RELIABLE COGNITIVE DIMENSIONAL DOCUMENT RANKING BY WEIGHTED STANDARD CAUCHY DISTRIBUTION
Categorization of cognitively uniform and consistent documents such as University question papers are in demand by e-learners. Literature indicates that Standard Cauchy distribution and the derived values are extensively used for checking uniformity and consistency of documents. The paper attempts t...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
ICT Academy of Tamil Nadu
2017-04-01
|
Series: | ICTACT Journal on Soft Computing |
Subjects: | |
Online Access: | http://ictactjournals.in/ArticleDetails.aspx?id=3013 |
Summary: | Categorization of cognitively uniform and consistent documents such as University question papers are in demand by e-learners. Literature indicates that Standard Cauchy distribution and the derived values are extensively used for checking uniformity and consistency of documents. The paper attempts to apply this technique for categorizing question papers according to four selective cognitive dimensions. For this purpose cognitive dimensional keyword sets of these four categories (also termed as portrayal concepts) are assumed and an automatic procedure is developed to quantify these dimensions in question papers. The categorization is relatively accurate when checked with manual methods. Hence simple and well established term frequency / inverse document frequency ‘tf/ IDF’ technique is considered for automating the categorization process. After the documents categorization, standard Cauchy formula is applied to rank order the documents that have the least differences among Cauchy value, (according to Cauchy theorem) so as obtain consistent and uniform documents in an order or ranked. For the purpose of experiments and social survey, seven question papers (documents) have been designed with various consistencies. To validate this proposed technique social survey is administered on selective samples of e-learners of Tamil Nadu, India. Results are encouraging and conclusions drawn out of the experiments will be useful to researchers of concept mining and categorizing documents according to concepts. Findings have also contributed utility value to e-learning system designers. |
---|---|
ISSN: | 0976-6561 2229-6956 |