From: Subject: Korean NLP at Penn Home Page Date: Sun, 2 Feb 2003 12:02:44 -0500 MIME-Version: 1.0 Content-Type: multipart/related; boundary="----=_NextPart_000_004A_01C2CAB2.FEC136E0"; type="text/html" X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 This is a multi-part message in MIME format. ------=_NextPart_000_004A_01C2CAB2.FEC136E0 Content-Type: text/html; charset="ks_c_5601-1987" Content-Transfer-Encoding: quoted-printable Content-Location: http://www.cis.upenn.edu/~xtag/koreantag/ Korean NLP at Penn Home Page

KOREAN NLP = AT THE=20 UNIVERSITY OF PENNSYLVANIA

Korean forms one of the major languages in multilingual NLP = research at=20 the University of Pennsylvania. = This site=20 introduces three main projects on Korean NLP currently being conducted = at=20 Penn: Korean XTAG, Korean Treebank, and Korean/English machine = translation.=20 These projects are partially funded by the Army Research Lab via a = subcontract=20 from CoGenTex, = Inc., and by=20 NSF Grant SBR 8920230.

    Contents

    Korean XTAG=20
    Korean=20 Treebank
    Korean/English=20 Machine Translation
    Papers
    <= A=20 = href=3D"http://www.cis.upenn.edu/~xtag/koreantag/#People">People
    <= A=20 href=3D"http://www.cis.upenn.edu/~xtag/koreantag/#Some Links to NLP = in Korea">Some=20 Links to NLP in Korea

Korean XTAG =


=20

Korean XTAG is an on-going project to develop a wide-coverage = grammar for=20 Korean using Feature-Based Lexicalized Tree Adjoining Grammar (LTAG)=20 formalism. For grammar development system, it uses the XTAG system = which we=20 have customized for Korean TAG development. The XTAG system was = originally=20 developed for English TAG and it consists of a parser, an X-windows = grammar=20 development interface and a POS tagger. We have modified the original = XTAG=20 system and incorporated a Korean morphological analyzer to handle rich = inflectional morphology in Korean and facilitate lexicon development = and=20 parsing. More on the Korean XTAG system description can be found in = our TAG+5 = workshop=20 paper.

LTAG is based on the Tree Adjoining Grammar (TAG) formalism = developed by=20 Joshi, Levy and Takahashi (1975). The TAG formalism in general, and=20 lexicalized TAG in particular, are well-suited for linguistic = applications. An=20 LTAG consists of a finite set of elementary trees anchoring lexical = items and=20 composition operations of substitution and adjunction. The elementary = trees=20 represent extended projections of lexical items and encapsulate=20 syntactic/semantic arguments of the lexical anchor. In the last = decade, the=20 LTAG approach has been applied to various NLP tasks such as parsing, = machine=20 translation, information retrieval, generation and summarization = applications.=20 More on the introduction to LTAG and the current status of our Korean = LTAG=20 grammar is documented in our technical = report.=20

[BACK=20 UP]=20

Korean=20 Treebank

A Treebank is a corpus annotated with morphological = and=20 syntactic information. Each word in the corpus is annotated with=20 morpho-syntactic tags and each sentence is bracketed to represent its=20 structural analysis. This kind of corpus has served as an extremely = valuable=20 resource for computational linguistics applications, and has also = proved=20 useful in theoretical linguistics research.

Annotation Format

For syntactic = bracketing,=20 we use a phrase structure annotation. Similar phrase structure = annotation=20 schemes were also used by the Penn English Treebank, = the Penn Middle English = Treebank and=20 the Penn Chinese=20 Treebank. This annotation is preferable to a pure dependency = annotation=20 because with a phrase structure annotation we can encode richer = structural=20 information than with dependency annotation, as illustrated below:=20
  • Phrase structure annotation has phrasal level node labels such = as VP and=20 NP, whereas dependency annotation does not have any node = labels.

  • Phrase structure annotation can explicitly represent empty = arguments,=20 but dependency annotation cannot.

  • Phrase structure annotation can distinguish between = complementation and=20 adjunction, but dependency annotation cannot.

  • Phrase structure annotation can make use of traces for displaced = constituents, whereas dependency annotation = cannot.

Corpus

The corpus for the Korean = Treebank=20 project consists of texts from military language training manuals. = These texts=20 contain information about various aspects of the military, such as = troop=20 movement, intelligence gathering, and equipment supplies, among = others. The=20 texts in the manuals were originally in printed form, and in order to = use them=20 for our Treebank, we converted the manuals into a machine-readable = form. This=20 corpus contains 54,366 words and 5078 sentences.=20

Guidelines and a Sample File

Applications

  • The linguistic information in the Korean Treebank will provide a = standard framework in which to train and evaluate tools such as POS = tagger=20 and stochastic parsers.

  • The Treebank will also be used to extract lexicalized grammars, = e.g. a=20 Korean Tree Adjoining Grammar, which can be used for other = applications,=20 such as natural language generation. There are already tools = developed at=20 Penn that train parsers and extract Tree Adjoining Grammars from a=20 phrase-structure based Treebank (Xia 1999), which will be equally = applicable=20 to the Korean Treebank.

  • Having an on-line corpus of parsed texts will be extremely = useful for=20 research in corpus linguistics and will lead to many interesting = theoretical=20 results.

[BACK=20 UP]=20

Korean/English Machine = Translation

This is a joint project with CoGenTex and Systran.

Basic Elements of our = Approach

Given that=20 Korean and English are very different from each other in structure and = morphology, many challenging problems arise, demanding sophisticated=20 linguistic analysis. The basic elements of our approach include:=20
  • Following the model described in Palmer, = Rambow=20 and Nasr (1998) for English/French translation, our system has a = plug-and-play architecture that is composed of state-of-the-art=20 off-the-shelf components in parsing (and morphological analysis) and = generation. These components communicate with each other via a = common=20 predicate-argument structure representation.

  • Our system is a hybrid system that profits from a stochastic = parser that=20 was independently trained on domain-general corpora and a = hand-crafted=20 linguistic knowledge base in the form of a predicate-argument = lexicon and=20 linguistically sophisticated transfer rules. The linguistic = knowledge base=20 plays an important role in handling structural divergences and = recovering=20 dropped arguments.

  • For defining transfer rules, we use the `lexico-structural = transfer'=20 framework, which is based on a lexicalized predicate-argument = structure. In=20 this framework, the transfer lexicon does not simply relate words = (or=20 context-free rewrite rules) from one language to words (or = context-free=20 rewrite rules) from another language. Instead, lexemes and their = relevant=20 syntactic structures (essentially, their syntactic projection along = with=20 syntactic/semantic features) are mapped. This framework was applied=20 previously in English/French and English/Arabic MT (Nasr= et.=20 al. 1997; Palmer, = Rambow=20 and Nasr 1998).

Corpus

The corpus for this = project is a set=20 of Korean/English parallel texts that consist of battle scenario = message=20 traffic and military language training manuals which contain = information on=20 typical military events such as troop movement, intelligence = gathering, and=20 equipment supplies, among others. Each half has roughly 50,000 word = tokens,=20 and 5000 sentences.=20

Presentations

  • Some=20 issues concerning Korean/English machine = translation
    Presented at=20 ARL's Federated Laboratories Symposium (FEDLAB), University of = Maryland,=20 College Park, February 2-4, 1999

  • Slides explaining the demo of our Korean/English machine = translation=20 system at the Army Research Lab, February 10, 2000
    Example = sentences=20 and the output of the MT system
    Predicate = argument=20 lexicon and transfer lexicon
    Overview = of the=20 corpus and current status

[BACK=20 UP]=20

Papers

  • Chung-hye Han, Na-Rae han, Eon-Suk Ko
    Bracketing = Guidelines for=20 Penn Korean TreeBank, Technical Report, IRCS-01-10 (2001) =

  • Chung-hye Han, Na-Rae han
    Part of Speech = Tagging=20 Guidelines for Penn Korean Treebank, Technical Report, IRCS-01-09 = (2001)=20

  • Chung-hye Han, Juntae Yoon, Nari Kim and Martha Palmer
    A Feature-Based=20 Lexicalized Tree Adjoining Grammar for Korean, Technical Report, = IRCS-00-04=20 (2000)

  • Chung-hye Han, Benoit Lavoie, Martha Palmer, Owen Rambow, = Richard=20 Kittredge, Tanya Korelsky, Nari Kim and Myunghee Kim
    Handling=20 Structural Divergences and Recovering Dropped Arguments in a = Korean/English=20 Machine Translation System
    Proceedings of the Association = for=20 Machine Translation in the Americas '2000.
    Published in = Lecture Notes=20 in AI series of Springer-Verlag,= ©=20 Springer-Verlag (2000).

  • Juntae Yoon, Chung-hye Han, Nari Kim and Mee-sook Kim
    Customizing = the=20 XTAG system for efficient grammar development for Korean=20
    Proceedings of the Fifth International Workshop on Tree = Adjoining=20 Grammars and Related Formalisms, TAG+ 5 (2000).

  • Chung-hye Han and Owen Rambow
    The=20 Sino-Korean light verb construction and lexical argument structure=20
    Proceedings of the Fifth International Workshop on Tree = Adjoining=20 Grammars and Related Formalisms, TAG+ 5 (2000)

  • Martha Palmer,Dania Egedi,Chunghye Han, Fei Xia, and Joseph=20 Rosenzweig.
    Constraining= =20 Lexical Selection Across Languages Using TAGs.
    Tree = Adjoining=20 Grammars: Formal, Computational and Linguistic Aspects (TAG+ 3 = Workshop=20 Proceedings)
    Eds. Anne Abeille and Owen Rambow, CSLI, Stanford=20 (2000).

  • Chung-hye Han, Fei Xia, Martha Palmer, Joseph Rosenzweig.
    Capturing = Language Specific Constraints on Lexical Selection with = Feature-Based=20 Lexicalized Tree-Adjoining Grammars
    Proceedings of = International=20 Conference on Chinese Computing '96 (ICCC = '96).

