07.09.2021 | News IntraFind involved in new research project

Together with the project partners Technical University of Deggendorf and DATEV, the enterprise search specialist IntraFind is investigating the topic of "Natural Language Semantic Search in Big Data" in the SEMIARID (SEMantIc seARch big Data) project.

SEMIARID LogoThe project, which is funded by the Bavarian State Ministry of Economic Affairs, Regional Development and Energy with a good one million euros, will run for three years. The three project partners, Deggendorf Technical University (THD), DATEV and IntraFind, want to jointly research how internal corporate search engines can be improved by using AI techniques.

A problem of our society is no longer the lack of information, but the excess of information. Therefore, we need search engines to find the right answers to our questions. This is true for the World Wide Web as well as for internal company information. Specifically, the project will develop semantic search engines for employee access to internal content, search engines for support databases, and search engines for access to specialized information.

A company-internal search engine must be able to provide helpful answers to questions such as "What is the telephone number of Ms. Monika Schmidt?", "Who can help me with printer problems?", "In which products is component xyz installed?", "What is the temperature tolerance of component xyz?".

To answer questions correctly, the search engine must be able to interpret normal sentences correctly. In the project, the latest research results in artificial intelligence are to be used and further developed for this purpose. The central research area here is "question answering". This is a research area at the intersection of search engine technologies and natural language processing.

The methods used must be scalable to Big Data applications with several hundred thousand users and several hundred million documents. In addition to scalability, the comprehensibility (Explainable AI) of the search results and answers of the search system are a key project goal.