Telefon: +43(0) 662 834 602 - 0

Follow RSA

Universitäre Partner

DI Florian Kleedorfer

Operational Studio Management

Florian Kleedorfer studied computer science at the Vienna University of Technology. He works at the Studio Smart Agent Technologies in the position of key researcher. His research interests include machine learning, text/web mining, information extraction, semantic Web and linked data.

He was a leading member of the project teams for Clouds of Music, SEMPRE. He led or participated in several R&D projects with,,, Smart Information Systems, and others.

Link to all his Publications


Abstracts of his recent publications

E-marketplaces on the World Wide Web are information and transaction silos, which in the general case don't allow transactions across their boundaries. The consequence is that the Web, often termed the global marketplace, is fragmented along the dimensions of geography, content domain, supply or demand, user base, and many more. This fragmentation makes it inefficient to buy and sell commodities on the Web. We propose a framework that serves as a foundation for a distributed, de-centralized e-marketplace on top of the Web, making boundaries between existing systems disappear from the user's perspective. The framework standardizes the creation and description of objects that represent supply and demand. In addition to this, it allows for independent matching services to connect objects suitable for a transaction and it defines protocols for the message exchange between such objects.
The Web as related to commerce suffers from a fundamen-tal asymmetry. While there is a great number of commercialoffers available, consumer needs are rarely represented ex-plicitly. Thus, the most widely applied process of connectingthe prospective consumer of a resource with its supplier isWeb search. In Web search, the user needs are implicit,driving the interaction, and therefore only the interactionpartners can try to deduce them. We present an approachfor a) publishing needs on the Web of Data and b) buildinga protocol that allows decentralized matching of needs andcommunication between need owners. Albeit inspired by theanalysis of marketplaces, the proposed framework allows fora much broader range of social applications, such as collabo-rative problem solving, help organizing the sharing economyor finding interesting people to meet.
Structurally, the Web resembles a marketplace: Web clients download documents from servers just like customers buy goods from kiosks. Interestingly, the marketplace metaphor for the Web goes further: most Web users are in the role of the customer and the documents downloaded in most cases actually represent commercial offers. This second resemblance leads to a fundamental asymmetry: in general, commercial offers are represented as first-class objects on the Web. Consumer needs, however, the very notions that drive the interaction, only materialize in the form of search terms or browsing behaviour. In this paper, we present an approach for explicitly representing needs as first-class objects on the Web of data in a way that allows interacting with them.
Telefon: +43(0) 662 834 602 - 0