Methods of Artificial Intelligence have long since taken root in our daily lives. Search engines are learning to optimise their search results from the search behaviour of thousands of users. We receive personalised recommendations not only for films, books and ads, but also for scholarly articles, and of course in social media.
A good recommendation requires that
- Contents are indexed and analysed semantically by a computer system and
- pertinent contents are provided to users at the right time.
Contents must be understood to assist users in their search for information.
This requires a semantic representation, for instance as knowledge graphs, topic models or contextual vectors. Therefore the research group addresses all aspects necessary for an intelligent processing and provision of information:
- Extraction of information from large document collections,
- Semantic representation of data,
- Analysis and processing in applications,
- Intelligent visualisation of the results.
- Text Mining
- Information Retrieval
- Recommender Systems
- Natural Language Processing
- Information Extraction
- Machine Learning
- Deep Neural Networks
- Topic Modeling
- Knowledge Graphs
Natural language is difficult to understand for algorithms, but that is exactly what is necessary for the processing of large data volumes. Art galleries, libraries, archives and museums (GLAM) have large volumes of text data whose automated analysis can offer many insights. One important step is the annotation of text data. Annotations are needed so that methods of Artificial Intelligence can be applied, such as deep neuronal nets. So-called Human-in-the-loop-approaches are used to include human expertise in the processing of natural language into the AI process, so that humans can be better integrated into the annotation process for these data.
In this project we collaborate with the working group of Professor Demartini at University of Queensland, Australia, with support from German Academic Exchange Service (DAAD).
The Connect & Collect (CoCo) project coordinates the network of regional competence centres for labour research. The research and development activities aim at a "cloud of labour research" that supports interdisciplinary cooperation, promotes technological and social innovations, and provides structures for sustainable knowledge transfer.
To support the competence centres, the CoCo project is developing an infrastructure with AI-supported tools and innovative methods that will enable a new approach to interdisciplinary labour research. A "cloud of labour research" will bring together the competence centres and other stakeholders, and provide opportunities for cooperation and networking. An important component is a data and knowledge repository that enables the joint use of research data, knowledge, and resources. In addition, AI-based tools will support linked-up working and the transfer of results to industry.
The project started in cooperation with three Fraunhofer Institutes, DIE, and ZBW in March 2021 for four years. ZBW is the leading technology partner. Funding is provided by the Federal Ministry of Education and Research.
More about Connect & Collect