| Summary: | 碩士 === 國立臺北科技大學 === 電機工程系碩士班 === 90 === Accelerated growth of the Internet and on-line news in English allow non-native English speakers to access on-line news in English more frequently. However, Chinese speaking Internet users cannot retrieve relevant topics from an enormous amount of news owing to the difficulty of making a precise query in English. Moreover, non-native English speakers cannot retrieve relevant news in English owing to limited vocabulary skills.
This study proposes a novel information retrieval system for Chinese-English cross-language when retrieving on-line news articles. Thus, Chinese speaking Internet users can formulate queries in Chinese and then retrieve relevant news in English via the proposed system. The proposed system first collects on-line news from Chinese and English news web sites daily. Sentence segmentation is then performed using the Chinese query. With sentence segmentation, the original Chinese query can be expanded to more Chinese queries. Finally, these Chinese queries can be translated into English queries and then the relevant English news retrieved. Additionally, the relation between the announcement date between Chinese news and English news for the same event is considered to enhance the precision of the proposed system.
|
|---|
