\documentstyle[a4]{article} \pagestyle{myheadings} \markboth{Author:kk}{Doc.No. AID/newsletter/4} \setlength{\textheight}{9in} \setlength{\topmargin}{0in} \setlength{\headheight}{.3in} \setlength{\headsep}{.5in} \parindent=0pt \parskip= .35\baselineskip plus .0833333\baselineskip minus .0833333\baselineskip \begin{document} Artificial Intelligence for Development\\ Document No: AID/newsletter/4\\ Last Modified and by whom: 29/8/91 kk \\ Distributed on: 29/8/91\\ \LARGE \begin{center} Artificial Intelligence for Development\\ Newsletter Number 4\\August 1991\\ \end{center} \normalsize \section*{Contents} \begin{enumerate} \item Knowledge Based Computer systems in India \item VITA news \item SatelLife \item Conferences \item Groups, bulletin boards and mailing lists \item Contact Personnel \end{enumerate} \section{Knowledge Based Computer Systems in India} {\ S. Ramani and K. S. R. Anjaneyulu\\} {National Centre for Software Technology\\ Gulmohar Cross Road No 9, Juhu, Bombay 400 049, India\\ Email: ramani@ncst.ernet.in, anji@ncst.ernet.in} India has a well-known tradition of study of thought, logic and language. There has been pioneering work by Indians in the areas of artificial intelligence and pattern recognition. In more recent years, the launching of the Knowledge Based Computer Systems Project by the Department of Electronics, Government of India, has contributed a lot to the development of national capability in this field. This Project, assisted by the United Nations Development Programme, has stimulated a lot of significant research in many areas. There are a number of other centres active in AI R \& D, in addition to the centres involved in this project. The paper presents a list of centres where major work is being carried out and the areas of their interest. A bibliography of over seventy five selected publications is included. \subsection{Early Work} India has a well-known tradition of study of thought, language and logic. Panini's elucidation of the structure of Sanskrit was one of the landmarks in the progress in this field. So, it is not surprising to see the tremendous attraction that the field of artificial intelligence has for thousands of Indians, within the country as well as without. An early paper by an Indian, which had a world-wide impact in the field, was on the use of syntactic methods in pattern recognition [Narasimhan, 1963]. The Computer Group at the Tata Institute of Fundamental Research has had a long tradition of work in the areas of natural language [Narasimhan, 1981] and speech recognition. There has been continuing work on pattern recognition and computer vision at the Indian Statistical Institute spanning a few decades. \subsection{Current Activities} There are a few areas in AI in which researchers in India have been working and have achieved some progress. These areas include: AI in Industry (Scheduling, Productivity Management, etc), Expert Systems, Intelligent Tutoring Systems, Machine Translation, Natural Language Understanding, Pattern Recognition, Search and Speech Recognition. In 1986, the Government of India initiated the Knowledge Based Computer Systems (KBCS) Project, with assistance from the United Nations Development Programme (UNDP). This five-year project is a major research initiative in the area of Artificial Intelligence. UNDP's support has been instrumental in enabling the visit of eminent AI workers from major laboratories around the world to visit Indian KBCS Centres. It has also enabled Indian KBCS workers, particularly those in the early stages of their careers to get trained in major laboratories abroad. The Ministry of Human Resource Development has started an Intelligent Systems Project with nodes all over India. Indian Institute of Management, Calcutta and the Indian Institute of Technology, Kharagpur have made significant contributions in the field of search. There is a project on Machine Translation at the Indian Institute of Technology, Kanpur. The Central University in Hyderabad offers a Masters Degree in AI and Robotics. The National Centre for Software Technology, Bombay offers a post-graduate diploma programme in KBCS. For a partial list of institutions active in AI in India, see Appendix A. \subsubsection*{Work at KBCS Project Centres} The Proceedings of KBCS '89 [Ramani et al, 1989] and KBCS-90 [Bhatkar and Rege, 1990] have reports from the coordinators of the KBCS Project Centres around the country. While these reports do not constitute a complete report of AI in India, they are good indicators of the current state of AI in the country. The Centres of the KBCS Project and their areas of work are listed below: \begin{itemize} \item Architecture for Parallel Computer Systems: Indian Institute of