0uI65f`ƞ.c1L(,@HC @p J; \-lM$@Td›< ÂOHk {9@ ']Ҍ";cզWƷ $\܀^g jꭠρ9͘@A,؆.@J@ 5L]^n06]]qFޞB)( V > \- 4p%{MP8D]"#C'  @@!yG3fK8:`\tQ*"L 5Wh> Sh@(SEQ @EOW9Ml+Zy ~)e}E$+'@&9YF_b[kj?,< =h?Th aZUU~5Zg68@: ȓcGO8awڼyMMs:=.1 @,% PxE^ @%J\ O!\U{Tb$k@ @ @+F!@@;  0@ g F E @L D"nt}[̵w@ @(9a#ؖE l{* @ @y$V%-E` @ ` |@ @@' yr '4n2/_Zn7[}w]c,_ryerƦwG%ed,@1hm^]S~dfSY27:t wOC{]*9 VM-+ב1+Xug8:Vґ~VIYY͟2|ٻOZft/mP!A^p~R @+c S3^`Z[[MKi:늵mskt덺lvo^o꾽f9AU 㪊~r]D~"/dnwlȞ9yp??t[޾MύEN!!@".n.\<_fD[$z#QkKKRS;۫w?̹-r5lm[r@h,"kykCQjWmJ8]b $$.NV Shw7qG-ii2͑kF65G[۬iQ Vwڼ-Cjq򈴣3>`wPkǮ";"rmECĺGcS>C~˯M A@ $WDA@f %\Jӛ֫}VxZ"-QI_mW#n) )'e)q_I֞=Ú+Wm8ø7&2+bEsS"}Al!|V^9bN^}u M&7Uw: @@%$?LK3q,)+-ث_+WLin6:~[ԯRy0gs^bYC4RVQrf)G#DZ[RL)3UDś9^YnsT/Rn꨾QJKο6ͷLs @HE o "ipo(n87`v{^Qb-Q{HP$@7J\k?oe[Gx;cP4 @@pCad%hmv(NMV77d`^J8YXÞ5W_;Yy\нi8Vhl-w9hsbWv3s i%*+"Z ^!Vl=Jhc^ڥU_w4Jސ5$ @ y w 4SkIfqF`%>|^"vdk;~򆛮:S i~Sg\J9?ݎ]E9YnsT/U=:[vTH!]2 >r6hSlp:˜b!@Rȫ,|^}͝  "77;`xvcKD,: Q7]W51vpLmBǴ3EOBin٥ (m+/wt_PmU۵runńwpF ~;&3"W9+e)t!2|߇Uq9&YI3#n @ /,t#VpvLWZz-kUFڷ P,btCx9ڑҾ:κQ>UJNpTOvVܕ=3v` ^e +2.V㔟O Wma*qq8*0 ڪi<τOmm^s]S~df\z=7]# @.ͣm觡Q =|.ݜV<=HM\qҨ?ݕRtfz6s~p|0oKF;gjsv([w#\i^n֯f7Lvivbe-"aCHEEa*87Wp@G Q):pgj0u)D쉟q4JCKP}@h?pBwvu' <\xyIbI҅͸IHe*u'!r%N|vz *.NƣOؒ]tN6tJXѕWΕ pYCJ(TƸfY>C'T{{jNf }Θ7508Aj@"sKKֿG]"X^Z]=%+.J[|Ӻtɲd$6[}K/cfL+`| dBp&Ԋ9O8Ap -@lm/ӣ.| lH; @(QC(A[I[[0' @@ ȇnv@%K l/@ @^!G^O`cK|W^~kcos`vw=4tPLZRm' @@ <9v􄉓Ov[*nG5u/np"X;o;:Ցj-(L? ؠ_߾i9v#vvie!q#07Lʶzߺut.}J@SillevPJT=~1/kkz!R}<^2MםOx}yy^_WmdX՗'dkrEz{v wC=nݺ}jwF?٨S_nږ=v/`](:L./JxF8MVk;1f~éw0nRf?h޽izk :8a{m;7-?ۜ}ߪQ䉯2>3C Sm; M3a3m8g 9Df^Z_+O8 ׵OyH&~S}mݟf2# ֛!'_P"#a>\7z;Sk&UWtQ{M!ړ'{Ѝz¨|"C pp~@X[B8.}\5OZ`Qxl? =)ړ=y/K: CxffD|rnsƐ߾Vf?7om@ouR Ii0W\qm=ק8S|fʌ -6Y[3kwZ0Į 7P) Zv ,xo6+F^̋̊@O~ />mkn[7m#͐NX0.yUZedȑNEw߱}@ SO8blnwQ`ξ sĶPĶRv*LBd;Z%}xޙP_quw+/HZ@y>''sJJV'uI ,h"QWڷ!olㆈtᒖ'^eկ9g2#|jzït .\'jvVskT3fquuu++W3b Nb#l3b6C̈8F>\6y?VpIO2 &=pYn>z߾YcxԹNiW:ztU>VQ @/#LKwq[6*b]hyfZCVq]8 Փ8/'F S\eWŢx4Zu'(FͶt}Oˈ4pF%Kr.;^o@2tI ,,a:!:7-$%y z-[;Ԫ;-=v\8]{7v۪WKXhT3䡒mb;+ͅgg6nJMS/턑ě:qVv'(:=tКcL/7eZ宁k?Mxwu__@ ,Ⱦwuȃh,ߏfwnn}G{_[[Sn78U&vp>ߦd~PŭgXy}Q!!iseͱK7WoOz9':~TM(P%VÝ;WB@ɔ8TEGkv6C.v<_';n҂Q#nq¿ELYԸxǍWւ:_2gdsMm+"R9xCK K{>V+MRuՌ)g3`Cq3.xl&C{Umeu[Y_d|i-{x=WƴټR5s»oe>Zqw4k>`! 'd_WaS2%;fᣆG u(o,;e׵[U^l;1K<ܸ{/G~hu͟rJ![;v΁YnSvJ<|ăG=ԍ JXA8n6C 6f̨']ꇙݏÕQڙ S6& @ !ݟ X׮ru8Mvo4z4I^C.1S/C2tpڿS:+`7c9ou\&!n`iAsk.[թ]%یlPv嗯^$1h8c>~免C9䫯RRWc} $559w~k׫vqjs[})bÇ%xuOMִnOg U^c]cU%ɂ(q5fГ9raCkj v͚V4z?2e ޘF>دU/C*uj7hx˶-Cޘ @$jw[6pؓo'wn'{ѱ${򒾵3B?++)g]22؂/Ijzwx) &Kצ,ֺ_мy0B2%hnkN<,oӯTtʲ@ɔ8aJԚm>kg㧌7e8-"U3f|͸q5vp-qݯ1c|{ZOˑDٶX 8￟}MϞ3/FW:e}! J*j7`@UV$Pwկh Li!Ą%1D.*kyơCR;ç~/vkmjKh~dSK!p$l @&õk;sow۴+,K(Niܔ 6z4Q{3pߪƚW啻dx(4暕{񞇩r4?\sU `ӭq;WnMi$-%e{ RD L|^6X] r^%!&Vk듿ТSc^ ?sݫgOyTPis]7 W:)֜vIյ#ꫵרԌ)E*Ug$we!c. ٷ*t' ]uU~6`&s~oolj=@ @ *^ 2M->%ڽgT_o/LYl/餛F'yorc ~OvqZMGjf`[Ǘwƾ{wBpbrx4>. `mq_yџ}뵐Yv;8oon_NZ`#i{*G>z7_&+N@ɔ8a #>v\֢Փj[^7;s[Lz}gfίhluoo}AD]ε9uMXm0pݘKUAu6R >q&Bo1}kLs0K*׀4GZS tIXeP wGd~# 5;ÕmM%Ǎ5[,o~,o?0i!CW\t&?v}szv=Wd &͎lσ}ٙi,r35͊V2tԘ%{Q;prVҲXizZ5+*.k $piug:9J|=oOzYoz'dY27:Y@@'&FF~W=|.ݜ١ʑrHNHWlZTB+/kz:_$G*V7:u?y)S'wdsʅÃx}_Qd㫙:\M-fϜڙߐ%9yS7j?oΙ}/rlFw4sztGLd,79Sè5wpUNVǞ;! ,\ХS:7Y >w=8֥`{p?:NKLM6v`MU>wՕ5 6j]I?#=<;ɪ_g-kG[%\r~@B# ?\G2 <=|Bخ=!؋)vC@C 1p7^FP/ݳ.+ETx1\U~c^FSRuM7:ȉ|)]nuVk]v$nݺ=uԪ*VcT=;ԣhP;:m^̬ĴolsbgI:E拴(5wa-`o[a\{sWJ8GY@}kW5v[ojKϿwOܼ*ӞU^<9J|͝ծ/=i6>I~dopcJDڭfA eK=/ 5wsGY5@CzŘ%#t| շK6gqk{rn0$I@HF\j}EK-ZJϊRk @@`@CzŘJ/_Vd@Zx{tA;B@ (= % %o!d[C@V "":ǭ( @scZd,—2ey% C@ߜ|{ @ 'ǎ0qIv3o޿<9 Eg& +)k .T "׍#f}gΉ/cfi̍=:BKpvy#yi",|xS9u FϑE;aJ[Xgt @ PELk ,\}h#7 @@H 0y-s )oKi'-h3qƄfי0H_{ޓ:Nx>x#Oi @F xz aݳc:+ܺ cJyv;-? M{(a]҂z>}vF;{lv jպIXS2X W~xI @ P R [ǩčۣBUǀ$+Q6}=w%/Leࡷ㏳OUe'p4a>Yi[>})eɓߝv @=R X /&SYk)x,6md~ץbG`,!_|{>l!wC!d?~\t\ɖm+=bV[sزr\>kBQ㭵Ǿw/sg(  X%g FhgWey=~%޷<o>=YL4WZ3]Zj]g2U,W578x28'  @hȓcGO8awF6_Y-ڽ>pooo^9i5P\ߨ=ɛ\'0BiU'FϑUgx6K7ҹѹ<%Bj0j^CwRiDu(V7vyYK + e!B@H;9@@@&J[IQ | eu#@4.0epz @%JGc[v  R!l/sKPO@ @^JLbpEYo^ _abp !@.C`EW nnTVY@Ȉ8#lEp5#ti\XbtsKu0:hDY%HRRY ϔsÔ#*CiJ+)QDjSt4D(%e?OyT]SboeSZ\~6.6K暅3͒X: 2@|% Fo0F_ S L)bvhqp”~MBvI.R .#ꧩ\Mu=D5_jKe#Ot#UK=%TndZMK_T@@@X[gJ> -l pI5*[(xϖГemm}*@ #d,C=)@ @@N Mɩ @ @"s @ <H-4028A @ ;-ޕ @ H!ef!@ @ ҈C @rJS @ B!.@ @)pNb @ PZ? @ @ B Hl˩ @ @(lS?0@ @)P8` @ <@@ @xo< @ <@2E@ @@@w| @ 8) @: @ @ yL @ R [ǻ @ ѕ  @ @ #@ @Mn @ b .VG@ @h7pb @Z !@ @F@ @(bh%| @ v%#ea @ aB hl0) @ npK @ @ ;( @CΔ@ @L @ @ p@ @`n @ H-݇44 @ @@.!ރ\M@ @@N xC @C _u@ @H uVA @ @ ?R A) @ @/!@ @P % @rJS @ B!.@ @)P>)~` @ S0@i` @@ ΃@ @)pNb @ PZ? @ @ 9ŋq@ @(Bi  @ @/!@ @P % @rJS @ B!ZGV q @ C c @ B!Z@ @hp; @x O!@ @GV@ @(i+< @ v@Y!@ @x  @A ĶvBV@ @@%=x@ @@;OV@ @@A@D3 @ 8ׄ@ @A\̀ @ k\> @ pA4N@ @@ sM @ PR [Ax @ 4 &@g, @ ':> @ @R`q`[{1@ @`wt$@ @@ .{ @   @ @'?sJ @   @ @'?sJ @  J'!uw @ @P'!e6f @ @C @D% @ &.;@ @p@b @  @ @ KY@ @(ln @ ,@g $f @ @ ®A @H-+o @ #k @ 0] @J$JB @>@ @%A\L%!@ @L @ .f @ JGb @ @K ƶ$C @B `0A @[!@ @ p(L$ @ b'.@ @E K@[($ @ X.n<Zw @ @ PfH@ @(l)!< @ @ 1`h@B @@js @R/*@ @RV%;A @  C}^V  @ @H-;A% @ @L @ .f @ @ @@I%퓐J @ NJ A @ P$HPI@ @y @ ,'/@ @@I@D3SI@ @@ @ @$ K$ @   @ P%T @R  @ @H c% @ Z @  ;A#R@ @HM) @ @@wF  @ 85#R@ @@' H @ @ 5pjF @ N@ ĶNP[@ @%K ƶeD!@ @%;A= @ 8pw@ @T K' @ '.@!@  0AJנ @:>ɻuS@ @()}} C*NA=!@ tF@;c7N @@}zYW~={Uٷ{]ڵ{EEʺD"vw+Z| @  };s~˔ږ'L}A/5i鸷|1Q0[-0`@_ @ _>]*vWUj}֧Utۻ*=~#JV9A @(I34̜2oa%%KEQuzm6F#6;ۥ0`dG @ @KkE769MLJHDg>%h*f?@ @Gh/++܍Dw#hkTi1DE`R H!@ @@wFJ @:=VVUc5Y^ vD[bCw@!@ @O0} @J@(lT<$ @ tR3ZkJA @ @NTU @ @@2C @M}foAoXUl9ن=@ @(}rˍ~.@ @H+?agU4q @ g_E"\͆ @  #WD"Qmn]CC% @ "/bK @ dB ?>/)Гe`|Q C NC]h;MHc9ZW)[f,++Vn~5+/kS>~!p@ ml-t{]*9 55#e:2r=튭IV[ / @ $"BK%ѽ@ @(j#zhnf @  L_ @ &Bá8@ @I d( @@HVۢ) ? @ &PCg?f @:CwS @ @()pIᠲ @ Y ;kR/@ @X@ @%A\L%!@ @ @ @ @x <x!@ @%@ @(vboA @ P0 @[!@ @ p(L$ @ b'.@ @E  @ @  @BH-#+o @ P`R [; @ PR PfH@ @Zzѳovյkn]uD*:WP @ d-Sj[&Oo0)~PK_|wK_{k3U%ͶXpc @H`ӯWYe.}Z.jntɾң8v(@ @2Ɔfii2͑hSs)b[HyD[7F8m9@ @ƦֆHtoKlhCYinoT,^քCP @ 䄀BF^ihkTso.kpd6"9i B @@N HපE}1"RѾ̑&zUu@!4p$uC @аgM~Hejh.-#35(Yg>pl  nS!% @ @ G nr]9N92X+)P8Gb @ Mm:.j`~*\qe083 @ $;MˡyAs8{@y` @ @КܫU5WbXs/R,㨑O-X +V! @ dKkώn6CϹ@HZ ZKY-_ʖBKY @ 4FQK4p1ft"XgtV$ @ w1MѦe-M [:n454U457mAӫ<=cӄ@ @) M}iwmwsO]{U6[?'|_u_k0&N>ino)TVR=g"M' 0C- ڑ*@ N@@6m-F~T*$p;G+ /6)۬黭nRQjnGNYӼ>-d`!pʶ e @(:kv:|-ӵХkO ԋ`ş-bH@ @() څVN @ @ ]ht @ @j\^ @ @Hv@ @(h' @ yJ>H@ @Mߔ8@* @ $ 0@*^D!@ ?L.G݃dNEPc\ @ R%!ХʐzC @@B.A @@#=CKPo@ @%F\b Nu!@ *p< @ Pb%T @@@jSo@ @%F  @ @@j]y$ @ PbR Bu!@ @s@wvV @ #K@ @@qX}v!@ 8^-]Uۂ50Ļ Շ @ Pu+0| .u@ @x="4@ @(z)կjJGb[ @ йaZUU~5ZgU1ֈj΅@ @@[ .b@ @t݃+*YQ?@ @! @ @u1 @ PR` 4AoK9 @ 'BşUYĻ Շ @ PRes60C @%Z+@ @N f @ #حá @X%. @ PXQx@ @X}ͽ{@5SMe&pg4  @@1tmq|]R b?>C @@7X*t5 @ Ϫ{NT  @@`t55 @ P9 zo @ @ r?hT w @ @( ) @ @M\w @ %,  @ 6ȓcGO8˃56RpdY2H_VJ@ @'ڼ(Bt٧r| hFG['ee];6}+H+Kx㤄 @ VWгw]c579ejZZMkUF[Z#ZLo[[oe㵻~7YfiPv9Љ3^'U}ra{&Y#= >(>P]&>M|m6_zِ=Wgt*ݖ[.$ԶUlZ[hDo$jkKKRS;堁?YfmQz>I?t?^RG|=GO8>Qp' OS14oޒ}ws#}ȺzKWjS-bpg^kEƽدԯnKK46GZïmiros&H|bqʢG)G=živC@(IF/d(*ݯ>PiӔɁVMmʱ˯ c `>`bôde$^`s#U$>=RdWVWv\ڌ޷q}*]to)h]c9Ҧ9amLim}㓃儩},^pv; u幜l@6="32mP-肝(>PGwdm4iʠ 51]|=z1W}օGbfs-Whט Y[R8]:]~W\q}m>$_Ұ*&_}u ΍.~O_R߯뒥 k46`%^g3I;347r {uY,/>P|D)1>M|24,Z Kн{oMWqrlw-hw dŜv,fb#^ }ݭUu&azWe}hpKknz%  @ e1c*po}dX! &ôlF q-}[_u{*aI^3~5Yu7~9tjZh8^AWX׎Q=1a@:0{"|$%NHI&Z(>P~)>M|24 ڊ- fv嘝S]ڥU&3}r,~%f[wjxXuB!ɦ{f;?Ǫt/ ?yi]I7< EMywQm{$w镄PCRY˧` ۧoAE A> A@: -K~?3s\ۻ\ ,eowfgwޱ3<~ΒS/ߍiz <[\>(X9ß3jݴ~]G ULMqG r?O)N_Fٓ|@ϘK7|֜c,#l/t띄 B^02>wɦ7Vsq@gܘ3PI8 0kZ/j_kS)b:g>ZlD<$5=p8G#p8`1#,C+z\pmXTd,&IA`"QR|"+ 9_{a20T#j {އ.=CԹ"C.:sjD-do~`1BEP;zs&еϙS8|뇥7T<BMn nr VQ[c}X`?8@ ,~e@~}/fEcbXŴ^:[+ =Rw.>M1)Po G/>͑na ts&K)? oP>S 0&t|I3>.Zm/{mw'L qZVG`!50.ƁW._]Z} lpުhy|k+[5|A7 ϙ nRp7 ??Ovph'<Ȥް'J.:*>RukU#1ǘ<{E"{ޝb.z~Kl p۬-v?.n(E?g~Gy: sPw¸mׂXhmw/i 7]IB.\YOHIG#p8G#j3W[*s9BPkYQ6!,akPuR(ڝv&JSؤxÄ3:hUH\4Q"p8G#p[2FKdұ=ŝ7#p8G#p;kj|p8G#p8e O'qTȩؼ1$W9*yYd*rhv |NE%}{%UնYqD6PJ_NA sY[s ;tMy?q:p'.p8mݸ vM_-^9G#@f`[jaf1>1MŔ p'.p8m& mrG#Pߓp ȓ]8|{r)K]񪤸 8GG<czr8\`u}BVKq8GH!DGwK&>,Sr,+KJrR[G!`sKo-/w,]đKg2+:tl1$t:"tT{Vǎ')GǍ<%稨yrL:'!]JNC s8@D&N"x۲-uJq8)4ܿo&clX)*@3=~>xK.hݝ*[ʢU?D(;MDtBsVnvZОEk/;sI~'(T㯽BR4DUj/R02POt p . :G#iM.撶Ei5EwKW;_U樯ؒ5^ruЩdb߶;H~#Nw\禩HJ2յoKrtiƁ1,s߃ O=}5Ww޿hAgsj o-vHIii23J\}ƿT&?