Applying Cognitive Technologies and Emulating Interpersonal Communication in Affective Tutoring Systems

Nielsen, Lars Myrhøj

4. term

Innovative Communication Technologies and Entrepreneurship, Master

2017

93

Denne afhandling undersøger, hvordan kognitive og affektive computing-teknologier kan anvendes til at efterligne interpersonel kommunikation med det formål at styrke Affective Tutoring Systems (ATS). Udgangspunktet er, at Intelligent Tutoring Systems (ITS) forbedrer læringsudbyttet, men stadig er mindre effektive end små- og en-til-en-tutoring, især fordi menneskelige undervisere bedre indregner elevens følelsesmæssige tilstand. Projektet adresserer derfor, hvordan emulering af den interpersonelle kommunikationsproces kan øge effektiviteten af digitale tutorsystemer. Metodisk bygger arbejdet på teori om interpersonel kommunikation og menneskelige emotioner, en gennemgang af kognitive og affektive teknologier, udvikling af en konceptuel ATS-model samt et forslag til implementering, suppleret af indledende og opfølgende interviews som beskrevet i rapportens struktur. Resultaterne viser, at interpersonel kommunikation kan emuleres til en vis grad i et ATS, idet moderne kognitive tjenester kan behandle komplekse multimodale signaler på linje med menneskelig signalbehandling, men de samlede evner er endnu ikke på niveau med en menneskelig hjerne. Afhandlingen peger på både potentialer og begrænsninger samt behov for videre arbejde.

This thesis examines how cognitive and affective computing technologies can be used to emulate interpersonal communication in order to strengthen Affective Tutoring Systems (ATS). It starts from the observation that Intelligent Tutoring Systems (ITS) improve learning outcomes but remain less effective than small-group or one-to-one tutoring, largely because human tutors better account for a learner’s emotional state. The project therefore asks how emulating the interpersonal communication process can enhance the effectiveness of digital tutoring. Methodologically, the work synthesizes theories of interpersonal communication and human emotion, reviews state-of-the-art cognitive and affective technologies, develops a conceptual ATS model and an implementation proposal, and includes preliminary and follow-up interviews as outlined in the report. Findings indicate that interpersonal communication can be emulated to a certain extent in an ATS: contemporary cognitive services can process complex multimodal signals in ways comparable to human signal processing, yet their overall capabilities still fall short of the human brain. The thesis discusses both the promise and the limitations and points to future work.

[This summary has been generated with the help of AI directly from the project (PDF)]

cognitive computing ; learning ; affective computing ; interpersonal communication ; learning technology ; icte

Download
View record in AAU Student Projects