Science, Bangalore \item Speech Recognition and Neural Networks: Tata Institute of Fundamental Research, Bombay \item Pattern Recognition and Computer Vision: Indian Statistical Institute, Calcutta \item Expert Systems: Indian Institute of Technology, Madras \item Expert System Applications for Government: Department of Electronics, Government of India, New Delhi \item Natural Language, Educational Applications, and KBCS in Planning and Scheduling, Logic Programming: National Centre for Software Technology, Bombay \end{itemize} \subsection{Comments on Areas of Work} {\bf Computer Aided Design using KBCS Techniques} As there is a large, well established, engineering industry in India, AI applications are bound to flourish. However, no major development in this direction has so far taken place. {\bf Computer Vision:} With a long-standing tradition of research in this field, India is bound to make significant progress in this area. {\bf Educational Applications:} With the enormous educational challenge India is facing, this area is bound to gather momentum. Multi-media techniques and Intelligent Tutoring System (ITS) techniques will come together to create impressive instructional systems. New peripherals such as CD-ROM drives are likely to make these applications far more cost-effective than they are today. One problem to be faced, however, is that of cost. Progress in educational applications is likely to be slow because the users' ability to pay for these applications is very poor. {\bf Expert Systems:} Expert systems have attracted wide attention. Holding promise of white-collar robots, this area is bristling with activity. Since the current paradigm for creation of Expert Systems has serious limitations, there is room for significant R \& D relating to the technology itself. {\bf Natural Language and Speech Recognition:} To the extent language is central to intelligent behaviour, natural language understanding will occupy a central place in AI research. Having fourteen official languages, India is bound to be very interested in machine translation! Knowledge Based Machine Translation is likely to be a reality by the turn of the Century. {\bf Knowledge Based Planning and Scheduling:} This is a very practical area of considerable economic significance. A few projects have been completed and many more are being undertaken. This area is bound to grow and contribute to the general credibility of AI among those who place importance on practical applications. {\bf Psychology and Linguistics:} An important weakness of Indian AI is that there is not adequate research in the country in the related sciences of Psychology and Linguistics. This will have to be rectified soon. \subsection{Conclusion} There is a fast-growing R \& D activity in KBCS in India. There are many potential and relevant applications. There are a number of educational programmes at the post-graduate level. There have been a significant number of international publications. The KBCS Project of the Government of India has provided a major stimulus to the growth of R \& D in this field in India. \subsection*{Partial List of Centres in India Working in AI} A partial list of centres in India is given below, along with areas they are working on. {\bf Expert Systems:} Dept of Electronics, Delhi. Indian Institute of Technology, Madras. Jadavpur University, Calcutta.\\ {\bf Image Processing and Pattern Recognition:} Indian Institute of Technology, Bombay. Indian Institute of Technology, Madras. Indian Statistical Institute, Calcutta.\\ {\bf Intelligent Systems in Management and Manufacturing:} Indian Institute of Technology, Bombay. Indian Institutes of Management.\\ {\bf Intelligent Systems in Medicine:} Indian Institute of Technology, Bombay. Indian Institute of Technology, Madras.\\ {\bf Intelligent Tutoring Systems:} National Centre for Software Technology, Bombay.\\ {\bf Logic Programming:} Indian Institute of Technology, Bombay. Indian Institute of Technology, Kanpur. Institute of Mathematical Sciences, Madras. National Centre for Software Technology, Bombay. Tata Institute of Fundamental Research, Bombay. \\ {\bf Machine Translation:} Indian Institute of Technology, Kanpur. National Centre for Software Technology, Bombay. Tamil University, Thanjavur.\\ {\bf Modelling Language Behaviour:} Tata Institute of Fundamental Research, Bombay.\\ {\bf Natural Language Understanding:} Central University, Hyderabad. Indian Institute of Science, Bangalore. Indian Institute of Technology, Bombay. Indian Institute of Technology, Kanpur. National Centre for Software Technology, Bombay. Tata Institute of Fundamental Research, Bombay.\\ {\bf Parallel Processing:} Centre for Development of Advanced Computing, Pune . Indian Institute of Science, Bangalore.