High-Performance Text Mining

SAS® High-Performance Text Mining

Quickly identify topics and contexts with in-memory text analytics

Analyze millions of social media posts to discover what topic is hot. Enrich customer segmentation with unstructured information. Distill important insights from large, diverse content sources. Big data technologies redefine the possibilities.

Benefits

Reduce time to decisions with fast, automated processes.

Using machine learning and natural language processing techniques, time-consuming activities that were previously done manually (such as theme identification, tagging, or building topic libraries and document indexes) are automatically generated and executed. High-performance capabilities mean even large collections can be quickly evaluated. Get comprehensive answers and insights faster than before.

Combine unstructured and structured data with advanced modeling techniques.

Apply sophisticated analytics against all of your data – not just subsets or aggregates – and you can improve accuracy for more targeted, high-impact decisions. And by using the best modeling techniques along with more model iterations, you can answer even your most difficult questions. Combining structured data with text data uncovers previously undetected relationships and improves decision making.

Improve predictive accuracy by including large-scale text documents.

Readily and automatically examining very large data sets – even billions of documents – can help you obtain more reliable results. With distributed, parallel processing, you can shrink analytical processing time. Analyzing more data faster can potentially improve modeling processes for more accurate predictive power.

Test more ideas and scenarios to optimize model performance.

Processing that used to take 30 minutes can be reduced to less than a minute in a muliticore computing environment. Reduced run time means you can build more models and get results faster. Then, easily retrain your models using different parameters to quickly optimize model performance.

Screenshots

Features

High-Performance Text Mining
  • Natural language processing
  • Text processing options
  • Text filtering
  • Topic generation
  • Graphs and tabular output
  • Available for Greenplum, Teradata and Oracle Exadata appliances, as well as on commodity hardware using Apache Hadoop or Cloudera

Recommended Resources

Back to Top