← Back to Publications

Interpretable Natural Language Segmentation Based on Link Grammar

Authors:

Year: 2020
Venue: SingularityNET / Aigents (2020)
Links: paper (PDF)

Summary

Presents a sentence-segmentation method that relies solely on Link Grammar's grammatical relationships between tokens rather than punctuation, enabling segmentation of unpunctuated text from speech-to-text engines or crawled web pages. The interpretable, grammar-driven approach generalizes across languages and feeds the semantic-query and natural-language-generation components of a question-answering pipeline (Aigents).

Relevance to Hyperon

A Link-Grammar-based, interpretable-NLP source in the legacy/interpretable pipeline lineage (Link Grammar → segmentation → QA) of the Semantic Parsing Deep Dive.

Key References

• RawData source
• Semantic Parsing Deep Dive