\\ {\bf Search:} Indian Institute of Management, Calcutta . Indian Institute of Technology, Kharagpur. \\ {\bf Understanding and Synthesising Speech:} Indian Institute of Technology, Madras. Tata Institute of Fundamental Research, Bombay.\\ \section{VITA news: DevelopNet News, July 1991} DevelopNet News is published by Volunteers in Technical Assistance (VITA) in Arlington, Virginia, USA. See below under `Groups, Bulletin Boards, and Mailing Lists' for how to subscribe to it. \subsection{AFRICA: BETTER DATA COMMUNICATIONS} by Gary Garriott Inexpensive software can now help to solve some of the telephone problems in Africa and other technologically remote areas. Making voice telephone calls from many African countries to Europe or North America is usually simpler than calling the country "next door." But it's even easier for people in North America or Europe to call Africa. In fact, Africa's problems of phone communications have become critical: they now seriously obstruct the flow of scientific information into the continent as well as the sharing of information among African scientists. Even in areas that have direct dialing, line quality suffers from "cross-talk" and other interference during busy periods. Africa's vast territory extends across many time zones, making it difficult to com- plete voice calls at convenient times. Finally, the costs of phone calls can be prohibitive, although calls initiated from the United States to Africa generally cost about one-third as much as similar calls from Africa. Inexpensive or free "mailer" software can help to send messages and data files. It runs on personal computers, even laptops, connected to voice- grade telephone lines. With it, computers in Africa can call each other satisfactorily, especially during the night when rates are cheaper and lower demand reduces interference. How can Africans take advantage of better international communications facilities? They can schedule calls from a "polling computer" in Europe or North America; it will pick up messages from, and relay them to, other computers in African countries, including those that are next to each other. Thus, one African country sends a message or data to another African country through Paris or London. This is more reliable than a direct-routed call if the foreign city initiates the calls both for receiving and retransmitting the information. Moreover, it is usually cheaper than any other form of data transmission and may save foreign exchange. VITA now uses mailers in this way to communicate with its projects in Pakistan and the Philippines. Of many available mailers, FrontDoor is highly versatile, but requires about 1 megabyte of disk space. It can schedule unattended message and file transfers. Other mailers (Fido, Seadog) are simpler to install, but much less convenient to use. Until regional and international communications improve, mailers and microcomputers can use the present telephone systems to meet African needs without resorting to prohibitively expensive public data networks. Product information: (Seadog) System Enhancements Associates, 21 New Street, Wayne, New Jersey 07470. Fido and FrontDoor, in noncommercial versions, are obtained from bulletin board system operators and are free of cost if used noncommercially. [The author is VITA's director of informatics.] \subsection{SWAZILAND: CHILD NUTRITION AND COMPUTER GRAPHICS} Better nutrition for young children is a critical issue in most countries. The governments of many Third-World countries try to improve public understanding of child nutrition through educational programs. The programs are aimed at parents, teachers, children, and the food industry. Swaziland's educational programs for child nutrition were almost unmanageably complex. Not only was a large variety of pamphlets, posters, flip-charts, and fliers needed, but ethnic and cultural variation from one region to another required different materials that could not be conveniently produced from a central location. According to Benedict Tisa (Westmont, New Jersey, USA), the problem was solved by setting up a large, computer-based "image bank" in the Ministry of Agriculture in the capital city of Mbabane, for use by workers in all areas. This was possible because the ministry already had Macintosh microcomputers. The Swaziland Project for the Promotion of Young Child Feeding trained an artist at the ministry to use graphics and scanning programs to cre- ate the image bank. Images from the bank can easily be made larger or smaller, or changed in detail to suit any regional or other special need. Tisa says, "The potentials of this system are just being recog- nized. . . . High quality, small editions of training materials can be