The goal of the DIESEL project is to develop a novel scalable approach to enable search through distributed enterprise data. Based on the natural language input DIESEL will understand the semantics of the user input, e.g. a question, and convert the search into a federated search over the large enterprise data. The result of the federated search queries will be converted into a comprehensive answer. The aim of the research project is to enhance the efficiency, reduce costs and improve the decision making process through a holistic answer based on the large enterprise data that has been queried.
On February 29th, we finished the first milestone including an initial feasibility study, technical architecture and a thorough requirements specification pertaining to our use cases. In particular, we want to cover three domains:
- Querying Enriched Encyclopedic Knowledge
- Search over Wikidata
- Medium-Large Enterprise Search and Knowledge Graph