DonNTU | Masters' portal of DonNTU | Artem Kushnir
Master of DonNTU Artem Kushnir
Artem Kushnir
Faculty of Computer Science and Technologies
Software Engineering
Theme of master's work: Implementation of associative search and
inference in knowledgebases by means of SQL-queries

Scientific supervisor: Docent, Ph. D. Alexander Andruckin

Abstract *

Introduction

A human often strives to make life better. There are many answers to the question why this is happening, and they are not the subject of this article. The human knows his needs and uses tools to satisfy them. The technical development of mankind is the development of tools. Research activities are aimed to opening new ways of meeting the needs.

The volume of knowledge increases with the development of mankind. "Knowledge is fixed and practically tested information that can be reused by people for solving various problems."[1] In ancient times a scientist could have a big part of all knowledge and use it in the process of invention. Now the possibilities of the development, memorization and knowledge usage have increased, but this increase is very small in comparison with the growth of knowledge level.

One person or even a group of people cannot possess or use the entire volume of human knowledge. At the same time the more knowledge is used in the inventive activity the more productive it is. For example, an engineer usually knows about 200 different physical and technical effects but may apply no more than 100. More than 3000 effects are described in the technical literature [2]. Outlook and the possibilities of an engineer during the invention process increase together with the number of physical and technical effects he can use.

Increasing life expectancy, raising mental abilities and speed of learning will improve the inventive ability of man. However, these processes are not largely controlled by human beings. Another solution is building of an artificial intelligence system, which, while having a considerable amount of knowledge, could synthesize ideas, physical principles of operation and technical solutions that will help to meet people need. This paper is dedicated to the principles of such a system and its development.

Subject of development and research

The methods of the theory of inventive problem solving (TRIZ) for generation of new technical solutions are researched within this work. The subject of development is a system that provides associative search and logical inference in the knowledgebase and generates options for possible physical principles of new devices.

Actuality of the theme

As shown by amount of conferences held in America and Europe at the present time there are numerous researches and developments on various aspects and modifications of TRIZ in the world. Experience of the practical use of theory in various fields of human activity is accumulated. Sometimes it goes beyond the engineering challenges. Industrial companies and research centers are turning to the organization of systematic training of specialists on TRIZ in their own businesses and are looking for systemic ways to implement TRIZ technology. Higher education in Europe and Asia began to develop new specializations to the title of Bachelor and Master of Design Innovation. Automation of inventive activity is an important task. The demand for knowledge-intensive software products is increasing. Based on this, we can speak about the actuality of the theme.

Purpose and tasks of development and research

The aim is to develop and implement the method for analysis and solving of inventive problems. It should allow getting a lot of alternative technical solutions, each of which meets the original inventive situation. The method is based on the associative search and logical inference in the knowledgebase. Operations with data are carried out using SQL-queries. The objectives are:
  • Implementation of the system that fills a knowledgebase and extracts knowledge from text.
  • Storage of knowledge in a real database, using existing software to implement the inference and search.
  • Construction of the mechanism of inference and associative search with uaage of SQL-queries.

Prospective scientific novelty of development and research

The novelty of the research and development is the implementation of the new analytical and solving tool of TRIZ as well as the method of filling and updating the knowledgebase used by the tool. In particular, the author finds the following specific proposals as new ones:
  • Usage of SQL language for dealing with the knowledgebase. This makes the system mobile and independent of the DBMS.
  • Dynamic update of the knowledgebase during the search of technical solutions.

Prospective practical value of the received results

The practical significance of the results is the implementation of the method of analysis and solving of inventive problems, which can be successfully used in industrial practice and in training.

Overview of current results

Currently the main approaches to solving engineering problems have been studied and the scope of TRIZ has been identified. The structure of the knowledgebase that holds the fund of physical and technical effects has been developed. The subsystem that fills the knowledgebase and extracts knowledge from text has been implemented.

Conclusion

Our world is changing rapidly. The new generations will live in completely different conditions, where new values will appear and new skills will be required.

We need to learn to live in the world we know nothing about, to solve problems that no one had ever encountered before. Artificial intelligence techniques and TRIZ are designed to help people perceive the world differently, to discard the usual limitations of thinking, to see new solutions. Automation of these techniques and their usage will bring many benefits and help humanity move forward.

References

  1. Данные — информация — знания // TreKo.RU — портал TРEнеров и КОнсультантов. [Электронный ресурс] / Режим доступа: http://msk.treko.ru/show_dict_336
  2. Половинкин А.И. Основы инженерного творчества: Учеб. пособие для студентов втузов. — М.: Машиностроение, 1988. — 368 с.
  3. Черный А.А. Принципы инженерного творчества: Учеб. пособие. — Пенза: Изд-во Пенз. гос. ун-та, 2005. — 43 с.: библиогр. 14 назв.
  4. Кудрявцев А.В. Методы логического поиска // Metodolog.ru — сайт изобретательских задач и методов их решения. [Электронный ресурс] / Кудрявцев А.В. — Режим доступа: http://www.metodolog.ru/00447/00447.html
  5. Чутко И. Красные самолеты. — М.: Сов. Россия, 1982.
  6. ТРИЗ — технология творчества // Сайт Тризисного Центра. [Электронный ресурс] / Режим доступа: http://www.trizland.ru/trizba.php?id=3
  7. Creativity & Innovation Changing the World // Сайт Creatividad e Innovación. [Электронный ресурс] / Режим доступа: http://homepage.mac.com/penagoscorzo/creativity/triz/index.html
  8. Истоки ТРИЗ // Сайт компании Трисолвер Украина. [Электронный ресурс] / Режим доступа: http://www.trisolver.com.ua/methoden/erfindungstheorie_triz.htm
  9. Хоменко Н. Теория решения изобретательских задач // Сайт центра ОТСМ-ТРИЗ технологий. [Электронный ресурс] / Хоменко Н. — Режим доступа: http://www.trizminsk.org/intr.htm
  10. Справка о проекте Джонатан Ливингстон // Сайт центра ОТСМ-ТРИЗ технологий. [Электронный ресурс] / Режим доступа: http://www.trizminsk.org/e/jl.htm
  11. История компании // Сайт компании Трисолвер Украина. [Электронный ресурс] / Режим доступа: http://www.trisolver.com.ua/news/news_1.htm
  12. ТРІЗ, Товариство Реалізації Інженерних Задач // Портал Бизнес-гид. [Электронный ресурс] / Режим доступа: http://triz.business-guide.com.ua
  13. Развитие и распространение ТРИЗ // Сайт Тризисного Центра. [Электронный ресурс] / Режим доступа: http://www.trizland.ru/trizba.php?id=4
  14. Филатов В.А., Щербак С.С., Хайрова А.А. Разработка высокоэффективных средств создания и обработки онтологических баз знаний // Блог PhD Щербака Сергея. [Электронный ресурс] / Филатов В.А., Щербак С.С., Хайрова А.А. — Режим доступа: http://shcherbak.net/razrabotka-vysokoeffektivnyx-sredstv-sozdaniya-i-obrabotki-ontologicheskix-baz-znanij
* — The work is in development for the current day and is planned to be finished by December 1st, 2010. The paper and relevant materials can be obtained either form author or its supervisor after that date.

ДонНТУ | Портал магистров ДонНТУ