Hey Md. Naeem Aziz,

That is an insightful project to work on.

You can work on an existing dataset and implement sentiment analysis with the help of Natural Language Processing (NLP) based on the string and data of texts which are the statements often used by people actually taking the test to a pre-defined set of questions.

Training the model on that and then testing on new samples of data can be a challenging and an effective task.

Reference for sentiment analysis is here:

Research Polarity Testing and Analysis of tweets in Twitter using Tweepy

I would suggest you work on an annotated data of image sets of facial expressions and emotion detection.

A reference is here:

Preprint Machine Learning based Lie Detector applied to a Collected a...

There are various methods that have been used to create lie detectors using machine learning. One of the common approaches is based on physiological signals such as heart rate, blood pressure, respiration rate, skin conductance, and facial expressions. These signals are known to change when a person is lying or under stress. Machine learning algorithms can be trained on these physiological signals to detect patterns or changes that are indicative of lying.

Article Machine Learning-Based Lie Detector Applied to a Novel Annot...

Article Lie Detection using Speech Processing Techniques

https://www.frontiersin.org/articles/10.3389/frobt.2019.00064/full

Article Deception detection using multimodal fusion approaches

https://www.sciencedirect.com/science/article/abs/pii/S0747563221003861

https://scholarspace.manoa.hawaii.edu/items/5aa7d880-1b26-4269-b7a1-4be708bea328

Lie tester must gather a collection of bodily reactions from people who are speaking the truth and those who are lying. The information would then be used to teach the machine learning algorithm, which would learn to recognize patterns in physiological reactions linked with lying where physiological reactions, popular method is to use supervised learning algorithms such as support vector machines (SVM) or artificial neural networks (ANN).