+'҅ W-~tWZ* 2[ϽcnxynXOΡ&2tP<ӄO$$&'&%+#q!C0Ba"FCY~\bM>jo,ehm ߶os\Y>u,L0?VNKd ]m_/ܻ[d*-+`wD⹡~Tz )[9/\IۍgLJ=f_阰hov@7Tv~p8@4ۯ(v+;Q/^W̙d xƂC*fЯ0k峾*HeU#+6'_h~`֫/=y?䅡$$@~eֶ--@`;b" @dF:d_ǎ:n`;X2oims/sg %6TOIf|ٯ<^W~7a^I-*V4,}r19777??_zr*4o 0{_% 7Ln=7n$;7?YaSJ:oXx,&2mAٸ}RkܯVW#Pp~PH̶n0DY~.&}{kkLl:kG]:{6=?; _#T~xۯuG#8[a!;w%ӝ`Nzv,ٻx5UjJg-+pJBC@!*6OʙOslaڻW?JCU Q.^ogFV@)@,P%Y)30 Hl0r"V:Y E(Mur)מpoposj~qҐ~BO_QFnK )~tPb9jVl.4dۖK3]fkw#,(5NO>`)MoVjR?Ta^_P1$+$r󄼂}c`wjJ=q^u8mMowWnwjʼ<)%xX2֪ G=oNl#{4"L'*q<7p8q_J49)wez~ 7c{̇&J vYꏭ[l<93g!4` Bs Tz>y~[;J>m<80/}Ng J|K zQJf,@vrZJeU}^8g0ZMOYYBԔ" PY1ӯ^{O$7^Uy5mq? Gb_,:5נ6 4O\#p8MXz2Qo^o+a]Uv!ア!fWuD/yLG62ʕxiG/2xq+vHI>qؙ|<֯N?BI~',Uz: %& 6l0*ƍWXX/f}HyC܄0v;Q494g4 wߙ>I,>Dp%{ofyپ2#?0&e% JgM5qH%`s2~ڄ^PqΜytDZBew^s_߸}/7gfew3l Ң2|5N6CF_V2z0 -7mˇWj).[5W3&\tӵ's)TLM)cD́.hA?_: -Ár}^p|_l);(t+R'G#74`v]R K*ON+YzOC+%W6,Q%' 3/xc/*k`-AOKQRH>2r+a_Y|[4yށ<3nKl S3"cם[)=QE^!K #OCA˂ W11/ݘ1t.-+`&@eUՉlees.դo{%%Ň^zCU -zgWrYh*V4ޔce$YA(~Eb4]lgSbO@񏪿I236b=<#=3 =I +}q?(#WSSxMNw ZTgYm !gn=;gup en勥]A=.Cq UxPo|ӯK®{k-$N[PP?jO)zfUO8G#8C~k,#=8+O=$)? /OE~g#+Cx^W]Q/IV|[F_hPhZTϷjHf1>3ްve~z4/^l?y#Hx-zNecDM 1+P5|ƞ3z0v,+3:{d1vHdeJArb,F^ylc_ yݧO?x͏Ѓ@Plii/z!'ǒ^$3`_Xʗ`$OrͅIKI)ny<80lDjF{=vIZʞ3vT'[?]/Vępnc CF r),{YyMWk=ￆ,ZnzK]~W= =▇^^ jg>Vjڱ`o2fӡ( ;DCT ,dphf*p8e YiZXώ(^2r/0@qטxig.ǖOpM&=#CY⟷>U- Ty@Ya_|tUoAa埿~7{' al;mZ; ` +γ5L=cWEܻ_~{8|&g%GeXJ }rϿf9ظ~sݷͻOR7ct/.*ȇDf{*Z-o9|{畋.r n~K弯$v>oF~^nN_D*B\DʞY[fϛL UD̩Ud)FRR)dNq!3S7WYO B5ύ~B41!җ vktgvڣ h}Ϙ%\:wx}%x3G~v[Ork YTN._.uTIr yFG#M85ѽ(Cw_~5k}su% }UyyJYOATג#wļuTMV1g_w0,{y'ݮ!* 'gP|B2ʭ3n߄jgT΅g^4%Pe6/6a\Q 's^ϼ b_~JF%u7aMiӐȹϐ껁=Î=>$ѸwgyZ|vO=#g 1z訓 _#xDBeb\%O>;ÀH~x9{z~r{=JVsz5Sȥ8G#vh"Q@IE?_Qv*[~Ag|>= ߭miPAO6*S9g=rFj0xSbjb2|nn| Qo<H-PK7Vv*{sMC}+fwHL+cJقpWHtC|؄nۏ_ܶqգ/7TyV|c ;K}k ^dmHc+Q?Nixy2&汯o=~;`u݋[ĥ8GM#D]bci*wũ_Q2<7GUؿsU|zӣG}ZT{s]F hΐ Cp7.R~-I o+Xa咉?+9/ =.=S߶ЮΗ+رHq`ÉKq8@FmVmrG#Pso*:xc.?{2NU$ pHpqaG#h4p&ȵVZq8qo@ҩΛ#%.xu"'!uNC s8@DW5V5$fEy8G#geEo^ z2R9@w}ًW=mG%-J+5$wZRڥ~o)-Gkp8@k ][U#p"/6JDq .p8GCcklr\4xj$Վ@[:LоT  K,l`¨步^wzٗoZqG#wn%3Vq[cTGcU`+Q1IHұWYr*xК-=ry(W^/=BC#BZ@ rYt(E"Ob!5 s8G#p#67lNI)RLK-*ܪG -Ui*BE[oUَ wVdBM̴PĨ[+_ٶY^)P:9Jlc_vK9Pc,\G#p8m0 V>/T22ϛ@mKۿR !=NلL|ץƊ| "\KKm.-G6𲇔+ [=WxPT7G]KO0gym}w;B}B  <(J9 \G#p%T`_T戼Qt!="+xD^è,~JJdRջ:Tye R$x/NlaֻOZ+5kԮ7eJa³p8G#p"{c=90c}0ػdjWT5RJ8d U#,IF7by&s?\?G#pFT6hy:ZqGd|h4ȅ0_B#Qo K-WPȿ%Ņ}"5|#Mf4"`?쀯v9G#p"ߥOI%[5$f[͑S_|j(K&û^a_0},r^E<UN$_@aKC" lW98jo4){P,ꖾy}a_V)ˎ)۝)3Ii#2X G#pjK#k t.Kpj"k -.]0Z%f+tA)\OU`j[[?·BUq8G DKo!vD%\_"p-g/|8Gy^Wr2PxUTh[[?' \G#"FOuy)GB ilёRk9>/#pZ@8s[^*p8G#p8ÌfyqG@01(y)]6)[tdڥ~Mv=^)G#Llo$?E'h$H(406ű(/+I^f@ޭJc:.9飈ji ci:I5(TNtKBpFpi]bkܓf 0Y9G#p8zG |.]ʓS IQ:bg;fH9=lYG!R* 5DE$#AP 'uA@Uitm"Ѿ &v0nL5aӉz56.;<Y*mYØ̬u5p8G#T-:~Ow7( ^V@@e't7b۞DDt XKؤ҈:RB(a2u¾ mjE]CfX5qM ,$ah5.%h_ ъ0SLKʂj r¥$ Q fWZ 5p&Bl ""p`8Rͅ*1ql٬gkUr8G#p#YQ)"F8njpLH&1hK]Gdf'EԖK (HFS2Hb#ݦVPWxL|˥tN\Yj^V+i u'&ZbhV#Z\VAS3, E(x2rNΌ]1J( MuYm MQͰێZsZ `$NA6^C:DC1Ԕ8pUy\Lt*)̑쑭@v';v:0 Xz~5ccE-oeW4 Cákm618 Ҁ20j´`.&Hp`: ؚҭN;-8z Iʂ_{l Khvv_v@?|)9^st##'/9jHlz4 nہ(}"kռk_G[IϜmnΜdbr~i2>퀕XhccSIm~L@zL*7ubn+$ߚ( j2eq3,fBأKX1W/P*첁@.y}\*J!?rWx@I\=]XpZv;YK"_f!vUJՉ/"ĴL-p.R>ya$)SAM6`Tn;,\NUҜ Z3&Kx̻f 5[|06omF`YB%zэ=+b2p[ey9!C 2n=]o,9&USiqihz3a-*lv7 PkmGm[ؤG#8)ع)3R7g!XF@^Z$&yIc|S2I`1 1h1[,. Ĭ넇zrlMs )YT-pB?,(tj0 즰܄ʺnfH՜.P&ӑY2qB0F]M*D ^2qxAuEl=\ۖh8}lY?9SOj'vf篥_Yr9Ek`~h"`>TAެÐf%19([4UM@c'W_IWw;lEu8nQJNWM}Y}MCPY )'zX쮏crk1$L*6kjw\9:-U۵-WUA.MVpǞ}>XjnrfQFSӈCM3R2bFRSgbsQ"$3$j9OZvF Bs-?z /1g̯9w*L3ΌXୌyʄ:I-a&X=Te.6|PLՌ.4YܱAx!GX䃴?$X5zxZkvO9m qGҭ(&۬CH{& 8{ࠏBCZYò:&U^t뿚՚>vC5ӭWEuϗԣ-C88PsAK8pձ}_MtK8As=bs(鉆0aDŽywvh7Յ&Rm,N3޼N!/MSaYv>DeO*/7J]E{m:too()R=/&~SǭI*K̘69*bϟ< 4vK+*C 2vx@aN\z硋IVM ZH] y}#@2WYKy4$yU¾^v%w౽zShV>&3 p|L|l!42GKh&9/b6uˎEK:9qN/ mqLX=_u>S@Np&uꕝ%8P^XfQo#*vZzpYNdf^P_{q.K־jV̵-sc)l+[ֹ6:6h\OhN9-rcqI/`yὝ^zs1j6PZ\VN=vZ Lgѯ4iMॷΘ1c֬Y9tK_)V9 i=W_}ꫯ?0(8>3H >Sm~]]AdzR@xK%B5J-Ӟgxg%Ÿ{^pFXk*m!Xfvzѱ[;ym X&zw<%=_ؖ[޻>W6=-,B 3*Q>~9r ݲ U9= #̙3^[—a7;8TaKAN Z!?0ZQnFqhf.;Hg 2cHaݯ%xP;5}82ut飍brZCMEÊu#e44وW kC8x.G_8Gjإ,QCҸeh8p/ _(XpO"SC{>015($$ASk0AD{^p0_G{DvEj,qQX!w\u5{m}.%QI>fuQuN/+w!Ac5Z7oGAee;+i+ IE?'$&El5Os8؜ͩ§K>1@יW:jtrA;b/\\|_ҫ['{=3#΁~8V?%+bh~ zh9Qivh]|@: Y2'fTȈ:UN\=v vMH[7r/);z>!OjZj~=c.[b|r:W5 /1x:d|˕gL}ܹs۵km۶K.$.7,eff]^qe`CBѨDⅤU\;krP}< XrGn%wh/C,3%=&x{PdڱxTNsooK VY,604u2{ּMѹ$O<5Z}wܼDΗmzox >uKHE’]VXXqFp`_=6CZP* 9xao&}lLF:دQ0ch7X()Z޳W y>8Ałuhۻxo512wN|>8#ZYIfW]ag7b.|d!ߩ pAo`YNr0ݾ"Er0U#M BSs?~hJ3D*]tj \wر~үhpů3odp/C*MbI̷8)/RaF$&% !Nd YK8EVeȒ)BiƊ6+lo.&{aul.϶; $`$fR Kpit$ ljtaR] gt[-g/xvѝ?ȟ??kOxT/?CM:etъ]l=y.\zJ^=w T9ϟҒp 'g[ >)Y:*q8 %`גijzMc K#(1y؄;d|sG2l%v<9ԵCA}Z z6JEnbv`ʹP J`_e.*i0W>;i*3Y5~syݱ.80,=<$#1::@.x2*Kqι랇0#;X±$L8쫞=7QAt 廚,! @^BȿJF@̪ɖ"=ӫw|! fnc* ^8vN8ypS:+ˏ?#|˙<2wї?ED"[B/;3g1쫇,m,̙X&BFh(}< [bň#y3"%]7 QфCe]Pu Ya#IewRnN'q]sFrOʃڝ7!Փn?ox5ksvHĮI]/;uj 3RMA֎AI/:`e;0pId0=Bn]Yja~ —&Ց]]3@LH8ʷs\kzf!3>;(Z |f' ?[Ej3q[)YA(sP{9ww6$"|v5Yvqp]z,&{tolܴ| /4\z9Nsxxlw2{;9v"xjR;t5YZCJ|"Y Ҏ Ix]9ʗ@A>$p8 _TWFЛ_9Y #>y2Kzȇp9"].G?&IIh1J+@_Moժg^tDo9pYgxE 4_=YJwҥʳ`C@2VWvLR$ 3 ,R8407K1>qrᝧd\-5NCPѯ{^ڊ(#.jf'qGx)O י97ϽG$AFJ\E;]ٓR%gE[xcc?uy:KJyr,Q. , n<ɱ Mݞ%=XvOE'oz>A1Mbs `ɘ7 y݀VS{j0gGGMڲiq$z&]J}v\aٙ[Mh- TZ۴vou0_TܘwŮ:B6"VD,PCվ{LW YXrƞ/>BwM~mV z饗b^,((Ʉ2]vnٲ)  堦WIxʕLS`80ׯ_du}}=qO֮] c 6wi̘15 C cs/ͤo?_' ۯTD$:7j{C1V@ksnW5W=X` v;)lKt_ n&Ƅ^N`M5iB9YD+Nmf3~ Sug*`hku*"Փ7#|4,}*;NXǞKt?u:fh A &R`e^u)ДWJR^m].fbmjXy:@U%9IhmyŎ{g6'/2.s-QGЕt[5q԰_GO;B&_wu֬qҷJ'O}߲OYLcq=PsLc4DU!sC3.~\0 LbJ_`ЉF .5BG1op'd䀸EGi5dV w=fFi'(#;3ׁ.l/gQݥCE;&Uu]OH@(CȓpĎ*!PVԳbpI+ z[Ťkڙö͈z.?U5䇡T822ƱignDϺMx/a/@:f%II=)PtJVMR ӣ> Ԟb#۪{a(r+"ૄSCr4O(I6l/lr3<+:(\WgQ)3-YYK^b{2w5*+]HI`|Jab\_-0ۙ3R좖.<;v"K$vKFk:-V$b;\yqr?E,)iەUkN,SD%ettDahFTטNIC& ]':ew.[ٰ{UQZML [ ףGw_kf=”쐙Sxh BA:7?/觯D>TAтtX;k4qڬ$MZ\d!$ޠI0OcQK6ŀ.5JXKt*Iqcpˮ&W7X{Nڧ^ʽ8+&X.&DCj@MW?cyZc[?T+7iLWeP~iĚ@U2޺gŁKo\]Sgc*zP}[*sb*kc7 ſ ''fN2ҳoDwz EKrożRH7҃V/</JG $=^|9tr4^c |<4-z󋞯V^p=&\oZt}m,Ee 㾋TϤ8g-{u;w/z}ˊ;nMx|Wizn} $FQ}uϹ(8=4#+N>E7yWѣg21em?A~YF'h pુvaڛ0TOR.X,( 3++ol)RX{$ܡtoȸJJ 2<&$$9ezVp۷/K)XAH%H2;U8p ,:5Jk)e1_%i]ߦ`vkqQU,ܶߘU㒼ok`HjꜨg._#'n+v+JK.O̬wqQFp[L} tY O&,XH]pӀI.90X>E,$c&>oe-C0}Kx$Vm^!oت N:/^Q:d4cJRc$+vRDͽ.iW#lǐߝ]`kNx$#HN} ELVw:uNWӥs+FGBb`Nh-}?cwCtbndcg) mnkbzoε_i{_|/ ZѨM\+K$&ȸ.;Nè+S`"g.\!=V)EףD O0hc4F`ѥ_~]=X"4=V`v 1Kd KY_WnXI6*NpnOg]8k7>W[~ZI]b+le5jж.vlfQsfM9 ""+'`5(k('~{t v̿-thQ{Jv/ڰ3Mmt6:YA૴cd=!29CoA197 =A /@VRjt`^Sly]kvۓ]wA+)\kOS#D?HZjM5/tWjbT|ڕppdOmӚeeW)gl`*E VVVEw#UKV9vAc޹j:Y]L "gbTً?N,CL;Xm&h@"W5ڄ(!% 6 56(PcEkRot`}C)75[:M,ldfvH|?ˌs.Q~N2lG`VuYEˆ {,xIY{m>TY aFK$?eu5\iFpd%_M=/7_zIBS/h$S]%&1*sn&o,ؖMz.i3$ZtYmDR{MOyEdO\ڽ4hV'gBl]v]> çשY!a\EOpo)xO,γٹqq鑘A`۾}{؊-66dW618b+[I~17lɰK 9-XEgV/)3]ō!hr|[cJX LߗzH ᑪ}NyԨSOž賐z,+ѭ=SSvm qc;JuÕo-)ĝ1Q 4|GpÍ$ X 7?@*h`l%tJ.m>؁[&PU`Z?T˪Ln(M{kNټQMF1wPI̿+{V_Vs5]] %vD s d%]R&t/*#qA3u[hg{]H2P['ԛ,fV.^O}+vt$ $;u }x h (ځ1_`zE߿3NMKwCLM6jcLOMBYc.'zQdn/*GX݇,.!'*P JC;hDeu9~'?2:О+zIh!*::&!Iti{e%2+G hLY&ݶ~ĭ'VbNKUVW>#詊e?JhJ@N3dg_{> _ãE?>}ɽ/F>sr:v.u* -=,OZ | *h._^<*'N=tg~2R?u&suRS\qRzFkt6u/wq 1{̻*M?xȳ'XxCxB>4M*mL'>F7f\S: 3/}{N7&Ô8C.P,A cbYBZ4}UUUm[>4@Ҙ#L2|ar,_Vb[FpEBv0_0wh9ӡilGd)`$G_t/1 I ߲ۧU,5ˁ }ckx6U'fjչdpj5kMraYY}ف}K[߷PF;t^?.[󈆾j7m T6h_&Z}P$gCBFe2o~B?eKa} K/{~n:噵.AızqI<0ޚ&|ЃZ1?gg7Oi~ynjeOr#;f]v^5q S0%!x~D٬mZ誴$k9`vϖ ]ŒG> 3R8ܡ w.^ӥ^nѕ5v:FpԮ;{CA7?dFɌ!w]gM$JK&k⫶{[*E^9JbJ.xDkF692 |`,y [t:\9vۜ .md&kUptJc1;A-sF4ڷ9Q't(~liŲ-kVB<k+S`WwH7A*hBb*w\~9gR>b3Z--cGT(..V?xH=ekQS ,|ehl+/<'}&_/m  몿 MgIBZLWf kGeK%qi=dk̳s'Ϙ=! 䧀L+@ '+}%=r(JRWWF=`1Ur+f^UدBBdM%(>tȾo&tGvioر=ع~OnYz_{~[Cq"-b ʍ9b5uWL5znzP:Q:Oz2?/bdmdwKVSu^a ~bH7IDAT`Cf)pfF+XtAnbܣG˲ SVDҰ{}0CT ͈73r0,,oibrӚA#GֳY1F»` 5 Wv/l.aŴd/, `4)`rFqՎ 6؅Eg`0kk+][TTnRԑw LKVQQo1'1mh0o=U:)>ͯvs`g;(4Y.ś}&=۵h9JԶFdp99YZ{wP$ ڜM(`A(4zptv^.Фn4sfINxV}ʛ3W[jދX]T@ŶCd935s&f'Msq_|ۧk?n\kdNLƹ*WqW k_;$DtT{sHtR0Wl88ls_7%.΂Ífc~!_:KV.SS[fY+wݲ߻v:V/Ui&/Vkuj`ԄҦJo|vJ'D\߽ZH &QbUOtR& z$}> 2bfmʉO$G%We_ukk?/tH(%MqoF%x]N؜vZg:kF0[M^6zWطVt#QZ3~/:1*ktrĮjИAJ#׾c\z|B|tlM?h'9Giē-j_c]ZKaVQ'A*ѩ`S83O;e!C/N1()~gO_< p:;˭%8GERfMN`nѪ MkO+ڲbwQރ %e e*+]qෝMFS3Y+ Ɨ 2 %Qt5>'e&+rIN.Lq!M{OsM&cX?6z;7|{kY} 2ͶV$$sql\ZJ\4RS,B~+MI7z؆5BD",mu75 $_4=h ͓ĿjĿыyoY8b|NZCXX[cЉz`'V&Qla`X".2e2Vk9uZVc<ר&tZmV KAKboa%-&ZX~CI%4X8ZOBnE;6dLIE,Rmlt–5X]tDG z®(H& ^Θ00o^LfK6fKeV =$ Pj\zU:8p^nD۫,_gWNԙ`eQ"~`c "#SEJ& .