Adoption of Opinion Mining in the Faculty Performance Evaluation System by the Students Using Naïve Bayes Algorithm

Abstract

Abstract-The paper promotes adoption of Opinion Mining in the faculty performance evaluation system by the students using Naï ve Bayes algorithm. The study may help the university administrators to identify the strengths and weaknesses of the faculty members based on the textual evaluation made by the students written either in English or in Filipino language. The system provides graphical representation of the evaluation result in pie chart with the percentage of positive and negative feedback of the students. Based on the result of the survey, the respondents highly recommend the proposed system. Thus, the school administrators and educators will be more aware about the sentiments and concerns of the students. The reports generated from the developed system could be utilized to enhance the performance of the concerned faculty. In addition, the output from the objective evaluation of the teacher's performance could also be used as a basis for merit, awards, and/or promotion. Furthermore, the said innovation may encourage as well as improve fact-based decision making for university stakeholders.

Adaptive Approach for Intelligent Web to Enhance Business Intelligence Applications

“…Attention to the value of user opinions made a great leap toward the approach of opinion mining and how it could enhance the web intelligence. Many researches used opinions from different points of views and in different fields as (Yuan et al, 2017) succeeded in using a sophisticated kind of opinions which is emotional features and social interactions to provide recommendations that meet user needs, ( Balahadia, & Comendador, 2016) depended on opinions in enhancing the performance and the level of services provided by a faculty and ( Jaskolski et al,2016) used opinion mining in predicting the relevant categories as well as the specific aspects in the field of distance education .…”

Sentiment Analysis of ASOS Product Reviews Using Machine Learning Algorithms by Comparing Several Models

“…The results demonstrated that the model could effectively estimate online feedback. A crossbred attribute-centric sentiment categorization was introduced to incorporate domain-specific knowledge and extract definite word relations [13]. Datasets were created to test the effectiveness of the method.…”

“…The emotion classification approach affirmed by [12] is not efficient, as it uses only a lesser amount of data for analysis. The hybridized attribute-centric sentiment classification stated by [13] discovered the most prevalent bigrams and trigrams in the corpus. The method has proved unsuccessful in detecting the attribution and demanded a prolonged computational effort.…”

“…The first model created for the three-class classification dataset is the Nu-Support Vector classification model. The nusupport vector classifier is akin to the Support Vector Machine algorithm with the only difference being that the nu-Support Vector Classifier has a new variable to administer the number of support vectors [13]. The second model created is the Linear Support Vector classification model.…”

A Novel Framework Using Machine Learning to Effectively Analyze the Faculty Evaluations

“…This provides a deeper understanding of the course that was not evident from quantitative data. Balahadia and Comendador (2016) performs opinion mining on students comment in Filipino or English, using Nave Bayes classifier. The results of the evaluation can be shown to higher management in the form of bar graphs and pie-charts.…”