This Workshop aims at stimulating multi-disciplinary discussions about the challenges related to corpus creation and annotation for social skills behavior analysis. Contributions from computational, psychological and psychometrics perspectives, as well as applications including platforms to share corpora and annotations, are welcomed.

Social skills represent a fundamental resource in any professional and personal situation for conducting smooth interactions. Methods in Artificial Intelligence have become increasingly popular in order to automatically assess social skills – by analyzing multimodal behavior in several contexts such as public speaking, job interviews, group interactions, in populations with dysfunctions (e.g., autism spectrum disorder, schizophrenia), and involving both human-human and human-machine interactions. These methods could facilitate planning interventions targeted at improving these competencies, for example by giving appropriate feedback and personalized training.

The first step to investigate multi-modal cues of social skills, no matter the context, often consists in the analysis of corpora, by automatically extracting behavioral features and manually annotating more subjective and psychological constructs. Finding an existing corpus that could be exploited for one’s research interest could be a difficult step, either because of difficulties in obtaining access, or because they are not adapted for different research goals. Often, researchers prefer to create their own corpus. This results in a large amount of existing corpora that are often not accessible to other researchers and are not fully exploited. The main challenges related to corpus creation include the choice of the best setup and sensors, finding a trade-off between eliciting natural interactions, limiting invasiveness and collecting precise information. The second issue in this context regards the process of annotation. The choice of the type of annotators (experts vs. nonexperts), the type of annotations (automatic vs. manual, continue vs. discrete), the temporal segmentation (windowed vs. holistic) is crucial for a correct measure of the phenomenon of interest and getting significant results.

The topics of CATS2021 will have a strong impact on researchers and stakeholders across different disciplines, such as Computer Science, Social Signal Processing, Psychology, Statistics. Leveraging the opportunities offered by such a multidisciplinary environment, the participants could enrich their perspective, strengthen their practices and methodologies and draw together a research roadmap tackling the discussed challenges, which might be taken up in future collaborations.