TAILIEUCHUNG - Báo cáo khoa học: "Graph Ranking for Sentiment Transfer"

With the aim to deal with sentiment-transfer problem, we proposed a novel approach, which integrates the sentiment orientations of documents into the graph-ranking algorithm. We apply the graph-ranking algorithm using the accurate labels of old-domain documents as well as the “pseudo” labels of new-domain documents. Experimental results show that proposed algorithm could improve the performance of baseline methods dramatically for sentiment transfer. | Graph Ranking for Sentiment Transfer Qiong Wu1 2 Songbo Tan1 and Xueqi Cheng1 institute of Computing Technology Chinese Academy of Sciences China 2 Graduate University of Chinese Academy of Sciences China wuqiong tansongbo @ cxq@ Abstract With the aim to deal with sentiment-transfer problem we proposed a novel approach which integrates the sentiment orientations of documents into the graph-ranking algorithm. We apply the graph-ranking algorithm using the accurate labels of old-domain documents as well as the pseudo labels of new-domain documents. Experimental results show that proposed algorithm could improve the performance of baseline methods dramatically for sentiment transfer. 1 Introduction With the rapid growth of reviewing pages sentiment classification is drawing more and more attention Bai et al. 2005 Pang and Lee 2008 . Generally speaking sentiment classification can be considered as a special kind of traditional text classification Tan et al. 2005 Tan 2006 . In most cases supervised learning methods can perform well Pang et al. 2002 . But when training data and test data are drawn from different domains supervised learning methods always produce disappointing results. This is so-called cross-domain sentiment classification problem or sentiment-transfer problem . Sentiment transfer is a new study field. In recent years only a few works are conducted on this field. They are generally divided into two categories. The first one needs a small amount of labeled training data for the new domain Aue and Gamon 2005 . The second one needs no labeled data for the new domain Blitzer et al. 2007 Tan et al. 2007 Andreevskaia and Bergler 2008 Tan et al. 2008 Tan et al. 2009 . In this paper we concentrate on the second category which proves to be used more widely. Graph-ranking algorithm has been successfully used in many fields Wan et al. 2006 Esuli and Sebastiani 2007 whose idea is to give a node high score if it is strongly linked with .

• Báo cáo khoa học
• Graph Ranking
• Sentiment Transfer
• Institute of Computing Technology
• báo cáo khoa học
• báo cáo ngôn ngữ
• xử lý ngôn ngữ tự nhiên
• Tweet Recommendation
• Graph Co Ranking
• Rui Yan
• ngôn ngữ tự nhiên
• Graph based Ranking Algorithms
• Sentence Extraction
• Applied to Text Summarization
• mô hình ngôn ngữ
• Bài giảng Xử lý 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 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