TextTrainer™
Accurately Feed Analysis Tools With Less Time and Effort
Many powerful analytic tools are available today including data mining software, link analysis tools and geographic information systems. Unfortunately, the data input to these systems must be relational (tabular or columnar), which prevents these tools from being applied directly to unstructured textual data such as reports, messages and news articles. To date, either an expensive and time-consuming effort to pull out the relational data from the text was undertaken, or the tools simply could not be used for the data in the text.
Accurate Cost-Effective Extraction of Relational Data
TextTrainer can help your organization extract relational data from unstructured text sources. TextTrainer is a trainable system that learns to pull the desired relational data from the target documents. Several government-run evaluations have shown that extraction systems using TextTrainer are as accurate as handcrafted systems produced by expert computational linguists.
As a result, TextTrainer makes relational extraction far more cost-effective without sacrificing accuracy.
The TextTrainer Advantage
- Expert computational linguists are not required
- All the work can be done by your existing staff
- Extractors can be up and running in a few hours
- No formal definitions of the desired results are required
- Analysts—not technologists—drive the extraction
- Analysts determine the level of accuracy and confidence in the extracted results
To learn more, please contact us at Intel-Info@ngc.com.
Additional Resources