Intelligent agent for e-commerce using genetic algorithm / Kok Sun Sun

Kok , Sun Sun (2000) Intelligent agent for e-commerce using genetic algorithm / Kok Sun Sun. Undergraduates thesis, University of Malaya.

Preview

PDF (Academic Exercise (Bachelor’s degree) Download (15Mb) | Preview

Abstract

This project report outlines the introduction, literature study and methodology, system analysis and design, system implementation, system testing and summary of the whole project. The major objective of this system is to develop an intelligent agent that can gather. analyze and categorize information from the Web on the printers that complying the parameters specified /identified by the web users, and to provide recommendations for online purchases for all types of printers. The recommended web sites are restricted to web pages that provide online purchase facilities and deliveries in Malaysia. The system is able to update the information from the database provided that the administrator monitors its activities periodically. This system is achieved using the Genetic Algorithm which is capable of performing information retrieval and learning algorithm. The basic concept of the GA is the natural selection and survival of the fittest, which is identical to the human genetic. In addition, the system is developed using the developing tools that are essential in web applications, namely. web server, database server, system server and web programming tools such as ASP and Visual Basic. In order to develop an intelligent agent, various programming techniques are used in achieving the property of self learning, information retrieval and searching algorithm. The system design is based on the features of a search engine which consist of "'Spider" to read through the HTML pages, .. Indexer" to decide and index the relevant web pages, and ''Search algorithm" to retrieve the data from the results of indexing. Whereas the process of examining through the web pages, retrieving and searching the relevant data in a liTML page, and selecting the best satisfying data are based on the features and operations of the Genetic Algorithms.

Actions (For repository staff only : Login required)