Information Retrieval and Machine Learning Methods for Academic Expert Finding - Algorithms 2024 17 51 - abstract

Aug 5, 2026

In the context of academic expert finding, this paper investigates and compares information retrieval (IR) and machine learning (ML) approaches, including deep learning, for identifying researchers who are experts in specific domains based on a user’s expertise request. IR methods build multifaceted textual profiles for each expert by clustering information from their scientific publications. ML methods formulate expert finding as text classification using experts’ authored publications. Experiments on biomedical datasets PMSC-UGR and CORD-19 show IR techniques are generally more robust and suitable, with select exceptions achieving strong performance.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages18

Reading Time45 min

AccessFree to download

Introduction

Expert finding overview and applications

Academic expert finding use cases

Paper objective: IR vs ML methods

algorithms Article Information Retrieval and Machine Learning Methods for Academic Expert Finding Luis M. de Campos 1, *, Juan M. Fernández-Luna 1, Juan F. Huete 1, Francisco J. Ribadas-Pena 2 and Néstor Bolaños 1 Citation: de Campos, L.M.; Fernández-Luna, J.M.; Huete, J.F.; Ribadas-Pena, F.J.; Bolaños, N. Information Retrieval and Machine Learning Methods for Academic Expert Finding. Algorithms 2024, 17, 51. [https://doi.org/10.3390/](https://doi.org/10.3390/)[ ](https://doi.org/10.3390/)a17020051 Academic Editors: Edward Rolando Núñez-Valdez, Vicente García-Díaz and Frank Werner Received: 18 December 2023 Revised: 11 January 2024 Accepted: 19 January 2024 Published: 23 January 2024 Copyright: © 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ([https://](https://)[ ](https://)[creativecommons.org/licenses/by/](creativecommons.org/licenses/by/)[ ](creativecommons.org/licenses/by/)[4.0/](4.0/)) . 1 Departamento de Ciencias de la Computación e Inteligencia Artificial, ETSI Informática y de Telecomunicación, CITIC-UGR, Universidad de Granada, 18071 Granada, Spain; [jmfluna@decsai.ugr.es](jmfluna@decsai.ugr.es) (J.M.F.-L.); [jhg@decsai.ugr.es](jhg@decsai.ugr.es) (J.F.H.); nestor.bolanos@ugr.es (N.B.) 2 Departamento de Informática, E.S. Enxeñaría Informática, Edificio Politécnico, Universidade de Vigo, 32004 Ourense, Spain; [ribadas@uvigo.gal](ribadas@uvigo.gal) * Correspondence: [lci@decsai.ugr.es](lci@decsai.ugr.es) Abstract: In the context of academic expert finding, this paper investigates and compares the performance of information retrieval (IR) and machine learning (ML) methods, including deep learning, to approach the problem of identifying academic figures who are experts in different domains when a potential user requests their expertise. IR-based methods construct multifaceted textual profiles for each expert by clustering information from their scientific publications. Several methods fully tailored for this problem are presented in this paper. In contrast, ML-based methods treat expert finding as a classification task, training automatic text classifiers using publications authored by experts. By comparing these approaches, we contribute to a deeper understanding of academic-expert-finding techniques and their applicability in knowledge discovery. These methods are tested with two large datasets from the biomedical field: PMSC-UGR and CORD-19 . The results show how IR techniques were, in general, more robust with both datasets and more suitable than the ML-based ones, with some exceptions showing good performance. Keywords: expert finding; information retrieval; machine learning; deep learning; recommender systems; academia; authorship attribution 1. Introduction In both our personal and professional lives, there are moments when our knowledge or ability to solve a problem falls short, necessitating us to connect with individuals whose experience or expertise can provide valuable assistance for the issue at hand. If a tool were available for such situations, it could assist us in locating these individuals, who may be represented within a system through various means, such as the services they offer (in structured or unstructured formats), CVs, blog entries, tags, and more. These matches would typically be identified based on our expressed needs, either through a natural language query or a simple list of keywords. This software would be responsible for determining the most suitable experts who can fulfill our requirements. This problem is called expert finding and in general refers to the process of identifying individuals who possess specialized knowledge or expertise in a particular subject or field. The goal is to locate individuals who can provide valuable insight, guidance, or assistance in relation to a specific topic. Expert finding

FAQs

What problem does the paper address?

The paper focuses on academic expert finding: locating researchers with domain expertise when a user requests help on a topic.

How do IR-based methods approach expert finding?

IR-based methods construct rich textual profiles for each expert by clustering information drawn from the expert’s scientific publications.

How do the results compare IR and ML methods on the biomedical datasets?

Across PMSC-UGR and CORD-19, IR techniques are generally more robust and better suited than ML-based approaches, though some exceptions show strong performance.

Aug 5, 2026

Upload by Emma Wilson