tag:blogger.com,1999:blog-3518334.post115890801845100240..comments2024-02-26T03:12:14.514-07:00Comments on About Translation: Foundations of Translation - Lesson 1Riccardohttp://www.blogger.com/profile/08033214185364578008noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-3518334.post-57450373055029890582010-05-30T17:30:38.398-06:002010-05-30T17:30:38.398-06:00A little typo:
'Need to summarize and shorten ...A little typo:<br />'Need to summarize and shorten the oiginal' (instead of original)Azzurra Camoglio [She/Her]https://www.blogger.com/profile/11184590433011188609noreply@blogger.comtag:blogger.com,1999:blog-3518334.post-1162395077228397242006-11-01T08:31:00.000-07:002006-11-01T08:31:00.000-07:00Hi everybody,TermExtractor, my master thesis, is o...Hi everybody,<BR/>TermExtractor, my master thesis, is online at the address http://lcl2.di.uniroma1.it !!!<BR/><BR/>TermExtractor is a software package for automatic<BR/>building, validation and maintenance of glossaries in<BR/>english language.<BR/><BR/>TermExtractor extracts terminology consensually<BR/>referred in a specific application domain. The package<BR/>takes as input a corpus of domain documents, parses<BR/>the documents, and extracts a list of "syntactically<BR/>plausible" terms (e.g. compounds, adjective-nouns,<BR/>etc.). Documents parsing assigns a greater importance<BR/>to terms with text layouts (title, bold, italic,<BR/>underlined, etc.). Two entropy-based measures, called<BR/>Domain Relevance and Domain Consensus, are then used.<BR/>Domain Consensus is used to select only the terms<BR/>which are consensually referred throughout the corpus<BR/>documents. Domain Relevance to select only the terms<BR/>which are relevant to the domain of interest, Domain<BR/>Relevance is computed with reference to a set of<BR/>contrastive terminologies from different domains.<BR/>Finally, extracted terms are further filtered using<BR/>Lexical Cohesion, that measures the degree of<BR/>association of all the words in a terminological<BR/>string. Accept files formats are: txt, pdf, ps, dvi,<BR/>tex, doc, rtf, ppt, xls, xml, html/htm, chm, wpd and<BR/>also zip archives.<BR/><BR/>--<BR/>Francesco Sclano<BR/>e-mail: francesco_sclano@yahoo.itAnonymousnoreply@blogger.com