Powered by OpenAIRE graph
Found an issue? Give us feedback
addClaim

Similaridade entre perfis sociais

Authors: Mendes, Rui Daniel Alves;

Similaridade entre perfis sociais

Abstract

Over the last decade, Internet usage has gone viral and become extremely important, positioning itself as an integral part of our lives and social interactions. Social networks are now one of the richest sources of information regarding user profiles. By dealing with social network data, an area that as recently seen growing interest, one can create methods to study profile similarity, using available user data. A possible tool to correctly identify similar users may be applied in a multitude of areas and contribute to the decision making process that may draw gains to people’s lives, either in financial perspective or life quality (such as health). In this dissertation, metrics that are applied when determining profile similarity were researched and discussed, giving a complete overview on the concepts involved and difficulties experienced. Moreover, it is possible to create profile clusters as well as correlate interests. As such, a performance analysis of different clustering algorithms is done, namely K-Means, Hieratical Clustering, DBSCAN and BIRCH. Techniques used in recommendations systems are also discussed. Finally, future work is proposed where this project would serve as the basis of a recommendation and profile analysis systems.

Na última década, a utilização da Internet tornou-se viral e de extrema importância posicionando-se, atualmente, numa parte integral das nossas vidas incluindo a parte social. As redes sociais online são uma das fontes mais ricas de informação sobre os perfis de utilizadores. Ao lidar com dados de redes sociais, a similaridade entre perfis representa uma área que tem tido algum destaque. Uma ferramenta capaz de identificar corretamente utilizadores semelhantes pode ser utilizada em diversas áreas e contribuir para tomadas de decisão importantes que poderão resultar em proveitos, sejam de cariz financeiro ou melhoria da qualidade de vida de pessoas (por exemplo, saúde). Nesta dissertação são realizados estudos com diferentes métricas de distância de forma a determinar a similaridade entre perfis. É possível, também, criar agrupamentos de perfis assim como correlacionar interesses. Posteriormente, é feita uma análise de performance entre diversos algoritmos de clustering, nomeadamente o K-Means, Clustering Hierárquico, DBSCAN e BIRCH. As medidas de similaridade foram também utilizadas para estimar valores associados aos interesses dos utilizadores, numa abordagem inspirada nos sistemas de recomendação.

Mestrado em Sistemas de Informação

Country
Portugal
Related Organizations
Keywords

Análise de dados, Redes sociais, Sistemas de informação, Sistemas de recomendação, Perfil social, Exploração de dados, Similaridade, Clustering, Interesses

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Upload OA version
Are you the author of this publication? Upload your Open Access version to Zenodo!
It’s fast and easy, just two clicks!