Network-enabled devices and the consecutive surge of internet users created a significant impact on the knowledge society. In the era of Social Media, the final user is the new reporter that embraces these tools to spread novelty and breaking news about specific subjects. Sometimes, this type of reporting is biased due to the level of the user's knowledge of the matter or to intentionally achieve a goal. The ability to share other users' post magnify this phenomenon and create a domino effect that can lead to the diffusion of biased information. Therefore, the reliability of the shared novelty about an event is becoming fundamental more than ever. The proposed method tries to deal with this drawback by cross-relating text streams with corresponding heterogeneous levels of reliability, for instance, Twitter and Google News. As Text Stream Mining technique, the method adopts the Fuzzy Formal Concept Analysis to incrementally carry out a fuzzy lattice for each text stream. These two fuzzy lattices are compared in time and context-aware manner to derive the trustworthiness of the relations among entities that are mentioned together in the same tweet content. Preliminary experimental results made on a real dataset show that the proposed credibility assessment system provides good performance depending on some parameters, like similarity threshold T (empirically fixed at 0.8) and time window, set at 1 hour.

Cross-relating heterogeneous Text Streams for Credibility Assessment

De Maio C.;Fenza G.;Gallo M.;Loia V.;Volpe A.
2020-01-01

Abstract

Network-enabled devices and the consecutive surge of internet users created a significant impact on the knowledge society. In the era of Social Media, the final user is the new reporter that embraces these tools to spread novelty and breaking news about specific subjects. Sometimes, this type of reporting is biased due to the level of the user's knowledge of the matter or to intentionally achieve a goal. The ability to share other users' post magnify this phenomenon and create a domino effect that can lead to the diffusion of biased information. Therefore, the reliability of the shared novelty about an event is becoming fundamental more than ever. The proposed method tries to deal with this drawback by cross-relating text streams with corresponding heterogeneous levels of reliability, for instance, Twitter and Google News. As Text Stream Mining technique, the method adopts the Fuzzy Formal Concept Analysis to incrementally carry out a fuzzy lattice for each text stream. These two fuzzy lattices are compared in time and context-aware manner to derive the trustworthiness of the relations among entities that are mentioned together in the same tweet content. Preliminary experimental results made on a real dataset show that the proposed credibility assessment system provides good performance depending on some parameters, like similarity threshold T (empirically fixed at 0.8) and time window, set at 1 hour.
2020
978-1-7281-4384-2
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11386/4750709
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 5
  • ???jsp.display-item.citation.isi??? 4
social impact