Date of Award


Publication Type

Master Thesis

Degree Name



Computer Science

First Advisor

Ezeife, Christie




The ability to quickly extract information from a large amount of heterogeneous data available on the web from various Business to Consumer (B2C) or Ecommerce stores selling similar products (such as Laptops) for comparative querying and knowledge discovery remains a challenge because different web sites have different structures for their web data and web data are unstructured. For example: Find out the best and cheapest deal for Dell Laptop comparing and based on the following specification: Model: Inspiron 15 series, ram: 16gb, processor: i5, Hdd: 1 TB. The “WebOMiner” and “WebOMiner_S” systems perform automatic extraction by first parsing web html source code into a document object model (DOM) tree before using some pattern mining techniques to discover heterogeneous data types (e.g. text, image, links, lists) so that product schemas are extracted and stored in a back-end data warehouse for querying and recommendation. Although a web interface application of this system needs to be developed to make it accessible for to all users on the web.This thesis proposes a Web Recommendation System through Graphical User Interface, which is mounted readily on the web and is accessible to all users. It also performs integration of the web data consisting of all the product features such as Product model name, product description, market price subject to the retailer, etc. retained from the extraction process. Implementation is done using “Java server pages (JSP)” as the GUI designed in HTML, CSS, JavaScript and the framework used for this application is “Spring framework” which forms a bridge between the GUI and the data warehouse. SQL database is implemented to store the extracted product schemas for further integration, querying and knowledge discovery. All the technologies used are compatible with UNIX system for hosting the required application.