在话题追踪研究领域,由于话题是动态发展的,在追踪过程中会产生话题漂移的问题。针对该问题以及现有自适应方法的不足,本文提出基于反馈学习的自适应方法。该方法采用增量学习的思想,对话题追踪任务中的自适应学习机制提出了新的算法。该算法能够解决话题漂移现象,并能够弥补现有自适应方法的不足。该算法中还考虑了话题追踪任务的时序性,将时间信息引入到了算法中。本文实验采用TDT4语料中的中文部分作为测试语料,使用TDT2004的评测方法对基于反馈学习的自适应的中文话题追踪系统进行评价,实验数据表明基于反馈学习的自适应方法能够提高话题追踪的性能。
Abstract
In the field of topic detection and tracking, since topics develop dynamically, topic excursion problem may appear in the tracking process. To overcome this problem and the shortcomings of current adaptive methods, we propose a new adaptive method based on feedback learning. Based on the idea of increment learning, the paper presents a new algorithm for the adaptive learning mechanism in the task of topic tracking. This algorithm can solve the problem of topic excursion, and remedy the deficiency of current adaptive methods. Time sequence of topic tracking task is also considered in the algorithm, and time information is introduced. In the experiments, we use the Chinese part in TDT4 corpus as test corpus, and use the TDT2004 evaluation metric to evaluate the adaptive Chinese topic tracking system based on feedback learning. The experimental results show that the adaptive method based on feedback learning can improve the performance of topic tracking.
关键词
计算机应用 /
中文信息处理 /
话题追踪 /
基于反馈学习的自适应方法 /
增量学习
{{custom_keyword}} /
Key words
computer application /
Chinese information processing /
topic tracking /
adaptive based on feedback /
incremental learning
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] James Allan1 Topic Detection and Tracking: Event-based Information Organization [M]. USA: Kluwer Academic Publishers, 2002, 1 - 16.
[2] Thomas Galen Ault, Yiming Yang. Information Filtering in TREC-9 and TDT-3: A Comparative Analysis [J]. Information Retrieval, 2002, (5) : 159 - 187.
[3] V. R. Shanks, H. E. Williams. TDT2001 Topic Tracking at RM IT University [A]. The Topic Detection and Tracking (TDT) Workshop [C]. 2001.
[4] 王会珍,朱靖波,陈文亮,等. 基于一元语法模型的中文话题追踪[A]. 第二届全国计算语言学学生会议[C]. 2004: 422 - 427.
[5] Aalbersberg, I. J. Incremental Relevance Feedback [A]. In: proceedings of the 15th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval [C] , 1992: 11 - 22.
[6] Tim Leek, Richard Schwartz, and Srinivasa Sista. Probabilistic approaches to topic detection and tracking. In James Allan, editor, Topic Detection and Tracking: Event-based Information Organization [M] , USA: Kluwer Academic Publishers, 2002, 67 - 84.
[7] Linguistic Data Consortium. Creating the Annotated TDT - 4 Y2003 Evaluation Corpus [H] , TDT 2003 Evaluation Workshop, NIST, 2003.
[8] The 2004 Topic Detection and Tracking (TDT2004) Task Definition and Evaluation Plan [H] , version 1.0, http://www.nist.gov/speech/tests/tdt/tdt2002/evalplan.htm, 2004.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
国家自然科学基金资助项目(60473140)
{{custom_fund}}
PDF(576 KB)