[BACK=20 UP]=20

People (click to = see a picture = of=20 us)

Faculty

Post-doctoral Fellows

Graduate Students

Staff

  • Joseph Rosenzweig

Thanks to

  • Mee-sook Kim (participated from Nov. 1999 to July 2000)=20
  • Nari Kim (participated from Mar. 1998 to Dec. 1999, now at Konan = Technology, Inc.)=20
  • Juntae Yoon (participated from Mar. 1999 to Mar. 2000, now at = Daum=20 Communications)=20
  • Jong-Cheol Park (participated at the very early phase of the = project,=20 now at KAIST)
[BACK=20 UP]=20

Some Links to=20 NLP in Korea

[BACK=20 UP]=20
=20
This web page is maintained by Chung-hye Han
Last = changed:=20 $Date: 2001/07/31 03:13:27 $
------=_NextPart_000_004A_01C2CAB2.FEC136E0 Content-Type: image/jpeg Content-Transfer-Encoding: base64 Content-Location: http://www.cis.upenn.edu/~xtag/koreantag/hangul.jpg /9j/4AAQSkZJRgABAgEASABIAAD/7QE0UGhvdG9zaG9wIDMuMAA4QklNA+0AAAAAABAAR/+0AAIA AgBH/7QAAgACOEJJTQPzAAAAAAAIAAAAAAAAAAA4QklNJxAAAAAAAAoAAQAAAAAAAAACOEJJTQP1 AAAAAABIAC9mZgABAGxmZgAGAAAAAAABAC9mZgABAKGZmgAGAAAAAAABADIAAAABAFoAAAAGAAAA AAABADUAAAABAC0AAAAGAAAAAAABOEJJTQP4AAAAAABwAAD///////////////////////////// A+gAAAAA/////////////////////////////wPoAAAAAP////////////////////////////8D 6AAAAAD/////////////////////////////A+gAADhCSU0EBgAAAAAAAgAC/+4ADkFkb2JlAGSA AAAAAf/bAIQADAgICAkIDAkJDBELCgsRFQ8MDA8VGBMTFRMTGBEMDAwMDAwRDAwMDAwMDAwMDAwM DAwMDAwMDAwMDAwMDAwMDAENCwsNDg0QDg4QFA4ODhQUDg4ODhQRDAwMDAwREQwMDAwMDBEMDAwM DAwMDAwMDAwMDAwMDAwMDAwMDAwMDAwM/8AAEQgAHwDoAwEiAAIRAQMRAf/EAT8AAAEFAQEBAQEB AAAAAAAAAAMAAQIEBQYHCAkKCwEAAQUBAQEBAQEAAAAAAAAAAQACAwQFBgcICQoLEAABBAEDAgQC BQcGCAUDDDMBAAIRAwQhEjEFQVFhEyJxgTIGFJGhsUIjJBVSwWIzNHKC0UMHJZJT8OHxY3M1FqKy gyZEk1RkRcKjdDYX0lXiZfKzhMPTdePzRieUpIW0lcTU5PSltcXV5fVWZnaGlqa2xtbm9jdHV2d3 h5ent8fX5/cRAAICAQIEBAMEBQYHBwYFNQEAAhEDITESBEFRYXEiEwUygZEUobFCI8FS0fAzJGLh coKSQ1MVY3M08SUGFqKygwcmNcLSRJNUoxdkRVU2dGXi8rOEw9N14/NGlKSFtJXE1OT0pbXF1eX1 VmZ2hpamtsbW5vYnN0dXZ3eHl6e3x//dAAQAD//aAAwDAQACEQMRAD8A9VSSXO9drdmdSr6XU9+T dYz7Sca0Y4oprb+h9f1MnBzrXWWW/o2MZ6r/AH2fzNSbOXCLq+jNy+H3Z8PEIAAylI7RjHd3MnMx MOsW5d9ePWTtD7XtY0uInbueW+72oy4Xqztn1Z6tgussruwDVXbhRR6LRZdVdTkUWYmLh7m3s9/u 2f4T1qF2DeoY7sZ+SGX+mw7S00XCyfb9DHdV9osb7/psqTY5OIkbUAf+lxf4vAzZ+SOLHGQuZlkn j9I9NCGHJhlH/axz/wDeNpJc7e/6nY1TrsjpzKamRuss6da1ok7Rue/FDfpFC+rpx8bqfU7hiX4V ebfXVjY7cS5tYrpb6bMn1G0MqqbkusdY9lnp+j/hEvc9QHp17S2V9yPt5MlZRwD0ieLg9yXFGMoR PHL5YS9x6dByczEw6xbl3149ZO0Pte1jS4idu55b7vastuVm9R65nYFWS/Cx+msqBNTa3Ptfe31t 7n5Nd7K66WN2emyv/hPV/wAGsHrfWrMv6rdTwssPfl4mR9ldkMqeK7PRup/TvexjsbGssb/gfX+n /wAbWxCeYAE9uKr2kYbhdy/w+eTJjiSPUcPuRhfuYsfNfzeT5eD5JRn8z3CSSoddt6lT0jJt6Wz1 M1rJqbG48j1HMYfp2Mr3vrZ+/wD4Oz+bUhNAnsL0aeOBnOMARHjkI8Ujwwjxem5S/db6C3MxH5L8 Rl9bsmsbn0B7TY1vt9z6p3tb72LmsfINmdj4/Qrz1jEvxj+0mZlz7KmtIjHsstsZf9nycn9My7Eb V9D/ALSU/wA6s76jf8o4X/pqt/8Ab61Re96oxA3O/T9H5f8AGdD/AEbWHLllM/q4gxhw8GT1e7/O wl68f8x/hwn7j3iSr5ufh9PoORm3MoqH5zzEkAv2MH0rLNrf5tnvWD07Pz8C3L6h1bHy2YWa/wBW ky65uNU0WvDcnE3XZGHY5jfVv9L1cb9LXV+q+j6aklMAgfb/AFf7zUxctPJCUx0oQj1zT/Shi/fn H5vQ9Mks7rdjX9Azb6LSAMay6m6l5aZaw3VWV20ub7fa3879IuWb1DCf9XrW5d1eRfk4LgGDMvzb TcKvX329Pex9WI6t1XrWX/8AaR7Ppps8oia8OLdl5bkZZocYMtMgxSjGPEY/1/ng9o3MxH5L8Rl9 bsmsbn0B7TY1vt9z6p3tb72Iy476tOvPXcH1mMZHQqRXseXyz1GbHv3V1enY7/R/pP8AjV2KOOfG LqtVnOcuOXyCAPF6RInx/wAFSSzPrDUf2XkZbHvZbiVWWs22XVtO0eo5tjcS/Fc/c2vazc/9EsfK cBT0S7FN+7qtlQtxnZeW+wVWNbddbS5mXV7cVn87Z6Oz3/4NKU+EkV2O/f0pwcqMsYy4iLMo/KCI mEPd344/oPVpKNbG11traSWsAaC4lzoAj3PeXPe7+W9cx08ZXUPtPo4z2/ZMizFs39VzRL69u9zN tTv0fuRlKqHU+f8ABZhwe4JysiOOrP6v9Pb+cy4nqUlkdF6f1PEys23MsnHv9L7NR9otyfT2Bzbv 0uWyt/6V7t6F1u/HtyW4eGbMjqzxtZRVk3VV1NHvOTntxLa/SpZ6rfzfXyf0dNP/AAY4/TZFeBSO XBze3GfHGhIzgARH08UuP18EY4v8pP3OB2L8ijGqddkWMpqZG6yxwa0Sdo3PfDfpFSrsrtrbbU4P reA5j2kFpaRLXNcPpNcvNsq69zsTHzWxn4mJ1KnKdbabckn7Pba37Ruqq/Qfpd+FsuyavRs/nfpr vOgf8hdN/wDClH/ntibjy8ciKqmxznw/7tihMz45TJ2rg4OLJGMo1xcXH7Xuf4bfSXIdHx+oZVnU g31MmvHzr6KnW9Ry6XNYwt2VbKW372t3fzj3+oug6Xi5OP6vr1+nu27f1u/LmN0/0xlfo/8AW/5z +wnQmZVpV+f/AHrDzHKxw8Q4+KUa0/V/peHu+5/4230kkk9qv//Q9VWL9a34+L00dTtxvtLsJ4LS 25+PYwWfq7/RyMcer7vUZ6lX0Hs/4tbSSEhcSNr8LZMGT28sJkGQifVESljM4fpw9yHqhxxeZ6hX 9V8nouR0nF6niYrcl4sfabmWuL/UZe+y11l3q3WWbNu+y1dJXZXbW22pwfW8BzHtILS0iWua4fSa 5SSQjGjemwGgr5V2XPxxEfXpKWT9ZIZPXk4IzlxcEJf5KDkZ31fd1DFfh5nUcu3HtjeyMds7SLG+ 6vEa/wCm1Q6NlZrutdXwMjJfk1YX2b0HWNra4erW62zd9nroa73fyVtJIcAsEEit9fmGvp/5y4c1 I4545xjKMh6PTCHs5DPHKWWHDD5pww+1JoX9IrfmOzse+3CybGCu99GyLQ3+b9avIqvqfZV9Gu3Z 6uz9H9BY/wBaenY3Tvqbm4+OCQTW+yx53WWWOup9S++z/CXWfnOXTpJSxgiQ24gRf95OHm8mPJik SZxxTx5OD5eP2JcUISn/AFf0P82pJJJPazSswMt9jnt6lk1tcSQxrcba0E/QbvxHv2t/lvWZjfaK PrgcW285LT071fVtrpFoPr7PT9bHoof6P/BfvroEk0wutTob3Z4cyYiYMIy44GHywgY3+lxRhxOd