ML Anthology

Empirical Knowledge Representation Generation Using N-Gram Clustering

Robin Collier
AAAI 1994 pp. 1434
/aaai/1994/collier1994aaai-empirical/

Abstract

The work discussed below enables the automatic genera-tion of structures similar to the key templates which are predefined in information extraction/retrieval conferences- this would be a significant development. The motivation is similar to that of AutoSlog (Riloff 1993) which generates a domain-specific dictionary of concepts, although the approach is quite different. System Overview The approach acquires a domain-specific semantic repre-sentation by carrying out stochastic analysis of a corpus. Sets of conceptually similar paragraphs are utilised. The corpus and semantic representation are used to generate schematic structures. These are used to concisely store the knowledge contained within existing texts. New texts are processed to dynamically update the knowledge base. Any novel concepts encountered are

PDF AAAI Semantic Scholar

Cite

Text

Collier. "Empirical Knowledge Representation Generation Using N-Gram Clustering." AAAI Conference on Artificial Intelligence, 1994.

Markdown

[Collier. "Empirical Knowledge Representation Generation Using N-Gram Clustering." AAAI Conference on Artificial Intelligence, 1994.](https://mlanthology.org/aaai/1994/collier1994aaai-empirical/)

BibTeX

@inproceedings{collier1994aaai-empirical,
  title     = {{Empirical Knowledge Representation Generation Using N-Gram Clustering}},
  author    = {Collier, Robin},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {1994},
  pages     = {1434},
  url       = {https://mlanthology.org/aaai/1994/collier1994aaai-empirical/}
}