TAILIEUCHUNG - Báo cáo khoa học: "A Combination of Active Learning and Semi-supervised Learning Starting with Positive and Unlabeled Examples for Word Sense Disambiguation: An Empirical Study on Japanese Web Search Query"

This paper proposes to solve the bottleneck of finding training data for word sense disambiguation (WSD) in the domain of web queries, where a complete set of ambiguous word senses are unknown. In this paper, we present a combination of active learning and semi-supervised learning method to treat the case when positive examples, which have an expected word sense in web search result, are only given. The novelty of our approach is to use “pseudo negative examples” with reliable confidence score estimated by a classifier trained with positive and unlabeled examples. We show experimentally that our proposed method achieves. | A Combination of Active Learning and Semi-supervised Learning Starting with Positive and Unlabeled Examples for Word Sense Disambiguation An Empirical Study on Japanese Web Search Query Makoto Imamura and Yasuhiro Takayama Information Technology R D Center Mitsubishi Electric Corporation 5-1-1 Ofuna Kamakura Kanagawa Japan hiro@ea . Abstract This paper proposes to solve the bottleneck of finding training data for word sense disambiguation WSD in the domain of web queries where a complete set of ambiguous word senses are unknown. In this paper we present a combination of active learning and semi-supervised learning method to treat the case when positive examples which have an expected word sense in web search result are only given. The novelty of our approach is to use pseudo negative examples with reliable confidence score estimated by a classifier trained with positive and unlabeled examples. We show experimentally that our proposed method achieves close enough WSD accuracy to the method with the manually prepared negative examples in several Japanese Web search data. 1 Introduction In Web mining for sentiment or reputation analysis it is important for reliable analysis to extract large amount of texts about certain products shops or persons with high accuracy. When retrieving texts from Web archive we often suffer from word sense ambiguity and WSD system is indispensable. For instance when we try to analyze reputation of Loft a name of variety store chain in Japan we found that simple text search retrieved many unrelated texts which contain Loft with different senses such as an attic room an angle of golf club face a movie title a name of a club with live music and so on. The words in Web search queries are often proper nouns. Then it is not trivial to discriminate these Nobuhiro Kaji Masashi Toyoda and Masaru Kitsuregawa Institute of Industrial Science The University of Tokyo 4-6-1 Komaba Meguro-ku Tokyo .

• Báo cáo khoa học
• A Combination of Active Learning
• Semi supervised Learning Starting
• Positive and Unlabeled Examples for Word Sense Disambiguation An Empirical Study on Japanese Web Search Query
• báo cáo khoa học
• báo cáo ngôn ngữ
• xử lý ngôn ngữ tự nhiên
• Bài giảng Xử lý ngôn ngữ tự nhiên
• Ngôn ngữ tự nhiên
• Kỹ thuật lập trình
• Xử lý nhập nhằng ngữ nghĩa
• Xử lý ngữ nghĩa
• Xử lý ngôn ngữ
• Natural Language Processing
• Ngôn ngữ lập trình
• Nhập nhằng cấu trúc
• Tri thức về ngôn ngữ
• Mô hình ngôn ngữ
• Mô hình n gram
• Dịch máy
• Phương pháp dịch máy
• Hiểu ngôn ngữ
• Phân tích ngữ nghĩa
• Phân loại tin tự động
• Biểu diễn vị từ
• Thuộc tính về sự kiện
• Thuộc tính về sự kiện
• Mô hình ngôn ngữ Google Book N grams
• Mô hình ngôn ngữ KenLM
• Phương pháp làm mịn
• Tách từ tiếng Việt
• Gán nhãn từ loại
• Phân tích cú pháp
• Nghĩa từ vựng
• Phân giải nhập nhằng từ
• Phân loại văn bản
• Lexical chain
• Phản hồi thông tin
• Bài toán PTCP
• Cấu trúc ngữ pháp
• Phân tích cú pháp xác suất
• CKY kết hợp xác suất
• Văn phạm phi ngữ cảnh xác suất
• Hình thái học