SUBJECT DISTRIBUTION USING DATA MINING

Abstract

The main objective of higher education institutions is to provide quality education to its students. The faculties employed by the educational institute's plays the dominant role to achieve highest level of quality in higher educatio n. The faculty having excellent subject knowledge and teaching skills have the majo r impact upon the performance of students resulting in good academic results, placements and hereby increasing the quality intake of students. This paper will assist the academic p lanners in distribution of subjects among the faculties in the department such that the students can make the optimum use of facu lty knowledge, experience and teaching skills to reach the new heights. A large number of self financing private institutes have opened over the last decade with the objective of p roviding quality education to students in various fields of engineering and other professions. The factors affecting the qu ality of education include faculty profile, placements, infr astructure, working environment and vision of the institute. Out of all the above factors the most important fac tor is faculty profile, Most of these self financing insti tutes compromise on the quality of faculty to cut down th e cost of salaries and recruit inexperienced, untrained and l ess qualified teachers. These less qualified and inexperienced te achers are not able to make the optimum use of the institute's resources and are not able to provide quality teaching to the students. As a result the performance of the students remains be low satisfactory; this also affects the placement of th e institute. This adversely affects the quality of intake in the institute and causes further deterioration in the performance of the institute and slowly the institute reaches on the verge of cl osure. A large number of foreign universities have also go t approval from the ministry of Human Resource and Development to compete with the Indian Universities. After the ent ry of these foreign universities, the survival of these self fi nanced private educational institutes has become further challengi ng. Since the motive of most of the self financed institution s is to maximize the profit, hence they are not able to com pete with the foreign institutes resulting in the closure of these institutes. In order to compete with the foreign universities a nd Government aided Indian Institutes these self finan ced private institutes should increase their budget on hiring e xperienced and qualified faculty so as to provide excellent su bject knowledge and make optimum use of the institute's resources. This will have major impact upon the performance of the students, resulting in good academic results, place ments and increased quality intake in the institutes. Like th is these institutes will be able to sustain their existence competing with the good Indian and Foreign institutes. An assessment about the faculty's subject knowledge and teaching skills should be made and based on which t he faculty should be allocated the subject for teaching the st udents. Apart from the subject knowledge, faculty's qualification , feedback and the result should be taken in to consideration while allocating the subjects in future academic planning . This paper uses Educational Data Mining Technique (EDM) to improve the distribution of subjects among facul ties in the department such that students can gain benefit from knowledge and experience of faculty members to improve their performance. Data mining, the extraction of h idden predictive information from large databases is a po werful technology with great potential to help head of dep artments in the institutes in distribution of subjects. It disc overs information within the data that queries and report s can't effectively reveal. After gathering data from the r esume submitted by the faculties at the time of recruitme nt and feedback form filled by the students over the year s, data mining technique need to be applied to determine se t of patterns for allocation of subject among faculties. With the help of data mining techniques, such as cl ustering, decision tree or association analysis it is possibl e to discover the key characteristics from the details of faculti es and possibly use those characteristics for future predi ction. This paper presents decision tree algorithm as a simple and efficient