ExpFinder: An Ensemble Expert Finding Model Integrating $N$-gram Vector Space Model and $μ$CO-HITS

Abstract

Finding an expert plays a crucial role in driving successful collaborations and speeding up high-quality research development and innovations. However, the rapid growth of scientific publications and digital expertise data makes identifying the right experts a challenging problem. Existing approaches for finding experts given a topic can be categorised into information retrieval techniques based on vector space models, document language models, and graph-based models. In this paper, we propose ExpFinder, a new ensemble model for expert finding, that integrates a novel N -gram vector space model, denoted as nVSM, and a graph-based model, denoted as µCO-HITS, that is a proposed variation of the CO-HITS algorithm. The key of nVSM is to exploit recent inverse document frequency weighting method for N -gram words, and ExpFinder incorporates nVSM into µCO-HITS to achieve expert finding. We comprehensively evaluate ExpFinder on four different datasets from the academic domains in comparison with six different expert finding models. The evaluation results show that ExpFinder is an highly effective model for expert finding, substantially outperforming all the compared models in 19% to 160.2%.

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

“…learning system) 2.2. We use generate_dp_matrix() to estimate the weights of documents given T in nVSM [1]. The function estimates nTFIDF of each topic t ∈ T by integrating the nTF weighting and the nIDF weighting.…”

“…Existing expert finding models can be classified into three categories such as vector space models (VSM) [2,3], document language models (DLM) [4,5,6], or graph-based models (GM) [7,8,9]. ExpFinder [1] is an ensemble model for expert finding which integrates a novel N -gram VSM (nVSM) with a GM (µCO-HITS)-a variant of the generalised CO-HITS algorithm [7].…”

“…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 as presented in [1].…”

“…We use run_expfinder() in trainer.py to handle this step. The function receives ETopM, DTopM, ECG, c x and c d , generated in (Steps 2 and 3) as parameters, and reinforces the estimation of expert weights given topics by integrating nVSM and µCO-HITS [1]. For each t ∈ T , we perform the three steps:…”