NELL: Never-Ending Language Learner Knowledge Base from Web Reading

Since January 2010, the Never-Ending Language Learner (NELL) has been autonomously reading the web to build a knowledge base. It has acquired over 80 million confidence-weighted beliefs, learned millions of features for reading, and can synthesize new relational predicates. The project is led by Tom M. Mitchell at Carnegie Mellon University.

Use Cases

• Training models for ontology learning based on the system's ability to extend its own relational predicates.
• Developing knowledge base population methods based on the web-scale belief extraction described.
• Researching continuous learning paradigms based on the never-ending learning framework.
• Analyzing confidence-weighted semantic relationships based on the 80+ million beliefs.

Strengths

• Contains over 80 million confidence-weighted beliefs, indicating substantial scale.
• Has been running and learning continuously since January 2010, suggesting long-term data collection.
• The system autonomously learns features and parameters for reading, indicating a degree of self-improvement.

Limitations

• Row count and specific data structure are unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

Carnegie Mellon University

Collection Method

Autonomous web reading and inference by the Never-Ending Language Learner (NELL) system.

Time Range

January 2010 onward