Physics Problem Solving: ISAAC-II
Proceedings of the Seventh International Joint Conference on Artificial Intelligence. Program Demonstrations. Pages 1063--1.
© 1981 International Joint Conferences on Artificial Intelligence
Abstract
- Novak, G. and Araya, A., "Physics Problen Solving Using Multiple Views", Technical Report TR-173, Conputer Science Dept., University of Texas at Austin (78712). .
- Novak, G., "GIRL and GLISP: An Efficient Representation Language", Technical Report TR-172, Co:nputer Science Dept., University of Texas at Austin.
- Novak, G. and Araya, A., "'Research on Expert Problen Solving in Physics", Proc. 1st AAAI Conference, Stanford, CA, August; 1980;-pp. -f78-180~
- Novak, G., ''Representations of Knowledge in a Progran for Solving Phys.tcs Problens", Proc. Sth IJCAI, Canbridge, MA, 1977, pp.--286-29f. •-- 5.
- Novak, G., "Conputer Understanding of Physics Problens Stated in Natural Language", Anerican Journal of Con~'O.<!ti~~l!..~ Lingu~~~c~. Microfiche s3; 1976.
- n£ L.En END c:F A IEA1 IS Sl.f'f'ffiTED BY A Sf'Rir«; WITH A cat3T~ c:F 50 LM"T > C n£ Sf'Rir«; I'R<ES A'i ~E c:F 70 OCGREES WITH n£ ~IzaiTA.. , WHILE n£ 13ER1 I'R<ES A'i ~ c:F 20 DEGREES WITH n£ ~IZCHTA..)(A BOY loEIGilr«; 50 lB STIHIS AT n£ RIGil END c:F THE 13ER1 , A'iD A loEIGil c:F 200 LB IS ATTIO£D 2 n FRQ-1 n£ LEFT END> c n£ IEA1 IS SlFF'ORTED BY A RCf'E WHICH I'R<ES A'i ~ c:F 120 OCGREES WITH n£ ~IZCHTA.. > (THE ~ IS ATTIO£D 3 F"T F"Ra'l n£ RIGil END> CIDI l'l£H IS THE SPRir«; STRETCHED , 1M> WHAT IS THE L[r«;TH c:F THE BER1> ~: 3.263518 n . 11.21Hi8 n Figure 1: Exanple Problen and Solution
References were extracted automatically from the PDF and may contain errors.
Full text extracted automatically from the PDF; figures, tables and equations may differ. The PDF is the version of record.
PHYSICS PROBLEM SOLVING: ISAAC-II
Gordon S. Novak Jr.
Conputer Science Departnent University of Texas at Austin Austin, Texas 78712
ISAAC-II, a progran which solves physics problens stated in English, will be denonstrated in real tine. A notebook of problens which have been solved by the progran will be displayed; the denonstratlon wtll include problens taken fran this set. In addition, new problens in elenentary physics (stated ent.trely .tn Engltsh, Le., not requ.tring a d.tagran as part of the problen statenent) contributed by conference attendees will be accepted, and w.tll be attenpted by the progran.
This progran, a successor to our earlier ISAAC progran, first parses the sentences of the problen statenent to build an internal nodel of the objects in the problen and their relationships; this nodel is stated in terns of real-world objects (cars, people, etc.). Using rules which exanine qualitative and quantitative features of objects and relationships, the progran chooses v.tews of objects (or groups of objects) as phys.tcal systens (e.g., "rig.td body" or "uniforn acceleration fran rest"). This set of nodels gives rise to a set of equations which can be solved to find the answer to the problen. The internal nodel of the objects is also used to generate a picture of the problen; an exanple of such a picture (generated by ISAAC) is shown in F.tgure 1.
The najor dHference between this progran and ISAAC lies in the choice of problen solving nethods. ISAAC could solve only rigid body statics problens (which linited the range of solution nethods needed); it also· wrote a large nunber of equations which were solved by brute force. ISAAC-ll has a larger repertoire of problen solving net hods (even for the "sane" type of problen); this enables it to select the best nethod for a particular problen. The resulting solutions are less co:nplex algebraically, and are closer to the solut.tons produced by hunan experts.
Physics problen solving is an .tnportant research area because the techniques used for solving physics problens, when fully developed, can aid teaching of physics and can be used for intelligent conputer-aided engineering analysis and design.
This research was supported by NSF award No. SED-7912803 in the Joint National Institute of Education - National Science Foundation Progran of Research on Cognitive Processes and the Structure of Knowledge in Science and Mathenatics.
@inproceedings{ijcai1981p210,
title = {Physics Problem Solving: ISAAC-II},
author = {Novak, Gordon S.},
booktitle = {Proceedings of the Seventh International Joint Conference on
Artificial Intelligence, {IJCAI-81}},
publisher = {International Joint Conferences on Artificial Intelligence},
editor = {Patrick J. Hayes},
pages = {1063---1},
year = {1981},
month = {8},
note = {Program Demonstrations},
url = {https://ijcai.org/proceedings/1981/210},
}
TY - CPAPER TI - Physics Problem Solving: ISAAC-II AU - Novak, Gordon S. T2 - Proceedings of the Seventh International Joint Conference on Artificial Intelligence, IJCAI-81 PB - International Joint Conferences on Artificial Intelligence PY - 1981 SP - 1063 EP - -1 UR - https://ijcai.org/proceedings/1981/210 ER -Copied
Machine-readable versions of this paper, for tools and AI agents:
Welcome
Future Conferences
AI Journal Division
IJCAI Proceedings
About IJCAI
IJCAI Awards