English-Lithuanian-English Lexicon Database Management System for MT .


136 views
Uploaded on:
Category: Sales / Marketing
Description
English-Lithuanian-English Dictionary Database Administration Framework for MT. Singleton Labs. Gintaras Barisevicius and Elvinas Cernys. Kaunas College of Innovation, Bureau of Programming Building. Circumstance in Lithuania. Circumstance in Lithuania. General electronic word references
Transcripts
Slide 1

English-Lithuanian-English Lexicon Database Management System for MT Singleton Labs. Gintaras Barisevicius and Elvinas Cernys Kaunas University of Technology, Department of Software Engineering

Slide 2

Situation in Lithuania

Slide 3

Situation in Lithuania General electronic word references http://www.fotonija.lt ; http://www.led.lt Morphological investigation apparatuses Text corpora (100 mln. words) http://donelaitis.vdu.lt Speech acknowledgment frameworks Machine interpretation examine

Slide 4

Previous lexicon Open to the client lexicon Rigid word reference structure Lack of traits Not all parts of discourse included Indexed records for lexicon stockpiling Polysemy excluded Phrases excluded

Slide 5

Requirement to new framework Open to the client word reference Easy administration of the characteristics All parts of discourse Big volume stockpiling Solution to polysemy and equivalent words

Slide 6

Project measure

Slide 7

Current framework Orientated to MT More qualities, simple to develop All parts of discourse included Database for word reference stockpiling Polysemous words, spaces Automatic era of morphological structures System can deal with different OS.

Slide 8

Development prepare From C++ to Java Rational Rose instrument CVS for adaptation control administration MySql database

Slide 9

Adding new dialects

Slide 10

System arrangement localy

Slide 11

System sending on the web

Slide 12

Future desire Phrases Text corpora Syntax run acknowledgment Additional components Possible other interpretation decisions WEB interpretation Video subtitle interpretation

Slide 13

Text corpora use in MT The pen is on the table . PEN RASIKLIS use with STALAS is all the more frequently! Look use with RASIKLIS STALAS TABLE LENTELE Look use with RASIKLIS

Slide 14

Conclusions Thorough examination of Lithuanian and English dialect directed Additional components to the word reference must be included (expressions, linguistic structure rules) Filling the lexicon can be begun Machine interpretation is in progress

Slide 15

Thank you for your consideration. Gintaras Barisevicius gintaras.barisevicius@stud.ktu.lt gintaras.barisevicius@singleton-labs.lt

Recommended
View more...