A Knowledge-based System for Data Analysis and Interpretation

Abstract:

Traditionally, statistical packages are employed to derive or infer facts about a Universe of Discourse through data analysis and interpretation. It is analysis that serves to transform data into information. Statistical packages provide the users with relatively easy-to-use and powerful mechanics of data analysis, but up to now they do not provide much help with the design and strategies of the analysis. As such, there is a risk of misuse of these packages by statistically inexperienced users. We propose the use of knowledge-based interfaces to support this category of users in statistical evaluations. This paper discusses our experiences from the implementation of a knowledge-based system called MAXITAB. It provides guidance in the processes of data analysis and interpretation and has been programmed as an interface to the statistical package MINITAB.

Key-Words:

Data Analysis - Knowledge-based Interfaces - Multivariate Analysis

• REFERENCES

• 1 Hook R. Getting people to use Statistics properly. Amer Statist 1980; 34: 39-42.
• 2 Darius PL. Expert Systems and Statistics in Processing Statistical Information. In: Proceedings of the SEAS Spring Meeting in Belgium, April 1984; Vol II: 529-38.
• 3 Fellegi IP. Data, Statistics, Information: Some Issues of the Canadian Social Statistics Scene. J Official Statistics 1988; 4: 53-65.
• 4 Chowdhury SI. Statistical Expert Systems – A Special Application Area for Knowledge-based Computer Methodology. Linköping Studies in Science and Technology, Thesis No 104. Department of Computer and Information Science, University of Linköping, Sweden. 1987
• 5 Chowdhury SI, Kallinikos G. Graphical Techniques for modelling activities – a comparison between ¡SAC (Information Systems and Analysis of Changes) and SADT (Structured Analysis and Design Techniques). PK: Report, [in Swedish], Department of Computer and Information Science, University of Stockholm, Sweden. 1981
• 6 Ross DT. Structured Analysis (SA): A Language for Communicating Ideas. IEEE Trans Softw Eng SE-3. 1977: 16-34.
• 7 Ross DT, Schoman jr. KE. Structured Analysis for Requirements Definition. IEEE Trans Softw Eng SE-3. 1977: 6-15.
• 8 Rudkin RI, Shere KD. Structured Decomposition Diagram: A new technique for System Analysis. Datamation, Oct. 1979: 130-49.
• 9 Tukey JW. Comment. Amer Statist. 1985; 39: 12-6.
• 10 Hahn GJ. More Intelligent Statistical Software and Statistical Expert System: Future directions. Amer Statist 1985; 39: 1-16.
• 11 Chowdhury SI. State of the Art in Statistical Expert System. In: Proceedings 7th International Workshop on Expert Systems and their Applications, Avignon 87, May 1987; Vol 2: 1095-1113.
• 12 Hand DJ. The role of statistics in psychiatry. Psychological Medicine 1985; 15: 471-76.
• 13 Hand DJ. Choice of Statistical Technique. Institute of Psychiatry, De Crespigny Park, England. 1985
• 14 Hand DJ. Statistical Expert Systems: Necessary Attributes. J Appl Statist 1985; 12: 20-7.
• 15 Sundgren B. How to satisfy a Statistical Agency’s need for general survey processing program. Amsterdam: Proceedings of the 1 SI Conference; 1985
• 16 Chambers JM, Pregibon D, Zayas E. Expert software for data analysis: an initial experiment. 43rd Session of ISI, I. P. 18.2: Buenos Aires, Argentina. 1981
• 17 Barr A, Feigenbaum EA. eds The Handbook of AI Vol. I. Los Altos, California: William Kaufman Inc; 1982
• 18 Kosko B. Letters to the editor. Al Magazine, Fall. 1984
• 19 Smith AMR, Lee LS, Hand DJ. Interactive user-friendly interfaces to statistical software. The Computer Journal 1983; 26: 199-204.
• 20 Gale WA, Pregibon D. An Expert System for Regression Analysis. In: Proceedings 14th Symposium on the Interface of Computer Science and Statistics . New York: Springer Verlag; 1982
• 21 Gale GA, Pregibon D. Student Phase 1 -A Report on Work in Progress: Artificial Intelligence and Statistics . Reading Mass: Addison-Wesley; 1986
• 22 Gale WA. REX review: Artificial Intelligence and Statistics . Reading Mass: Addison-Wesley; 1986
• 23 Darius PL. Building Expert Systems with the Help of Existing Statistical Software: An Example. In: Proceedings COMP-STAT 1986. Heidelberg: Physica Verlag; 1986: 277-82.
• 24 Blum RL. Discovery and Representation of Causal Relationships from a Large Time-oriented Clinical Data Base: The Rx project . Lecture Notes in Medical Informatics. New York: Springer Verlag; 1982
• 25 Ryan TA, Joiner BA, Ryan BE. MINITAB Reference Manual – MINITAB Project . Pennsylvania State University, University Park, Pa. 1981
• 26 Bengtsson P, Kärki I, Lindgren M, Molin E, Wallgren J. MAXITAB-Users Guide (in Swedish). Internal document. Dept. of Computer and information Science, University of Linköping, Sweden. 1986
• 27 Davis R. Consultation, knowledge acquisition and instruction: a case study. In Szolovits P. ed Artificial Intelligence in Medicine . Boulder, Colorado: Westview Press; 1982: 57-78.
• 28 Hand DJ. Patterns in Statistical Strategy, Artificial Intelligence and Statistics . Reading/Mass: Addison-Wesley; 1986
• 29 Wallgren A, Wallgren B. Undersökningar och Statistik, (in Swedish). Stockholm: Natur och Kultur; 1983
• 30 Goldberg M, Doyon B. SARI: A user-oriented data bank system for medical applications. Meth Inform Med 1976; 15: 69-74.
• 31 Chambers JM. Some thoughts on expert software. In: Proceedings 13th Symposium on the Interface of Computer Science and Statistics . New York: Springer Verlag; 1981: 36-40.
• 32 Hellevik O. Introduction to Causal Analysis . London: George Allen & Unwin; 1984
• 33 Kenny D. Correlation and Causality . West Sussex, England: John Wiley and Co; 1979
• 34 Suppes PA. Probabilistic Theory of Causality . Amsterdam: North Holland Publishing Co; 1970
• 35 Ida Annual Research Report 1987, ISSN-0281-4250. Dept. of Computer and Information Science, Linköping University, Sweden, Spring. 1988: 39-54.
• 36 IntelliCorp: KEE Technical Manuals. 1985
• 37 Epitec:. Epitool development environment reference manual. Epitec AB, Linköping, Sweden. 1986