Inspiration
Oftentimes when citing a source or simply trying to find information, we encounter many websites and articles that may not be trustworthy. There are also a lot of articles that promote fake news, which are often disguised as being published under credible newspapers. To solve this issue, we wanted to create a platform which could analyze the credibility of a source.
What it does
Sourcerer analyzes and rates how credible a source (particularly an article) is based on four criteria: the published date, last updated date of the website, use of references, and the domain name. Sourcerer also spreads awareness about the difference between good and bad sources, and disinformation and fake news.
How we built it
In order to build our web application, we split up the process into two main components, one being the code aspect and the other being the user interface or the design element. For the code aspect, we started off with brainstorming all of the key features that we wanted the project to include. This includes all the categories that factors towards computing the final reliability score along with the weightage for all the categories. After we had the main plan set in stone, we began with writing pseudo code for the program, including how to organize our code, what to implement in each method, and what data structures to use for certain calculations. Finally, we turned our solution into code in the Java language using Eclipse. Similarly, for the user interface aspect, we brainstormed ideas based on what we wanted to include in our interface. We sketched out a low-fidelity prototype, and after revisiting our vision for the project, we developed a more detailed prototype on Figma. Lastly, we added various interactions to showcase the user experience of Sourcerer.
Challenges we ran into
Throughout the process of building our project, we worked past the hurdles of developing functioning code under a short time constraint, figuring out the user interface for Figma for the first time, and technical hurdles with bugs and API issues. In order to get past these, we had to adapt our initial approach multiple times in order to come up with a more reasonable problem solution. For instance, our initial approach was to connect our code to the user interface in Figma but with the time constraint and technical challenges, we adapted our plan to complete both components separately for presentation purposes. Also, we had a hard time scanning a web article in an html formal so we converted it to a text file and performed our analysis from there instead.
Accomplishments that we're proud of
We are proud of creating a final product with both functional code and user interface. This is also our first hackathon, so it was definitely a learning experience but we were able to learn and progress a lot along the way.
What we learned
We learned a lot about how to use a scanner to access different parts of a file, as well as how to access keywords and key data points using the words around it. For example, in order to access the year the website was updated, we had to search for the © symbol.
What's next for Sourcerer
With more time and access to additional tools, we hope to reconstruct Sourcerer to work in a more efficient and convenient manner and appeal to a wider range of uses. In particular, we would implement a way for the web application to directly pull the data of a web article from a URL and convert it to a scannable text file. Also, we would use the help of machine learning to read through a broad range of articles and test for more categories that would factor towards the final reliability score (such as checking for peer reviewed sources). In addition, we will connect the user interface component to our code in order to make a fully cohesive and functional web application. In the future, we also want to design a software that searches for credible articles on behalf of the user.
Log in or sign up for Devpost to join the conversation.