= g{U֮-q> 6#OGޒԾUbRmmQK?9R?,H5Yh,~CKj]h$4 PTW&FNԷv ^` ML?> n8;(Q'bVb%aDs]0F 8[5%2Jj5 :LK, Et6h(ffƒ)$+}5QL`e$i$3AwHt/O9_2Sls805b]vϹj9v\hg& )UގܭRLJQF8ͦ]+ae8!C ^߾A`MnXфAf`Ի-s.WRapRX:UWoHS[uwCTZw8fzNK% T^$Ԡ\u)?GT Í<hwF,d$k̖q \K/0E*j)sjd% ki{NXR<@{"wh&2%XRr"+NL vs$x48pRa@pIzXcC}1l#cJ PS ,Ҿƾ&nწTfoR-)W#]FM>2@sЖ XEX%DA`1]2VlJ08[mSS Lß1Ql#gtقIGMzLm.`<s`Dat2Έgϖ@"X"K@2lsN]thSp6gD[HoGIga 22A ;PUZ+{ćDRI4C%_nj4(~YbGF}"yI=%o5*z/j"5N{V4}?5!d{W VRB~W!N DgF$o@MjDA,n  54w uy;UCj/o#{=Z5 (3.-A@X_SbeNbcFHwsgGBP\FuF$ݦTwa1_s2 #~tdbA%UdDe!E/ XxgL<+Cisw1rVap؟)?%&fDuw(rRLf3ۺ-`h81 MkaaF8ml(z<#p8G#Pi%yG2طVɐ7qo] H]Y&#Q2Ƹ\56Ud4s摒)G@bw`+4@6X2Ir+JJo(ۚW=J1+mz6 nZCdZ5  $$Challenges in the Translation Market%% Increased demand for translation Globalization New products  Information Society Shorter deadlines Online resources Simultaneous shipment, or  Simship Shortage of translators  Babyboomer generation Lack of graduates from translation programs Result: Translators are turning to tools for help, especially Translation Memory (TM) tools !Z1ZZ5ZZDZ]ZZ!15D\, !Translation Memory (TM)2What is a TM? CAT tool which stores previously translated texts with corresponding source texts and allows for these to be  recycled' in new translations. Organization and Storage Segmentation Alignment Retrieval of Information Exact Match Fuzzy Match Sub-segment Match Term MatchZZZZZ5Z5  #Impact of TMs on Translation Market$$ :Translators  want to use TMs to accelerate the translation process Clients  want translators to use TMs to save time and money Current Situation Ownership Translators   intellectual property Clients  TM is a value to them Payment Clients  Demand discounted rates Translators  Expect some compensation Other issues Cost of technology, experience with specific system Potential solution? Pre-translation .PP PFPPIP P5PPPP F 7    4     Methods of working with a TM:Interactive translation Translator works one sentence at a time, consulting and evaluating matches proposed by system Pre-translation The entire ST is compared against the TM database and matches are automatically inserted to create a  hybrid text , which the translator must then editTZ^ZZZ^ Hybrid text  Pilot StudyObjective: To compare interactive translation vs pre-translation to determine which approach is more beneficial with regard to: Productivity Quality Translator satisfaction Limitations TM system Participants Languages Texts  ZuZ.Z Z3ZZ #P.    . Hypotheses;Productivity Pre-translation < Interactive translation It will take translators longer to decipher hybrid texts and find formulations to fit in with pre-translated bits Quality Pre-translation < Interactive translation Translators are obliged to work with translated portions that may reflect different styles, and they cannot consult all solutions contained in the TM Translator Satisfaction Pre-translation < Interactive translation Translators are obliged to adapt their style to the pre-translated portion, and they have less control over creating a holistic text Z*ZsZ Z*ZZZ*ZZ *s    **   Experiment PreparationParticipants Domain Building the TM and producing pre-translations Documents and training for translators Evaluation criteria for evaluators Pre-testing and refining Pilot Study Experiment ExecutionTranslators: Blue, Green, Red, Yellow Translation order: Blue - ST2 using interactive translation & Red - ST1 using pre-translation Green - ST1 using interactive translation & Yellow - ST2 using pre-translation Time and Resources Questionnaire ! IG! Data Analysis - Productivity$Factors influencing results ST2 more difficult than ST1 (Green) Results of Blue and Yellow argue this Technical difficulties (Red) Familiarity with software Number of results in interactive mode Quality of finished translations Method of translation has no significant effect on productivity P$P&P~P@ZPP$&~A   Data Analysis - QualityIT - 5/8 higher scores PT - 2/8 higher scores Factors influencing results Difficulty of texts Quality of finished translations Amenability of one text to IT IT produces slightly higher quality than Pre-translation KZSZ9ZZZZZ/S9NData Analysis  Translator Satisfaction(( pInteractive Translation Not enough screen space Faster and more efficient Better productivity Principle resource*YY gPre-translation Time consuming Too many windows open at once Inconsistent style Not principle resource *XX  ConclusionRTranslation Memory systems keep translators interested and prove to increase the quality of translations produced Conclusions reached are preliminary but merit further research Increase number of participants New languages and language directions Different text types and TM tools For more information please contact: jwall035@uottawa.cahZhZZ9Zh9/   0` ___ff3ff` ___33f̙ffff` CCC` d___a%qw` *(Z___fFQt` 6CCCff3` CCCff` CCCf3f` f333ff` f333f3̙` ff___ff3ff` ff___ff3ff` ff___3ff` ff___333` ff33` ff33>?" dd@$?Fd@l A@lF`< n?" dd@   @@``PP   @ ` `(p>>   P P0 (  P4T ``  P "`` B P C HWt@AB}C}EvFGxGH`I{QUNVWX}} }}}} `$ }} }}  G H I }} }} }}           `  `` "(`  00B P C HXt@AB}C}EvFGFH`I}QUNVWX}} }}}} `$ }} }}  G H I }} }} }}           `  `` " hB P s *D1"`` P < "_^  T Click to edit Master title style! !* P 6t "_^  RClick to edit Master text styles Second level Third level Fourth level Fifth level!     S P 6t #" `` `  F*0   P 6 Vp #" ``  t H*0   P 6p #" `` ` t H*0 H P 0޽h ? ff___33380___PPT10. Eclipse   0 _ W ` T (  T:T ``  T "`` hB T s *D1"0`0B T C H̑@AB}C}EvFG3/HI}QUNVWX}} }}}} `$ }} }}  G H I }} }} }}           `  `` "``  00B T S NP@AB}C}EvFG2JHÊI}QUNVWX}} }}}} `$ }} }}  G H I }} }} }}           `  `` "   T < "m]  T Click to edit Master title style! ! T 6 "o]   W#Click to edit Master subtitle style$ $ T 6 #" `` `  F*0   T 6 #" ``   H*0   T 6F #" `` `  H*0 H T 0޽h ? ff___33380___PPT10.:   0` 33` Sf3f` 33g` f` www3PP` ZXdbmo` \ғ3y`Ӣ` 3f3ff` 3f3FKf` hk]wwwfܹ` ff>>\`Y{ff` R>&- {p_/̴` ff33` ff33>?" dd@,|?" dd@   " @ ` n?" dd@   @@``PR    @ ` ` p>> p(  p p 6 " `}  T Click to edit Master title style! !$ p 0(  " `  RClick to edit Master text styles Second level Third level Fourth level Fifth level!     S p 0X+  "^ `  >* p 0䂴 "^    @* p 0 s "^ `   @*H p 0޽h ? ff3380___PPT10.Jk Default Design 0 zr (    0p P   p P*    0p    p R*  d  c $ ?  p  0p  0 p RClick to edit Master text styles Second level Third level Fourth level Fifth level!     S  6p _P  p P*    6\t _  p R*  H  0޽h ? 3380___PPT10.pS>}  0  $(  r  S ؘppE p r  S pps  p H  0޽h ? 33___PPT10i..U+D=' = @B +%$   0 <V(  <x < c $2tP_^  t  < S 3tP/tsO<$@ 0 t H < 0޽h ? ff___̙33ff"!___PPT10!. &+(0/D!' = @B Dn!' = @BA?%,( < +O%,( < +D ' =%(%(D ' =%(D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*<!/%(D' =+4 8?dCB0-#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*<!/D' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*<!/D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*</<%(D' =+4 8?dCB0-#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*</<D' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*</<D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*<<R%(D' =+4 8?dCB0-#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*<<RD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*<<RD' =%(D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*<du%(D' =+4 8?dCB0-#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*<duD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*<duD' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*<u%(D' =+4 8?dCB0-#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*<uD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*<uD' =%(D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*<%(D' =+4 8?dCB0-#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*<D' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*<D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*<%(D' =+4 8?dCB0-#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*<D' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*<D' =%(Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*<R%(D' =-g6B fade*<3<*<RD' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*<RD' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*<R+   0L0 <(  ~  s *XtP_^  t ~  s *tYtP_^ t H  0޽h ? ff___̙33ff___PPT10i.m+D=' = @B +   0L0 DV(  Dx D c $gtPH"G!  t  D S htPG^<$@  0 t H D 0޽h ? ff___̙33ff___PPT10.P$B+E!kDT' = @B D' = @BA?