25-08-2017, 09:32 PM
1461401886-paper.docx (Size: 11.43 KB / Downloads: 7)
Abstract: Organizations are generating, processing, and retaining data at the rate that often exceed their ability to analyze it effectively, even though this data is often key to the success of organizations. It gives them competitive advantage and helps them solve hard problems. Traditional database techniques are mostly inefficient and impractical in maintaining the fast moving and voluminous data. Hence, this requires intervention of new BigData technologies and architecture that will explicitly address issues of scalability, big data search and performance that relational database were never designed to support. One such technology is ElasticSearch which is a search server based on Lucene. It can be used to search all types of documents with the help of scalable and real time search. This project focuses on creating an application which queries and searches for relevant research papers. The users will be able to customize their search and authorized users can even upload their own research papers. In addition to this, application will also suggest list of research papers related to the initial query and display list of papers referenced by the paper or written by the same or similar set of authors. The recommendations will help users find more relevant research papers and create better understanding of their area of interest.
Introduction:
Lucene was initially written in Java to support free and open source information retrieval software library. But as it turns out, it was quite difficult to use since it’s just a library and requires Java to work with it. Hence in early 2000s, a developer names Shay Banon started to work on an abstraction layer over Lucene that made working with search applications for Java programmers easier and called it Compass. After some years, Compass libraries were rewritten to provide real-time, distributed and high-performance search engine. The standalone server was released with name of Elasticsearch.