Inproceedings658: Unterschied zwischen den Versionen
K (Added from ontology) |
K (Wikipedia python library) |
||
(3 dazwischenliegende Versionen desselben Benutzers werden nicht angezeigt) | |||
Zeile 1: | Zeile 1: | ||
− | {{Publikation | + | {{Publikation Erster Autor |
− | | | + | |ErsterAutorNachname=Stojanovic |
− | | | + | |ErsterAutorVorname=Nenad |
}} | }} | ||
{{Publikation Author | {{Publikation Author | ||
Zeile 19: | Zeile 19: | ||
|Abstract=Although several approaches have been proposed for modelling an on-line shop assistant, recent customer's analyses show that they miss some assistance in the buying process. The common problem is that the behaviour of an on-line shop assistant is modelled on the procedural level, i.e. like a workflow. In this paper we present an approach that models this behaviour on the knowledge level, i.e. it takes into account not only which actions (questions) a shop assistant will perform, but also which goals he wants to achieve by taking an action. As a generic reasoning pattern of such an e-shop agent we use the cover-and-differentiate problem-solving method, a method very successfully applied in various diagnosis and classification tasks. In that way, we can (i) model the question-answering process such that the minimal set of useful questions will be provided to a user, (ii) easily reinterpret and fine-tune shopping strategies that exist in other e-shop portals and (iii) design and integrate new methods into generic reasoning pattern. We present an evaluation study which illustrates these benefits. | |Abstract=Although several approaches have been proposed for modelling an on-line shop assistant, recent customer's analyses show that they miss some assistance in the buying process. The common problem is that the behaviour of an on-line shop assistant is modelled on the procedural level, i.e. like a workflow. In this paper we present an approach that models this behaviour on the knowledge level, i.e. it takes into account not only which actions (questions) a shop assistant will perform, but also which goals he wants to achieve by taking an action. As a generic reasoning pattern of such an e-shop agent we use the cover-and-differentiate problem-solving method, a method very successfully applied in various diagnosis and classification tasks. In that way, we can (i) model the question-answering process such that the minimal set of useful questions will be provided to a user, (ii) easily reinterpret and fine-tune shopping strategies that exist in other e-shop portals and (iii) design and integrate new methods into generic reasoning pattern. We present an evaluation study which illustrates these benefits. | ||
|VG Wort-Seiten= | |VG Wort-Seiten= | ||
− | |||
|Projekt=SemIPort, | |Projekt=SemIPort, | ||
− | |Forschungsgruppe= | + | |Forschungsgruppe=Wissensmanagement |
+ | }} | ||
+ | {{Forschungsgebiet Auswahl | ||
+ | |Forschungsgebiet=Ontology Engineering | ||
+ | }} | ||
+ | {{Forschungsgebiet Auswahl | ||
+ | |Forschungsgebiet=Semantic Web | ||
+ | }} | ||
+ | {{Forschungsgebiet Auswahl | ||
+ | |Forschungsgebiet=Electronic Commerce | ||
}} | }} |
Aktuelle Version vom 16. Oktober 2009, 17:38 Uhr
On the Knowledge Level of an On-line Shop Assistant
On the Knowledge Level of an On-line Shop Assistant
Published: 2004
Oktober
Buchtitel: EKAW 2004
Seiten: 110-125
Verlag: Springer
Referierte Veröffentlichung
BibTeX
Kurzfassung
Although several approaches have been proposed for modelling an on-line shop assistant, recent customer's analyses show that they miss some assistance in the buying process. The common problem is that the behaviour of an on-line shop assistant is modelled on the procedural level, i.e. like a workflow. In this paper we present an approach that models this behaviour on the knowledge level, i.e. it takes into account not only which actions (questions) a shop assistant will perform, but also which goals he wants to achieve by taking an action. As a generic reasoning pattern of such an e-shop agent we use the cover-and-differentiate problem-solving method, a method very successfully applied in various diagnosis and classification tasks. In that way, we can (i) model the question-answering process such that the minimal set of useful questions will be provided to a user, (ii) easily reinterpret and fine-tune shopping strategies that exist in other e-shop portals and (iii) design and integrate new methods into generic reasoning pattern. We present an evaluation study which illustrates these benefits.
Electronic Commerce, Ontology Engineering, Semantic Web