%,( < +O%,( < +DF' =%(D.' =%(D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*D%(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*DD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*DD' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*D%(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*DD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*DD' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*D%(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*DD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*DD' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*D6%(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*D6D' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*D6D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*DCw%(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*DCwD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*DCwD' =%(D`' =4@BB#BB%(D' =1:Bvisible*o3>+B#style.visibility<*D%(D' =-g6B fade*<3<*DD' =+4 8?HCB4DBCBB*k3>'Bstyle.rotation<*DD' =+4 8?RCBBCB#ppt_hB*Y3>B ppt_h<*DD' =+4 8?RCBBCB#ppt_wB*Y3>B ppt_w<*D+R   0 p\(  x  c $tP_^  t   c $tP_^<$@ 0 t H  0޽h ? ff___̙33ff.&___PPT10.}0n+;Y"vD' = @B D' = @BA?%,( < +O%,( < +D' =%(%(D& ' =%(Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*%(D' =-g6B fade*<3<*D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*v%(D' =-g6B fade*<3<*vD' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*vD' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*vD& ' =%(Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*v%(D' =-g6B fade*<3<*vD' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*vD' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*vDc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*%(D' =-g6B fade*<3<*D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*+   0L0 R(  x  c $ tP_^  t   ZIfFA PW0  tH  0޽h ? ff___̙33ff___PPT10i.}ª+D=' = @B +   0L0 0(  x  c $$tP_^  t x  c $tP_^ t H  0޽h ? ff___̙33ff___PPT10i.}м +D=' = @B +,   0L0 0TV(  Tx T c $tP_^  t  T S tP7Y<$@ 0 t H T 0޽h ? ff___̙33ff**___PPT10u*.V+;Y"vDI*' = @B D*' = @BA?%,( < +O%,( < +D;)' =%(%(D ' =%(Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*T %(D' =-g6B fade*<3<*T D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*T D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*T Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*T 7%(D' =-g6B fade*<3<*T 7D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*T 7D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*T 7Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*T7%(D' =-g6B fade*<3<*T7D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*T7D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*T7D ' =%(Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*T%(D' =-g6B fade*<3<*TD' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*TD' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*TDc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*T%(D' =-g6B fade*<3<*TD' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*TD' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*TDc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*Tt%(D' =-g6B fade*<3<*TtD' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*TtD' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*TtD ' =%(Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*Tu%(D' =-g6B fade*<3<*TuD' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*TuD' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*TuDc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*T%(D' =-g6B fade*<3<*TD' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*TD' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*TDc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*T<%(D' =-g6B fade*<3<*T<D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*T<D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*T<+   0L0 @X\(  Xx X c $dP_^    X c $`P_^<$D 0  H X 0޽h ? ff___̙33ff~___PPT10^.+h3D2' = @B D' = @BA?%,( < +O%,( < +D$' =%(%(D' =%(D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*X %(D' =+4 8?dCB0-#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*X D' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*X D' =%(D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*X %(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*X D' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*X D' =%(D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*XC%(D' =+4 8?dCB0-#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*XCD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*XCD' =%(D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*XCj%(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*XCjD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*XCjD' =%(D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*Xj%(D' =+4 8?dCB0-#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*XjD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*XjD' =%( D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*X%(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*XD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*X+#   0 \(  x  c $"P_^     c $%P_^<$@ 0  H  0޽h ? ff___̙33ff!!___PPT10w!.*/+;Y"vDK!' = @B D!' = @BA?%,( < +O%,( < +D= ' =%(%(D' =%(Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<* %(D' =-g6B fade*<3<* D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<* D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<* Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*&9%(D' =-g6B fade*<3<*&9D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*&9D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*&9Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*%(D' =-g6B fade*<3<*D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*%(D' =-g6B fade*<3<*D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*D ' =%(Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<* &%(D' =-g6B fade*<3<* &D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<* &D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<* &D' =4@BB*BB%()))D' =1:Bvisible*o3>+B#style.visibility<*9%(D' =-g6B fade*<3<*9D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*9D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*9Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*%(D' =-g6B fade*<3<*D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*+   0L0 P^0(  \x \ c $/P_^   x \ c $0P(     t]#  ^# #"&3p( 3   S^ <6?I p #  T 46 min (ST2), ( #  R^ <l;?[p I  T 46 min (ST1), ( #  Q^ <<@?tp [  NYellow, #  P^ <4>?I 8# p  T 60 min (ST2), ( #  O^ <dI?[8I p  T 51 min (ST1), ( #  N^ <xK?t8[p  MGreen, #  M^ <O?I # 8 T 55 min (ST1), ( #  L^ <W?[I 8 T 64 min (ST2), ( #  K^ <\?t[8 KRed, #  J^ <8a?I #  T 49 min (ST1), ( #  I^ <(f?[I  T 49 min (ST2), ( #  H^ <c?t[ LBlue, #  G^ <o?I ]#  VPre- Translation* #  F^ <l?[]I  ]Interactive Translation* #  E^ <v?t][ Q Translators*   # ZB T^ s *1 ?t]# ]ZB U^ s *1 ?t # ZB V^ s *1 ?t]t ZB W^ s *1 ?# ]# ZB Z^ s *1 ?t# ZB \^ s *1 ?[][ ZB _^ s *1 ?I ]I ZB b^ s *1 ?t# ZB i^ s *1 ?t8# 8ZB p^ s *1 ?tp # p h ^ 0 GP  Blue and Yellow  same amount of time to translate both texts Red and Green  show exact opposite resultsk kH \ 0޽h ? ff___̙33ff___PPT10i.