produced easily and quickly for workshops and seminars. Designer mater- ials can be made by the field workers themselves since the use of the program is basically point and choose -- very simple to use." The total cost of equipment and software is between \verb+$+10,000 and \verb+$+12,000. Although the cost seems high, Tisa states it can be justified if the project is large and the equipment can be used also for word processing, data collection, and desktop publishing. In Swaziland, its use has already spread to another complex effort, the Weaning Project. Source: Benedict Tisa, CompuServe 71650,23 . \subsection{PC GLOBE} by Donald Ditter and others PC Globe displays maps of the world, regions, groups of countries (e.g., ASEAN, developing countries) and 190 individual countries. You can select country maps that show principal cities, elevations, or other features. Tables cover demographic, climatic, socioeconomic, cultural, health, educational, and tourist information. Bar charts (for which you select the information) compare countries. You can add cities to the 2,000 already in the database and update currency-exchange rates. Coun- try flags are shown and national anthems played. The manual is well written. The program is easy to install and use and occupies about 1.6 Mbytes on a hard disk. The displays are uncluttered and easy to read; they can be sent to your printer or a disk file, or to many other software programs. The publishers plan to send update notices to registered users. PC Globe would be highly useful to anyone interested in international development, and to students, teachers, writers, and editors. Because of its high quality and reasonable price, it is an outstanding value. PC Globe, version 4.0. PC Globe, Inc., 4700 South McClintock, Tempe, Arizona 85282 USA (1990). Requirements: IBM(tm) PC/XT/AT/PS2 or com- patibles, 640 K RAM, floppy drive or hard disk, DOS 2.0+. Supports Hercules(tm) monochrome, CGA, EGA, VGA. Available in French, German, Spanish, and other foreign-language versions. List price, \verb+$+59.95; mail order vendors, \verb+$+39. \section{SatelLife} Taken from SatelLife update 4/5/91, Julia Royall: SatelLife has designed an affordable telecommunications system called Healthnet, through which health professionals, academics, researchers and activists can have access to each other as well as their colleagues abroad. Initially SatelLife is focussing on Africa, where the need is greatest, before expanding to countries in Latin America, Asia nd the Pacific Rim. We have chosen `store and forward' satellite technology to achieve these objectives. A new application of tested technology, it is low cost, dependable and appropriate for locations where telephone service is poor or non-existent. SatelLife is presently setting up demonstration sites in collaboration with universities in Zimbabwe, Kenya, Zambia, tanzania and Uganda and is working locally witha variety of organisations to develop the support systems and programs necessary to ensure maximum utilisation. these include skills training and technical support for the user community. The ongoing technology transfer will be accomplished as African users train new users. HealthNet users will be able to query colleagues, conference others in the field, order and receive relevant medical literature or request a dtabase search. As the turnaround time may be only a matter of hours or days, they will be able to collaborate more effectively - perhaps for the first time - with others in the profession. We are able to circumvent the necessity for telephones by using simple ground stations composed of personal computer and radio (similar to `ham' radio) which receives signals from and transmits to a low-earth-orbit satellite about the size of a beach ball. Although access to information is a compelling part of SatelLife's mission, of equal importance is providing the means by which people can communicate with one another when telephone, telex and fax are too expensive or simply not available. The resulting isolation can be devastating. More information on SatelLife is available from:\\ Dieter Klein\\ SatelLife\\ 225 Fifth Street\\ Cambridge\\ MA 02142\\ Phon: 617 868 8522\\ FAX: 617 868 2560\\ Email: dklein@wpi.wpi.edu \section{Conferences} \subsection{Kenya: Social Implications of Computers in Developing Countries} \begin{description} \item[Name:]International Conference on the Social Implications of Computers in Developing Countries \item[Date:] March 23-25 1992 \item[Venue:]International Centre of Insect Physiology and Ecology, Nairobi, Kenya \item[Hosted:] International Federation for Information Processing (IFIP WG9.4) and Kenya Computer Institute \item[Foci:] Appropriate technology, IT indigenisation, Social/cultural