원문정보
초록
영어
In this paper we propose a model of classification based on the principle of the fuzzy proximity of the terms within the documents. Given the heterogeneous nature of the Arabic documents in our possession, we have studied for this purpose the research model based on the semantic proximity of terms and inspired from the classic Boolean model.
Our approach is based on the assumption that more the occurrences of terms in query are close with good connectivity in the extracted semantic graph from the set of document , more this document is relevant to this query.
We propose a measure that provides a contextual and semantic search. We used not only a semantic graph to highlight the semantic connections between terms, but also an auxiliary dictionary to increase the connectivity of the graph and therefore the discrimination of documents relevant to the query.
목차
1. Introduction
2. Measure of proximity
2.1. Semantic proximity
2.2. The model of fuzzy proximity
2.3. Relevance of a query against a document
3. The semantic resources
4. Text preprocessing
5. Radial basis function
5.1. Definition
5.2. Approximation
6. Construction of the auxiliary dictionary
7. Construction of the semantic graph
8. Choice of the radial basis function
9. Our new model: The fuzzy model of the semantic proximity with radial basis
9.1 The local semantic relevance
9.2 The fuzzy model of the semantic proximity with radial
10. Results
11. Conclusion
References