Szczegóły publikacji
Opis bibliograficzny
Creative expert system: result of inference and machine learning integration / Bartłomiej ŚNIEŻYŃSKI, Grzegorz LEGIEŃ, Dorota WILK-KOŁODZIEJCZYK, Stanisława Kluska-Nawarecka, Edward NAWARECKI, Krzysztof Jaśkowiec // W: Database and Expert Systems Applications : 27th international conference, DEXA 2016 : Porto, Portugal, September 5–8, 2016 : proceedings, Pt. 1 / eds. Sven Hartmann, Hui Ma. — Switzerland : Springer International Publishing, cop. 2016. — (Lecture Notes in Computer Science ; ISSN 0302-9743 ; LNCS 9827). — ISBN: 978-3-319-44402-4; e-ISBN: 978-3-319-44403-1. — S. 257–271. — Bibliogr., Abstr. — D. Wilk-kołodziejczyk - dod. afiliacja: Foundry Research Institute in Krakow
Autorzy (6)
- AGHŚnieżyński Bartłomiej
- AGHLegień Grzegorz
- AGHWilk-Kołodziejczyk Dorota
- Kluska-Nawarecka Stanisława
- AGHNawarecki Edward
- Jaśkowiec Krzysztof
Słowa kluczowe
Dane bibliometryczne
ID BaDAP | 102009 |
---|---|
Data dodania do BaDAP | 2016-12-13 |
DOI | 10.1007/978-3-319-44403-1_16 |
Rok publikacji | 2016 |
Typ publikacji | materiały konferencyjne (aut.) |
Otwarty dostęp | |
Konferencja | 27th International Conference on Database and Expert Systems Applications |
Czasopismo/seria | Lecture Notes in Computer Science |
Abstract
This paper presents an idea of a creative expert system. It is based on inference and machine learning integration. Execution of learning algorithm is automatic because it is formalized as applying a complex inference rule. Firing such a rule generates intrinsically new knowledge: rules are learned from training data, which consists of facts stored already in the knowledge base. This new knowledge may be used in the same inference chain to derive a decision. Complex rules may also represent other procedural activities, like searching databases. Such a solution makes the reasoning process more creative and allows to continue reasoning in cases when the knowledge base does not have appropriate knowledge explicit encoded. In the paper appropriate model and inference algorithm are proposed. The idea is tested on a decision support system in a casting domain.