Defused News

In a world filled with negative news, positive news can be scarce. We have devised a platform that separates positive and negative news so you can stay updated, without sacrificing your mental health.

Comment

Inspiration

Since the start of the pandemic, many people have had trouble getting used to the amount of negative news appearing in their everyday lives. This issue especially attacks youth, most of whom spend a lot of their time on social media. Additionally, many children and teens are also not used to the amount of bad news they are seeing, especially since they are younger, and as a result, usually more emotional.

What it does

Defused News scrapes relevant news articles from Google, which are then analyzed and evaluated as positive or negative news. This news is then formatted and presented to the user in a way where good news is just as accessible as bad news or neutral news.

How we built it

The projects consist of three main parts: the web scraper, the news filter, and the webpage. The web scraper was made using an API named SerpAPI, which permits developers to scrape Google search results. The news filter was mainly made using a package named NLTK, which stands for Natural Language Toolkit. We used NLTK to format each headline to be properly analyzed, by removing stopwords (such as and, in, but, etc.) and reducing every word into its root form. This allowed us to search for each word in a document of positive and negative words to determine whether or not the word is positive. If a news headline had more positive words, it would be considered positive, and if it had more negative words, it would be considered negative. If the headline had equal amounts of positive and negative words, it would be considered neutral. Finally, after getting the core of the project done, we formatted the news results into a webpage, using a package named PyWebIO. The page contains four buttons that display news in four categories: positive, negative, neutral, and all.

Challenges we ran into

One challenge that we ran into was when figuring out how to scrape Google Search results. When using more traditional web scraping methods such as BeautifulSoup and Requests, the search results would not be consistently scraped. This is because of the way that Google limits the amount of information that can be scraped without using a certified API.

Accomplishments that we're proud of

We're really proud to have at least created a working version of the project in such a short time. In the beginning, we were worried about whether we should go forward with such a project since we did not know how possible it would be when considering our skill levels and the amount of time we had.

What we learned

Through this project and the struggles we've faced, we've learned so much, not just about web scraping and language processing, but also about how to collaborate effectively in a team environment and how to work independently.

What's next for Defused News

We would like to implement a more accurate system of filtering news, as the project will never reach its full efficiency without some sort of deep, machine learning algorithm. Additionally, it is a goal of ours to make this project accessible on PCs and smartphones through an app, instead of through a webpage. This would definitely provide more opportunities for the expansion of our project.

Share this project:

Updates