Szczegóły publikacji
Opis bibliograficzny
Combining knowledge graphs with semantic similarity metrics for sentiment analysis / Piotr Swędrak, Weronika T. ADRIAN, Krzysztof KLUZA // W: Knowledge Science, Engineering and Management : 15th international conference, KSEM 2022 : Singapore, August 6–8, 2022 : proceedings, Pt. 1 / eds. Gerard Memmi, [et al.]. — Cham : Springer Nature Switzerland AG, cop. 2022. — (Lecture Notes in Computer Science ; ISSN 0302-9743 ; LNCS 13368. Lecture Notes in Artificial Intelligence). — ISBN: 978-3-031-10982-9; e-ISBN: 978-3-031-10983-6. — S. 489–501. — Bibliogr., Abstr. — Publikacja dostępna online od: 2022-07-19
Autorzy (3)
Słowa kluczowe
Dane bibliometryczne
ID BaDAP | 143031 |
---|---|
Data dodania do BaDAP | 2022-10-12 |
DOI | 10.1007/978-3-031-10983-6_38 |
Rok publikacji | 2022 |
Typ publikacji | materiały konferencyjne (aut.) |
Otwarty dostęp | |
Wydawca | Springer |
Konferencja | 15th international conference on Knowledge Science, Engineering and Management |
Czasopismo/seria | Lecture Notes in Computer Science |
Abstract
The paper proposes a new semantic similarity method with an asymmetry coefficient. The motivation behind this idea is that in some cases it is justified to break the symmetry while comparing certain entities. Such semantic similarity in some cases might be desirable from the psychological point of view. It allows us to enrich embedding methods with knowledge graphs by adding additional information about the specificity of a concept. For the evaluation of the proposed solution, the method has been used as a component for determining a sentiment of reviews. The values of the asymmetry coefficient for the selected set of chosen pairs of words were computed and compared. We present the results of series of experiments comparing the accuracy of different methods in the context of sentiment analysis in various configurations. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.