Smart Student Grievance Redressal System with Foul Language Detection

Abstract

Grievances are quite common in educational institutes and organizations and the majority of the institutes find it difficult in solving the students' problems. It might be that the students are diffident in approaching the management or the management is inefficient and most of the times the complaints remain unattended. The work in this paper aims to solve all these problems and presents a web portal created using Django, HTML, CSS, SQL, etc. where students can put complaints on the portal from the categories mentioned and the complaint goes to the concerned department. The student can track the complaint completely and know whether the complaint has been viewed, in progress, transferred, rejected, and solved. There is complete transparency in the proposed system and if the complaints remain unaddressed for a few days, then automatically the system redirects the complaint to the person above in the hierarchy. Special authorities have been provided to the principal/head of the organization which would help in expediting the solution providing process to the students. The quality of language is an area of concern in any redressal system since it can be used for fake propagandas by commenting on topics like religion, gender, race, sex, individual, etc. This is taken care of in this work by implementing Foul/Hate Detection using technologies like Machine Learning and Deep Learning. SVM, LSTM and Bi-LSTM models are used after training over 11,325 tweets and upon evaluating the results using evaluation metrics like recall, precision and F1 score, ‘LSTM model’ outperformed the other models by achieving ‘0.884’, ‘0.84’, ‘0.86’ values for the metrics respectively. The system also includes a page where all the complaints of the organization are listed, and a support feature is included which enables students to support other complaints which would in turn help the authorities in analyzing the major issues in the organization.