This research aims to design a Grievance Filing System built using automatic text classification without any manual interruption. Various methodologies are followed to achieve it and are implemented. Performance of the different algorithms is discussed. People are less aware about the lengthy methods for lodging complaints. We propose a simplified process of enrolling grievances to ministries. The system accepts grievances in recorded voice form. The system is designed for Marathi language. Input in the form of speech will ease people’s comfort for lodging grievances. We present a model where voice is first preprocessed, followed by text classification using deep learning approaches such as CNN and LSTM, the grievances will be sent to respective ministry. This system can be used by government ministries to get grievances from common people through a simplified process. User will be notified on the progress of their lodged complaint and on its successful resolution by respective ministry.
Read full abstract