Inspiration
According to a study by Pennsylvania State University, University Park, which estimated the size of Google Scholar at 100 million documents, suggesting that it covers about 88% of all scholarly documents accessible on the Web in English. With this overwhelming quantity of literature, surverying literature to stay abreast of the latest developments in a field is becoming increasingly difficult.
What it does
ScholarSphere is a Gemini-enabled application that is built to assist researchers in the various stages of research. Powered by Gemini, it is able to extract the search term from a user's input, perform web-scraping for Google Scholar results for the term, process the scraped data and display the same in the chat.
How I built it
ScholarSphere is built using HTML and stylized with Bootstrap. The backend utilizes Flask.
Challenges I ran into
The first focus in this stage of the project was to create a link between real-web data and the chat, such that the responses generated by the application are links accessible to the user, unlike the typical responses generated by AI-chats which are typically fictitious. Hence, web-scraping was incorporated into the project. Web-scraping naturally results in a large quantity of HTML source code, of which we are to extract papers and its various components such as links, cite count, snippet, etc. The initial stage of this project was heavily focused on this preprocessing of the data obtained by web-scraping.
Accomplishments that I'm proud of
ScholarSphere is able to identify the topic a researcher wants to search, perform web scraping for the term and then return real links in the chat to the references.
What I learned
ScholarSphere has proved to be a great opportunity to analyse and truly understand the various stages of research. This has led to ideas that I look forward to incorporating into ScholarSphere to make it a one-stop solution to every problem encountered during research.
What's next for ScholarSphere
The next stages of development for ScholarSphere will focus on improving its current functionality as well as incThe next stages of development for ScholarSphere will focus on improving its current functionality as well as incorporating many more features to turn this into a one-stop solution for research assistance. These include: - Fine grained control into the quantity of literature to be surveyed - Building a research tracking mechanism that is able to capture the various terms and organise them as breadth-level survey and depth-level survey - Extending the capabilities of ScholarSphere to assist in later stages of research such as assistant drafting papersorporating many more features to turn this into a one-stop solution for research assistance. These include:
Log in or sign up for Devpost to join the conversation.