Challenges

Learning to use the Flask framework and all packages in such a short time was the primary challenge during the development of this project. Voice input in particular was hard to integrate. The web API that we used in project 5 for our project was too old to be integrated into our project, so we need to look for other tools. After some research, we tried the Google SpeechRecognition API, but this API has its own problem. We ended up using PyAudio, a package that is able to detect voice from microphone, but it is not installable through pip. After some digs in the internet, we ended up using a wheel file to install it. However, PyAudio is not functional on MacOS.

It was hard to find any participants because of COVID-19. People were harder to reach out because a zoom meeting needed to be set up and the user needs to setup the installation of the application, which is not an effective way. In the end, we asked people around us, like our family members or roommates. If we were not under quarantine, the process would have been much easier. We only need to walk up to people in the library and ask if they are interested in testing our application out.

Future Works

In the future, we would like to work with a team of therapists and psychologists to improve our scripts for our chatbot. If we could get some professional advice, we could make the chatbot smarter in giving back responses to the users. We could also use face recognition to detect the user's emotions as they are talking to the chatbot, so the chatbot will be able to adjust its response based on their facial expressions. More machine learning algorithm could also be used to improve the responses of our chatbot to make the application learn from the user’s previous sessions.