Self-Tuned Descriptive Document Clustering using a Predictive Network

Document network is defined as a collection of documents that are connected by links. Document clustering become ubiquitous nowadays due to the widespread use of online databases, such as academic search engines. Topic modeling has become a widely used tool for document management because of its superior performance. However, there are few topic models differentiate the importance of documents on different topics. In this survey, can implement text rank algorithms of documents to improve topic modeling and propose to incorporate link based ranking into topic modeling. Text summarization provides an important role in information retrieval. Snippets generated by web search engines for every query result is an application of text summarization. Existing text summarization techniques shows that the indexing is done on the basis of the words present in the document and consists of an array of the posting lists. Document features such as term frequency, text length are used to allocate indexing weight to words. Specifically, topical rank is used to compute the subject stage rating of files, which indicates the significance of documents on special topics. By taking flight the topical ranking of a file as the opportunity of the record concerned in corresponding subject matter, a generalized relation is created between ranking and subject matter modeling. In this thesis, can implement topic discovery model for large number of medical database. The datasets are trained and extract the key terms based text mining and fuzzy latent semantic analysis (FLSA), a novel approach in topic modeling using fuzzy perspective. FLSA can maintain health & medical corpora redundancy problem and provides a new method to estimate the number of topics.