Taught by Mrs. Taricco, Computer Science is a course that focuses on various skills commonly used in the real world. Specifically, we learn about the basics of JAVA, including topics such as Primitive Data Types, Decisions, Arrays, and Iteration. Additionally, we learn the Web Languages of HTML, CSS, and JavaScript. In fact, this entire website was made using those three languages! Later, we will use our skills to solve a problem in the community in the Apps for Good project.

Shown below is the program and output of the Stars program. This code uses a combination of the Graphics object, loops, and conditional statements to generate an artistic canvas. To go above and beyond the basic requirements, I added user input that allows the user to choose how many stars are displayed, and I implemented randomness so each star has a different number of points.

Shown below is the code and output of the Line Art program, which creates a unique pattern only using straight lines. Although the program consists of just a few lines of code, it produces a highly detailed design. By using only eight loops, I was able to generate a repeating pattern in which one box is inside another.

The internet is an essential source for students, researchers, and educators to find factual information, but the reliability of what they find online has become extremely difficult to assess, especially for research articles. A major source of the problem is the rise of predatory journals, which are publications that charge authors a fee to publish work without peer review or results. Then, this flawed or inaccurate research is allowed to circulate online like real articles are, degrading the value of the internet as a research assistant. These biased and inaccurate sources may lead to the students’ future works being based on misinformation. The inaccuracy of articles online only continues to rise, with a clear gap between the amount of unreliable material and the tools available to identify it.

The Infoguard extension is designed for three overlapping user groups, each requiring verification for the credibility of online information. The first group of target audiences is students, specifically those older than 13 years old. These individuals often conduct research for essays, projects, and assignments that require credible sources. During this process, they are highly vulnerable to misinformation and often lack the skills or time to evaluate their credibility. Our tool provides the students with an accurate, objective method to examine sources without deep research. The second group of target audiences is teachers and educators, who can use InfoGuard to examine online material before exposing it to students. Moreover, they can examine student work to ensure their responses are accurate and well-researched. The last group that this product would be targeted towards is research professionals who benefit from cross-referencing other sources in their text. This extension allows for a quick verification of their sources to ensure that their writing is credible.

• Given a selected piece of text, the product should output a percentage representing how likely the information is accurate. This percentage will be based on comparison with reputable academic sources and should be at least 70% accurate. The percentage will be calculated by comparing whether the phrases and terms in the selected sentence match those from the reliable source. Then, for each sentence, the total accuracy is calculated by dividing the number matching phrases by the total number of phrases in the sentence. This will allow for each sentence to have an objective accuracy when compared to sources that are 100% accurate.
• Allow the user to highlight any text on a webpage and submit it for analysis, directly from the browser.
• Highlight the analyzed text on the page with a color coded overlay displaying accuracy as well as the exact percentage.

Infoguard is a browser extension that addresses these issues by allowing users to highlight any passage of text and receive an accuracy score generated by cross-referencing the selected claim with peer-reviewed literature from trusted databases. It uses a Natural Language Inference model, combining multiple fact-checking functions into a single tool. The scalable nature of the extension allows for the addition of several features in the future, such as a citation panel, full page scans, and a local personalized user history. The extension offers a practical solution for the students, educators, and researchers who face this challenge in information literacy.

The Infoguard extension focuses on three features that define its value: accuracy detection, access within browsers, and color-coded credibility organization.
Firstly, the extension must be able to take a highlighted passage of text and return a percentage representing how well the claims are supported by reliable, academic literature. This is done through a backend that cross-references the text against real academic papers from publicly available databases, such as Google Scholar and PubMed. A source is only considered reliable when it contains over 5 peer reviews to ensure reliable and accurate data. Using a pretrained Natural Language Inference (NLI) Model, the application should classify each fact with evidence from non-biased sources, achieving a detection accuracy greater than 70 percent.
Secondly, the application must allow for in-browser text selection and submission, with users interacting with the extension inside the browser tab. Through an icon in the browser toolbar, the user should select the extension and the text to send to the backend for analysis. This feature is built using standard WebExtension APIs, making it compatible with Chrome and Firefox. This feature only needs the user to select the extension and the text to analyze.
Once the analysis is complete, the results must be displayed directly on the page through a color-coded system. The color of the highlighted text corresponds with the accuracy and reliability of the information. The colors of the text will range from green for the most accurate information to red for the least accurate. To see the exact percentage, the user will simply hover their mouse over any highlighted section, and a floating label will appear next to the text. This system will allow for both a detailed view and a summary of the entire text.

There are various possible additions that enhance the value of the product and provide some important user interface improvements.

• Source Transparency (Citation Panel): Instead of showing only the percentage, our extension could display a side panel listing the exact sources that contributed to the accuracy score, with links, as well as a brief explanation of whether each source supports or contradicts the highlighted claim. This is a direct solution to one of the limitations of our MVP, as without this, users would not know why a passage scored the way it did.
• Full Page Scan: Users may find it inconvenient to highlight individual passages. A “Scan Full Page” feature could be added to automatically separate the entire document into claims, then run them all through the NLI pipeline in parallel and produce a colour-coded overlay of the whole page at once. This would save time for students or researchers who are studying multiple sources. It could also provide a whole document score, so users would be able to know if an entire source is reliable or not.
• Personalized History: The extension may also maintain a local history of previously analyzed content, allowing users to bookmark accurate sources and share them with others. For example, educators may flag a source as unreliable and share it with other teachers.