During the last decade the World Wide Web has evolved into the prime platform for accessing and exchanging information and is often seen as a global information system. Whether searching for infor-mation using Web search engines, collecting or commenting on information or entities using Web 2.0 techniques, or performing everyday transactions like e-shopping, the Web has developed both considerable flexibility and a usually satisfying quality. Thus the Web is often not only seen as the most disruptive technology of recent years, but has also a strong societal impact. But for really benefitting from the vast variety of possibilities an intimate knowledge of how to interact with huge spaces of information and services is needed and especially untrained users still encounter severe problems. The reason is that interfaces to access the Web have largely stayed at the level of simple keyword searches and to finally retrieve the necessary information, usually several turns of query reformulations and refinements have to be taken. The aim of the International Outgoing Fellowship ‘Analogy Queries by Ontology-based Data Analytics’ (ANAQONDA) is to research, develop, and evaluate methods for efficient access to both information and services on the World Wide Web using the intuitive concept of analogies directly derived from human cognition and communication practices.
Analogies are often considered the core concept of human cognition processes and form intuitive tools for aiding efficient communication. In brief, analogies form typical patterns to state or explain in-formation needs and even complex concepts. Moreover, analogies are valuable for transporting concepts also including connotations between different domains and thus are a prime candidate for trans- or interdisciplinary information exchanges. Hence the outcome of this project will not only provide essential insights into large-scale information system development in general, but also for tailored applications like shared scientific working spaces or industrial strength collaboration infrastructures. This endeavour opens up a number of scientific challenges: How can information systems understand what a user provided analogy actually means? Based on domain-specific data and ontological knowledge, how are analogies discovered? What discriminates given items, what are commonalities on a perceptual level? How do the context and the domain affect inference processes?
Furthermore, this project explores related problems like research into crowdsourcing, perceived similarity, and intelligent queries in general.
This project is partially funded by DAAD FIT.