Show simple item record

dc.contributor.advisorBalog, Krisztian
dc.contributor.advisorKostric, Ivica
dc.contributor.authorHenriksen, Rune
dc.contributor.authorMosand, Sindre
dc.date.accessioned2022-07-23T15:51:20Z
dc.date.available2022-07-23T15:51:20Z
dc.date.issued2022
dc.identifierno.uis:inspera:93568650:50622953
dc.identifier.urihttps://hdl.handle.net/11250/3007938
dc.description.abstract
dc.description.abstractThis paper is concerned with creating conversations around the ECIR conference with multiple modalities. We use a Furhat robot from Furhat robotics to have a physical presence and handle voice. Rasa addresses the AI component of the system. First, we develop a framework for managing messaging between Rasa AI with the Furhat frontend and a display frontend. Second, we create a protocol that describes how to format messages for this connection. The results of the thesis discuss the conversations from the ECIR conference. We evaluate the log from the conference and do both statistical analysis and a discussion of the qualitative results. The thesis also describes the general principles of conversational AI in the background chapter.
dc.languageeng
dc.publisheruis
dc.titleA conversational framework for multiple modalities
dc.typeBachelor thesis


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record