Doctoral thesis

A finder and representation system for knowledge carriers based on granular computing

    2017

1 ressource en ligne (163 p.)

Thèse de doctorat: Université de Fribourg, 2017

English In one of his publications Aristotle states ”All human beings by their nature desire to know” [Kraut 1991]. This desire is initiated the day we are born and accompanies us for the rest of our life. While at a young age our parents serve as one of the principle sources for knowledge, this changes over the course of time. Technological advances and particularly the introduction of the Internet, have given us new possibilities to share and access knowledge from almost anywhere at any given time. Being able to access and share large collections of written down knowledge is only one part of the equation. Just as important is the internalization of it, which in many cases can prove to be difficult to accomplish. Hence, being able to request assistance from someone who holds the necessary knowledge is of great importance, as it can positively stimulate the internalization procedure. However, digitalization does not only provide a larger pool of knowledge sources to choose from but also more people that can be potentially activated, in a bid to receive personalized assistance with a given problem statement or question. While this is beneficial, it imposes the issue that it is hard to keep track of who knows what. For this task so-called Expert Finder Systems have been introduced, which are designed to identify and suggest the most suited candidates to provide assistance. Throughout this Ph.D. thesis a novel type of Expert Finder System will be introduced that is capable of capturing the knowledge users within a community hold, from explicit and implicit data sources. This is accomplished with the use of granular computing, natural language processing and a set of metrics that have been introduced to measure and compare the suitability of candidates. Furthermore, are the knowledge requirements of a problem statement or question being assessed, in order to ensure that only the most suited candidates are being recommended to provide assistance.
Faculty
Faculté des sciences économiques et sociales
Language
  • English
Classification
Economics
Notes
  • Ressource en ligne consultée le 31.05.2017
License
License undefined
Identifiers
Persistent URL
https://folia.unifr.ch/unifr/documents/305388
Statistics

Document views: 37 File downloads:
  • DenzlerA.pdf: 2