eContext is a text classification technology and general taxonomy which allows you to:
- Pinpoint what’s being searched online to identify deeper insights and make better business decisions
- Isolate the newsworthy topics that readers are engaging with to make better recommendations
- See which conversations are trending on a given story to predict a user’s future actions by understanding their past and present browsing and research habits
How the eContext API Works
We’ve created a general taxonomy of the commercial and social web. This consistent framework groups conversations, behavior and digital interactions into topical hierarchies up to 25 tiers, with clear connections between those topics. Push large amounts of text data from multiple sources through eContext to create a single data set for measurement and analytics to get meaningful insights in less time.
Feed eContext Any Text Data
As you feed text data through the eContext API, our custom pre-processing parses the text into linguistically appropriate units that eContext can work with efficiently
eContext Goes to Work
eContext checks each unit against positive and negative rules to determine the correct topics the content exhibits, processing high velocity data streams in real time, including the full Twitter firehose
The Result: Deep Insights
The result is a high-resolution picture of the interests, actions, or insights embedded in the data for direct analysis, or deeper learning or programmatic action
Bolster Machine Learning and AI with Deep Text Classification
eContext is the only high-performance, multi-class classification service capable of structuring any text data into hundreds of thousands of topical nodes with no training.
Machine Learning Alone
Maintain a litany of separate models to process a variety of data types and topics
Requires specialized knowledge and skill to implement
Maintenance is difficult as new data must be added constantly
Customization is difficult, requiring staff time to produce training sets and build new models each time new research is requested
ML Enriched by eContext
No modeling required. As a general taxonomy, eContext has 500,000 nodes so you can pivot as needed
Requires only a few lines of code
Rules-based eContext understands all topics of interest on the commercial and social web. API is refreshed daily
eContext’s architecture allows you to have fine-grained control over enhanced customization with the ability to tailor classifications by vertical, client, or channel without building a new model each time