An Improved Random Walk Restart Algorithm for Multisimilarity Enhanced Academic Recommendation Systems

OUR PROPOSED PROJECT TITLE:

AcademicRec: Personalized Research Paper Recommendation System Using RWR

IEEE BASE PAPER ABSTRACT:

In the academic research field, identifying suitable collaboration partners and selecting appropriate journals for publication remain significant challenges for researchers. Existing academic recommendation systems often fail to provide person- alized, accurate, and efficient recommendations. To address these issues, this article proposes an innovative academic recommendation system that incorporates multisimilarity features. By constructing an academic collaboration network and optimizing the transfer probability matrix to reflect scholars’ relationships, the system captures scholars’ collaborative tendencies and potential connections. A key innovation of this work is the proposed Muls-IRWR algorithm, which improves traditional random walk with restart (RWR) by integrating various similarity measures. Using a subset of the DBLP citation data, we develop our academic collaboration network to calculate precise scholar similarities. Experimental results demonstrate that our system significantly outperforms existing models in terms of recommendation accuracy and efficiency, highlighting its practical value and potential for use in real-world academic applications.

PROJECT OUTPUT VIDEO:

OUR PROPOSED PROJECT ABSTRACT:

The rapid growth of scholarly publications across diverse domains has created a significant challenge for researchers in identifying relevant and high-quality research papers aligned with their interests. Traditional search-based systems rely heavily on keyword matching, which often fails to capture the deeper semantic relationships between research works and user preferences. As a result, researchers may overlook important papers or spend excessive time filtering irrelevant results. This has led to an increasing demand for intelligent, personalized recommendation systems that can effectively guide users through vast academic repositories by leveraging advanced computational techniques.

The need for such systems is further emphasized by the dynamic nature of research interests and the interconnected structure of academic knowledge. Conventional recommendation approaches, such as content-based or collaborative filtering, have limitations in capturing complex relationships among research papers and user interactions. In contrast, graph-based approaches provide a more holistic representation by modeling papers and their similarities as interconnected nodes. Random Walk with Restart (RWR), a powerful approach, enables the system to explore these relationships while maintaining personalization by biasing the walk toward the user’s interaction history. This approach ensures that recommendations are not only relevant but also contextually meaningful within the research landscape.

AcademicRec: Personalized Research Paper Recommendation System Using RWR, presents an intelligent recommendation framework developed using Java for backend processing, JSP, CSS, and JavaScript for the frontend interface, and MySQL for data management. The system constructs a knowledge graph where research papers are represented as nodes and their similarities are stored as weighted edges. User interactions, such as viewing or engaging with papers, are captured and used to initialize a personalized restart vector. The RWR algorithm is then applied iteratively to propagate relevance scores across the graph, effectively identifying papers that are closely related to the user’s interests through both direct and indirect connections.

The proposed system provides a user-friendly interface for researchers to explore personalized recommendations, search for papers, and interact with the academic content. The recommendation results are ranked based on computed relevance scores, ensuring that users receive prioritized and meaningful suggestions. Additionally, the system supports administrative functionalities for managing papers, users, and the underlying graph structure. By integrating graph-based learning with an intuitive web interface, AcademicRec demonstrates an effective solution for enhancing research discovery, reducing information overload, and improving the overall efficiency of academic exploration.

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS:

System : Pentium i3 Processor.
Hard Disk : 20 GB.
Monitor : 15’’ LED.
Input Devices : Keyboard, Mouse.
Ram : 8 GB.

SOFTWARE REQUIREMENTS:

Operating system : Windows 10/11.
Coding Language : Java.
Frontend : JSP, CSS, JavaScript.
JDK Version : JDK 23.0.1.
IDE Tool : Apache Netbeans IDE 24.
Tomcat Server Version : Apache Tomcat 9.0.84
Database : MYSQL 8.0.

REFERENCE:

Yihang Huang, Tengfei Li, Yu Wu, Liangtian Wan, Hainan Wu, Xiaojie Wang, and Zhaolong Ning, “An Improved Random Walk Restart Algorithm for Multisimilarity Enhanced Academic Recommendation Systems”, IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS, VOL. 12, NO. 5, OCTOBER 2025.

java ieee projects

An Improved Random Walk Restart Algorithm for Multisimilarity Enhanced Academic Recommendation Systems