SoDa-TAP v2: Social Data Analysis Made Simple

Abstract

Social platforms are the mirror of our society’s values, beliefs, and activities and have become the subject of study of many disciplines; researchers often study the themes and sentiments of social-platform discussions and attempt to understand how the various aspects of these discussions correlate with people’s influence and the spread of ideas. This type of research requires substantial software engineering work. We developed SoDa-TAPv2 (Social Data - Toolkit Analysis Platform, Version 2) to automate many useful tools and to make them available to scholars of all disciplines. SoDa-TAPv2 integrates (i) a data ingestion and analysis pipeline, and (ii) a visual query language through which to review data and evaluate hypotheses. The pipeline enhances datasets with lexical analyses, sentiment analysis, humor detection, and identification of personal values and Big Five personality traits from text and images. The visual query language features an intuitive drag-and-drop interface that enables users to filter and slice datasets to create distinct sample sets and save them for future use, perform aggregations, categorize data into buckets through classification, clustering, and natural breaks, and compare these buckets using statistical analyses and visualizations.