Comb

Challenges Targeted

We built this hack as a solution to both Noibu's and Cohere's challenges. It is a Chrome extension that makes online e-commerce more delightful through the use of web scraping and NLP through Cohere's API. We use sentiment analysis to categorize reviews that are more informative and less biased for the shoppers.

Inspiration

How do we streamline an online shopping experience that is already so robust and simple? Buyers are presented with more options than ever, further exacerbated by the inclusion of easy payment methods, coupon code finders, and reverse image searching. However, one thing that is missing from the consumer experience: The ease of access to numerous high-quality reviews that provides additional education into the purchase of their products.

What it does

This is where we devised the concept of Comb, a lightweight browser extension that surveils the internet for various forms of reliable reviews in a quick and visually appealing manner. Upon its activation, it queries a web scraper to search the internet for high-quality, informative reviews that otherwise require the consumer to leave the online shopping environment. Having to leave a website to get more information disrupts the consumer shopping experience and reduces engagement with e-commerce websites, which is an especially important component of establishing strong business-client relations. Comb aims to put the essential information in the palm of the consumers' hands, enhancing the shopping experience as a result.

How we built it

This project has both a front-end and back-end component, the former consisting of HTML/CSS and Javascript, and the latter consisting of Python. Through the use of the Selenium framework, we are able to scrape relevant information (ratings, reviews, videos) from a selection of trusted storefronts and media outlets, and compile that info neatly in the extension front-end. The Cohere NLP toolkit is also utilized to determine the general sentiment of a given product review, whether it be positive, neutral or negative.

Challenges we ran into

One major challenge that we ran into when constructing this project was the connection between the front-end and back-end sides of the implementation. Since our browser extension relies heavily on web-scraping and NLP APIs, we ran into some issues taking the back-end information, processing it, pushing it to the front end in a clean and simple way that is easy for customers to understand. However, after several hours of deliberation and work, we were able to accomplish what we needed to finish this project.

Accomplishments that we're proud of

We are proud that we have successfully implemented web-scraping into the back end and having it return valuable information that can prove very useful in future iterations to potential customers.

What we learned

Throughout the planning, implementation, and building of this ambitious project, we have learned many things along the way, ranging from new ways on how to collaborate and work together, new coding practices and technologies, as well as innovative ways to think about user design and experiences.

What's next for Comb

Although Comb is in its infancy stage, we believe there is a lot of room for growth should we decide to go forward and pursue this further. There is many opportunities to expand and improve functionality so that it can become a well-known browser extension that assist many buyers in their purchasing decisions. Some examples of these improvements include a faster processing times, functional design choices, as well as better API models that provide better and more accurate search results.

Built With

• cohere
• html/css
• javascript
• python
• selenium