Options
2022
Conference Paper
Titel
CISQA: Corporate smart insights question answering system
Abstract
The development of semantic information systems is a highly topical issue in various domains, e.g., financial news, which both ordinary users and experts deal with. Until now, queries in natural language on open vocabulary from, e.g., query.wikidata.org were hard to answer. The development of such software offers the possibility to make semantic information systems more effective, efficient, and user-friendly. This paper proposes a semantic knowledge-based question answering system called cisqa. It develops entity linking and relation linking approaches that have been experimentally proven to be superior to the state of the art. cisqa also handles question ambiguity, translates natural language questions into SPARQL-queries, and delivers answers to the user in an appropriate manner.
Author(s)