In the past few years there has been a significant rise in the  ranking function development of functions and top retrieval algorithms. This enables the users in the search of ad hoc and database retrieves that includes the data  buyers who are searching for products and etc. apart from this we also introduce a gratis problem like how to properly guide a seller while he is looking for the best products that matches his attribute. This will help our product to stand out in the crowd as a top notch product that can compete with the other existing products in the market.

Being an extraordinary product it will automatically grab the attraction of the pool of esteemed buyers. Several formulations are also developed by us and although these are NP complete but we still use various exact and approximate algorithms. One of the exact algorithm is that which based on integer programming or  formulations of IP. Again another exact method is based on algorithms of maximal frequent item set mining. On the other hands greedy heuristics are the base materials for approximation algorithms. A detailed study can help us to know the advantages of our synthetic and real data methods, used by us.

In this project work the problem of selecting the most suitable attributes is focused greatly. The product must be highly ranked and stands out in the crowd. In order to produce good ratios of approximation greedy algorithms are shown experimentally. The problems considered here are important to the exploration point of the ad has data but the definition of the problems does have its own limitations. In some applications nether query log nor the database is available for the purpose of analysis. In such a case we have to make guesses assumptions about the users preference and type of competition.