Language analysis aims to convert text into given expressions. The text is input as the form of sentence, and the output expressions include word segmentation, POS tagging, name entities recognition, dependency parsing, word sense disambiguation and sense role labeling. Some achievements in our group have been made. For example, an unsupervised method of Chinese word sense disambiguation based on equivalent pseudoword has been proposed; our SRL finished with the 5th place in the shared task of CoNLL 2005; word segmentation and other 6 technical modules have granted a license to Kingsoft Corporation; Now we are undertaking the project “language technology platform” from the National 863 Program. |
|
|
-
Text Retrieval
( 1 items )
The study of this group focuses now on Question Answering, automatic Ontology construction, Topic Detection and Tracking, and personalized search technologies. Up to now, this research group has undertook a number of Chinese national projects, some have been completed such as "Research on the technologies of open domain Question Answering system" as the National Natural Science Fund Project and "Chinese information fuzzy matching technologies" as the Chinese 863 plan project. The group has also participated in several national information security projects, and implemented some application system for corporations using mature technologies.
-
Text Mining
( 1 items )
The main research directions are information extraction, automatic text summarization, text classification and clustering. We focus more on information extraction, including co reference resolution, relation extraction, event extraction and integration. The TM group has won the first in automatic text summarization of national 863 evaluations in 2003, and has finished the IBM university cooperation project “automatic multi-document summarization”and the TENCENT fund with text classification, clustering and summarization.
-
Paraphrase and Translation
( 1 items )
Paraphrase and machine translation was found in Dec.2006. We believe that there are close relationship between them, say the reformulation of the same sentence. In the field of paraphrase, we focus on the research of context specific lexical paraphrasing and paraphrasing for relation extraction. When it comes to translation, we emphasize it on the application level, aiming to help people with learning foreign languages. At present, we are undertaking the project “Chinese Phrasal and Sentence-level paraphrasing” supported by National Natural Science Foundation of China.
|
