Opinion Sentiment Analysis

Abstract

An automatic system to analyze the opinions/feedbacks of the people working in any organization is proposed. The proposed system does two major jobs: the first part extracts the most relevant/important aspects from the textual feedbacks provided by clients/managers/co-workers for all the employees having similar job profile and the second part measures the degree of positivity or negativity for all the extracted aspects for every employee and then an overall performance of everyone is evaluated. We have implemented the system to study the feedback of teachers. The proposed system is flexible and versatile than the existing feedback evaluation system of teachers where students evaluate the teachers on the predefined aspects decided by experienced and senior teachers. Our system, Faculty Evaluation System (FES) identifies strengths and weaknesses of teachers on all aspects which are important to students. This information may be used by higher authorities of the institute in identifying suitable faculty members for different academic and administrative activities of the institute in addition to the teaching responsibility.