Authors: Ishan Khare, Michael Nixon, Soham Konar
A copy of our website's code can be found in this repo.
Here is a link to our final project video and demonstration: https://youtu.be/ARwPzbgVJm8.
As students, it is important that we can plan out the courses that we will be taking over our undergraduate career and so we need a system or document that makes that process easy for us. The current system for computer science majors is a set of program sheets, one for each track in the computer science major. While these program sheets are functional, they have a few issues: they are not interactive, they are unclear about or donʼt list all courses for each requirement, and they only list the course numbers for the majority of classes. Our goal is to make the BSCS program sheets as interactive, informative, and accessible as possible while also making the user interface clean and dynamic. The features we are introducing in our new BSCS program sheets are that they will be interactive, they will list all courses for each requirement, they will list all course names and numbers, and the web page will update to reflect your course selections and unit counts as you fill out the sheet.
Our user interface balances ease of use, flexibility, and extensibility. Through the employment of color as a visual aid, users are guided through the worksheetʼs various steps towards the completion of their track requirements. Incomplete requirements are made salient by the color red and return to black once they have been met. At a glance, users can immediately gauge their progress and identify areas that require completion.
The flexibility of the interface allows users freedom in how they provide information and engage with our interactive worksheet. With all fields displayed and available to the user at all times, there is transparency in the length of the worksheet, how the results will be presented, and the overall unit progress towards graduation requirements. Progress updates automatically as users input and change data, giving immediate feedback.
Excitingly, this interface design is easily extensible to other applications. The minimal visual design, tailored to the brand language of Stanford University, means the worksheet design is relevant to other subjects and/or departments at Stanford. Our general, reusable predicates and rules in the underlying logic make it simple for administrators to customize the worksheet to match their specific degree requirements.
We decided to use much of the data obtained from the MSCS data sheets as a starting point. Then, we scraped additional information from the Stanford Course Catalog (Explore Courses). We stored this information/dataset using the course(id,department,number,title,num_units) predicate. For example, this is what the predicate for CS 151 would be in our dataset: course(cs151,"CS","151","Logic Programming",3). In addition to the data, we used logic programming to represent the different requirements users would have to meet. These included subgoals that had to be satisfied as well as view definitions that calculated a minimum number of units. One problem we had to overcome was to ensure that students could not apply two classes to the same requirement. For example, students are not allowed to take both CS107 and CS 107E and use both to satisfy the requirement. Instead, they must choose one or the other. To ensure this occurred, we made use of exclusive(). Most of the other technical logic programming details can be found in the source code itself.
Overall, it is not necessary to utilize logic programming for the implementation of program sheets. However, logic programming simplified the development process quite a bit. Since the majority of our project consisted of validating requirements, using Epilog was a perfect fit. Styling the various HTML elements was quite easy as we rarely had to worry about the underlying system and could work at a more abstract level. If we used typical programming, we would have had to include numerous if-then statements. As a result, fewer lines of code were needed to implement the core functionality of the project with logic programming.
Although logic programming made developing our project easier, we still faced a considerable number of challenges. First and foremost, the large datasets made our coding very cumbersome. In certain situations we had to create hundreds of options for a single HTML dropdown and implement logic for each of those dropdown options. In general, the smallest of typos could sometimes cause the functionality of the worksheet to fail. We also had moderate difficulty allowing for multiple-select to satisfy electives as shown below:
Our current project provides students with a way to check their progress for the Bachelorʼs degree in Computer Science — AI track. However, there are several other tracks that we should also consider, namely, Biocomputation, Computer Engineering, Graphics, HCI, Information, Systems, Theory, and Unspecialized. Upon having completed one program sheet, creating very similar ones for the other tracks should not be too difficult. Additionally, our current program sheet has no real backend for data collection and submission. This would be another feature to implement to increase the utility of our project. A possible extension would allow students to specify which quarters during their Stanford experience they would take each course. Depending on course conflicts and scheduling, the program sheet would then provide feedback on whether or not the proposed four-year plan was feasible. In general, there are various ways to expand upon our work, a few of which we have just elaborated upon above. We are excited to see where this leads us in the future!
