An important issue in applying machine learning algorithms to Natural Language Processing areas such as Named Entity Recognition tasks is to overcome the lack of tagged corpora. Several bootstrapping methods such as co-training have been proposed as a solution. In this paper, we present a different approach using the Web resources. A Named Entity (NE) tagged corpus is generated from the Web using about 3,000 names as seeds. The generated corpus may have a lower quality than the manually tagged corpus but its size can be increased sufficiently. Several features are developed and the decision list is learned using the generated corpus. Our method is verified by comparing it to both the decision list learned on the manual corpus and the DL-CoTrain method. We also present a two-level classification by cascading highly precise lexical patterns and the decision list to improve the performance.

This paper describes a practical Japanese natural language Question Answering system adopting effective selection of dynamic passages, Lexico-Semantic Patterns (LSP), and Predictive Answer Indexing. By analyzing the previous TREC QA data, we defined a dynamic passage unit and developed a passage selection method suitable for Question Answering. Using LSP, we identify the answer type of a question and detect answer candidates without any deep linguistic analyses of the texts. To guarantee a short response time, Predictive Answer Indexing is combined into our overall system architecture. As a result of the three engineering techniques, our system showed excellent performance when evaluated by mean reciprocal rank (MRR) in NTCIR-3 QAC-1.