Customized Mining of Big Web Resources using an Intelligent Mobile Agent: A Case of Product Comparison, Evaluation and Recommendation

Web has become a large and important source of information in routine as well as spectacular decisions related to one life. Web is a collaborative platform containing big amount of data which are really unstructured and unorganized. This research work proposes a light weight fuzzy mobile agent developed in java that works in conjunction of a multi-layer architecture. The agent utilizes user profiles containing information about prospective users in fuzzy manner. Later these values are converted into crisp values and sent to the mobile agent. The agent mines Web for the required information and presents the information to users after fine tuning with the underlying fuzzy rule based system along with justification. Using the architecture a prototype system is developed for product comparison, evaluation and recommendation. The paper discusses the problem, objectives of the research work, literature survey, architecture of the system, fuzzy user profile with membership function and java code segment of mobile agent. Sample outputs are future extension (such as mobile app with image processing interface and big data) on the work are also presented here.