3-547msg1.txt

运用贝叶斯网编的一个识别垃圾邮件的程序

Subject: proposal for a morphological database of classical greek

i be write to seek input on a proposal that we tentatively plan to submit to the neh at the end of the summer . the idea be fairly simple : we want to use the morphological parser that we have be develop for the past eight year or so to generate morphological analysis of every unique string in the thesaurus lingua graeca ( tlg ) , the database of greek text available on cd rom from uc irvine . the tlg be large - - 42 million word at present and a new version be due out later this year with 57 million word . greek be a highly inflect language - - not as bid as georgian and some other , but a verb can , with prefix , have million of different form . the tlg corpus extend over a thousand year and include virtually all literary greek , and thus would support diachronic as well as synchronic linguistic analysis . i would like to know if there be anything we could do that would make this work on greek useful for the linguistics community in general ? classicist need this database , but it would be very excite if it could stimulate additional work . the work summary of the project follow . the proposal outline be fairly succinct ( c . 7 page ) but it be full of greek and do not lend itself readily to transliteration . if you would like to see a copy , please send me your us mail address and we will send one to you . casual reaction to just this summary be , however , more than welcome . note : reactions need not be positive . if this do not seem a worthwhile thing to pursue , i would love to know why . thank ! gregory crane department of classic boylston 319 harvard university cambridge ma 02138 crane @ ikaro . harvard . edu a linguistic database of classical greek this project will extend an exist parser for classical greek , expand its database of stem to cover the majority of all word attest in the literary record , and will use this database to create a morphologically parse database of more than 1 , 0 , 0 unique string available in the tlg : in the end , we will publish the database of analyze string , the database of stem and ending which drive the parser and the parser itself . the result database be an essential piece of scholarly infrastructure that will ( 1 ) revolutionize current search technique for the tlg and other greek database , ( 2 ) make it possible to apply more sophisticate retrieval / text analysis to greek text , and ( 3 ) provide a basic but crucial lookup tool that will aid non-specialist in other field ( e . g . , philosophy , political science , religion ) who seek to work directly with the greek database . note : this document be a sketch for a possible proposal to be submit to the neh at the end of august 1992 . it be , in effect , a proposal for a proposal and be thus open to revision on any and all point .