SkillCrest: A Skill Assessment System Using Deep Knowledge Tracing and User Feedback Sentiment Analysis
K V Vaisakh, Akash Ravi, K.O. Aakash, Achyuth Sai, Jasmine Bhaskar · 2021 2nd Global Conference for Advancement in Technology (GCAT) · 2021
Skill Development on individuals holds an important role in the progression of a developing country. With the advancement of Artificial Intelligence, tutoring systems are becoming more automated. Especially during these days of the COVID-19 pandemic, there has been a sudden rise in demand for online tutoring services. Our work helps the user to track their skill using the Deep Knowledge Tracing System. It eases the assessment of a student with accurate results giving emphasis only on the skill development factor. This paper proposes a skill assessment system which adopts a hybrid model that consists of deep knowledge tracing and user feedback sentiment analysis. DKT is an algorithm that uses the Recurrent Neural Networks to track the improvement of each individual on a particular skill. The project also delivers the findings on a user level, by the creation of a website named “SkillCrest”. This website shows the primary information on skill development and helps the tutors to identify which skill a student lacks using DKT. This feature can also be used for self-assessment of a student. We also incorporated a sentiment analysis component in our website to analyze the feedback given by each individual. The implementation of sentiment analysis is proposed to be done using the Recurrent Neural Networks. This result is used further to retrain the model.