Using Feedback Tags and Sentiment Analysis to Generate Sharable Learning Resources Investigating Automated Sentiment Analysis of Feedback Tags in a Programming Course

Abstract

This paper demonstrates how sentiment analysis can be used to identify differences in how students and staff perceive the opinions contained in feedback for programming work. The feedback considered in this paper is conceptually different in that it is given in the form of tags that when associated with a fragment of source code can be considered as a sharable learning resource. The research presented investigates the differences in perception of whether feedback is positive, negative or neutral according to students and examiners. This paper also investigates the adequacy of an automated sentiment analysis engine with a view that sentiment information when combined with the feedback tag and source code may create a more informative sharable learning resource. This paper describes the investigatory technique and presents the initial results. Results indicate that there are important differences between the sentiment of feedback perceived by students and examiners. This paper highlights the benefit of including sentiment data along with feedback.