kua3OF46TZk31DbXmMGNu2katrsvyKshrf0j2/RUv2lmf+VWX/nYv/vYr6SXD4kf4q33hQBxxlwj huRy/wDqxq9ODG4wrrwz0+thIZQRUBB95exuJZdU1rnOULMDLfY57epZNbXEkMa3G2tBP0G78R79 rf5b1dSR4dANdPp/0Ue6eKUgI+rfiHuf+lfcefxvtFH1wOLbeclp6d6vq210i0H19np+tj0UP9H/ AIL99dAkkhGPDet2bTmze6YnhEeGEYHhocXD+lUeFzOotpvtFeR0d/UG1fzdrm4z2+4NL/T+1ZFd rf5f6NVPsXTv/nZ/8Dwf/elbySBgCb/ZH/vV8OalCIiBQH9fNH/o5WnTfYzD3MwLafRhleIDQHFo 2tb6Wy/7M2trf3rq/wCbWB9YelHqPTsmrD6Cas+8tc3JcMRrt29tlrnXV5L7d1jd66tJKWMSHCSa qv0U4OblhyDJCEeISE9Tl/R9XDKsnqj/AHnIwq8fGvDsboT8Nz/Y65jcRkNJG7eaMk2+npv9rVo5 NFt9YZVkWYzgZ31CsuIj6H6xVezb/YRkkRGhX9n/AEWOeYzkJkCx34sn2+9LI4vV8PMo6Tm3ftLI t9LHtf6dleK5jtrHO2WM+x+6t356s9LyLR0DBvbSbrDjUO9KkVsJLmMn02vdj49bWz9DdX/IWikl wVK7O1d10uY4sYhKEbE+PiAjCxXDwH24wcOzFwrbHW2/VsvseS573V4RcXEy5znHK9znIuFXj414 djdCfhuf7HXMbiMhpI3bzRkm309N/tatdJDgF3f4R/71J5qRiYmIoiq4835e6pJJJPa7/9k= ------=_NextPart_000_004A_01C2CAB2.FEC136E0 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://www.cis.upenn.edu/~xtag/koreantag/parsethis.gif R0lGODlhewBlAPAAAP///wAAACH5BAEAAAAALAAAAAB7AGUAAAL+hI+py+0Po5y02ouztqGHDYYi 1hnfiKaU55lMqcZyA9fn0s76fMNQvguObL/DTYjU2E5HRC8JzbBavkSpGs1qt9yu9wsOi8ffJvns MKPXVrYb944b5XL144Ol8ziuGEsPYJfWl1IDKEgDhFiRV8cXaLLI2OgmqcBkeQHkyBip8qSX6eQJ KQJKJ2qUQ6mJl1r2mLVaCvbKuoMXmBtm+2rqwiTmuxX8xzu2Ccibe6uc5NMz7Pw76iQ9TWKm1owN QinI3a05mHYtbrVoGQ51has+sS4UvCfRW2v+kzocL8O/4nvNHwpDQ0hsEBhiSkGD3s45fAgxosSJ FCvKS5bN4p2xeQfxYUOY6BBGDiBHldTiKuHJKx6j7JIyB6auQ4QY0mo1M5SxjCt3avwJNKjQoRRP Er2T8+iQYi2PVommVGa1mFHhbbvUdCIrO0a1IqVRFWuRcmGTft0YVR88ov70ZV2DEOBbZFI7lYVj 8y5VvQ35JvT7F7BglYOlzA3b1a/CZy/PsSQpkKNjyREIrtBKOZ/hm+LaQc52+GPisqcK19tb0zQ6 sz5VVw6tNKXr2bRrFy4AADs= ------=_NextPart_000_004A_01C2CAB2.FEC136E0 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://www.ling.upenn.edu/gifdir/linguistics-sml.GIF R0lGODlhQABRAPcAAP///94IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4I AN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4I AN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4I AN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAN4IAJSUlJSUlJSUlJSUlJSUlJSUlJSU lJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSU lJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSU lJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlJSUlAAA hAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAA hAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAA hAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAAhAAA hAAAhAAAhAAAhAAAhAAAhAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACwAAAAAQABRAEcI/wCDCBxI sKDBgwgTKlzIUOC/gQ+DRIRIMOI/iw4nNtzIsSPDixo9dgxwkKRAkwoxGlQZcmVLkQhBcnxZkSZM ihljOtx5s2JGmwUvbhRKENBAoz0/8gSKMylMkFB7TpzalKhElU6zat0qESfLmkuxPs3JtWPIl2e1 okSZculBqDJHllzI9upHuEytcsXLdChaiET7vk0YtyBSwy5TFq4JN+nholv5lp0s+SZVnl3dZhY8 uTNlr1ajPhy9k7Tn01I1o9bpczBmtTPxss6ctO7J1R5tB9GtGyzurBZpBr77OzJnkcOVKias93VD oM1jtmy+WHHjlUGdTo+e+CbSx4+L+/82eDj8581Zwxs1n5r5dfHw45uNStst6eDc5evff35zaJmX XXZagKrZ5Z9z/YlVX3Bh4VbYWQB61RR/wMmGHXr6pfWWgvEJSCFw2cHW24Uc1mciTHWNKOGJgB2X EG+7WUdfUBaKBCNBKmIYYm1zIZQjYR/SCB1Z0s24F5GtlTjeXi76FWRsv6H1HnOemSZYXlLml19s RkpnXWsyVmfZXy02qZRpRfJFnZgLaVmgZYy1eeWEY214pZlBrHcQew7WmNB3TE55YUOARvbmT13u edSifQrKKHt8PoUnYuQZ59GkT1KIaYbtLaiXhokm6ByDZG1q1ooS3ofggBMu9tVVW2YG6hRAQQIC ADs= ------=_NextPart_000_004A_01C2CAB2.FEC136E0 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://www.cis.upenn.edu/~ircs/pix/s-ircs.gif R0lGODlhdgCWAPf/AP///xAQECEhITExMUJCQlJSUmNjY3Nzc4SEhJycnK2trb29vc7Ozt7e3u/v 75yUlIyEhHtzcykhIQgAAJyEc62Mc2NaUkI5Mb21rZyUjDEhEBAIAO/n3q2lnJx7Uox7Y3trUsat hLWcc5R7UmNKIVI5EDEhCOfezsa9rb2cY3tzY97OrVpSQqWUc+fOnM61hDkxIa2UY9a1c2NSMVJC IVpCEBgQADEhAEoxAGNCALWtnIyEc2tjUkpCMda9hO/OhN69c5R7Que9Y//OY62MQoRrMcacQvfG Upx7Md6tQue1QrWMMf/GQoxrIc6cMaV7Id6lIZRrEK17EM6UEP+1EGtKAJxrAKVzAN6cAOfWrb2t hP/nrd7GjOfOjM61c3trQq2UUta1Y//Wc8alUu/GY961UrWUQv/OUtatQu+9Qr2UMcacMZRzIf/G Ma2EIbWMIWtSEO+1IZxzEKV7EL2MENacEO+tEDkpAHNSAIxjAK17AMaMAM6UAO+tAKWchIR7Y+/e rc69jGNaQsa1hO/WjP/ehEI5IYRzQsatY969Y6WMQt61QntjIZx7Ib2UIe+9KcacIdalIXNaEEo5 CGtSCO+1ECEYAEIxAFpCAHtaAJxzAKV7ALWEAL2MANacAPe1AO/nzt7Wvb21nK2ljJSMc//vvd7O nP/nnFJKMYx7QnNjMa2UQjEpEGtaIYxzIffGMSkhCK2MIYRrGPfGKc6lIWtSAIRjAOfevca9nO/e nJyMUrWUIVpKEOe9KdatIffvzq2lhGtjQjkxENa1KffOKZyEGGtaECkhAP/3zu/nvXNrQvf379bW zsbGvZSUjIyMhP//797ezmtrY87OvVpaUvf33pSUhNbWvWNjUkJCMRgYECEhEAgIAEJCrVJS3kJC vRgYSiEhazk5vUJC3jExtTExvSkppRAQQhgYYyEhjBgYazEx3hAQSikp3iEh3kpC1kpC3kI53jkx xjkx1jkx3kI5zjEprTEptTEphCkhc1pKvXtrlJR7jJyEjAAAAMDAwCH5BAEAAP8ALAAAAAB2AJYA QAj/AOnFGydwoDdv8egpXCcQ3kFv9BjSGwcv4cCB8MbFs0hQ4TiKG71RhBgPnkJ4FSlqXCexHUuF MAUS1Igyo8aYMce9pOdSYsyBE0ui7PjSn9GjSJMqXcq06dFjnz7lMZrnU59jTrNq3Yr0Y0SfLLmK 