Conference: Home | Paper/Poster Submission | Organizing Committee | Pre-Conference | Speakers | Archive: Conference

***

Data Mining for Limited Resource High Demand Countries

Mohammad Kaykobad*
Abstract: The data mining is the set of techniques that enable us to extract useful information and knowledge from the raw data for later use in more efficient governance, management and administration. For countries like ours resources are limited and not adequate for satisfying high demand. This prompts us to allocate resources for satisfying demand  having high utility/return, be vigilant to avoid misuse of resources. Inadequacy of resources causes fraudulent activities, crimes that should be efficiently detected and measures to be taken. Some of the areas in which data mining techniques can contribute significantly to strengthen and consolidate our walk on the path of progress are:  analyzing customer and market trends for optimal utilization of resources. All these can be done using data mining techniques.
To present a concrete example we take the education sector. We know Government investment in education is a meagre 2% compared to recommended figure of 6% of GDP. For a country with population being the only surplus our progress depends on our ability to transform the huge population into human resources key to which is world class education. For a developing country like ours it is not a small task to educate about 40 million kids and young people attending educational institutions. Only 35 countries of the world out of 200+ countries have more population than we have students!!! To make our education more effective with scant resources we must analyze data now available with computerization of different education related activities. For some 20+ years we have been using computers in tabulation of public examination results. A lot of data have already been generated but unfortunately remaining unutilized. With a bit of change in registration information for examinees, say including educational, economic conditions of parents and locality we are able to extract many useful information, dissemination of which will inspire our population to plan being more informed and set useful targets. From the huge data we have been accumulating over the years we can find the trend of the population, their income, skill and education. We are able to appreciate correlation between parents education and financial capability  and quality of education students are getting. If it comes out that quality of education depends on the quality of education of parents then possibly would-be parents will plan for their own education creating an opportunity for the country to move ahead with a better qualified human resources. Data mining tools can be used to   generate statistics as to the performance of students in examinations of different subjects like mathematics, science, language, possibly can help us decide about  effectiveness of text books, necessity of training of teachers of different localities for better results, create a healthy environment of competition for excellence in skills of different subjects. It is also possible to rank localities based upon performance at different scales of aggregation like institution-wise, upozela, district,  division wise to inspire healthy competition.  This is possible not only in education sector but also in health analyzing pattern of diseases population suffer from region wise, season wise to take measures so that unproductive time due to illness is minimized, allocation of health resources can be done optimally. Data mining tools can also be effectively used to optimize performance of hospitals, city corporations, educational institutions and other agencies, in crime detection, in agriculture and in all other sectors as well.
It is said that data are more valuable than computer systems. Unfortunately in Bangladesh we are yet to appreciate usefulness of data. Appreciation of importance of data can be improved by applying data mining techniques on huge data available with the government and other agencies.
03_Prof. M KaykobadF
*Professor
. Department of Computer Science & Engineering (CSE)
. Bangladesh University of Engineering and Technology (BUET)
. &
. Fellow, Bangladesh Academy of Sciences

************