Behavioral Monitoring using vocal sentiment analysis

Project Title

Behavioral Monitoring using vocal sentiment analysis

Project Area of Specialization Artificial IntelligenceProject Summary

Communication through voice is one of the main components of affective computing in human computer interaction. Humans produce vocal sound which can be characterized by acoustic features like pitch, timbre, loudness and vocal tones. Human emotions can be differentiated on the basis of varying characteristics. Detecting human emotions from speech and analyzing them is potentially beneficial for improving human conversational skills. Using the current knowledge of sentiment analysis available, this proposed model handles speech data and extracts human emotion information like happiness, sadness, angriness etc. As the importance of providing services increases, vocal sentiment facilitates to monitor the human behavior. There are certain scenarios where human behavior leaves an imprint on others. This proposed model focuses on an application which takes human voice into account, processes it, extracts sentiments and give complete emotional insights to relevant higher authorities. 

Project Objectives

The objective of this proposed model is to automate the time consuming task of supervisors i.e. to monitor the interactions of their employees with customers by extracting sentiments from their vocal data. We aim to develop a mobile based application which implements this model and returns a detailed insight on person’s emotion over different instances.

Project Implementation Method

This project requires following methodologies and tools:

Data Collection:

A voice dataset is collected that contains some basic but important features. With the help of which more accurate classification of the sentiments will be done. Tools will also be selected on the basis of programming language that will be used. Most probably, JUPITER NOTEBOOK, GOOGLE COLABORATORY and VISUAL STUDIO CODE will be preferred because their interfaces are quite descriptive that may help in demonstration.

Data Cleaning and Feature Extraction:

Data cleaning and feature engineering will be required for a good model with high accuracy. All the unnecessary data and redundant features will be removed from the dataset. All those features will remain in the dataset that can be added in sensitivity list.

Modeling and Training:

An efficient model will be chosen after testing different models/classifiers like SVM, HMM, CNN and some other neural network models. Predictive analysis is totally based on the accuracy of model that will be predicting the sentiments.

Report Generation using Data Analysis:

Based on the sentiments extracted by the deep learning model a detailed report will be generated using data analysis and visualization techniques. This report will provide enough insights to monitor the emotions of relevant people

Benefits of the Project

Call Center Supervisors:

Interaction between employees and customer has to be monitored by the supervisors to ensure quality of every call. Supervisors can use this model for this task and based on the insights provided by our model, they can act accordingly. • Emergency Scenarios (hospitals, police stations, fire stations etc.):

In case of emergency, caller agents should use a considerate tone to avoid any chaos. Supervisors can use this model to assess their behavioral compatibility in critical scenarios

. • Any organization where behavioral monitoring is needed

Technical Details of Final Deliverable

Final Deliverable is a mobile application featuring a voice recording system. It records human voice, extract vocal features which will be processed by the model to detect the emotion. 

Final Deliverable of the Project Software SystemCore Industry ITOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Industry, Innovation and Infrastructure, Peace and Justice Strong InstitutionsRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	250
Mic for collecting Data	Equipment	1	250	250