(07.03.2022 - 20.03.2022)
As part of project closure, we had a code review session with the clients to walk them through our codes and receive feedback. This is to ensure that the clients are familiar with the structure of the portfolio and the functionalities of the system. During the session, some minor improvements were suggested and we met with the clients during the following week before officially sealing the project. Overall, the clients are satisfied and happy with our work and effort. We were thankful for their patient guidance and appreciate the experience as well as knowledge learnt from this project.
For the past couple of weeks besides finishing the functionalities of the project and final touches we have been working on proper documentation and the project website.
The project website is a very detailed source of information about the project. It has many subsections such as research, implementation, UI, system design, testing which are crucial to get familiar with in order to thoroughly understand the work we have been putting our effort in for these past few months.
For people interested in continuing developing tools with Tapas in clinical use we recommend going to section Future work and Appendices->Manuals, where all the necessary information about installation process, prerequisites can be found.
Our project has a number of future works that would improve some aspects of the project. A lot of our time has been researching and implementing features - some that have improved the model and some that haven’t. With the time we had, we couldn’t explore all possible solutions to the problems we encountered and therefore future work would be to continue trying to find solutions to improve the bot's ability to answer questions in the medical field.
Train TaPaS-Base with reliable clinical data
Adapting the clinical BERT training approach, possible future experiments include preparing a huge amount of reliable clinical tabular data of all types and training the clinical TAPAS from the TAPAS-Base model. Following the pre-training process, fine-tuning tasks could be done to allow support and increase accuracy for SQA, WikiSQL, WTQ and TabFact. Other than that, another experiment that could be done includes concatenating tables into appropriate sentences such as “Header” + “Grid”, then passing the resulting sentence into the clinical BERT model. Both of the approaches stated above have yet to experiment and therefore, the final outcome remains uncertain.
Expand Post Processing Database with More Clinical Data References
As the bot interacts with more reliable data, more references would be explored to expand the database that we use for post-processing so that it would become more robust and accurate, thus, improving the bots ability to predict if a TaPas response is precise enough to be sent back to the user.
More File Extensions Support
Furthermore, currently, the chatbot can only input and analyse .csv files. In the future, file extensions feature will be extended and clients will be able to use files in the .txt, .pdf or other formats they wish to facilitate their interactions with the bot.