Auflistung nach Autor:in "Mehli, Carlo"

Gerade angezeigt 1 - 4 von 4

A genetic algorithm for optimizing parameters for ant colony optimization solving capacitated vehicle routing problems
(ACM, 2020) Faust, Oliver; Mehli, Carlo; Hanne, Thomas; Dornberger, Rolf
04B - Beitrag Konferenzschrift
Decision support combining machine learning, knowledge representation and case-based reasoning
(Sun SITE, Informatik V, RWTH Aachen, 2021) Mehli, Carlo; Hinkelmann, Knut; Jüngling, Stephan
Knowledge and knowledge work are essential for the success of companies nowadays. Decisions are based on knowledge and better knowledge leads to more informed decisions. Therefore, the management of knowledge and support of decision making has increasingly become a source of competitive advantage for organizations. The current research uses a design science research approach (DSR) with the aim to improve the decision making of a knowledge intensive process such as the student admission process, which is done manually until now. In the awareness phase of the DSR process, the case study research method is applied to analyze the decision making and the knowledge that is needed to derive the decisions. Based on the analysis of the application scenario, suitable methods to support decision making were identified. The resulting system design is based on a combination of Case-Based Reasoning (CBR) and Machine Learning (ML). The proposed system design and prototype has been validated using triangulation evaluation, to assess the impact of the proposed system on the application scenario. The evaluation revealed that the additional hints from CBR and ML can assist the deans of the study program to improve the knowledge management and increase the quality, transparency and consistency of the decision-making process in the student application process. Furthermore, the proposed approach fosters the exchange of knowledge among the different process participants involved and codifies previously tacit knowledge to some extent and provides relevant externalized knowledge to decision makers at the required moment. The designed prototype showcases how ML and CBR methodologies can be combined to support decision making in knowledge intensive processes and finally concludes with potential recommendations for future research.
04B - Beitrag Konferenzschrift
Re-Engineering a Knowledge-Intensive Decision Process to support consistent decision making
(Hochschule für Wirtschaft FHNW, 2020) Mehli, Carlo; Hinkelmann, Knut
Knowledge and knowledge work are essential for the success of companies nowadays. Since decisions are based on knowledge and better knowledge leads to more informed decisions, the management of knowledge and support of decision making has increasingly become a source of competitive advantage for organizations. While previous research proposes many different approaches such as Ontologies, Knowledge Management, Artificial Neural Networks, Case Based Reasoning etc. to support various decision making, there is not one single best solution outlined by research that applies for all kind of knowledge intensive processes and its decisions. Therefore, there are no specific decision support methods outlined by research which fit best for specific knowledge or decision patterns. The aim of the research was to determine appropriate methods to improve decision making; identify decisions and analyze knowledge used for those decisions in a specific process scenario; propose a system design to improve knowledge management in said scenario; evaluate the impact of the proposed system design on the knowledge intensive process. This thesis contains a profound literature framework and follows the design science research (DSR) strategy. In the awareness phase of the DSRprocess, the application scenario of student admission process was acquired using the case study research method....
11 - Studentische Arbeit
Virtual bartender: a dialog system combining data-driven and knowledge-based recommendation
(2019) Hinkelmann, Knut; Blaser, Monika; Faust, Oliver; Horst, Alexander; Mehli, Carlo; Martin, Andreas; Hinkelmann, Knut; Gerber, Aurona; Lenat, Doug; van Harmelen, Frank; Clark, Peter
This research is about combination of data-driven and knowledge-based recommendations The research is made in an application scenario for whisky recommendation, where a guest chats with a recommender system. Preferences about taste are difficult to express and the knowledge about taste is tacit and thus can hardly be represented and used adequately. People or not aware of how to describe flavors in a standardized way and how to do a justified choice. This is because knowledge about taste is mainly tacit knowledge. To deal with this knowledge, data-driven recommendation is adequate. On the other hand, in particular experienced customers use knowledge about distilleries, locations and the distillery process to express their preferences and want to have arguments for the recommended products. This shows that a combination of data-driven and knowledge-based recommendations is appropriate in areas where tacit knowledge and explicit knowledge are available.
04B - Beitrag Konferenzschrift