Using InterSystems UIMA

What problem does it solve?

This Skill addresses the challenge of extracting meaningful information and annotations from unstructured text data by leveraging the Universal Natural Language Processing Architecture (UIMA) framework. It enables the integration of various text analysis technologies to process and store annotations efficiently.

Core Features & Use Cases

• UIMA Integration: Facilitates the use of InterSystems IRIS Natural Language Processing (NLP) and third-party UIMA-compliant analysis engines.
• Annotation Generation: Creates annotations (e.g., named entities, concepts, relations) for source text, storing them separately from the original data.
• Pipeline Architecture: Supports the creation of analysis pipelines where multiple UIMA engines process data sequentially.
• Annotation Store: Provides a persistent, SQL-based storage for annotations, allowing for easier querying and analysis compared to traditional XML formats.
• Use Case: Analyze customer feedback from support tickets to automatically identify product issues, sentiment, and key topics, storing these insights in a structured format for reporting and product improvement.

Quick Start

Configure a UIMA functional index on the 'FullArticle' column of the 'NYT.Articles' table to process text data using the 'iKnowEngine.xml' descriptor.