基于维基百科的短文本语义扩展方法研究

摘要

Facing the challenges of the sparsity of short-text content and the difficulty in extracting linguistic context,we propose a solution based on the structured information feature of Wikipedia to expand the semantics of short-text using NMF algorithm.It enriches the semantic feature of the short-text information by automatically identifying the concepts in Wikipedia that are pertinent to it.Thereby,it can effectually improve the effects of data mining and analysis on short-text information.Experimental results demonstrate that compared with existed meth-ods,the use of method is able to further improve the efficiency and accuracy of the semantic extension of short-text information.%面对短文本信息内容稀疏、上下文语境提取困难的挑战，基于维基百科的结构化信息特征，提出一种利用NMF算法来扩展短文本语义的方法。通过自动识别与短文本信息语义特征相关的维基百科概念来丰富它的内容，从而有效提高短文本信息数据挖掘和分析的效果。实验结果表明与已有方法相比，应用此方法可以进一步提高短文本信息语义扩展的效率和准确率。