Information Processing in Electronic Markets: Measuring Subjective Interpretation Using Sentiment Analysis

Abstract

Information availability plays an important role in the efficient resource allocation of electronic markets and e-commerce. Most of this information is of qualitative nature containing essential facts that are, however, difficult to decode. Currently, the information processing capabilities of human agents facing such qualitative news is mostly unknown. Accordingly, it is crucial to understand how different decision makers process qualitative information. In this paper we show that sentiment-analysis facilitates research in qualitative information processing. We use a capital market example to demonstrate how investors and analysts perceive novel information. We find that their interpretation is different from one another: investors rapidly translate novel information into transactions, whereas analysts take more time to respond. We further observe that analysts emphasize different parts of information than investors, and are less put-off by complex information. The approach can be applied to other electronic-markets and the e-commerce industry where individuals react upon textual information.