`(+D=' = @B +G7   0L0 F6>6pYf5(  dx d c $}P_^   x =f c $tPP B   4  Y f #"F: |||{ Y  e <J? Y I25%* #  e <@9? Y HIT* #  e <ě?j Y I40%* #  e <? jY I65%* #  e <D? Y LEV 2, #  e <?   H7%* #  e <?  HIT* #  e <?j   I14%* #  e <غ? j  I21%* #  e <<?   LEV 1, #  e <?_   PYellow.(#  e <? _  I28%* #  e <? _  HPT* #  e <?j _  I73%* #  e <<? j_  I45%* #  e <? _  LEV 2, #  e <?  J3%,(#  e <d?  JPT,(#  e <?j   K60%,(#  ~e <? j  K57%,(#  }e < ?   NEV 1.(#  xe <?f  OGreen.(#  we <? f  K53%,(#  ve <? f  JIT,(#  ue <d?j f  K25%,(#  te < ? jf  K78%,(#  se < ? f  NEV 2.(#  re <\?   J8%,(#  qe <?  JIT,(#  pe <?j   K27%,(#  oe <? j  K35%,(#  ne <,$?   NEV 1.(#  ie <(?m   MRed.(#  he <P&?  m J4%,(#  ge <1? m JIT,(#  fe <4?jm K27%,(#  ee <,;?jm K31%,(#  de <??m NEV 2.(#  ce <@D?   I0,(#  be <H?  I0,(#  ae <pM?j J0%,(#  `e <0R?j J0%,(#  _e <O? NEV 1.(#  Ze <,[?9  NBlue.(#  Ye <L]?  9 WMargin of Imp., #  Xe <Hb? 9 WHigher Quality, #  We <pi?j9 JPT, #  Ve <n?j9 JIT, #  Ue <r?9 I, # ZB e s *1 ? ZB e s *1 ?Y YZB e s *1 ?YZB e s *1 ?  YZB e s *1 ?9 9ZB e s *1 ?9ZB e s *1 ?jj9ZB e s *1 ?9ZB e s *1 ?  9ZB e s *1 ? ZB e s *1 ? ZB e s *1 ?mZB e s *1 ?jjmZB e s *1 ?mZB e s *1 ?  mZB e s *1 ?m mZB e s *1 ?  ZB e s *1 ? ZB e s *1 ? f ZB e s *1 ?j jf ZB e s *1 ? f ZB e s *1 ?  f ZB e s *1 ?f f ZB e s *1 ? ZB f s *1 ? ZB f s *1 ? _ ZB f s *1 ?j j_ ZB  f s *1 ? _ ZB  f s *1 ? _ ZB f s *1 ?_ _ ZB f s *1 ?  ZB (f s *1 ? ZB *f s *1 ? YZB ,f s *1 ?j jYZB .f s *1 ? YZB 0f s *1 ?  YH d 0޽h ? ff___̙33ff___PPT10i.@eU@+D=' = @B +   0L0 @ :(  x  c $vP_^   x  c $oP_.   r  S PjP~ ^     0HS f< ( General consensus = Interactive mode superior to Pre- translation mode All translators preferred working in Interactive mode Interactive mode is superior to Pre-translation mode in terms of job satisfaction8 oS H  0޽h ? ff___̙33ff___PPT10i.Կp,+D=' = @B +   0L0 ``0(  `x ` c $pP_^   x ` c $HP_^  H ` 0޽h ? ff___̙33ff___PPT10i.{++D=' = @B +  0 (  X  C    t  S t 0  t  H  0޽h ? 3380___PPT10.P>rpYA\ #P׻G)=y|@?u c.`S|nc,=xI ~$1[,( F/ 0|DArialgsRomanTTX-ܖ 0ܖDTimSymbolEclipseDefault DesignInteractive Translation vs. Pre-Translation in the Context of Translation Memory Systems: Investigating the Effects of Translation Method on Productivity, Quality and Translator Satisfaction%Challenges in the Translation MarketTranslation Memory (TM)$Impact of TMs on Translation MarketMethods of working with a TM Hybrid text Pilot Study HypothesesExperiment PreparationExperiment ExecutionData Analysis - ProductivityRoot EntrydO)@Z:Pictures,Current UserC>SummaryInformation(  0ܖ A .  @n?" dd@  @@`` x_   r $ '(*+,.  5 6 ?,b$dMwP \=.], 0AA F8$@T" # $ Pʚ;ʚ;g4LdLd|f 02ppp@ <4dddd w 0T4- g4;d;d(h 04p@ pp0___PPT10 pp___PPT9 4 ?  %O  =!#Interactive Translation vs. Pre-Translation in the Context of Translation Memory Systems: Investigating the Effects of Translation Method on Productivity, Quality and Translator Satisfaction(Zg=Julian Wallis Localisation Research Centre October 25th, 2006,>Z5  $$Challenges in the Translation Market%% Increased demand for translation Globalization New products  Information Society Shorter deadlines Online resources Simultaneous shipment, or  Simship Shortage of translators  Babyboomer generation Lack of graduates from translation programs Result: Translators are turning to tools for help, especially Translation Memory (TM) tools !Z1ZZ5ZZDZ]ZZ!15D\, !Translation Memory (TM)2What is a TM? CAT tool which stores previously translated texts with corresponding source texts and allows for these to be  recycled' in new translations. Organization and Storage Segmentation Alignment Retrieval of Information Exact Match Fuzzy Match Sub-segment Match Term MatchZZZZZ5Z5  #Impact of TMs on Translation Market$$ :Translators  want to use TMs to accelerate the translation process Clients  want translators to use TMs to save time and money Current Situation Ownership Translators   intellectual property Clients  TM is a value to them Payment Clients  Demand discounted rates Translators  Expect some compensation Other issues Cost of technology, experience with specific system Potential solution? Pre-translation .PP PFPPIP P5PPPP F 7    4     Methods of working with a TM:Interactive translation Translator works one sentence at a time, consulting and evaluating matches proposed by system Pre-translation The entire ST is compared against the TM database and matches are automatically inserted to create a  hybrid text , which the translator must then editTZ^ZZZ^ Hybrid text  Pilot StudyObjective: To compare interactive translation vs pre-translation to determine which approach is more beneficial with regard to: Productivity Quality Translator satisfaction Limitations TM system Participants Languages Texts  ZuZ.Z Z3ZZ #P.    . Hypotheses;Productivity Pre-translation < Interactive translation It will take translators longer to decipher hybrid texts and find formulations to fit in with pre-translated bits Quality Pre-translation < Interactive translation Translators are obliged to work with translated portions that may reflect different styles, and they cannot consult all solutions contained in the TM Translator Satisfaction Pre-translation < Interactive translation Translators are obliged to adapt their style to the pre-translated portion, and they have less control over creating a holistic text Z*ZsZ Z*ZZZ*ZZ *s    **   Experiment PreparationParticipants Domain Building the TM and producing pre-translations Documents and training for translators Evaluation criteria for evaluators Pre-testing and refining Pilot Study Experiment ExecutionTranslators: Blue, Green, Red, Yellow Translation order: Blue - ST2 using interactive translation & Red - ST1 using pre-translation Green - ST1 using interactive translation & Yellow - ST2 using pre-translation Time and Resources Questionnaire ! IG! Data Analysis - Productivity$Factors influencing results ST2 more difficult than ST1 (Green) Results of Blue and Yellow argue this Technical difficulties (Red) Familiarity with software Number of results in interactive mode Quality of finished translations Method of translation has no significant effect on productivity P$P&P~P@ZPP$&~A   Data Analysis - QualityIT - 5/8 higher scores PT - 2/8 higher scores Factors influencing results Difficulty of texts Quality of finished translations Amenability of one text to IT IT produces slightly higher quality than Pre-translation KZSZ9ZZZZZ/S9NData Analysis  Translator Satisfaction(( pInteractive Translation Not enough screen space Faster and more efficient Better productivity Principle resource*YY gPre-translation Time consuming Too many windows open at once Inconsistent style Not principle resource *XX  ConclusionRTranslation Memory systems keep translators interested and prove to increase the quality of translations produced Conclusions reached are preliminary but merit further research Increase number of participants New languages and language directions Different text types and TM tools For more information please contact: jwall035@uottawa.cahZhZZ9Zh9/    0L0 DV(  Dx D c $gtP  t  D S htPG  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~     !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~%      !"#$&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|PowerPoint Document( ˿DocumentSummaryInformation828es New RomanTTX-ܖ 0ܖ DVerdanaw RomanTTX-ܖ 0ܖ"0DWingdingsRomanTTX-ܖ 0ܖ@DSymbolgsRomanTTX-ܖ 0ܖ A .  @n?" dd@  @@`` x_   r $ '(*+,.  5 6 ?,b$dMwP \=.], 0AA F8$@T" # $ Pʚ;ʚ;g4LdLd|f 02ppp@ <4dddd w 0T4- g4;d;d(h 04p@ pp0___PPT10 pp___PPT9 4 ?  %O  =!