Students’ Communicative Competence Prediction and Performance Analysis of Probabilistic Neural Network Model N.Ayyanathan 1 A.Kannammal 2 A.Bavani Rekha 3 1 Department of Computer Applications, K.L.N.College of Information Technology, Sivagangai District, Tamilnadu, India 2 Department of Computer Applications, Coimbatore Institute of Technology, Coimbatore, Tamilnadu, India 3 Department of English Language Studies, Madurai Kamaraj University, Madurai, Tamilnadu, India Abstract The purpose of the research paper is to classify and assess the communicative competence of a class of thirty students of higher secondary school in India, before and after a well formulated training with the specially designed syllabus for the reinforcement of communicative competence. Regression statistics is applied to analyse the impact of different predictor variables for the total performance. The affecting variables or predictors are modelled applying Probabilistic Neural Network (PNN) techniques to classify the category of communicative competence performance of the learners. Learners are trained and tested in each component of communicative competence, based on the guidelines of Common European Framework of Reference for Language, Teaching and Assessment (CEFR). Keywords: Education Data mining, Communicative Competence, Common European Framework of Reference for Language, Teaching and Assessment (CEFR), Regression Statistics, Probabilistic Neural Network (PNN), Learning Management System (LMS) 1. Introduction The student‟s Communicative Competence plays an important role in getting their first job and further helpful in sustaining their position they hold and move successfully in their career ladder. The authors adopted Common European Framework of Reference for Language, Teaching and Assessment (CEFR) [10] for analyzing the various parameters of Communicative Competence and to evaluate the teaching learning process by applying Regression Statistics and Probabilistic Neural Network model. Michael Canale‟s [6] theory of Communicative Competence comprises of the four different components, Linguistic, Discourse, Socio-cultural, and Strategic. The first two sub-categories reflect the use of language itself. Thus Linguistic Competence includes “knowledge of lexical items and of rules of morphology, syntax, sentence- grammar semantics and phonology”. The second sub-category is Discourse Competence the ability to connect sentences in discourse and to form a meaningful whole out of a series of utterances. While Linguistic Competence focuses on sentence- level grammar, Discourse Competence is concerned with intersentential relationships. The last two sub-categories define the most functional aspects of communication. Socio Cultural Competence “requires an understanding of the social context in which language is used: the roles of the participants, the information they share and the function of the interaction”. Strategic Competence is the way we manipulate language in order to meet communicative goals. The importance of communicative competence is discernible as it permeates virtually in all the interactional activities of human beings. The contemporary researchers work on communicative competence and the relevant assessment techniques and how the methodologies keep changing over the years are reviewed in section 2. The detailed description of the data source and the descriptors taken for the neural network model, the independent variables and dependent variable for the statistical regression model is discussed in the section 3. The experimental results are discussed in section 4. The authors conclude the work in section 5 with future research direction. 2. Review of the Literature Noam Chomsky [9] was the first one to coin the term „competence‟ and differentiated competence as „an idealized capacity‟ and performance as „production of actual utterances‟. Initially traditional concept of communicative competence focused only on linguistic competence. Later Dell Hymes [11] alluded to the limitations of Chomsky‟s and zeroed in the sociocultural factors. Further he developed the communicative competence model based on the degree of possibility, feasibility, appropriateness and occurrence focussing the heterogeneous IJCSI International Journal of Computer Science Issues, Vol. 9, Issue 4, No 2, July 2012 ISSN (Online): 1694-0814 www.IJCSI.org 312 Copyright (c) 2012 International Journal of Computer Science Issues. All Rights Reserved.