A decision support engine: Heuristic review analysis on information extraction system and mining comparable objects from comparable concepts (Decision support engine)

摘要

Decision making involves comparing solution with each other in Decision support but it also necessary recommend which objects are comparable and in what way. This is challenging Question in data and knowledge processing, which urges for better pattern mining. Today's web is web of document where we find reviews, complaints, feedbacks posted on blogs, e-commerce websites and social networks which are rich source of knowledge for pattern mining. Research presents analyzing comparable question and then extraction of information for two objects as comparable and if not recommendation on objects comparable, if comparable answers. We propose a Decision support Engine that Answers queries asked by finding comparable objects if not comparable identifying user search intent find comparable object and answer with key values comparing them. Current state of art research system check if objects are comparable if not they don't provide recommendation to user for comparable objects. Supervised methods are limited to set of input and expected output, whereas unsupervised system output at times is false positive. In order to overcome this limitation semi-supervised methodology is used to develop algorithm for mining. This article is outcome of Methodical summary of literature on current research scope in data mining and NLP. Precisely it is analysis on abstract methodology and research scope on 24 appropriate manuscripts retrieved as per our research domain. The search contributes to field of Information retrieval and web search by solving five Research Question and major issues and challenges with procedures. Types of patterns that have been extracted in previous approaches with new learned method to extract complex patterns. In General review consequences demonstrate as many scholars have worked on pattern mining and decision support system there is need of precision and accuracy in pattern mining. Evaluation of research needs to be tested with various parameters this research evaluates decision engine with Mean average precision (MAP) and feedback rating of user to answers produced by decision engine, with regular evaluation of precision and recall.