ABSTRACT
Understanding the meaning of texts, capturing the intention of users, and precisely providing information services are the crucial techniques for manifold Web applications, however, they are not trivial problems. On the Web, texts often have the characteristics such as diversified structures, multifarious noises, variant language expressions, data sparseness, and so on, which make difficulties for textual understanding. To this day, although many researchers have employed multiple resources (synonym lexicons, etc.) and methods (exquisite designed patterns and natural language processing techniques) to understand the texts, the problem is far from being solved.