conflicts, human resources development, opportunities and risks \item[Format for papers:] 2 copies, A4, double space, not more than 20pp \item[Proceedings:] to be published by North Holland \item[Language:] English \item[Registration:] KSh 1,800 (approx USdollars 100) \item[Accommodation:] participants arrange, Kenya computer institute assists on request \end{description} \subsubsection*{Deadlines:} Abstract (1 page): 15/9/91\\ Full paper: 15/11/91\\ Acceptance: 25/1/92\\ Final manuscript: 25/2/92\\ \subsubsection*{Authors' correspondence:} and fuller information available from: Dr. Mayuri Odedra\\ 16 Avon Court\\ 34 Kenswick Road\\ London SW15 23U\\ \subsection{WORKSHOP FOR DEVELOPING COUNTRIES INET'91 } If anyone was at this workshop, part of INET '91, Copenhagen 15-20 June 1991, could they post a summary? Thanks. \section{Groups, Bulletin Boards and Mailing Lists} \subsection{I.F.I.P. working group 9.4: "Social implications of computers in developing countries" } (See also the note about the Nairobi conferece in March - ed.) KORPELA@fi.UKU writes: the International Federation for Information Processing (IFIP) is a multinational federation of organisations concerned with information processing. Currently it has 45 members representing 63 countries. One of its main objectives is to bring together computer professionals to stimulate research, development and the application of information processing in science and human activity. IFIP has eleven technical committees to which various member countries nominate the members. The technical committees cover a broad range of computing themes like information systems, education, and communication. Each TC has one or more working groups affiliated to it. The objective of a working group is to bring together a group of computing professionals interested in a specific area of work. Each WG is expected to act as a forum of exchange of ideas by organising conferences and publishing monographs. TC 9, 'Computers and Society', incorporates the working groups 9.1 Computers and Work, 9.2 Social Accountability, 9.3 Home Oriented Informatics and Telematics, and 9.5 Social Implications of Artificial Intelligence. The TC 9 working group WG 9.4, 'Social Implications of Computers in Developing Countries', was formalized in September 1989. The starting point for WG 9.4 was a scientific conference in New Delhi, India, in November 1988. The proceedings of the meeting were published by Elsevier Science Publishers in 1990 as 'S.C. Bhatnagar and N. Bjrn-Andersen (eds.): Information Technology in Developing Countries'. The next conference will be held in Nairobi in March 1992. Another major scientific event for the Working Group will be the 12th IFIP World Computer Congress in Madrid, Spain, 7-11 September 1992. WG 9.4 will organize one session within the track 'Diminishing the Vulnerability of the Information Society'. WG 9.4 invites researchers and practitioners of developing and industrialized countries to join its activities. As the WG is still in its making, one of the first tasks will be to establish a strong membership on all continents, and to organize interaction amongst the members. The aims and scope of the Working Group are reprinted below. For more information, please get in touch with the officials at the addresses below. Requests for membership should be sent to the secretary. Chairman: Prof. S.C. Bhatnagar Indian Institute of Management Vastrapur, Ahmedabad 380 056, India Tel. +91-272-407241 Fax +91-272-467396 Internet: root@iimahd.ernet.in Ag. secretary: Mr. Mikko Korpela University of Kuopio, Computing Centre PL 1627, SF-70211 Kuopio, Finland Tel. +358-71-162811 Fax +358-71-225566 Internet: korpela@uku.fi Bitnet: korpela@finkuo AIMS \begin{enumerate} \item To collect, exchange and disseminate experiences of developing countries \item To develop a consciousness amongst professionals, policy makers and public on social implications of computers in developing nations \item To develop criteria, methods, and guidelines for design and implementation of culturally adapted information systems \item To create a greater interest in professionals from industrialized countries to focus on issues of special relevance to developing countries through joint activities with other Technical Committees. \end{enumerate} SCOPE \begin{enumerate} \item National computerization policy issues \item Culturally adapted computer technology and information systems \item Role of transnational corporations, regional and international cooperation and self-sufficiency in informatics \item Social awareness of computers and computer literacy. \end{enumerate} \subsection{Subscribing to VITA} VITA's public, on-line discussion list, DEVEL-L, provides a forum for