Connecting data based on implicit knowledge
We seek to build an ontology that can connect various data from the other research units.
Takuichi Nishimura
Co-investigator
Professor, Japan Advanced Institute of Science and Technology
Link to members
Our Project
We are using the therapy textbook to structure experts’ implicit knowledge and link them to multimodal data.
Psychotherapy is a complex treatment that involves highly advanced decision making depending on the patient’s state. We are writing down the methodologies and risks of such decision-making in concrete terms so that it can be used to semi-automate the improvement of the quality of therapist training. Specifically, we can link multimodal sensing data (such as sound, facial expression, bio-physiological data, motion capture) to the structured knowledge to help trainees to deepen their understanding of interactions in psychotherapy.
We are also seeking to construct a hierarchically organized ontology from analyzing the vocabulary used in the structured knowledge. This ontology can clarify the meaning of the structured knowledge and allow us to define and annotate the meanings of each aspect of a given multimodal data to associate it to another data modality. This annotated sensor database can then be used to create an automated recognition AI.