Identification of Chinese Personal Names in Unrestricted Texts

Identification of Chinese Personal Names in Unrestricted Texts. Lawrence CHEUNG, Benjamin K. TSOU Language Information Sciences Research Ctr Tat Chee Avenue, Kowloon Tong Hong Kong {rlylc, rlbtsou} @cityu.edu.hk Maosong SUN National AI Lab., Tsinghua University, Beijing, China lkc-dcs@tsinghua.edu.cn Abstract Automatic identification of Chinese personal names in unrestricted texts is a key task in Chinese word segmentation, and can affect other NLP tasks such as word segmentation and information retrieval, if it is not properly addressed. This paper (1) demonstrates the problems of Chinese personal name identification in some IT applications, (2) analyzes the structure of Chinese personal names, and (3) further presents the relevant processing strategies. The geographical differences of Chinese personal names between Beijing and Hong Kong are highlighted at the end. It shows that variation in names across different Chinese communities constitutes a critical factor in designing Chinese personal name identification algorithm. Keywords: Chinese personal name identification, Chinese word segmentation, Chinese IT applications, Chinese linguistic differences 1� Introduction Automatic identification of Chinese personal names in unrestricted texts plays a key role in Chinese NLP tasks such as word segmentation. Personal name identification in Chinese involves many different issues that are not found in similar tasks in English. Despite extensive research in the issue in recent years (Wang et al. 1992, Song et al. 1993, Sun et al. 1995), many NLP applications still suffer from the weakness of current available performance. The paper has four sections: Section 1 illustrates some IT applications in which Chinese personal name identification plays important roles, to promote an appreciation of the significance of the research. Section 2 introduces the structure of Chinese personal names, and Section 3 discusses the relevant processing strategies. Lastly, we will highlight the significant differences of Chinese personal names between Beijing and Hong Kong in Section 4, showing the added difficulty caused by variations in different Chinese communities. 2� The Role of Automatic Identification of Personal Names in IT Applications 2.1 Information Retrieval (IR) * This research is supported in part by the National Basic Research Scheme of China, under grant no. G199803050 and by Language Information Sciences Research Centre, City University of Hong Kong. 28 IR systems, including search engine on WWW, retrieval system of digital library, text mining etc., search and retrieve the requested information from large bodies of data stored in electronic form. With the advent of the Internet, search engine techniques have become most widely influential. However; the effectiveness of current systems is not satisfactory especially for retrieving Chinese web sites/pages. We have examined how personal names interfere with the correct retrieval of relevant results in Sinai, a popular Chinese search engine, and Google [Big5 Chinese] 2. For instance, suppose we want to get some information about "EN:" (lieutenant general) from WWW, and enter that word as a query to. Sina, then some of the "relevant" web pages will be returned, as in (1)—(4): (1)Mitigt4111 (2)AiktiStli. .fr1A-11-SAMP ITIXti (0.10 (3)ISIRMVVIIMR40fig143/C.P32a10N (4) irtiFIVVE� UM4INITIgitPirlArt -** ° (Sina) (Sina) (Google [Big5 Chinese]) (Google [Big5 Chinese]) These webpages …

Content truncated to 500 words