What is DiscoverText?
DiscoverText provides a collaborative platform for text analytics that blends human judgment with machine-learning capabilities. Users can collect, clean, and analyze unstructured text data from sources like surveys, social media, customer feedback, and public comments. The software includes tools for deduplication, clustering, and interactive machine classifier histograms to accelerate the training of custom classifiers.
With features like document redaction, near-duplicate detection, and patented CoderRank for annotator ranking, DiscoverText supports rigorous research and e-discovery workflows. It is designed for teams and offers free access to students and professors, with project support from the founder.
Features
- Human-Machine Collaboration: Combines human annotation with machine learning for iterative classification.
- Custom Machine Classifiers: Build reusable 'sifters' to sort items by relevance, topic, or sentiment.
- Document Redaction: Remove sensitive information and produce Bates-stamped PDF collections.
- Deduplication and Clustering: Automatically group near-duplicates to understand data landscape.
- Interactive Classifier Histograms: Identify high-value items for human coding to accelerate training.
- CoderRank: Patented method for ranking human annotators to ensure accurate results.
Use Cases
- Analyzing open-ended survey responses
- Monitoring social media (e.g., Twitter) for trends
- Processing public comments to government agencies
- Conducting academic text analysis research
- Managing customer feedback from CRMs and chats
- eDiscovery and legal document review
- HR sentiment analysis from employee surveys
- Market research text classification
FAQs
-
Is DiscoverText free for students?
Yes, students and professors get free access, training, and project support directly from the founder.
