Efficient Frequent Pattern Mining Techniques of Semi Structured data: a Survey

Semi - structured data are a huge amount of complex and heterogeneous data sets. Such models capture data that are not inte ntionally structured, but are structured heterogeneously. These databases evolve so quickly like run time report generated by ERPs, World - Wide Web with its HTML pages, text files, bibliographies, various logs generated etc. These huge and varied become d ifficult to retrieve relevant information User is often interested in integrating various formats (like in biomedical data text, image or structured) that are generally realized as files, and also wants to access them in an integrated fashion. Users not o nly query the data to find a particular piece of information, but he is also keen in knowing better understanding of the query. Because of this variety, semi - structured DBs do not come with a conceptual schema. To make these databases more accessible to u sers a rich conceptual model is needed. Traditional retrieving techniques are not directly applied on these databases. Unfortunately the tools and methodologies used for RDBMS do not give efficient results and so fail to bridge the gap. Hence efficient and scalable methods for mining the semi - structured data is needed, via discovering rule or patterns from the huge semi - structured databases. These databases are modelled by trees and graphs