3XonqlmztqiePXtnrNusC+N5W+dyoNYFAPICWNAUgd68Ry/16WMH2V8AW7D0iZrWX9VPp/5CO7O2 LdIACgAjZaA3QdMADQAQUEoxIzyG8dq1u/uXL9MIhzNh9cethIu80MRE5YPHklFu/vz+/cEJkw3a NnJwWrwWx1IBh6PnZfBMANN5MYc6ZK3XNVIciwvF/57NbVK1wwRsYNKz2IpRN9FH+7OEZ8/a+2Zz BAewFW/eBkmlI09CEg311jHH+ZNcJ8yZ1QcfmfhGGw7LnYXFFTgkOIFRtjQYlSdWXLIBAQUgQAow L/zQyyseWtUHFp7woZgqJUaAAALUyJdVN++8g5NCWmWC35BEFnkWJ0ZxYuSSTEaVCVcKHVSTgUk9 1uSVTCLpj5JYdllkY0pt9NFG9GREz1topqnmmkzBdFNMSemx1hUsADANDp4IAQgH1FwDAAfZ5MDH GNeEUg0XtvSQ1zNgpOBCNCzkoSgAjCJiCqQCPBPKPziUEFoHwFjRCTDMQJMMKjlgEdUU1uTVzC2d 0P8hSjSgVNPCDY10cQsGKgBQhapnXbHUQw5NxNI6TnGzhydYqNqHJ8wOFu2znXTCB3OecNJJszA6 i4W2qmKxhx6bJDkYFp1w4qwnenAi41owRmtVttZ6yyywn2CxIVzrrPQVPd5wBdgldywDgCAMYvEI HVXksQdz3CoW1YvNNthtJ0Zty61ZWPDBySXNAMDCFdBKHBUWeAhDxLX5RjzbWGVGmRBMWknwQF64 SOAPAXm1cEkPHAAACihA5JFLMi/ocA3DrIQCQDW3AJJJnQBIk5cP2mAAADQ/7HFHDwaD8osRHxgM yBN1QiMEtmUDAAgk0hisgAA8L2MAzArJFROyR9n/8MgrvHAVQIkISEPAJXjg2yUWRilWBgCjJJ4v J8w+HrniUenB5rAbMaQRgRsQ80gAYkEHQAFLOQBAANzYAMAydxwzazM7RHEFhtOIXEWdC9xRxTOU 5nBJ7ixkUkWrENR3xwEAKIDBUXlZt7k/ea8zkFevvLLB9FrdUHHLijPuD+b33cC9WDutgxI9FRAT y/kFPPOPXqhzdtRosO1QRSsANLPBBnmpgiXysoFj+AUB/igAABC4M38kQDT7YeD5/IGamHiDH6+Q xQTXpLktbXBze/tXRADmkYwMxCfj8IZ2hvKQknzEITOpSQlLsxMpYURMJ1nfONoRER2WJicz04mb /yaSQ5PssIdG/OEQPTKOD4rlBmtpjC3W8jInboVMQZwZUwJAANItZQBdNAoBCCC9pHBxjAQIBhPi QCQpEokJYrBGMxawgH8soAMIMMAAtjKAA8ivjgt4QAR0pJSgnGRm8OBOXryjFOHoxQYJ4sYNMqEH LvwFAxviRgGNkoLDnOFFncjEJSyxARvcIQd5MN9RDHA36P3Fi26Rx08cchBF7qUpwBjPfv5SiBfp IQcIugMennSBw9ziBsCZTy1YRiRMZIUAB0hA0KSjgKRcZGYVtCUjk/OwT4jnL7L5DYX6IIbDOEAb x8jEk/yhiagMIToOeAYltoWfOBSiGpSJSinyIv9BpYAmLw4o41HOYY53EKQkCxHLJWyBhyrk4KE5 qAIe8CAqGFkhE3ioBUQjWgs8OAxffdhDHnJgmTvUwgp8uM+L+KAHW+QAB3e4wQ2OIdMb3AEHmMCD Jra1GFBewi1lMskIR9gUIXnpqGbREpeQelQwMUWoEGGJCptoFCsw9apHShJWtzoVpFzEWAwJiz+m WE80KIIFX6gQlljaLEIAAARWGAxbk6pVjlGOSX2QQzQAoIkWdcmpezsWkJZinz7QCQCc6gQO5seC 3dnpDjXQQA9UsNdguA4AltDEYkVmiwECwAaaZezOEHuJ4YkME3VqwAZuUIVlmcUTnrWEpAAQjXT/ 1gkFjm0GDlJqlj04hUwaeQgPt2IDxQxmMFvFDx+MwlsiUcyvFjLZfbCQoK200ILxEBh/OBMBE+xC FjDo3//ycodrLGMHLIDBTYPWupClIhNvwAEAPzvaf1iCE1Ho3x0yQYdI4CBkBLhDnTgFLE9sNlKb HQ3P/uEWk7xkhUuZQDGyZ4ytWOmq4iPfUbs6PR7Swx0/MookdvG+rTzwGbiMRiigMQhh1EcLzDgB NHTwjFQgYk99AkADBoCJJHAhFNaoQgmCBipF/NgaeEiEq0hAAA7MLYEMaEADULcm1bTDIgTahyNc YUWj2CAPy7DGDSYKCwB4oBPT8q0/mCUvdKmr/1n3qm6XjRKPsLbDTCIgBiS45wDPHKWfSJFpKqaA r2YZZTHcehGznDPnpvRkhAfpRwanJ4AoPwMBdPRiADjTgAUggACcYcAF/HGBoIGCAQfoRAby4oz9 MGA0BlgkA0gHPAcsgBmwPN9C8jYQDI7gHPkgRz0IApGDkrAnLwkiTFwIFJxYT4gHqXNY6cISZAtE ItP2iUzmUpBpK6QdDznIOuZSErlM9U0dYeJOPicUmfC6JFIi4kReSCw3DcVzMlyISjJyEI3cpCII 2ZtKbsLsEA7VJwCfmb5tYhETwkQ1qGm0WDTR1TxoQuJjSfcIO4dxp8gpKl0l6yc62PGmpGQjKP+R UsnjdB83roXkKz/KQTU+2JIv9T4hJ5KWYj7Cl6SbKdwYI/AO84wxepEA1ECA6qTDdAAMIYocOgsT 9imZZNwIAc9YwNL1coCkDCA0evmHjTC99QUWUoYhgYg2mwKbvxjlBqXFgdxRcJibVsEKVeDQGQ7z dKugywp4wIEqNwOgo2RmkVnxy1LqYRONryYr/kE8UxyZF9lk8hJ6iMNfchGiZAInAIcpRUjxcIMN tO4SKGXOcpkC9u5QWSyyDIpIaAb51vQlNjQoOwCY4KBO5OA4m6Q8AIDwe6NsAAd60PAn+LCvpBxA 902vZlIIcqxyl2TtSpmAJ76pl3D6YwCHOcH/JWiaA5Jh4eL+OM9fVGCUCaSzuURi/lgOr5ek1EMg EcHIdmrfnaTgoJvlBE7HcHl6sAWH4QwbcAdWACJVRQXQcBi4tgGYp3z34QnAs0dZsXWvZxTgYA7Y 0S/UxzdKoQ03Qn/dcSMCoB4/8AMGaE6kgAACwA2YtxivEB3M8AyWYVUO8gMG03Shx3tREQfq9xcM YEettxcChRToQA7YUSY6cX1isQG2IF8bIoFWsAcV8yx7gCE2kEk3YAvLUmickAOW0Hx3cAXMNDF8 oAmYcAM2wA2adAyXkAOZkAd5YAV5wFCYIHf1oRjC8hbndkJ1ll1asQHGdVyIeFwOkoiMiB+D/+Ee /hBX0LUW5xIvfMAH94JcQ4IF28MVs0cR1iMRTlFcydUkSlWKTMKJWmFE+AcnSkGKqGiKdRWLS0Jd TSEXocgQKEGIRmEJFEiL+HGKwNgkfSBnRtERaVdniWQJkziMOjeLzohXEkJBC6E+peENOPCLLqKJ TPU9wQiNR9UsW4UFjCZtnuOK89EinpAHgtA/2eh33CIv8NgsfPCA91UFwqAIMsIHBmMJWPgif3gF cJYFImMFwAJn0tUHerBXG5B8iJZoJxMxQzKNRtEQM4GOR1EWE6MHdXInijU/BHADHXkHwQAAKMAD MGAIpUdAenBgOTBfDXlg9RU7eEEAecA/0/8wXgBgC+GyCZdlA5yAk5eAA72CASLZPzZgH5XBFCqk QmSSN02BCVhgBnixDF3zAapzNmmDCEtwKKSCCHtwDCEDAIiAlW5DC2KZF4iAAxYwHQoEDYmAA60C AKbwM6ZGNPiyB6sGAISABj8DDb8ADMsQAnzwBFX5A/jiTL/lb9gEhU6BB9G4FqunlJHpJFzxEizh DvAmgp8xRiwgCEFABKpwDKlyFs3YJKsHf5SoJxwwMtCFBW4FV34FiWJhbtEWiryYFZoRXriwB4/w CJBgDcywDLeAKnSAC9cgNk8AIx4QDdVQDSBwB3xQB05TDdJwA5kSCrziK3vABn4CADFwB7n/k2Mc 0AOlmS9k8Cgjw5oWkDs7BlT9Im1fQVVaQUCW0Cr/kICptVpV0Al5QAPbwAJ+AADXMF8ScAyYoC1m AAAd4A95cRwLhnp1sgw2kAl8sAc4QDw5MGB4cjIHlgcJVl8NNnvqsz4B0xQl0Au9EAudqBQbMC7u Ei580AlYaFec8I8htS0lo2bqMjHs4i4UMy0b4yDdooqbg2ylMSZKEQuOMAywMBa+eJq12DgTI44q ZaUqVUWbg5lukhBHwQuv8AjGyBQP1HVHEQAIAEXgg6URk4XckpqacAODBiMEyQJ4KKeEJi3RYhkb FIr4dyzIUgKvQAtigRcothShMQFVYAiv/7MBxGMDIaMMnHBgxzBgl1AFpEApbQhgeYAHrUIKtYAJ xeQA/iAAfhYaRlGTVSYRFvEVw0AMk/AWBcBpdvRq/rBpDoAADIACU6AHPBAdPKACZoM2AOAAYRAE zAAAv7AIcpkXppAKyfoLSrBOCkCqZ4oXHLAA0qcmEUF92DQGrxCrjSYMXpAMyYALuqCa+UKlRoJ+ HYdtMaEPrzAM5zM4NmIASWguliM5J8Ou+IExMUdBCPcRKaBn0wN+lFIASRgAQ5cABOAMQJYJHUMq 0LAAPPYDybAMpWYnsEQAoeEAezQAC8AAuepE7kZE44BBGjQ9/cQz1OCgAICBRnFA/rCfqv/UCZ71 hrCBQLDxPEeBF61kRT53EvSgsgGrFJagB6sHI1p6tEfBa+OmEPxADB4ADuFADu8Qb2YSES5xbdXY reoTcLiYPvhHbFiETYBKIA8HcR7GQ1eWbAc3bt7AQweFGqFIS7O3bB1hbLg5EUMRgh4Rbht3TaXx t0T7biBREQg1Qg6xtR2REdNGEV06Js02QkLkb/IGufMpVEFBuelWICdxUEORfwlBSxdRuCmnQu4m FyaxEdcDEfoWbq6LuSU6urrYugbReJJrcERUu2ZyuwWRQrqbE06bfYqJCclUvNa0bNfGEMp7FDnw CW6kH8/7tNX4aJx5tNzgCYwRdZ/gCcn/q7wOZz1mkpsBWwVn0RgXlnfPO3MJARJncrSGuBYVBy8t erSY23M1t3KQiXPeuxZ4ULwM57cVEb8lN7/44XLTdb8lZ0Ff4bwlZ1RDor5Fsk4rBxM7QXtKwTM+ CEEw28HRsQVAeBb1uxZMcAoP2MF9toFnigBHCE8JILNHcY7y9hH1SYRN8UCvdBSuMARDqBdpQML/ y315UQpAaMFGYQCtR0gDsHUKkGtJDHajlhSm8YQwkUhZMT9hd3tu5w8bcAWcoAd6AB+SAcZY6EaR oRfQ8AocIy5XwMAwixQKpBdQnBTAU8elcUIyYb5LocV5wWCTdxi0YXqldFMDqhfLoEmW/yB3RhEF P1wKJ7MHVkBSlmAJx0CR/mBHSDFNiJV4/KEU8vC5+KYVftzJr3EYtUAeE9iCeXEIXTjIRnECh5Ev nUB6pmcDOJAHbzamrtQZaDJsKiESdUHKfwHISyF83ieDnMDKp4AuxUcbo7V563gH2yNJYOghHJYU OmxO//AMBkBI14EaCldnxLzFgQxOqHAjGGAwueEge9Ab3DABx9F2evEDtoBMtHEDeeAJftUHmDyz IKwXh5oU1aiLKoTFTlHKxqwU9Fx5loCwelENleB3IfI/N+AcwmcK8uVlmNAJRUKbScENBxB5PgjO N8Gl+5fQxcwUx/ADuuQPyfoXTqAJtf8QU5hgC53gHqbzFwNoFMeAB2noiE3LFAJwAC+8DAT9VY3r mE2h0EthVQHYfT3NDYesFwj40zWKfoFwGNOwL/QR1EOCfnSjFaDXxUdhD/Owt8F1oiptzkexAQzi TS/NDThgBIfRAOhUH32AfpqQxnrBAaSjHnG9JFcBG1vhx9tqFEt4f/O5EeX8x0ixATnQTUQMAMl8 CZxgBymsFwRgCesBie3UBtHBAAe6z1dyCwBgpk3BwbeUFATlI9aDwVrBya+DFAuyGFSnF00w1Zi9 GIbxF8wggYqpg5/wCj+sFyvwA0fABGkAR4Tw29BwFn/RAAaQa2i6dCC7FB2Y1pILFEz/wTMMQEd0 ZEd0pDrpkQNPUwrqvd7JEAoGQwDKzBynUA3tbUeqkyDEfTJ60AJjyXTQcAojfBZHgMLRB85JgQ+M nUMa4Rae3Ykb0NGYM3oSMgGXIJCmyQf3HL6p4iEhBaJv6A/yjAN4cAXWci/Qsgc5bQt5wCD+jCbf MLooIYpigQlVcAeWAEmWcEq2oAky4glXkAdVMEqQZAM3gAmZkHocEyI20ImkuYAczi5T6Ib/swGW bFN3EFM2hQO1sAncG1KMNhajW7b76xQ3gOJ7cOZnTqPVUqMdg+ZoruZnjjkwsgfrlAk0qnyKNi4U ZwvqlAm2YAWaEKMt8ixDbXL+5q0Q/5wVNxeNwogliiildMUV5tY5Q5UVGlmZUdHomM6nThHbNiQT TuHRmB7pi16ZANvpB7FEBpwUlzDqa6Hprv5TTfGE46ZxhOXqrw6Orq5mS+G6biIS5osDuH4fsD7s X44UM0RLZYIUlDnspO7skllIQTEzCEFVwg7tub4l2N4cyK5wB10mE6Cu2F7s2C5/1JMdZWIX1bvu TtG71qikrU4k6CLG3dQlEGMLFc4cHgLrkO7OOSCVWCXr5x4X1XgU6uoJkOAFf6IM3LtWHyCdtkB3 XqAqnvDw+NLoWJAKNMAeRtIHatAAERAG/U4kvG4UgoW2j2cUmAAvh0VgClkFOIB3F/+tJOeXUTlg C3dAQHVoC1WACZtwCTq/CXqAB7Le8w+VO6iACQa5jjCvHBAD9ADw4DzJGw+FAzy5jjSeB5jQ8Geh mN3+OXuTFAhsFZvAoZpQBbnTDCyQOxiAA40AANRgCTngA3kwdBCQArXgJ8UzdM+QAjmwVwr2WCpA ZIPACaHgANtAAgCAAXcQLlEwKQSwBnvgJ9uAAysAANhwB1/wJ87wC3IOx/4QbutzLEtBVn1Q9nby fxmqOwMWO99JoDeeF5agWMTTWbJP+wCACqM1DaXFWJnQjg4wDf8wDV2AL3wQW1fQjv+wX5bKoZjj VFRcEBg85n0zGFeACnZyCSCaO4L/gAPY/w+XoAirYONt2QSexQqXcAPcT4Z5IQwEkzvWgBmPdQfz wwPHsAFB8wEX3QRXABB9Pn3yZAkAgBs5LHEAUOKOCgC6rKACMO2SJiwD+9jw19Hjx3jwxtGLR4/e unX0Pq70ZwvPjYtY9Fy6kYnPzEtXOlWxRSSVHCxYchxTtkQTzUyesGAiqgaLphslvgTwd0MPljzH ZijypAfHnUNB5ggciMXKjRt5+MjEkSPIF1uePGWCaYXspyssWZYcJxLlunjj9H68MdDwYcSf+tzV yPjw4sd8OvJ5/NhxYsuYD98YvLJdvJTw4MX7DK/zx06aVa9eLdkfZdaxZXc6vdJb/8mS67x5g7eu tsdLevRw4jQ8o2zkn1zD1txHTxUcnFRjsZXz8idMvz/yhrebnjuT8X4HKHBwx6BqADAc4ySwj6c9 SpMnXt6cU45oAFjouUwdBQAZjjvMkwm0A+kz7+BpxyQDD/IHh2P+AYART+J4ZRg4LLkhOiz4yASP HKqgzKwcMMHBrj74wOOSSzLpCA9MqsgDh6A4wQMHmiRkIZMqNmzPMD5sKZETLDiBzopjDFwJsJFM ctI37Q4aAIEGpKFkGGJi+WcZbWhQ7xhcotnghkD04GMQAAi4g4Nl0lpBGFIAWIYjAgBoIINf+EgB gAvu8GEMHCRshoX/MKBRMRuv0f8vkxymAaAZAhYAQAElO+qNJJRM4q1BAPy5AABoYLDkEhbknOaf abjAhIdlDloGGxsOagZVU6DoZBUdMDiIzkeDCwaAf6wKKlD9qih1mkM/wQIHR1mwgtk0/alzgUr9 6e6klEBbx7QoATBBGG0aAOCAYzZg6IMNGclD1Rs2iEYUSzoAAJcNW8GkBQAeGOCgJOu0qAobWiXl BkxyuENCQXAo9ZrgBFrWUUFquaMZPqUFgNpKR+uLL75qO0YWS0yARa9MvgprE09caiSIIvboYw9M blAlFTa6uoGGRu64wYq8qkJxphvC4mQTpG66o2GNgLbJxWr1wjakvsbRrTNjiHn/pZHacICsssSu MwyLjgRsbD7DLml6sJMW3HacwLhdyZFXHilBSebIPgxsf8S2eyDXztbrL5SaPGklXoh5ZORqLyFy 77LCVuyKS2w5jqw+IJccsyr87uzS3saJWiWPYnklFgOfASABAggwgJlm8tA7ObyxuCLOLDxRNhUa rpgdgNoTw2IDzTuLZ/iSwiupIxseIYYX7epsgCUCWvIkjxBxsKkPrEq8JIc8OtnkyBw60sQTK2AU 0tEeMLElDxitkOmSY6rIPPjO/pIacCh7eQXx3yREYDAEAIAalxBBvm4ADABIYyEA6EAQGvUoFjhK AXfAAwKtkQMViKsDWsiEBb+i/wMGdoQBAvBHAKPXAAfQLyXtQEluUOIP0RlDSQFkxmAcpJCDbMB0 ftiGA6LBCU/ggGIsyEGpmnGJKlgDAKSowiV0pAk87AAA1sDDFWzAkAIkYAAdOcgCFvCPBJBQc73R VmByA4fRVcsAAHBAATxCHn8kAABu3Nco9HAIAaKCBYJAxTFugDAcUGQBd8BBnKyBIwnx4GgQmGIO 9GCJABzEAB6RYwI6IsYx0gMeJGGbSfTxiF5oTgCpw2RHBpANEnJCEzxohgUEIQgQTPEYx7CJHu7Q kU3EDBN7wIpWFKGHHCSkE7Q5zQAIUEq/YeskgTFJDF4BB/r9xgahqAYRcsALHf/gwnZ3cxxitmaY 7ETTQPhj5joWsT9x1mYCOcBDHq6CmdixBgvcSKeBFkSPewbmHYar528yoRHLCSieqmFaP3+TKZNs ch38IAYk0lmAZTwjdQlIIUtsYAXabbNxedMMFjhiUO0glB68GQcF0hjNOj3jI9roDA7al5Hphc8f RLoCDk4Us1uCdJxPGg09UvAKR6RTjgdh4//8MQAHHMBTDyghAJ6RiU6EwgHbIMGkRnUQZ/CAIUqN YwMCEAAFBKBO/5PjFg1KEmaybRwMlQVIAyDH/znoIwFEwCUEYap/NOMZezAIACyBgwBGoKmC9YiE