Agent Coalition Formation of Chinese Question Answering System Based on Improved Ant Colony Algorithm

Author

Di, Shuling ; He, Pilian ; Li, Huan

Author_Institution

Sch. of Comput. Sci. & Technol., Tianjin Univ. Tianjin, Tianjin, China

fYear

2010

fDate

22-24 Jan. 2010

Firstpage

98

Lastpage

101

Abstract

In order to enhance the performance of Chinese question answering (QA) system, this paper presented a model for Chinese question answering system based on multi-agent. Optimal coalition was formatted by using improved ant colony algorithm. This paper described the problem, introduced solving steps based on ant colony algorithm (ACA), and improved ACA considering the characteristics of Chinese QA. The capability of agents and users´ satisfaction were best shown in the process of optimization. The experiment demonstrated that the improved algorithm could avoid falling into local optimums, accelerate the convergence rate and improve the ability of searching an optimum solution.

Keywords

information retrieval; multi-agent systems; natural language processing; optimisation; Chinese question answering system; agent coalition formation; ant colony algorithm; multiagent system; Electronic learning; Chinese question answering system; ant colony algorithm; coalition; improve; multi-agent system;

fLanguage

English

Publisher

ieee

Conference_Titel

e-Education, e-Business, e-Management, and e-Learning, 2010. IC4E '10. International Conference on

Conference_Location

Sanya

Print_ISBN

978-1-4244-5680-2

Electronic_ISBN

978-1-4244-5681-9

Type

conf

DOI

10.1109/IC4E.2010.19

Filename

5432376