Alessandro Arneodo
Expert-Finding System for Matching Academic Researchers to Funding Calls.
Rel. Daniele Apiletti, Simone Monaco. Politecnico di Torino, Corso di laurea magistrale in Data Science And Engineering, 2026
|
Preview |
PDF (Tesi_di_laurea)
- Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives. Download (811kB) | Preview |
Abstract
Finding the academic researchers that are best suited to answer a research funding call is a frequent and time-consuming task in large technical universities. This thesis presents an expert-finding system that receives the free text of a call and returns a ranked list of researchers of Politecnico di Torino, ordered by how relevant their scientific production is to the topic of the call. The system follows the document- centric formulation of expert finding: it first retrieves the most relevant publications with a hybrid pipeline that combines lexical retrieval (BM25) and dense retrieval (BGE-M3) through reciprocal rank fusion, with an optional cross-encoder re-ranking step; then it aggregates the relevance of the retrieved documents at the author level, using configurable weights for authorship credit and publication recency.
The corpus contains more than 60,000 unique publications from the last fifteen years, extracted from the institutional repository and enriched with abstracts from OpenAlex, IRIS, and Scopus
Relatori
Anno Accademico
Tipo di pubblicazione
Numero di pagine
Corso di laurea
Classe di laurea
URI
![]() |
Modifica (riservato agli operatori) |
