Data Mining and Knowledge Engineering

Abstract

Recommendation systems provide appropriate solutions to the users to reduce their decision complexity. This has become very popular today in the Internet World. The design and evaluation of such systems are the essential challenges for the researcher and online professionals. But the critical task is how to obtain the user preferences. This paper focuses on the recommendation services for flats availability within Chennai city limits and its surroundings (both urban and rural areas), and how the people from Chennai city choose a flat. The introduction of different flat (options) with different modern facilities, different areas, variety of amenities, and different budgets have made consumer’s decision making more complex. Online personalized recommendation systems help to improve consumer satisfaction. Usually a recommendation system is considered to be a success if the consumer buys the recommended products. But the act of purchasing itself does not guarantee satisfaction, and a truly successful recommendation system should be one that maximizes the customer’s after-use satisfaction. Employing an innovative MultiCriteria Decision Making technique (MCDM), such as the Analytical Hierarchy Process (AHP) lays the foundation for supporting complex product comparisons and evaluation of consumers. In this paper, we demonstrate the approach of the AHP method to develop a web-based recommendation system and experimentally evaluate the system by 111 participants. All the participants are internet users. This paper focuses on the simplicity and effectiveness of the AHP algorithm and system satisfaction. This systematic study contributes to research, and practically shows how the recommendation systems helps the consumers in reducing the decision complexities and the cost (by avoiding brokerage charges), introducing a variety of products in particular and improving the strategy of business. Based on the consumer’s behavior, a product will be recommended to the prospective buyer if our model predicts his/ her satisfaction level as high. The achievability of the proposed recommendation system is validated through the system. The paper includes: I.Introduction, II. Need for the Recommendor System, III. Objective of the Experiment, IV. An Overview of AHP, V. Architectural Design of System Implemenation, VI. Experimental Evaluation of the AHP based System, VII. Experiments Results and Discussion, VIII. Conclusion and References.

Keywords

Full Text