Affective analysis

Abstract

The aim of this work was focused on data collection and annotation for Aspect-Based Academic Review Analysis. The goal is to understand the sentiments expressed in course feedback regarding different aspects of a course. The feedback is annotated with aspect terms, category, sentiment, and opinion terms, and the overall summary for positive and negative sentiments, creating a rich dataset for analysis. While collecting data, a name-removal tool was used to ensure anonymity. For data annotation, a personalized annotation tool was used that incorporates a detailed category taxonomy tree. This taxonomy organizes aspects into broad categories like structure, evaluation, course material, lecture delivery, and general. Each category is further divided into finer categories, ensuring coverage of all possible categories for aspects mentioned in the feedback. This report includes the data collected, annotation process, and the insights collected from the data using graphs, to highlight the most reviewed categories in both positive and negative sentiments. A comparison between mid-semester and end-semester feedback is also presented using graphs from the data.