#Interactive Translation vs. Pre-Translation in the Context of Translation Memory Systems: Investigating the Effects of Translation Method on Productivity, Quality and Translator Satisfaction(Zg=Julian Wallis Localisation Research Centre October 25th, 2006,>Z5  $$Challenges in the Translation Market%% Increased demand for translation Globalization New products  Information Society Shorter deadlines Online resources Simultaneous shipment, or  Simship Shortage of translators  Babyboomer generation Lack of graduates from translation programs Result: Translators are turning to tools for help, especially Translation Memory (TM) tools !Z1ZZ5ZZDZ]ZZ!15D\, !Translation Memory (TM)2What is a TM? CAT tool which stores previously translated texts with corresponding source texts and allows for these to be  recycled' in new translations. Organization and Storage Segmentation Alignment Retrieval of Information Exact Match Fuzzy Match Sub-segment Match Term MatchZZZZZ5Z5  #Impact of TMs on Translation Market$$ :Translators  want to use TMs to accelerate the translation process Clients  want translators to use TMs to save time and money Current Situation Ownership Translators   intellectual property Clients  TM is a value to them Payment Clients  Demand discounted rates Translators  Expect some compensation Other issues Cost of technology, experience with specific system Potential solution? Pre-translation .PP PFPPIP P5PPPP F 7    4     Methods of working with a TM:Interactive translation Translator works one sentence at a time, consulting and evaluating matches proposed by system Pre-translation The entire ST is compared against the TM database and matches are automatically inserted to create a  hybrid text , which the translator must then editTZ^ZZZ^ Hybrid text  Pilot StudyObjective: To compare interactive translation vs pre-translation to determine which approach is more beneficial with regard to: Productivity Quality Translator satisfaction Limitations TM system Participants Languages Texts  ZuZ.Z Z3ZZ #P.    . Hypotheses;Productivity Pre-translation < Interactive translation It will take translators longer to decipher hybrid texts and find formulations to fit in with pre-translated bits Quality Pre-translation < Interactive translation Translators are obliged to work with translated portions that may reflect different styles, and they cannot consult all solutions contained in the TM Translator Satisfaction Pre-translation < Interactive translation Translators are obliged to adapt their style to the pre-translated portion, and they have less control over creating a holistic text Z*ZsZ Z*ZZZ*ZZ *s    **   Experiment PreparationParticipants Domain Building the TM and producing pre-translations Documents and training for translators Evaluation criteria for evaluators Pre-testing and refining Pilot Study Experiment ExecutionTranslators: Blue, Green, Red, Yellow Translation order: Blue - ST2 using interactive translation & Red - ST1 using pre-translation Green - ST1 using interactive translation & Yellow - ST2 using pre-translation Time and Resources Questionnaire ! IG! Data Analysis - Productivity$Factors influencing results ST2 more difficult than ST1 (Green) Results of Blue and Yellow argue this Technical difficulties (Red) Familiarity with software Number of results in interactive mode Quality of finished translations Method of translation has no significant effect on productivity P$P&P~P@ZPP$&~A   Data Analysis - QualityIT - 5/8 higher scores PT - 2/8 higher scores Factors influencing results Difficulty of texts Quality of finished translations Amenability of one text to IT IT produces slightly higher quality than Pre-translation KZSZ9ZZZZZ/S9NData Analysis  Translator Satisfaction(( pInteractive Translation Not enough screen space Faster and more efficient Better productivity Principle resource*YY gPre-translation Time consuming Too many windows open at once Inconsistent style Not principle resource *XX  ConclusionRTranslation Memory systems keep translators interested and prove to increase the quality of translations produced Conclusions reached are preliminary but merit further research Increase number of participants New languages and language directions Different text types and TM tools For more information please contact: jwall035@uottawa.cahZhZZ9Zh9/    0L0   !"#$%&'()*+,-./01D3456789:;<=>?@ABEDV(  Dx D c $gtP  t  D S htPG^<$@  0 t H D 0޽h ? ff___̙33ff___PPT10.P$B+E!kDT' = @B D' = @BA?%,( < +O%,( < +DF' =%(D.' =%(D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*D%(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*DD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*DD' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*D%(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*DD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*DD' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*D%(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*DD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*DD' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*D6%(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*D6D' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*D6D' =4@BBBB%(D' =1:Bvisible*o3>+B#style.visibility<*DCw%(D' =+4 8?dCB1+#ppt_w/2BCB#ppt_xB*Y3>B ppt_x<*DCwD' =+4 8?\CB#ppt_yBCB#ppt_yB*Y3>B ppt_y<*DCwD' =%(D`' =4@BB#BB%(D' =1:Bvisible*o3>+B#style.visibility<*D%(D' =-g6B fade*<3<*DD' =+4 8?HCB4DBCBB*k3>'Bstyle.rotation<*DD' =+4 8?RCBBCB#ppt_hB*Y3>B ppt_h<*DD' =+4 8?RCBBCB#ppt_wB*Y3>B ppt_w<*D+   0L0 P^0(  \x \ c $/P_^   x \ c $0PP $7    t]#  ^# #"&f3343]   S^ <6?I p #  T 46 min (ST2), ( #  R^ <l;?[p I  T 46 min (ST1), ( #  Q^ <<@?tp [  NYellow, #  P^ <4>?I 8# p  T 60 min (ST2), ( #  O^ <dI?[8I p  T 51 min (ST1), ( #  N^ <xK?t8[p  MGreen, #  M^ <O?I # 8 T 55 min (ST1), ( #  L^ <W?[I 8 T 64 min (ST2), ( #  K^ <\?t[8 KRed, #  J^ <8a?I #  T 49 min (ST1), ( #  I^ <(f?[I  T 49 min (ST2), ( #  H^ <c?t[ LBlue, #  G^ <o?I ]#  VPre- Translation* #  F^ <l?[]I  ]Interactive Translation* #  E^ <v?t][ Q Translators*   # ZB T^ s *1 ?t]# ]ZB U^ s *1 ?t # ZB V^ s *1 ?t]t ZB W^ s *1 ?# ]# ZB Z^ s *1 ?t# ZB \^ s *1 ?[][ ZB _^ s *1 ?I ]I ZB b^ s *1 ?t# ZB i^ s *1 ?t8# 8ZB p^ s *1 ?tp # p h ^ 0 GP  Blue and Yellow  same amount of time to translate both texts Red and Green  show exact opposite resultsk kH \ 0޽h ? ff___̙33ff___PPT10i.`(+D=' = @B +   0L0 @ :(  x  c $vP_^   x  c $oP_.   r  S PjP~      0HS f< ( General consensus = Interactive mode superior to Pre- translation mode All translators preferred working in Interactive mode Interactive mode is superior to Pre-translation mode in terms of job satisfaction8 oS H  0޽h ? ff___̙33ff___PPT10i.Կp,+D=' = @B +r K ='$1[,( F/ 0|DArialgsRomanTTX-ܖ 0ܖDTimes New RomanTTX-ܖ 0ܖ DVerdanaw RomanTTX-ܖ 0ܖ"0DWingdingsRomanTTX-ܖ 0ܖ@DSymbolgsRomanTTX-ܖOh+'0 hp<LX ht   Interactive Translation vs. Pre-Translation in the Context of Translation Memory Systems: Investigating the Effects of Translation Method on Productivity, Quality and Translator Satisfactionjulianjulian63Microsoft Office PowerPoint@ Ԕ@o/@p/'G g  pO  yff--$xx--'@Arial-. 62 Interactive Translation vs. Pre."System89-@Arial-.  2 `-.-@Arial-. $2 bTranslation in the .-@Arial-. @2 &Context of Translation Memory Systems:.-@Arial-. O2 -0Investigating the Effects of Translation Method .-@Arial-. 32 4+on Productivity, Quality and r.-@Arial-. *2 <2Translator Satisfaction.-@Arial-. 2 V@ Julian Wallis.-@Arial-. 2 \.LRC XI .-@Arial-.  2 \=.-@Arial-. $2 \@European Foundationt.-@Arial-. 2 b: October 25.-@Arial-.  2 aOth.-@Arial՜.+,0    On-screen Show ˿ ArialTimes New RomanVerdana Wingdings^<$@ 0 t H D 0޽h ? ff___̙33ff___PPT10.P$B+;Y"vDU' = @B D' = @BA?