Kernel- based Chinese recognition with ontology

Since the Chinese Websites have increased in the explosive Internet era, making efficient information retrieval systems has become one of the major endeavors, especially in fields of Chinese recognition. In this paper, the authors study the integration of subsequence kernel function based on ontology. Using the vector space model (VSM) to create subsequence kernels, the kernel methodology described here not only overcomes the VSM ignoring any semantic relation between words, but also results both in functional similarity and in sequence/words similarity by gap-weighted subsequences kernels, and the most important is that semantic character is also taken into account, which is very useful for Chinese recognition on Internet. Experiments show that the method has more exact retrieval results, and its cost is under the accepted tolerance.