Compute Query and Document Similarity using Explicit Semantic Analysis

There is a lot of data and information that exists all around us. With the increase in data around us, it is essential to find a way to retrieve and organize the data in an efficient manner. Using generic search engines would not be the best use case for domain specific data. Thus, customized search engines based on different domains are required. The submodule used in the search engine like spell checkers can be used in various other real-world problems like grammar correction in documents and so on. In this paper Natural Language Processing based technique is proposed to achieve efficient and accurate information retrieval. It calculates nDCG scores based on several analyses such as Spell Check, Information Retrieval, Query Autocomplete. The IR System in this paper reaches a nDCG score of 0.32. Since the concept space of the Semantic Interpreter is spanned by only 100 articles. It is mainly focused to compute scores based on the query being entered and results in the form of a graph.