GOBFBUTvrPRgm25Gw9Z0htX/I1SiigIm9ZEIAMCo5+KMAPLQV+DR9ahS8sgaGUCVr4J0HSxcbUng 0Q9itDWdoyQAVVZiTGR6JGhBoGcebpCkwYhVsZckgFlB+iTdeIMekNVptarwiY58Yn7NbVpK+nKS k8B2BOwIxzdE05eepgRbLNSWSQAjGm+sQ2rXYi1g1Msb0IjEO+OoB0rqocxOjpeF5lVm2pzUDvZ+ JlMaA403xrEbwfFXcGpd6CxGAI5wlCMk3gFNX8QLmP1mKlttG55CtQUY83YHoRhuL/7Ma+L+argk 7RjJgnhjYE2Bd6QvFgl/T5LgvjB0u92Fr3JJEuInBXmkImlbd7zRDtYuKL+7J9mkhdOGEtYy9ser ba+TGHubwHgOxKRh4W687J3biKannhOceMcREAA7 ------=_NextPart_000_004A_01C2CAB2.FEC136E0 Content-Type: image/gif Content-Transfer-Encoding: base64 Content-Location: http://www.cis.upenn.edu/img/poster.gif R0lGODdh/gDCAKUAAP7+/iAgIAAAACgoKNWTk+7V3bxJUd2kpKwgKMVqcqwwOLRBSbxZYvbm5u7N zea8vN2stM16es2Di83d7ou0zYOkxVGDrA9Zi3KbvABJg+729pu81d3u9qzF3WKTtEF6pCBikzBq m7zV5nqkxYODg3JycklJSWpqallZWUFBQaysrJOTk/b29ry8vDAwMObm5s3NzdXV1Zubm8XF1d3d 5u7u9gAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACwAAAAA/gDCAAAG/kCAcEgsGo/IpHLJ bDqfyIB0Sq1ar9isdsvteruCsDgMLZvP6HT5y26733DwWKyu2+945GAg3fMDe31/gYCEhoOIiIKL hYqFjH6QjpGPlYeTfHN0eZydnk2MlphTh5Kko5WipqulrI6qhZoCn7S1tJeugqi4sLytvru/mKiy tsbHdcCuwbq5jc6Uws29iayaSQQEAAUGRAcITt8ACUsICkII6dpnCgtM7sah0qnzhNTyzNB/9/SY xUjZtjFo0MABgAYPGghRWEChQiEGCzgoMJDhkIIHERRA98ABuAYbCRokuBHitoXtQio0uA0jS5af lFHLFy0YzXrV+NVk9O9I/sACCxYoQMAAAoIIHRE0MPfAAAOKCBJ8G7oAHDh0RBmkE6LAIIIHQ5ca QHBAK9EGVSUoOGduKACtERAsWJrOgbsH5mjhc6ZzH85TfP/qA2xJ1iyA2ihq9XigawIJRAmsZbDg 21LJFAlYFUKAgcelLBFAaLyWQAJzjRcj0LxUawMG7eYSRcD4q4QE7hZ41NsvcL6+hOcBH2ytp5Gf Qa0aBRB0bVkEBgoebSBuIDgGQgYyB3CVO2qi34gCUKCVu+QIADQ/kLyAAN4DBTSLe+sufvdOMoXd 3K+Pf99RhmGTGGVWiZMOZAwkCI9mZCmFHQMPHCBEdNsphQ4BEsrFwGnk/pC3WVwKNKCABOPNlcA5 CpwmzmvuMGDAOZ/sxc9wNApW42+/BJjEWrRVxV1jzHVzQFUG3HVUg+NlNEQBs3H3kEbZQQdZh+Ud lSI44rTDnUEKxLUibCJCdktOOPlXZn9o/reKcUZEsGF66EVQADkPPCDEYw2s8wCEDqBHDndERGCA QegJEcFDANwGQAQSRkBAnACUFVqiJBZ6G6OFBrRnoZ7kN0yaZ6opaqitsGlLATB6gtCfyIDS24w2 xorjrL6Zck2rCJD4SWMwtbqEp5KYaROow47qzxi+9orGBBRUQEERGPnqqj2y1gorrddaW8qt0qZh wQUYWGBBBhVo0C0U/jIKV2227KqL7SGmAjDBBhRQQG+9+OZL7731ciBEvgA/e0QH+PIrMLhHTOCB ExdkcMYEDWcgcQYfELGBuUYAK4qwyxDbsbE8cUvEuBc0XPLJJ2dgssonVwCABijHfIHLRqDMcsMH CXwECE1QILEZE0wMQggRXyCEBhl0cES6qNyobdPruguJjkVgUPLEWMeM9cQ0X731xP4WQXTWJQNQ cRIThK3EBxIr/QTSFE8wBMQZgCAC224XofHeZBZLqt+AL0L1EV87TMQEFWxN80EfRLx1CEh0IDHk Qxj978kjMzGxBU/QbXgRW+dNBNOrON0u1O+iPnW8Q3xteRFBT4wB/uiFf25E4kUsLETiP7eu+cRP THw27FiLPgTpsJgu9emlRz3I4EYUjgQGstNeuNrEwz6Ez70LQbkSWOusBPcZyH2E48YLwfc0f3/c /kweV8K6ENIjEYLEnC+pcu1IBA0t/RdoAAd8RrP8JYF6WGOC5ICXBO6lDwDIa4XyVMe85DlPEPMD gOvQhj8iIO1+hTNf9jxIP8NJTmAGRAIItsYEx9mOcEk7wvqeATL41VA/8TOEyKL3NSWMK4VIuwDb vva9w70QY707oRBSCMMELoGF4IthxvrmvsDZ8H04vOEfMlg/JDSAYh6smwYLN7sRDgGJJiTXEp+4 NREeYQRZW8K4/h4YwVdc0IKpa14eY7FD6zkxCWA8oxhd+Ef9FQGNQlAiAJjIQ6yVMWGPWwIHpKg3 Kl7RilnE4qe0GADo+ZGBSbhAEZHGMwAQUmKPFIL/SDjGRGYAhT7sYiOr90Q6vuoQJCgB+xRhghPc EgUpwOPT9DhMQHiSCLLchvga4EZSujKZqzxjCV0JS0AmE5lbGwETMIC9492SDypYwSNYAEFYpAAA wDhBOSVYrRZAMJOYzAQXe0gEnCXBmSWk59yOOE0ASK6AUdRnERaItcWZgWnk9IsLxAkLd8roBe8M FqhaYAIKLk+H84Ri6154SDEK4ZQZ0J0QCsDPVvrzlWsE5CnF/jcEEIYvDcqAaB8WIQwAvAAYAIAB PFcBABVMkJiiyGghTYoEfMqLf0MQQUmTiNJFJuGLCMTa8LAZujTIyAQAiIEi+FBTX74KACYA6i8C sM49svNYcwjoULt3BKOO8WuImuQh+/nPlL6xfNd8a/Gs+tWIpqIPKYABCQJxzkqsAAYlICwESwCD FCACBXsgwSNWoAJFyACdg9BlIU4wAMZS6wQtCKYJICvRXvSRqoX0AFuN4NbYbe0Cbovm0ejaVEaq Uo3X/OLXHugEX5C1rKEo5wyCOYAW0OAPJmCBFJQLiBa0QAZ8AIBmA9CCGCQ2pwNwQQwAoYKeTgEA kv1uC75L/t3o7gOsnR3tO+2IVmRZU6NvVYJb9Yq1Usr2ZbQF6BGop5DrFWEDIeQrtf6gvkm4YJ0v cCxZ+4DVQkQUAKQdAAwwO4Ce7oMGmJVCOQNwTkpsWJ0VDUCD9wCAFezBBDfdqfxOu1GN8o6jYSzl EIZIT5LO1aSKtK0pDbfCraUSAHD8mhvL4NsCh4KcUhgxVgMBAM4GoATlLCyBMVtYQAAiq30gAYXd GQkoTyGhfHhBCwjshxbcVKzPE+rEOlCBcYGyrR4dwiQLp7T7IvKk1eTh7r4m4yUWriRnkFFOTTHo mf5WCijYMFcHAGYNlxPMfEj0I8Dq4BRHN6wkhm50Z7rh/gwLszBqltgp5Rvn7RXOaHbOr11pJ4Q5 b00ERKAx1oaMrr4eIpiJ9qogsByAFXjawcwlsWMr/AJS+LoQVSZxWPvg1w1f1h6DbfKiS7utUCO1 qKVuMQtTjePaIiEDBnTp5mIdYDQAIwae3oOJyYoCK1DaBUv2S4U13evtRpezgRjxAGRKYCu74LJT 6LCj+wBdXVg0zSzO5x/dDGNB9nkIAHadUm/M1DyP7HUHkSXDsQbog75KxKTl8GD5UALmmnjTMJCC T8Hph3irYLz99jAgTKDlAbS7wsXlw4QHEO11k9W60W3BYWGwcjRjMOH0ZeubWZttbc96qWnUL+iY KEvV/goZpn3bQwpaAIPxpgIGDCW5L7nK2GCeYgYwB+zN7bECcQ5ABcPurJP5EN4AtHuyXK1BEVhw 8KOn9b2FdDOpH165wsGR4mm0eOuMVziMAeDFsxawvKmw00CMdVSWeIQfXqH5PVSWEvbeZKmQLkvd Dj5ytSupBhDleNtytHCLI2jkzW1JFU++EesKhZV3j/tWeJ0QyrX9FkkvUKLCmfBD2PhQHY8ERsq1 CLKW2PB020bJB6fvySPWJaZdk1Zslx5GN6a1WX1Plal1+XI0AvWMQL5C0tn6xPmpMLdqaH3QQJd7 QEFFsT+AYzp9tfY0BBhgUE1HBFE1VFojMxhHPyGA/gEO6IAeYHXw9X5oUEcW+E1ApXuDUAXucnvA 4n8Kt3T3AzqPNF+ztHQTA1KrlXSpVz7/R0lKQGvpJhg7cUnuk3mWV4PwdFEI93ffVnz4hUwiZYKf tFrCUzsLqEEfsIRMuISnJFIH2DZM4DPpc4EDVkxntXuWV3+mU0Uh44NNNFSEszhESARBtlb4k3rM JzlIIIEpuE9bw1JHwDa0NkM6KHqAs32dd4e7xIN+514/CF/fRobmx0ZoSDGpVwTjgnoC9TVyyEPp