%,( < +O%,( < +DG' =%(%(D' =%(Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*D%(D' =-g6B fade*<3<*DD' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*DD' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*DDc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*D%(D' =-g6B fade*<3<*DD' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*DD' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*DDc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*D%(D' =-g6B fade*<3<*DD' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*DD' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*DDc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*D6%(D' =-g6B fade*<3<*D6D' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*D6D' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*D6Dc' =4@BB*BB%(D' =1:Bvisible*o3>+B#style.visibility<*DCw%(D' =-g6B fade*<3<*DCwD' =+4 8?\CB#ppt_xBCB#ppt_xB*Y3>B ppt_x<*DCwD' =+4 8?bCB#ppt_y+.1BCB#ppt_yB*Y3>B ppt_y<*DCwD`' =4@BB#BB%(D' =1:Bvisible*o3>+B#style.visibility<*D%(D' =-g6B fade*<3<*DD' =+4 8?HCB4DBCBB*k3>'Bstyle.rotation<*DD' =+4 8?RCBBCB#ppt_hB*Y3>B ppt_h<*DD' =+4 8?RCBBCB#ppt_wB*Y3>B ppt_w<*D+r< .$1-( F/ 0|DArialgsRomanhhX- 0DTimes New RomanhhX- 0 DVerdanaw RomanhhX- 0"0DWingdingsRomanhhX- 0@DSymbolgsRomanhhX- 0 A .  @n?" dd@  @@`` x_   r $ '(*+,.  5 6 ?,b$dMwP \=.], 0AA F8$@T" # $ Lʚ;ʚ;g4LdLdf 0(ppp@ <4dddd4w 0h4- g4;d;dh 04p@ pp0___PPT10 pp___PPT9| 2 ?  %O  =#Interactive Translation vs. Pre-Translation in the Context of Translation Memory Systems: Investigating the Effects of Translation Method on Productivity, Quality and Translator Satisfaction(ZgzJulian Wallis LRC XI  European Foundation October 25th, 2006,>Z5 $Challenges in the Translation Market%% Increased demand for translation Globalization New products  Information Society Shorter deadlines Online resources Simultaneous shipment, or  Simship Shortage of translators  Babyboomer generation Lack of graduates from translation programs Result: Translators are turning to tools for help, especially Translation Memory (TM) tools !Z1ZZ5ZZDZ]ZZ!15D\, !Translation Memory (TM)2What is a TM? CAT tool which stores previously translated texts with corresponding source texts and allows for these to be  recycled' in new translations. Organization and Storage Segmentation Alignment Retrieval of Information Exact Match Fuzzy Match Sub-segment Match Term MatchZZZZZ5Z5  #Impact of TMs on Translation Market$$ :Translators  want to use TMs to accelerate the translation process Clients  want translators to use TMs to save time and money Current Situation Ownership Translators   intellectual property Clients  TM is a value to them Payment Clients  Demand discounted rates Translators  Expect some compensation Other issues Cost of technology, experience with specific system Potential solution? Pre-translation .PP PFPPIP P5PPPP F 7    4     Methods of working with a TM:Interactive translation Translator works one sentence at a time, consulting and evaluating matches proposed by system Pre-translation The entire ST is compared against the TM database and matches are automatically inserted to create a  hybrid text , which the translator must then editTZ^ZZZ^ Hybrid text  Pilot StudyObjective: To compare interactive translation vs pre-translation to determine which approach is more beneficial with regard to: Productivity Quality Translator satisfaction Limitations TM system Participants Languages Texts  ZuZ.Z Z3ZZ #P.    . Hypotheses;Productivity Pre-translation < Interactive translation It will take translators longer to decipher hybrid texts and find formulations to fit in with pre-translated bits Quality Pre-translation < Interactive translation Translators are obliged to work with translated portions that may reflect different styles, and they cannot consult all solutions contained in the TM Translator Satisfaction Pre-translation < Interactive translation Translators are obliged to adapt their style to the pre-translated portion, and they have less control over creating a holistic text Z*ZsZ Z*ZZZ*ZZ *s    **   Experiment PreparationParticipants Domain Building the TM and producing pre-translations Documents and training for translators Evaluation criteria for evaluators Pre-testing and refining Pilot Study Experiment ExecutionTranslators: Blue, Green, Red, Yellow Translation order: Blue - ST2 using interactive translation & Red - ST1 using pre-translation Green - ST1 using interactive translation & Yellow - ST2 using pre-translation Time and Resources Questionnaire ! IG! Data Analysis - Productivity$Factors influencing results ST2 more difficult than ST1 (Green) Results of Blue and Yellow argue this Technical difficulties (Red) Familiarity with software Number of results in interactive mode Quality of finished translations Method of translation has no significant effect on productivity P$P&P~P@ZPP$&~A   Data Analysis - QualityIT - 5/8 higher scores PT - 2/8 higher scores Factors influencing results Difficulty of texts Quality of finished translations Amenability of one text to IT IT produces slightly higher quality than Pre-translation KZSZ9ZZZZZ/S9NData Analysis  Translator Satisfaction(( pInteractive Translation Not enough screen space Faster and more efficient Better productivity Principle resource*YY gPre-translation Time consuming Too many windows open at once Inconsistent style Not principle resource *XX  ConclusionInteractive Translation keeps translators interested and proves to increase the quality of translations produced Using the pre-translation option and giving translators a hybrid text may not prove to be as beneficial for clients Conclusions reached are preliminary but merit further research Increase number of participants New languages and language directions Different text types and TM tools For more information please contact: jwall035@uottawa.cah$PhPP9P#h9, / }  0  $(  r  S 6tpE t r  S 7tps  t H  0޽h ? 33___PPT10i..U+D='  = @B +   0L0 ``0(  `x ` c $"P_^   x ` c $#P_^  H ` 0޽h ? ff___̙33ff___PPT10i.{++D='  = @B +r.[f^ .a$1I,( F/ 0|DArialgsRomanTTX-ܖ 0ܖDTimes New RomanTTX-ܖ 0ܖ DVerdanaw RomanTTX-ܖ 0ܖ"0DWingdingsRomanTTX-ܖ 0ܖ@DSymbolgsRomanTTX-ܖ 0ܖ A .  @n?" dd@  @@`` x_   r $ '(*+,.  5 6 ?,b$dMwP \=.], 0AA F8$@T" # $ Lʚ;ʚ;g4LdLdf 0(ppp@ <4dddd w 0T4- g4;d;dh 04p@ pp0___PPT10 pp___PPT9| 2 ?  %O  =#Interactive Translation vs. Pre-Translation in the Context of Translation Memory Systems: Investigating the Effects of Translation Method on Productivity, Quality and Translator Satisfaction(ZgzJulian Wallis LRC XI  European Foundation October 25th, 2006,>Z5 $Challenges in the Translation Market%% Increased demand for translation Globalization New products  Information Society Shorter deadlines Online resources Simultaneous shipment, or  Simship Shortage of translators  Babyboomer generation Lack of graduates from translation programs Result: Translators are turning to tools for help, especially Translation Memory (TM) tools !Z1ZZ5ZZDZ]ZZ!15D\, !Translation Memory (TM)2What is a TM? CAT tool which stores previously translated texts with corresponding source texts and allows for these to be  recycled' in new translations. Organization and Storage Segmentation Alignment Retrieval of Information Exact Match Fuzzy Match Sub-segment Match Term MatchZZZZZ5Z5  #Impact of TMs on Translation Market$$ :Translators  want to use TMs to accelerate the translation process Clients  want translators to use TMs to save time and money Current Situation Ownership Translators   intellectual property Clients  TM is a value to them Payment Clients  Demand discounted rates Translators  Expect some compensation Other issues Cost of technology, experience with specific system Potential solution? Pre-translation .PP PFPPIP P5PPPP F 7    4     Methods of working with a TM:Interactive translation Translator works one sentence at a time, consulting and evaluating matches proposed by system Pre-translation The entire ST is compared against the TM database and matches are automatically inserted to create a  hybrid text , which the translator must then editTZ^ZZZ^ Hybrid text  Pilot StudyObjective: To compare interactive translation vs pre-translation to determine which approach is more beneficial with regard to: Productivity Quality Translator satisfaction Limitations TM system Participants Languages Texts  ZuZ.Z Z3ZZ #P.    . Hypotheses;Productivity Pre-translation < Interactive translation It will take translators longer to decipher hybrid texts and find formulations to fit in with pre-translated bits Quality Pre-translation < Interactive translation Translators are obliged to work with translated portions that may reflect different styles, and they cannot consult all solutions contained in the TM Translator Satisfaction Pre-translation < Interactive translation Translators are obliged to adapt their style to the pre-translated portion, and they have less control over creating a holistic text Z*ZsZ Z*ZZZ*ZZ *s    **   Experiment PreparationParticipants Domain Building the TM and producing pre-translations Documents and training for translators Evaluation criteria for evaluators Pre-testing and refining Pilot Study Experiment ExecutionTranslators: Blue, Green, Red, Yellow Translation order: Blue - ST2 using interactive translation & Red - ST1 using pre-translation Green - ST1 using interactive translation & Yellow - ST2 using pre-translation Time and Resources Questionnaire ! IG! Data Analysis - Productivity$Factors influencing results ST2 more difficult than ST1 (Green) Results of Blue and Yellow argue this Technical difficulties (Red) Familiarity with software Number of results in interactive mode Quality of finished translations Method of translation has no significant effect on productivity P$P&P~P@ZPP$&~A   Data Analysis - QualityIT - 5/8 higher scores PT - 2/8 higher scores Factors influencing results Difficulty of texts Quality of finished translations Amenability of one text to IT IT produces slightly higher quality than Pre-translation KZSZ9ZZZZZ/S9NData Analysis  Translator Satisfaction(( pInteractive Translation Not enough screen space Faster and more efficient Better productivity Principle resource*YY gPre-translation Time consuming Too many windows open at once Inconsistent style Not principle resource *XX  Conclusion]Interactive Translation keeps translators interested and proves to increase the quality of translations produced Using the pre-translation option and giving translators a hybrid text may not prove to be as beneficial for clients Conclusions reached are preliminary but merit further research For more information please contact: jwall035@uottawa.ca6$:#:/    0L0 ``0(  `x ` c $"P_^   x ` c $#P_^  H ` 0޽h ? ff___̙33ff___PPT10i.{++D='  = @B +rSa /aM$1I,( F/ 0|DArialgsRomanTT*ܖ 0ܖDTimes New RomanTT*ܖ 0ܖ DVerdanaw RomanTT*ܖ 0ܖ"0DWingdingsRomanTT*ܖ 0ܖ@DSymbolgsRomanTT*ܖ 0ܖ A .  @n?" dd@  @@`` x_   r $ '(*+,.  5 6 ?,b$dMwP \=.], 0AA F8$@T" # $ Lʚ;ʚ;g4LdLdLe 0(ppp@ <4dddd w 0T* g4;d;df 04p@ pp0___PPT10 pp___PPT9| 2 ?  %O  =#Interactive Translation vs. Pre-Translation in the Context of Translation Memory Systems: Investigating the Effects of Translation Method on Productivity, Quality and Translator Satisfaction(ZgzJulian Wallis LRC XI  European Foundation October 25th, 2006,>Z5 $Challenges in the Translation Market%% Increased demand for translation Globalization New products  Information Society Shorter deadlines Online resources Simultaneous shipment, or  Simship Shortage of translators  Babyboomer generation Lack of graduates from translation programs Result: Translators are turning to tools for help, especially Translation Memory (TM) tools !Z1ZZ5ZZDZ]ZZ!15D\, !Translation Memory (TM)2What is a TM? CAT tool which stores previously translated texts with corresponding source texts and allows for these to be  recycled' in new translations. Organization and Storage Segmentation Alignment Retrieval of Information Exact Match Fuzzy Match Sub-segment Match Term MatchZZZZZ5Z5  #Impact of TMs on Translation Market$$ :Translators  want to use TMs to accelerate the translation process Clients  want translators to use TMs to save time and money Current Situation Ownership Translators   intellectual property Clients  TM is a value to them Payment Clients  Demand discounted rates Translators  Expect some compensation Other issues Cost of technology, experience with specific system Potential solution? Pre-translation .PP PFPPIP P5PPPP F 7    4     Methods of working with a TM:Interactive translation Translator works one sentence at a time, consulting and evaluating matches proposed by system Pre-translation The entire ST is compared against the TM database and matches are automatically inserted to create a  hybrid text , which the translator must then editTZ^ZZZ^ Hybrid text  Pilot StudyObjective: To compare interactive translation vs pre-translation to determine which approach is more beneficial with regard to: Productivity Quality Translator satisfaction Limitations TM system Participants Languages Texts  ZuZ.Z Z3ZZ #P.    . Hypotheses;Productivity Pre-translation < Interactive translation It will take translators longer to decipher hybrid texts and find formulations to fit in with pre-translated bits Quality Pre-translation < Interactive translation Translators are obliged to work with translated portions that may reflect different styles, and they cannot consult all solutions contained in the TM Translator Satisfaction Pre-translation < Interactive translation Translators are obliged to adapt their style to the pre-translated portion, and they have less control over creating a holistic text Z*ZsZ Z*ZZZ*ZZ *s    **   Experiment PreparationParticipants Domain Building the TM and producing pre-translations Documents and training for translators Evaluation criteria for evaluators Pre-testing and refining Pilot Study Experiment ExecutionTranslators: Blue, Green, Red, Yellow Translation order: Blue - ST2 using interactive translation & Red - ST1 using pre-translation Green - ST1 using interactive translation & Yellow - ST2 using pre-translation Time and Resources Questionnaire ! IG! Data Analysis - Productivity$Factors influencing results ST2 more difficult than ST1 (Green) Results of Blue and Yellow argue this Technical difficulties (Red) Familiarity with software Number of results in interactive mode Quality of finished translations Method of translation has no significant effect on productivity P$P&P~P@ZPP$&~A   Data Analysis - QualityIT - 5/8 higher scores PT - 2/8 higher scores Factors influencing results Difficulty of texts Quality of finished translations Amenability of one text to IT IT produces slightly higher quality than Pre-translation KZSZ9ZZZZZ/S9NData Analysis  Translator Satisfaction(( pInteractive Translation Not enough screen space Faster and more efficient Better productivity Principle resource*YY gPre-translation Time consuming Too many windows open at once Inconsistent style Not principle resource *XX  Conclusion]Interactive Translation keeps translators interested and proves to increase the quality of translations produced Using the pre-translation option and giving translators a hybrid text may not prove to be as beneficial for clients Conclusions reached are preliminary but merit further research For more information please contact: jwall035@uottawa.ca6$:#:/    0L0 <(  ~  s * tP_^  t ~  s *tP_^ t H  0޽h ? ff___̙33ff___PPT10i.m+D='  = @B +r!ڼe$1Root EntrydO)?Pictures,Current UserC&SummaryInformation(       !"#$&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|%  !"#$%&'()*+,-./01D3456789:;<=>?@ABEData Analysis - Quality(Data Analysis Translator Satisfaction Conclusion  Fonts UsedDesign Template Slide Titles'_㧿EFEFjulian--. 2 bR, 2006.-."System8-@ArialData Analysis - Quality(Data Analysis Translator Satisfaction Conclusion  Fonts UsedDesign Template Slide Titles'"_㧿 Karl KellyKarl Kelly-. 2 bR, 2006.-."System8-@Arial