CUbiC Student and Interdisciplinary Team Runners Up at AffCon 2019

CUbiC Student and Interdisciplinary Team

Michael Saxon, a CUbiC Master's student, and teammates Samarth Bhandari, Lewis Ruskin, and Gabrielle Honda of the Luminosity Lab, created a runner-up system in the CL-Aff Shared Task at the Second Workshop on Affective Content Analysis at AAAI 2019. The task challenged teams to create "happy moment" analysis systems that could, given a sentence describing a happy moment that a speaker had experienced recently, identify whether the moment was one in which the speaker had control (agency label) and whether the event involved interacting with other people (social label). The team's system, the Word Pair Convolutional Model, leveraged a preliminary analysis of common trends in the data in its design to deliver the second best performance of all 47 system runs submitted. Michael presented the system design to workshop attendees along with the winning and other runner up system.