ResearchGram is a AI based tool that leverages Machine Learning, NLP and cloud computing to generate research paper recommendations from over 500,000+ research papers in our dataset (subset from the Arxiv dataset) based on the input provided by the user (input can be provided in various image,video,audio and document formats).
Finding similar papers to a publication is one of the most crucial and challenging aspect of the research as researches need to manually inspect similar papers due to the inability of softwares like google scholar and Microsoft academic to accept non textual input making the process extremely inefficient.Adding to this, lack of access to university resources and libraries due to COVID-19 have amplified the existing obstacles faced during research. These challenged paved way for our project.
• Generates a textual summary of the input file provided by the user regardless of the format of submission(video/audio/image/PDF).
• Generates top 10 recommendation of similary papers.
• Generates summary of the top 3 matches research publications to facilitate effortless experience for user.
• Video: MP4 • Audio: MP3 • Document: PDF • Image: JPG/JPEG/PNG
• Further explore concepts discussed in class lectures recordings.
• Find research papers similar to a perviously read publications.
• get papers that can be used for citations of your essays.
• HTML5 •CSS3 •SCSS •Python • Django •JS •JupyterNotebook •PowerShell
•Google Cloud Vision API
•IBM Watson Speech-to-Text API
•Bert-Extractive-Summarizer (Generates summary)
•Specter (Generates Embeddings)