---
title: "Operationalizing Heuristics: Some AI Methods for Assisting AI Programming"
authors:
  - "Jack Mostow"
  - "Frederick Hayes-Roth"
doi: ""
url: "https://ijcai.org/proceedings/1979/133"
pdf: "https://ijcai.org/proceedings/1979/0133.pdf"
venue: "Proceedings of the Sixth International Joint Conference on Artificial Intelligence"
year: 1979
pages: ""
track: ""
keywords:
---

# Operationalizing Heuristics: Some AI Methods for Assisting AI Programming

Jack Mostow, Frederick Hayes-Roth

*Proceedings of the Sixth International Joint Conference on Artificial Intelligence*.

## Abstract

Machlne"'-aided heuristic programming is a paradigm for incorporating domain knowledJe In lntelltaent tftk performance programs. In this paradigm, a system Interactively assimilates a natural lfniU'Ie descrtptlon of 1 teekt advice on how to perform it, and definitions of the domain concepts In terma of which the adVIce Ia ewpre...d. The system translates this input into an internal representation, operationallzes the assimilated knowlectae to Make It effective, Integrates different pieces of advice, and applies them to the performance of the task. A typed appllcatlve LISP-like language Is used for the Internal representation of domain knowledp. Operattonallntlon Ia defined In terms of transforming well-defined but non-effective expressions Into effectively ewecutabte Onet. Several techniques for performing this process mechanically are presented and eppllfd to an IXItnple ctre.wn frotn the domain of the card game Hearts. A system to operatlonallze Hearts advice Is currently belnt lmpltmenttd It en Instantiation of the advocated paradigm.