The PURE API and my internship
I have just reached the middle point of my twelve-week internship with the UWP tech team working on extending the PURE API. It has been an amazing journey both to obtain skills and to get familiarised with work environment.
I am heading for the third year study of Artificial Intelligence and Computer Science at the School of Informatics and my area of interest lies within Machine Learning and Natural Language Processing. What surprised me most in the team was that the work environment is very relaxed even when the jobs are busy. My colleagues also made me feel very welcome and they are really treating me as a member of the team. I am grateful to have this summer experience that allows me to explore something out of the University. Well, the team is in the University anyways.
The PURE API
My project is to extend the current API of the University Website Programme, which extracts the researchers’ academic records from the Research Explorer of the University. The data then supplies the PURE widget that allows academics easily embed their profile in their CV or on their websites. The EdWeb can also adopt the data for future development.
The team developed the current version of the API a few years ago. It extracts all data of the academics from the PURE server and saves them all in the UWP database. Back then, the volume of the data was small so it took only a few hours to finish the process. Over the years, however, the academic records in the server has expanded tremendously. The team has to divide the workload into three to finish the process.
The team has decided, then, to replace the current API with a new one that utilises the latest “Change Endpoint” from PURE. The “Change Endpoint” returns all changes that took place from a specific date. For example, new academics have entered the University, some publications have changed, and someone left the University. With this endpoint, we can update only the changed records, instead of hundreds of thousands of records every time. This improvement also speeds up the process from once per week to once per day! What’s more, it saves up so many computing resources for the University to use.
My progress so far and what’s next
I have completed most jobs of getting data from PURE server, and I have done some testing to update our temporary database. It is working as expected.
For the future, there are still some improvements to make, such as adding new features and saving more kinds of data. What I am going to work on next is to write some codes that automatically transfers the data from our temporary database to the widgets and EdWeb. Hopefully I can finish it within the next few weeks. Alongside with doing my project, I would like to get more exposed to User Experience. Combining these experience I am very confident for my future career.