Linked Data Integration for Semantic Dialogue and Backend Access

A dialogue system for answering user questions in natural speech presents one of the main achievements of contemporary interaction-based AI technology. Modern dialogue frameworks function as middleware between the interface component and the backend where the answers to the user questions are stored in heterogeneous formats. We implemented an interface to linked data sources as part of a complex natural language understanding and semantic retrieval process, thereby integrating the querying and answering task into a common framework. The semantic backend system integrates multiple linked data sources to allow for an advanced multimodal question answering (QA) dialogue.