Study on Semantic Similarity Computation in E-Commerce Semantic Search

Through the study on semantic similarity in e-commerce semantic search, this paper comes up with an approach for construction of upper-level ontology structure in the field of e-commerce and a new method for computation of semantic similarity between information content and concept node link distance. Semantic feature vectors of information from web pages are extracted by making use of the upper-level ontology structure and relevant feedback-based retrieval approaches. The semantic similarity relation between query conditions and document information is determined through the semantic feature vectors of documents and the similarity function which is based on the upper-level ontology structure, and this approach is applied in commercial information search.