Introduction

What is DATS?
The Discourse Analysis Tool Suite (DATS) is a machine-learning-powered web application designed specifically for multi-modal discourse analysis.
Modern discourse appears across diverse channels, including text, image, audio, and video. Furthermore, with the ongoing digitalization of public life, there has been unprecedented growth in the amount of data available online. The vast amount of available, potentially relevant data often exceeds the capacity of conventional manual discourse analysis methods.
DATS was developed to address these challenges. It provides a unified, highly scalable digital environment that supports the typical workflow of a discourse analysis project—from data collection and management to exploration, annotation, qualitative and quantitative analysis, and final interpretation.
By integrating advanced Artificial Intelligence (AI) and Natural Language Processing (NLP) technologies with intuitive, user-friendly interfaces, DATS empowers digital humanities researchers to process massive datasets without requiring extensive computational expertise.
Key Features
DATS offers a comprehensive suite of features to support every stage of your research:
- Multi-modal Support: Analyze diverse data formats, including 📝 text, 🖼 image, 🎵 audio, and 🎞 video documents, all within a single platform.
- Automated Pre-processing: DATS automatically transcribes audio/video, recognizes objects in images, and extracts keywords and named entities from text, converting all your multimodal data into a unified, searchable format.
- Advanced Data Management: Organize your corpus with customizable hierarchical Tag systems and flexible Metadata fields.
- Powerful Exploration: Go beyond simple keyword searches. DATS offers cross-modal semantic search, allowing you to find conceptually similar documents, images, or sentences, even if they don't share the exact same words.
- Flexible Annotation: Create and manage your own hierarchical code taxonomy to annotate text spans, sentences, and image bounding boxes.
- AI Assistance: Leverage state-of-the-art Large Language Models (LLMs) and custom-trained classifiers to automate and accelerate the annotation and tagging process.
- Dynamic Analysis: Utilize Word Frequency, Code Frequency, and the advanced Concept-over-Time Analysis (COTA) to uncover temporal patterns and discursive shifts in your data.
- Visual Interpretation: Use interactive Whiteboards to visually map out relationships between documents, annotations, and concepts, facilitating theory building and interpretation.
- Seamless Reflection: Capture your interpretive thoughts at the point of discovery using the omnipresent Memo system and maintain a comprehensive project Logbook.
Target Audience and Use Cases
DATS is designed primarily for researchers and scholars in the digital humanities and social sciences who employ qualitative, quantitative, or mixed-methods approaches to discourse analysis.
It is particularly well-suited for projects that require:
- Handling "Big Data": Analyzing large-scale corpora (e.g., thousands of news articles or extensive social media archives) that are too vast for traditional manual coding tools.
- Analyzing Multimodal Content: Researching topics where the visual or auditory components of the discourse (images, videos, speeches) are as important as the text.
- Collaborative Research: Working in teams where multiple researchers need to code, analyze, and discuss the same dataset simultaneously in a shared, cloud-based environment.
- Privacy and Data Sovereignty: DATS is built for local or institutional server deployment, ensuring that sensitive research data remains entirely within your controlled environment, making it ideal for privacy-sensitive academic research.
Typical Use Cases:
- News Media Analysis: Tracing how specific topics (e.g., climate change, public health crises) are framed and evolve over time across different news outlets.
- Social Media Discourse: Analyzing the narrative structures, key actors, and public sentiment surrounding political events or social movements.
- Historical and Archival Research: Digitizing and analyzing large volumes of historical documents to identify shifting cultural concepts and terminology.