ML Anthology

MOLE: A Knowledge Acquisition Tool That Uses Its Head

Larry J. Eshelman, John P. McDermott
AAAI 1986 pp. 950-955
/aaai/1986/eshelman1986aaai-mole/

Abstract

MOLE can help domain experts build a heuristic classification problem-solver by working with them to generate an initial knowledge base and then detect and rernedy deficiencies in it. By exploiting several heuristic assumptions about the world, MOLE is able to minimize the information it needs to elicit from the domain expert. In particular, by using static techniques of analysis, MOLE is able to infer support values and fill in gaps when a knowledge base is under-specified. And by using dynamic techniques of analysis, MOLE is able to interactively refine the knowledge base. 1. Int reduction MOLE assists domain experts in building expert systems that do heuristic classification [Clancey 84, Clancey 85, Buchanan 841. MOLE is useful in domains in which the expert can preenumerate

PDF AAAI Semantic Scholar

Cite

Text

Eshelman and McDermott. "MOLE: A Knowledge Acquisition Tool That Uses Its Head." AAAI Conference on Artificial Intelligence, 1986.

Markdown

[Eshelman and McDermott. "MOLE: A Knowledge Acquisition Tool That Uses Its Head." AAAI Conference on Artificial Intelligence, 1986.](https://mlanthology.org/aaai/1986/eshelman1986aaai-mole/)

BibTeX

@inproceedings{eshelman1986aaai-mole,
  title     = {{MOLE: A Knowledge Acquisition Tool That Uses Its Head}},
  author    = {Eshelman, Larry J. and McDermott, John P.},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {1986},
  pages     = {950-955},
  url       = {https://mlanthology.org/aaai/1986/eshelman1986aaai-mole/}
}