Text Summarization using weighted Archetypal Analysis

Present research held on the Matrix Factorization approaches to Query-Focused Multi-Document Summarization searched by hard/soft clustering or Low rank approximation methods. We utilised distinct method called matrix Factorization to Query focused multi-document Summarization. Query Focused Summarization means set of sentences are integrated to graph and weighted by similarity between query and sentences in the document. Data set boundaries are weighted by values on positive or negative sentences. This project mainly focused on using Archetypal Analysis algorithm in existing matrix factorization method to show the important sentences of the document in the summary also studied model called multi-element graph and its effects on query focused summarization through weighted archetypal analysis.