the exchange of ideas and information on a wide range of issues and topics related to technology transfer in international development; for example, technologies, communications in development, sustainable agriculture, women in development, the environment, small enterprise development, meetings, book reviews. Subscribers to DEVEL-L automatic- ally receive DevelopNet News. To subscribe, send this command or mes- sage: SUB DEVEL-L your-full-name to (BITNET:)LISTSERV@AUVM or (Internet:) LISTSERV@AUVM.AMERICAN.EDU You can also subscribe to DevelopNet News without joining the discussion list. Just send an electronic message to VITA. DevelopNet News is an electronic newsletter published monthly by Volun- teers in Technical Assistance (VITA), a private, nonprofit international development organization located in Arlington, Virginia. Your redistri- bution of DevelopNet News is encouraged. Information on the approximate size of your mailing list is useful to VITA. President: Henry R. Norman Editor: Patricia Mantey Editorial Assistant: Rafe Ronkin (VITA Volunteer) VITA specializes in information dissemination and communications tech- nology. It offers services related to sustainable agriculture, food processing, renewable energy applications, water sanitation and supply, small enterprise development, and information management. VITA is cur- rently involved in long- and short-term projects in 10 countries in Africa, Asia, and Latin America. 1815 North Lynn Street, Suite 200, Arlington, Virginia, 22209 USA. Telephone: +1 (703) 276-1800, BBS (VITANet): +1 (703) 527-1086, Fax: +1 (703) 243-1865, Telex: 440192 VITAUI, Cable: VITAINC, BITNET: VITA@GMUVAX , Internet: VITA@GMUVAX.GMU.EDU \subsection{Guide to the Latin-American Networks:Index and Retrieval Instructions} PEDRO@ohstpy writes: Hi! As you probably know, over the last year I have been sending out periodic updates of a list of latin-american mailing lists, with addresses, purposes, and useful tips about how to use them. Fortunately, the list grew pretty big, and by now almost every country of the region is included, in one way or another. So, I accepted the generous offer of the LASPAU administrators to put it on a listserver, and make it accessible to everyone interested in geting it. This will avoid the intense traffic every time I send a list out, and, more important, will save me the time to have to reply to many requests for copies, and will help those with quota limitations, or communication charges. I will, of course, let you know when we are incorporating new updates. HOW TO CONTACT THE LISTSERVER Send an e-mail message to LISTSERV@HARVARDA.BITNET, with the following lines as the only text of your message: GET LISTS GUIDE LASPAU GET LISTS-P GUIDE LASPAU You will automatically receive a copy of the current edition of the Guides. LISTS GUIDE includes lists of general interest, while LISTS-P GUIDE has those lists of professional interest. You can request each of the files individually. To see what other information could have been stored, include in the message the following line: INDEX LASPAU If you have any problems with the listserver, please send a note to the administrators, Cesar Galindo-Legaria (cesar@husc9.bitnet) or Alberto Oliart (oira@bu-pub.harvard.edu). \pagebreak \section{Contact Personnel} Please send contributions to the newsletter or requests for addition to the mailing list to KK. Email is the communication method of choice (it takes so long to type the gubbins in) but communication through any medium is welcome, especially if it contains contributions to the newsletter! Incidentally, if you receive this newsletter in elctronic form and do not have LaTeX at your site to print it out in a nice format, get in touch and we'll send you either a hardcopy form or an electronic form without the commands, whichever you prefer. \begin{tabular}{l|l} {\bf Software library:} & {\bf Newsletter, overall co-ordination, meetings:}\\ Howard Beck & Kathleen King\\ Artificial Intelligence Applications Institute & Department of Artificial Intelligence \\ University of Edinburgh & University of Edinburgh\\ 80 South Bridge & 80 South Bridge\\ Edinburgh EH1 1HN & Edinburgh EH1 1HN\\ 031 650 2747 & 031 650 2726\\ hab@uk.ac.ed.aiai & kk@uk.ac.ed.aipna\\ \hline {\bf Addresses and contacts, funding:} & {\bf Literature resource and bibliography:} \\ Robert Muetzelfeldt & Ehud Reiter\\ Department of Forestry and Natural Resources & Department of Artificial Intelligence\\ University of Edinburgh & University of Edinburgh\\ Kings Buildings & 80 South Bridge\\ Mayfield Road & Edinburgh\\ Edinburgh EH9 3JU & EH1 1HN\\ 031 650 5408 & 031 650 2728\\ R.Muetzelfeldt@uk.ac.edinburgh & reiter@uk.ac.ed.aipna\\ \hline \end{tabular} \end{document}