Publication Type:

Conference Paper

Source:

2016 IEEE International Conference on Computational Intelligence and Computing Research (ICCIC) (2016)

ISBN:

9781509006120

URL:

https://ieeexplore.ieee.org/document/7919533

Keywords:

Clustering algorithms, concept generation, Concept lattice, Context, Data mining, Dimensionality reduction, Document Clustering, document extraction, document handling, domain specific keyword, FCA, formal concept analysis, Information Retrieval, keyword extraction, Lattices, Matrix decomposition, pattern clustering, Singular value decomposition, Sparse matrices, SVD

Abstract:

Nowadays Information Retrieval (IR) is difficult because of huge amount of information published on the Internet. So it is very relevant to organize documents based on its content. The proposed work address this issue by generating concepts from the documents and these documents are grouped based on a data mining approach. To generate the concept, keywords are extracted from the documents but the extracted set is very large. So for dimensionality reduction, SVD is applied. This paper proposes a novel approach for document clustering based on Formal Concept Analysis (FCA). Concept generation and dimensionality reduction are the two issues addressed here. FCA approach leads to give a fast searching result based on the domain specific keyword. The test result shows that the dimensionality reduction is attained after applying SVD.

Cite this Research Publication

Jisha R. C., Hari, S., and Shyba, S., “A novel approach for document extraction based on SVD and FCA”, in 2016 IEEE International Conference on Computational Intelligence and Computing Research (ICCIC), 2016.