Expert finding aims at identifying knowledgeable people to help in decision processes, such as eliciting or analysing requirements in Requirements Engineering. Complementary approaches exist to tackle specific contexts like in forum-based communities, exploiting personal contributions, or in structured organisations like companies, where the social relationships between employees help to identify experts. In this paper, we propose an approach to tackle a hybrid context like an Open Source Software (OSS) community, which involves forums open to contributors, as well as companies providing OSS-related services. By representing and relating stakeholders, their roles, the topics discussed and the terms used, and by applying inference algorithms based on Markov networks, we are able to rank stakeholders by their inferred level of expertise in one topic or more. Two preliminary experiments are presented to illustrate the approach and to show its potential benefit.
Scheda prodotto non validato
Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte di FBK.
Titolo: | Expert Finding Using Markov Networks in Open Source Communities |
Autori: | |
Data di pubblicazione: | 2014 |
Abstract: | Expert finding aims at identifying knowledgeable people to help in decision processes, such as eliciting or analysing requirements in Requirements Engineering. Complementary approaches exist to tackle specific contexts like in forum-based communities, exploiting personal contributions, or in structured organisations like companies, where the social relationships between employees help to identify experts. In this paper, we propose an approach to tackle a hybrid context like an Open Source Software (OSS) community, which involves forums open to contributors, as well as companies providing OSS-related services. By representing and relating stakeholders, their roles, the topics discussed and the terms used, and by applying inference algorithms based on Markov networks, we are able to rank stakeholders by their inferred level of expertise in one topic or more. Two preliminary experiments are presented to illustrate the approach and to show its potential benefit. |
Handle: | http://hdl.handle.net/11582/232821 |
ISBN: | 9783319078809 9783319078816 |
Appare nelle tipologie: | 4.1 Contributo in Atti di convegno |