An open-source framework for ExpFinder integrating $N$-gram Vector Space Model and $\mu$CO-HITS

TL;DR

This paper introduces ExpFinder, an open-source ensemble framework combining $N$-gram vector space and graph-based models to improve expert finding accuracy in academic and research settings.

Contribution

The paper presents a novel ensemble model, ExpFinder, integrating $n$VSM and $O-HITS$, advancing expert finding methods with improved performance.

Findings

ExpFinder significantly outperforms existing expert finding models.

The framework effectively combines text and graph-based approaches.

It demonstrates high accuracy in identifying domain experts.

Abstract

Finding experts drives successful collaborations and high-quality product development in academic and research domains. To contribute to the expert finding research community, we have developed ExpFinder which is a novel ensemble model for expert finding by integrating an $N$-gram vector space model ($n$VSM) and a graph-based model ($\mu$CO-HITS). This paper provides descriptions of ExpFinder's architecture, key components, functionalities, and illustrative examples. ExpFinder is an effective and competitive model for expert finding, significantly outperforming a number of expert finding models.