| Autor: |
Kumar, Akshi, Abraham, Ajith |
| Předmět: |
|
| Zdroj: |
Journal of Information Assurance & Security; 2017, Vol. 12 Issue 4, p146-153, 8p |
| Abstrakt: |
The collaborative and participative facets of both Software Development and Web 2.0 has compelled researchers and practitioners to probe the integration among these two distinct, evolving areas of study. Opinion mining as a subtask of text mining, automatically extracts knowledge from loosely unstructured and often ungoverned human-sourced big-data information available through Social Networks. This paper proposed a model for mining the opinion with a deeper emotional implication that can assist as a supporting tool for User Acceptance Testing. The idea is to gauge the acceptance of an open-beta release version of software by initially extracting the opinion from tweets and consequently assessing finer-grained levels of emotions using a hybrid approach (lexicon + machine learning) that can quantify acceptance criteria attributes such as usability. The tool has been implemented & evaluated for supervised machine learning variants, namely Naives Bayesian, Multinomial, Gaussian and Bernoulli Naives Bayesian along with Support Vector Machine. The effectiveness of the proposed tool is presented with a sample set of tweets based on a case study and initial results demonstrate that it is a motivating technique to test the business objective of the system developed. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
