Efficient Search Solution on Unstructured Data

Abstract

Markup languages are used to store and share data, but due to the undefined structure, searching may lead to ambiguous results. Keyword searching is generally used for retrieving the relevant data from large web database. Web databases may be in any one of the forms XML, JSON, AdsML, AIML, APM L, ATML, BeerXML, SGML etc. XML is a markup language used to represent data structures like records, lists, trees. Redundant data occurs in XML and it leads to an increase in storage along with an increased cost for data transfer and manipulation. Since JSON parses unstructured data much faster as compared to XML and is more compact. The ambiguity that results from searching unstructured data can be avoided by using baseline and anchor pruning algorithms. As a Baseline and anchor pruning algorithms are used on XML to enhance search results and to improve accuracy. These algorithms are applied parallely on XML and JSON to get top-K desired and accurate results. Baseline algorithm uses diversification method to enhance search results. Anchor-pruning algorithm is used to improve accuracy. These results may help user to select relevant queries and modify them as per requirement.