Y4c0OIkbaAh7KCx++AjEJ4hMR4jIVwSEZD3Kt1Y6tmMTqDi/Q2sWOGGK9hsTNnaKoFndhwiMtU4c /khBXghqgBiGS2cE96MzZSg2hziKb8Y2SiB7EiM+LjRVR8A7dVh7iCBd11c6B6aFf8BlHqgINXcK f5OJezB+UQSMBWgErvZCm4OE2GSIwmNEb6gEEVOFGKh5G0Yj3SVB3MgKACADvWctuUgI4GhNG+Bw TjCMibhR6jgxiEJfSxAxqvhN6xUAKnBm8QcAwUZ/NMRk+shk08CPNrhiYHiCKzgyEhOQrZOESEBj tCNE6FiSBykx2jQEZ9hwzwSDRnaRJIdOzAaL+yFd7eNpWsZ72sd//beJYjgEEpg3wQgtKNiC3TNJ zFgzRORHHDQxDXmFATADAEADMABZhYUC/1aR/ijQAtW4MYtFVmYHAxMGCDKQlr6WAhGZUyn3W8VV jXeUhapAlL0IABG3ZjMWks23gg1IP6bUUthkUANzTUVzV7NHBMowj+6kaV42c263AviHCCTAAvun ToHwApXlaPpIaTXZUwoGAGNmVuxlKx8pLxBYOB6AARHoAY2zWwAQgRETAq1pmEzHUSFAMw5jOA0A Oa7pOB4wAobZAa1JjBGYkG7GUtwDAlYHj1ZZj1MWCEgWAGO5b8QVjTNQCInmB82mVaRAmoRQYkkm baZpgX3kWkyZegVQjlhTigYpNrypQUcDObXDjJBXO0MGQuBCAR4AQgIjMVV5e0Kwa/p4ZZ9X/mAs EF6HgJN7gG6RVVZlRWKP4GUk1or9iIdfuIuqFAJNGAIe6qFNOKJNKDckuoS4aQSLOFBh05tJZTYn +mMUcKIj2k2zWTg6gwEfMKCEwZNkpnKdFpfWmGXNVk5GFm9EOk4HSlYwd57x+I1Idy5C8In0s6Kt gjgfAAIf4AEJmQR7IWWeJ5EUyWR3N22TlmEV5gIadnJkJYvKFmkLxlV+pUnUpgggKKVNkJd46qW9 IQP2Bl3SpmGJJVk++mSlkJaAsAIZSVZ8IE4PeWWAMKgOqk6kgIWnqYupuaeeUCd76gsu8ALhdZ0V JnRTcFMpgGlbJQOcBQDDtqZqSWFfFpYX/sppGemkVolRUaqpdhABrHIue8FVGdgLVmCJ+QEYwJpD GqqJGaSrduAWePqrGVqnNEU60WBlfPikl4qrmcqseLAVz4qtVpiN93iRV5CDlIisUrCs3JoG3iql 4TqNn0astWesvEenG7MmuaoEBrCvzREU/eqv/9oOAtuvZtClRhAWeCqJHENDvbF51QqvdumN8sSh UOARazEU6ZCxGYuxGtuxGmsG2GEEDjAUHNsOyiIEEsIbtyqUyGOTGzisi3aunNRJ+YoE4dGxPKKx F2sO6YCxHMspT2AhRkAXbSG0RuAZY7JL0XqvqpCtMTtTT1tBEYsI6poEaoGzPcuzOuux/hp7skvg EetwHBqrK0eQK0lrk0vrKbd3iQ8bf3V5p0tgFhy7tSWbtSSbsSELBXHRDUiQtUogFyorrvHKf0xr idMquAqLmjV7BHebs1prtx5LskPhtUtQFfdRBFr7t5ebB+8qBb5Gp+yFAvRWrVErsREEt387t4+7 tVqrupubBB6xsRsbtgAQu2thJ0QAGzsbG7RrB52rXOAVTy6rYWp6i7iHroVrp1Xbtzw7t87LtY8L D08AGRqbt0OwAJkivUMAIl3bCe/KqpTHsvRAqZSgefUqvMkrOMvLuFkbubJbtz2rACkLBY3LuESQ KvfbvL2KBzPUXRALjRL1ArUKCeFr/ql1tL5G0L6Pe7ftC7+vywRXqyVnwCBrESPg6mBNasD0MKf0 h7gAnL4eSbFP4Lqr68Cy+8CV+7FncBoZ23F4EK7AJbWvIGnwc60rG36BsLhFYLENvLo+3LraG7Ry kRdnoLO9ewfpkgLi9ALFi5ZwyQhsuWAcJgMViVgpkGIE50vGVVYwYAIrwHcSBgNa1ZYwAKqdFQDM dcE3HMKHYQY3e8JF27ytq7XzCwVy0QCngQZMcRrW+8JfpVWbxgcs4FOiIAQVdWCahqgV9ghi6gLe qWlSxqjRlchXdnKLnLj8sAkg27iQq8A++7hM4MIHARMIcQbzSxFtUgeCdl7fVbwN/vpOAOdos2pl OPkCobfIGgbGkTmWfQBlgaCZnZvJItwErqu6n6zAbcEER9yloswEBpvAz/wEe1Gg6lZODSYJAODI hxYAl4Vv2KVz6ES+UkC+FUYCaNkIkHahTKZrazy16jvMTNDADEzCPVzHSBDEvrIaajDNkjXL/gsJ YEqe20xWy8akEOlpCwoIVVYJYzqeDBWnwSwMCEwEzeu8Fl3CgBvPzXwM4gF/k7ZoJUZi32cJ+UjL tKyPjrVevsxzGdZhPwoI4ZXNlbpeMkUY8gel2xrPG+sQBOEQC/EQPX0QQXvExxC7+9wbspgCoYde pbBOsSzFFKphnYnLNgdBmKbS/t75DNt4aPSGth2ZwxM9BM1LtnmQDY9yINlAuXggIsacImYADG7p di9dCSkQYV75ByeAbycmWFagAp25AiFncxRlZSaAqpGmzYnqzoPbB6h7BERLG55wIvHbs9GcB3Gh wGQtzfE4Uz63MccaCYED2tm4sMJHszo8BLGbDpWtBo47FPZMC41ht7SntJaQdqbAfcdqaKXgF3so rV+drGB92pzRvLSQ2umAz7Vgueaw2kqAPFyHxqW7D9bK2x1ILfB6007L2GE9DhxbC2ORDtrRKiOb sR7dCiwQA4Q82p+9hTZ0j3dI2sA9fDmtrxub3DzCt61CvRVcgdgq2pOghVAr/toEDNr1p9gGnt18 NN9IYNz42wlDARnym8/nwBqz7dUqFuCVWt28jcMIfuCmrclPENvgXQuAixf7awtQwhy463HtnN24 DZSjQN3RTdulDcL+KNyXPRREjQeHcico7AlkDbRQgMnEqtvHO1YEfr4RbZphrRUku+KdAOWRgqdq vQRLbnDWuN6l5d41TuSOENZzy9zr6q5qLK4vLt0xnuHRXZnw7ds3ruBlm7H4PeZj7uWUB34aXqmL sNXYjZ5wnsAY+9p4UOXnstFWXubTeOZ33rRqfpG5ZOHx7eY9SAZlwLGGXgfIvacSIOi/8sHF+rIx a6by+rQyUGwejui5INwq/vwJP34uXcLiHsw07J3kWaTnFYaqVy7DYd28B9ARl84OdN6umr2yJqAC /fwHh5VYg6CmxTVYAUAChDwFkiUDv/cHiUUYlPoIXpV/eyBYhmACXWydyo68E227kcsAEtARdWC0 zOqstWaVAZByFcZc14l/z57N7QZB+hiWUyB01/6Q3vWg4BkANR1daSwEGelp0ugCwRTD3tjYRCAZ cLy12CvmRwC2Yy7sTiBoJ0cCdxdRuCyXA6BObpePTIbYBSaqElqakLoHiExegPAC8o5zNwGxED8E sIGz9IyxO14Ejpu5Uqrc7dvHSeALNNwI/EZiH89cNS3FHAbw5YSNT9Zs/o7MBwvtTsX7zxVGXD/5 v6c70cb8vljbJcxNtBxL6LYg2Vrb81PEPlmpCOgFCN2pYbR8XPtGYepDy5hlZEa21eXsYAPf0AT/ XS3+9ac92e5LwgirBBCOt7rKwJkeBb0R0n2gpvPIbz+pbLJMYkLKbiSmWb52oOU8CM2Gquapcmi8 bjb8wRM99lzLyT2r0+rArKiSsRbvTVf4AiG3cusUyRDGB0umYWEVVr9PYvhGly6Qltr8Ton1uQq9 TtXYbhI5bWnJ5IcPvQssxxw752VLspzeLX777rdn6v0sYssfYhCtlf2mjwe2cjFQbLQ88i6AlSJG 91X/uT51zg5mnRUG/nZo6XVAEAAAAoNA0Yg8DpJKJ1MQlQqGVeu1ikAotF0Fdwv+arljRQRb5RK0 afcbHq+yIwzEQ25dNplFZLLvr8+IsPDoiGXlcG8x8DBw8O8R0kVl5XIFANDwaXOTMWBKKs+qi8zU FAwVtSGtAcEAIOGOtNZ2yGBh6MwWEtLPd1F4yemPqBN0UlLSsZDJxUlGKRgZFFJ0tNZha5W7WzU1 FguCoGoh4TZdDqEVoEC3NjnyU5DJPpK5hIje157Y2hMjFn6QzChGrR/AUNio1HpA5ksZL94mnurS zooEKw3QqfOY5oAVNLUSAqsmjBBBQEVYaPJX7V4TmMFiqByEDyEf/n4MG5IiMOabNzHcJHJZgPFj UqVySqakNuxRpwEoTCgsxohfQCcqUBwp4SmnvJ0Mbc0K0w0VuKLcxC11+1bPQacAY5q86cjqr5li /6U8uHcZNbK1LBI9Be5sKlPl4DZe2hQYPZQmCe7JSjev1puRLdu0OnMwqaGIKaJV28Zxao98/faL 6Q+sMsCAm8WWLTvn5SLYbGlZYCBBAgYLCgutaPEL0ioFVKtz4BEyJ7xRWzcKKzVz52lOA+quDTB0 npBpHEQwIDFtYlQasbBvbgteuugnoXKe5F0zfsmCLF//3im8pR4gbiLEhtoijfjeIwWB1eSizDXY