Multilingual Marine Thesaurus Construction Based On The Category System Of Wikipedia

The ocean as the cradle of life, the treasure house of our resource, and its importance is self-evident. However the sea is vast and mysterious, for example, in our country there are 180000 kilometers coastline. The only way which we use to open the mystical veil of the sea is the scientific management and investigation method.With the development of science、technology and economy and with the continuous input of funds and human resources, the ocean research career of china have get great development. On the one hand, our country not only active participates in international cooperation in research, but also organizes the establishment of international research projects, attracting a number of countries. On the other hand, we have developed a lot of voyages independently using advanced experiment equipment and have accumulated a lot of precious data. Based on the present situation of the above two aspects, how to carry out scientific and effective management of data is particularly important.This paper is dedicated to the research in Wikipedia as semantic resources to build multilingual marine thesaurus, the main work includes:1. Researching on the characteristic of the organization of Wikipedia, analysis the feasibility of taking Wikipedia as semantic resources.2. Researching on the preparation standard and microstructure of thesaurus and format the mapping relationship between Wikipedia and thesaurus.3. Researching the algorithm of computing the semantic correlation based on Wikipedia.4. Reaching on the characteristics of Wikipedia dump data and extracting the information.5. Reaching on the characteristic of support vector machine classification and classification the information that has extracted using semantic correlation algorithms.