Comparative study of name disambiguation problem using a scalable blocking-based framework

In this paper, we consider the problem of ambiguous author names in bibliographic citations, and comparatively study alternative approaches to identify and correct such name variants (e.g., “Vannevar Bush” and “V. Vush”). Our study is based on a scalable two-step framework, where step 1 is to substantially reduce the number of candidates via blocking, and step 2 is to measure the distance of two names via coauthor information. Combining four blocking methods and seven distance measures on four data sets, we present extensive experimental results, and identify combinations that are scalable and effective to disambiguate author names in citations. Categories and Subject Descriptors H.3.3 [Information Systems]: Information Search and Retrieval General Terms Algorithms Keywords Name Disambiguation, Blocking, Measuring Distances

• Type: Conference
• Year: 2005
• Where: JCDL
• Authors: Byung-Won On, Dongwon Lee, Jaewoo Kang, Prasenjit Mitra
• Keywords: Ambiguous Author Names, Author Names, JCDL 2005, Scalable Two-step Framework