OLsHN91o06kv/s+I8e+6AN2KAL3iSkuDnQUJg04z+6bbUCXr8NOOw0/m6S67zDyEK8QDvemoFAZM zEMCBJhLhzUNX6vLtgsr/Mw161DSLy/emntlFdK4YO6BAx6QYI0DDhjyxyEa8JKABRRggAACxiOl yKyQnPAlJWv8byXJpHMzSlEW/Mk0izQ6TygulPvxGwWZSvHO+qiT6aUY53yKv3pqBE3K9x4KFDkE ATAOAcbCrGIboLQYFI426TqytTg5XPKkJm97VLcbHVNFx1OGuLQLQz8FwI5TPCVlPkipa/E+WI3d TkWcLESIpx+3Ka24KhoQg9RPgyRDPkTpg4oTGKG8cD8KJZ2N/llZHYsAovQEHeJZ1Ha9ggGJiHww 1f4kNERcOfXVq94Nvw2m2TChNUwBPAD4CV1a3pVWCzYavCXY/YZdqVhwj93szn6fKrdSE9MrjLHz ADhAAR4XJlkjO6p9w1SpULWJzmVfPNXJmnOzylzH7ADKQDSmbacABRauIoH4CDA4HnqNfM0uZzAk d+YnWPzL0WpybgxXHblA44Ff3Vs4AqS+ziPiZOrrNrB/1XYRrKhzuroxA4xDz+ShU2s5QqaTXPXR CvGarmrBOjaRSrQQ8NFuaRsr+6qJ//E2s5kbxRc7t8ca3EQzqqw7DTBVW9Mtxlec7G++J4/cbHHr gtqagD/N/lQVzttbkAHPlcK77Xm2i5lJyzO002JkXA/TwFNGemPABREYOynRZeL2qtNnkhxS1a3v vXW4GzvOG+avmEV5XT/C3V69+2N1Y98jNTt4wfXcVXPEkN6oAAcc4ML+Ala+ZQEzI9piAb9Sh/Og V52KBW5tM+qM+nyhPRwZTgG2G8IBFKMFxC2lcKl4jLYG4IIUNKEFKFhBCoYxABXgRQUqOAHfokFC P8AABS2QSQgVMYlKjBB9droM5hZEIIK5CwtmWkVjEmYKCc4rZi04BhOWWAIWnbAQK6gB9ZShRCSY YInHEMIhtDiAFZiAae3Lyvs+5cMCPcwNq3gO1sQgOyQ+/k0aViwCAE7gBxcoog8yMAQNvpLDZLgg ji1AAh374BI6DrKGAOjjkyYVFh6+hzikASIWSMaNC15hfm444hBs5Ra++GEIhVQCCmCAhK/Q4Bgo WKLpbgKADx6hi/vYByxLqcrGsfBtj2xOrbAFB6AIUA1yAGYWurDJW3ySiWAcQAo0gUgjVEUI0hAC AEiAghe8MnjM7IM2j8DMAGgzCQBYwQBcSYITvABy08OZLlVzmImIrwroAkoa41AAIYEEAQ6YxSU/ 0pRwKsElAcAiGF0AjS0aAQAsKKQLcDkAGARUCOFsZkAHegIAKDMACvXjg54wPELRyhTHSxACDkAl Nyhg/n9s8F5wJojGpCBTBTUJxAnGqYJUokCEAEABQ/fBxFViAqhyAUAMHHECEpAzoDQ1IRFiEFBy pg11wnPgWw6kijWmEXEkwyfogogA8SmInw6i00X5MYRpXgGWAIVBIICKCaFC0xc67YtcQ/nPjfJu N1N1i4i8mlIreI4AEpCA3BYgWAK0wwEHCCw30kSANTbAmP3MjwtWuQcWDEQYzTxCKZnQ1Pssi7LI YEFVXnLZI8iglEWgwVrv+rRr6HUp6kKASPOAETOKgWGb299SPmlT6bggtU2QwSwPSs6v4DKmk4sB a/0B3EcElLj7GiM7U7OWLlw1HVQSFVIoeKAfNQWn/jhNAglkAI3WhLcr9jDBCo7qWjehdwkkWIF5 41teZXgRh3hNH3io25hrnTG2XeCqmHrl1TB9smZ9MaC/rrcta9iHMsNK582s1l+4yC1T8OwNNzJp Dm5ouDFFStaE2Deu60QmIbXRi4KpFysLu0W7qxjmLYgSVgBk0KXv8af1NgMnfNjMTddLFXfupV8d Zu/FS3HYD/d6B01doZLXOjBHE+wZijFYesJ6cLf8ImGoqlOqSVaKEN0pNKXYEw1B6vAsnrONMEVH Yy3KGyPVySg+aAcrsHEv9joK248UBgwgJoUErroA2gKgZPGc8n+qLKPHoTjLEtsy1Uh3wLUpwaNT /jJN4haNoTo5WlklbPDGIqXnzz46P1EFmJ/VESTN9ZLTC9rx5HqMLwnROTdChhmkidzaHLI6HaLy xoBj7RgEv8rRleF11GwG4S5P7csUdt8UBFYlGxd7cYiK86fLh2s82Tkqy8hzj6XLXzK+J8aq2C22 lTLrJv3iLs+Gla5p9CZVMXATwC6LsLnA7gWJOFGUa5yJl1VqcquuERPGU5jPvcvj1MICEf+ABSrQ ATh0wAITtzgAQmAFDWgAAB+QwwdAUAWQe4ADJNG2keScO29jBtwsdhKFjIxMQGTaMf9VhaCtkAGf +/wDFIADBX6egSEYvQo+BwDS38CBDFxgCBgw/joINp6HY6cOK6iuB7Nx42xlRzjSmGE4td/TrlRE 9goT6EAGKIDxCgAA5EOYgBUqEIIJTADpcxeT0o2e8iv4nQIlH0IFpl71Ifg9DQBHhsBXtLoTO22B OlnxrsXICJzjSC0KwG4enN6KiF/A6A34QAYwUAUKiHzpQ7iABi5QgdFnYO0gyADqVf90AIAe6rfP QAhA4AGlTwD0Qk+Du2MDb8g/D1z0XqCknOHrPIk52JzKbgZSboHZw3701q9CBS7gAd8ffend93kD dq/9IWzgAqM/fe+HEIIMVED21mf9BcyPhatHL+vKdhoLm83lr3c57LBj7LKhnRTDzG6h8wDA/gKE bgF9Tva2z+7wrhUyQAMocOmMDumY7gLmDu86QPAAgPBub+428OkecPhWjh9artuADOYSTCvGjcvw bQ/0LQ4eQi2ITQ6crvoYkPBAAANKbwhO7+hEYOkqEOT4DvyOjgNhr+OGYAQKDwBAAO980APcgPjO x/j4A/liRPmYgfnEraHMjexSg1NyrBYKgPoU8O0a0OjebvBQLwPmjgItUOlyj+k+oAoxIAQ6oAlB EOo28PacruPc0P6UZsQYbwtHDXh4TC4mL9XAbNoIsDEoiCI0zyOcbu4WUAFPLwPqj+h+7uiM8ALx LgkBYO3cTwT4cPsywAOoLgoLAPTq7wrg/ozluG3rXq5yXBDPdicGG2m6Gs4t/ssUNs8WOOACWsED hC4ZFZD0TO8CnhEI/3AIPGADc+8DOW73AGACaI8DQiAEPoA5PmACOED93OD+/ib/GOUWZ6P/KC3h JO9iPuHykiJT2sLfOu2+/CXrQu1JArAZDC4GHVHhxG7VoC8OusswvCcPOvEKKpDp0sEDGqADcm8h gRBi8oPHpIMXvy75TmfIeA3FnA8m5jG7TuMA04EVr8ADCRElx7EKSSEDrjFpvsMWFUjSHK/gGFHF rozyEMjyaNAKKgkcDGDd3qATRYADLKD0LAAEhA4DLMDiKqABIhLjItIC5O4D3o7wMKAA/oTPA9yQ AiYAKq1gIuNwGqEy5SjgKq0QBSctWfLFWGKO8mCQox7vF8cwKe4nLU5SHVgRA0CP+n5u7erwAy5A BABT9koPMN3P50Sg5KwP9i4Q9KxAKVPyE6kOMmWyCs5RHadB68LoxNrxH58HAFnJxYAxHeypEvNJ KTrxL4GvAk6vAXpvIikAGvEOFkPgL1PP5ypwAhxz7TCu5C6AE62ABP/QApxu7S7A6czREAPOMwcO J5kEIBuRJx9R2uQRKM3OG2bMFvyy9L6SAjoO6Z7yAjbAFEsu6IDvxggPA9UOBMwvA9AQAIiu58Iv 9W7vFG2TLb0lI6Um3jiSCz1y1/Yv/iT75vlQMw+CBj3KAO1qwS/fTjxFzgI9wAIA0QMVcANgkTcx sAA88PuWjvzqk+nI7+cqMOUuYO0soAHirhAZDdnS8R2n08EA1B29zB+x0whIUpOE8TCKEiY9AANG QA3HMwqDTkUvgDnYcwHZEwMxMCxV9P1I7vbqkyJt8+5gDwRCgPCWc0RPEEaiE970cR1b69n+Udw2 subq8ub0qgESJkeu7SNY8SsBYEK1ERQx9DGj9AL5TksBQOrM0ujscwjwsFAtAO+esQMCVTPryj9p TSMDNBEjpwuLzFJFMpcUlJKKCKQOzS3CcgKWVH82jgIqYO6ETwMsrgNaFD07oOKE/o4CKCBVB0/4 hK4BhK8+20HtTHE8VdUiX9TTyNQmS4xGSU0nJe86B1IAIzEK8uDVvMASFyTuXDRxQEAq5/NQwpQf q2POWDAXb+MF1XTPWuUuJfEN3GkLaucew8Q9H5JlqCxGxXRGFbHEbHQ0K01Z/QkUtKexHgBI1zVx FC8YEFHUcnRRUgxZtS7aFq4gNRVgHzZb0+Y/s5DmRI1SCbTeKvbefBHJHBZiPzbx4BXr5PVgDdZi m8bLoM1kw45HQdZlHTUwxLROhHV1zFT/7ixNJXVc96vPDPJl7/EKtZDbKhYX8wbUls/e1tTIgPJn OU1gQa1f2MZtqnMnFXZlO8RnLJsW22gxBWuyZuNSF3P2+Hb2yAZQa892M0UW/0gWy/iv6/yPNCvj avFDFIIAADs= ------=_NextPart_000_004A_01C2CAB2.FEC136E0--