Web Based Item to Item Recommendation System
From the past few decades, Recommendation systems (RS) are ubiquitously used on the World Wide Web and are increasingly being use in more application scenarios. Traditionally, these systems have been closely linked with information retrieval systems. As they used similar models to identify relevant
2025-06-28 16:36:45 - Adil Khan
Web Based Item to Item Recommendation System
Project Area of Specialization Computer ScienceProject SummaryFrom the past few decades, Recommendation systems (RS) are ubiquitously used on the World Wide Web and are increasingly being use in more application scenarios. Traditionally, these systems have been closely linked with information retrieval systems. As they used similar models to identify relevant data. However, in recent years with the rapid development of these systems, the two fields have come to grow more separate. Currently, RS are categorized as collaborative, content and hybrid-based models. From the Empirical study of the literature, it has been found that in collaborative filtering limited factors will be used to construct the user and item profiles. Moreover, in our proposed project we firstly investigate the lacking factors of user and item profiles and proposed and implement the hybrid recommendation algorithm. The basic modules of the proposed software will be searching, extraction, refinement and recommendation. The proposed solution will not only facilitate the users by recommending as accurate and effective items but also provide a leading edge to the researchers. So that, they can enhance their researches in the respected discipline.
Project ObjectivesThe basic objectives of the proposed software will be
- Empirical analysis of the literature has been done where the proposed algorithms and research factors have been found.
- A testing web site have been developed for sales and purchase of clothing items
- In the extraction module user purchased behaviour will be extracting using research factors (i.e. using rating, user-query and user search history)
- Item and customer profiles will be refined by the extracted factors using cosine similarity Algorithm
- At the end the proposed system will recommend the item using collaborative filtering and matrix-based hybrid approach
For the successful implementation of the proposed project we used unified process modelling approach the whole system is divided into two modules entitled as client side and server side. Extraction, Searching and Refinement are the part of client side while Item and customer profiles management will be done on the server side. Php and Xampp are used as development tools for the successful implementation of the proposed project.
Benefits of the ProjectFollowing are the advantages gained by our proposed software:
- From the market survey and literature review it has been found that currently in Pakistan, very limited websites have been providing recommendation of their items to their customers. This website will become a leading edge for such websites.
- The proposed software will provide companies to boost their revenue by engaging multiple customers in a more deeper level.
- The proposed software will facilitate the customers to eliminate and reduce the serendipity issues.
- This system will provide acheap and effective platform for selling and purchasing of numerous items on a single platform.
At the end of the project we will submit the proposed software code and report to the university authorities.
Final Deliverable of the Project Software SystemCore Industry ITOther IndustriesCore Technology OthersOther TechnologiesSustainable Development Goals Industry, Innovation and InfrastructureRequired Resources| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 80000 | |||
| Mobile | Equipment | 2 | 35000 | 70000 |
| Documentation | Miscellaneous | 1 | 10000 | 10000 |