Text Semantic Representation
The classical text representation strategies aim to numerically represent the unstructured text documents to make them mathematically computable. With the rapid growth of information retrieval and text data mining research, the semantic text representation is attracting more and more attention. The problem is how to represent the text documents by explicit or implicit semantics instead of word occurrence in the document. The goals of semantic text representation are to improve the text clustering, classification, information retrieval and other text mining problems’s performance.
In the past decades, semantic text representation has attracted much attention in the area of information retrieval and text data mining research. There have different ways for categorizing various semantic text representation strategies. This entry generally classifies the previous efforts for this problem into two categories: explicit semantic text representation and implicit...
- 1.Alter O, Brown PO, Botstein D. Singular value decomposition for genome-wide expression data processing and modeling. Proc Natl Acad Sci USA. 97:10101–6.Google Scholar
- 2.Deerwester S, Dumais ST, Landauer TK, Furnas GW, Harshman RA. Indexing by latent semantic analysis. J Soc Inf Sci. 41(6):391–407.Google Scholar
- 3.Gerard S, Michael J. Introduction to modern information retrieval. New York: McGraw-Hill Companies; 1983.Google Scholar
- 4.Thomas H. Probabilistic latent semantic indexing. In: Proceedings of 22nd Annual International ACM SIGIR Conference on Research and Development in Information Retrieval;1999.p. 50–7.Google Scholar