Saga Natural Language Understanding (NLU) Framework
CHALLENGES WITH EXTRACTING INSIGHTS FROM UNSTRUCTURED, NATURAL LANGUAGE CONTENT
Nearly 80% of enterprise content is unstructured. It consists of fast-growing human-generated content, including memos, emails, text documents, research and legal reports, voice recordings, videos, social media posts, etc. Unlike structured content (tables, forms, log files), it is difficult to search for, let alone analyze, meaningful information from unstructured, natural language content.
As enterprises increasingly become insight-driven, they are seeking to leverage the vast unstructured data to improve business operations and accelerate speed to outcomes. But existing natural language processing / understanding technologies are not fulfilling enterprise demands – they are too narrow (chatbots), too shallow and generic (cloud-based natural language processing solutions), or too costly to develop, deploy, and maintain.
FILLING THE GAPS IN CURRENT NATURAL LANGUAGE PROCESSING SYSTEMS
As part of our collection of technology assets, Saga Natural Language Understanding (NLU) was developed to provide a scalable, easy-to-use framework that fills the gaps in existing NLP/NLU technologies. Saga delivers:
- Business-friendly, code-free user interfaces for creating and maintaining language models
- Reasonable, predictable cost
- Flexible, scalable deployments on-premises or in the cloud
- Automated NLU pipeline construction and management which substantially reduces development costs
- State-of-the-art handling of language ambiguity
- Integrated machine learning with many prepackaged models available out-of-the-box
- Ability to support a wide range of enterprise use cases
- Extensible with machine learning and knowledge graph technologies
Saga can be used as a standalone NLU framework or together with our other technology assets to optimize search and analytics performance. Learn more about our technology assets and reference architecture.
SAGA NATURAL LANGUAGE UNDERSTANDING BENEFITS
- NLU capabilities are powered by both Patterns Matching (for precision and ease of editing) and Machine Learning (for broad coverage and automatic learning).
- Patterns are simple to understand, fast, accurate, easy to get started, quick to show value, and work best when no training data is available. NLP output contains business object IDs which are easily integrated into business actions.
- The process of testing and deploying Machine Learning and language models is easily done and managed by non-data scientists as it does not require coding.
- Non-data scientists can perform 95% of the NLP/NLU work, providing prepared and “ready to go” data for data scientists to focus on creating better models where needed.
- Companies have full control over solution performance, deployment, and cost predictability.
- Multiple NLP development efforts, algorithms, and language resources are managed and coordinated entirely within the Saga framework.
EXAMPLE USE CASES
Saga NLU is flexible, easy-to-use, and scalable to support a variety of enterprise NLP/NLU needs. Some example use cases include:
- Intelligent enterprise search and question/answer systems
- Detecting insider threats
- Electronic communications monitoring for compliance and fraud
- Intelligent document analysis
- Sentiment / news analysis
- Data search and analysis for oil and gas (for both exploration and drilling operations)
- Recruiting search and match
- E-commerce semantic search
- Semantic search and question/answer for support portals
Contact us to request a Saga demo and learn more about how Saga can support your NLU use cases.