A Collaborative Method On Association Semantic Knowledge Base Construction

A great deal of information is distributed on the Internet with various forms,how to find comprehensive and accurate information has always been the goal ofmany network applications. Search engines could satisfy users'information need insome certain degree by implementing simple key words for retrieval. However,information need itself is usually too complicated to be expressed with words,sentences or even paragraphs. In real world, some means such as classification andcomparison could help people find out what they really want, which is unavailablefor search engines. In the traditional search engine, we always use the keywords asthe content of the query but not to meet the user large grained requirements. Itbecame a hot topic to how to meet the user requirement in all face as much aspossible.With the development of natural language process technology, semanticknowledge bases have been more and more widely applied, which is the emergenceof the basic resource of semantic analysis and computing in recent years. Now thereare mainly two methods of constructing semantic knowledge bases: one is themanually constructed method based on linguists and the other is the automaticallyconstructed method based on specific framework. However, the first method isaccurate and rigorous,but it also has some disadvantages such as long cycle andhardly-enlarged scale. And the second method has to face difficult problems such asaccuracy and rational validation. In addition, most semantic knowledge bases areindependent with each other, which has the defects such as insufficient knowledgegenerality and low resource utilizing rates.This paper presents an approach for the construction of an open semanticknowledge base. The knowledge base satifies the requirements below:(1) As the main describing objects, categories of the knowledge base all havesemantic attributes and basic attributes.(2) Categories determine the contents and the organization form of basic units—items whose existence and description depend on categories. Items not only cancontain information of the words themselves, but also can contain the categoriesinformation. And in our knowledge base each item might belong to differentcategories according to its meaning. Extracting attributes of categories, especially the basic attributes, is the core content of constructing knowledge base.Firstly, in order to extract the semantic attributes, the paper presents anapproach for the construction of an open semantic knowledge base based on thefusion of Wikipedia and HowNet. In this knowledge base, each item in Wikipedia ismerged into the HowNet semantic category. At the same time, the semanticattributes of each HowNet semantic class are auto-extracted from Wikipedia itemsthat belong to the same semantic class. Secondly, the paper proposes an autoexractionapproach of basic attributes for the starting point of categories'basicattributes and the premise of determining the itmes'content.Experiments show that the proposed algorithm reaches high accuracy and isfeasible and effective.