Improving Content-Based Filtering for NGO Recommendations to Volunteers

Authors:
J. Juslin Sega

Addresses:
Department of Computer Science and Engineering, SRM Institute of Science and Technology, Ramapuram, Chennai, Tamil Nadu, India. 

Abstract:

Volunteerism plays a crucial role in social development, yet finding the right NGO that aligns with a volunteer’s interests, skills, and location remains a challenge. Content-based filtering (CBF) has been widely used in recommendation systems, but its application to NGO-volunteer matching has been limited due to data sparsity and inefficiencies in feature selection. This paper proposes an improved content-based filtering model specifically designed for NGO recommendations, addressing the limitations identified in prior studies. Building upon previous work that applied K-Nearest Neighbours (KNN) with cosine similarity for movie recommendations, this research adapts and enhances CBF by incorporating advanced feature selection, enriched user profiling, and contextual metadata. The proposed approach integrates volunteer skills, causes of interest, past experiences, and geographical constraints into the recommendation process. Additionally, TF-IDF with semantic embeddings is utilised to improve text-based NGO profiling, enhancing the accuracy of similarity computations. The experimental results demonstrate that the proposed model provides more personalised and relevant NGO recommendations compared to traditional CBF approaches. This work contributes to the field of volunteer management systems by offering a scalable and efficient framework that fosters greater engagement between NGOs and volunteers.

Keywords: Content-Based Filtering; Recommender Systems; NGO Matching; Semantic Similarity; BERT-Based Embeddings; K-Nearest Neighbours; CBF Approaches; Volunteer Skills; NGO Recommendations.

Received on: 15/11/2024, Revised on: 10/02/2025, Accepted on: 28/03/2025, Published on: 07/06/2025

DOI: 10.69888/FTSTPL.2025.000441

FMDB Transactions on Sustainable Technoprise Letters, 2025 Vol. 3 No. 2, Pages: 115-126

  • Views : 30
  • Downloads : 4
Download PDF