CollabNet

Inspiration

In academia and industry alike, research thrives on collaboration. Yet finding the right collaborators—those who share overlapping interests, complementary expertise, and compatible working styles—remains a challenge. We were inspired by the idea of harnessing the wealth of information in open academic databases to build a smarter collaboration tool. By visualizing co-authorship networks and research trends, we wanted to make it easier for researchers to discover potential partners and accelerate innovation.

What it does

CollabNet is a web app that helps you discover collaborators and understand research landscapes. You can search by research topic or by author name to see:

Trending topics and scientists: an overlay suggests popular topics and notable scientists when you open the search interface.
Interactive dashboards: bar, donut, and line charts show global research trends, top contributing institutions, and co-author network sizes.
Topic details: select a topic to view active authors, a force-directed co-authorship network, and a map of contributing institutions.
Author profiles: view an author’s works, citations, affiliation, and their co-author network, all visualized in real time.
Match evaluation: when you request to connect, CollabNet computes a compatibility score and shows a detailed breakdown of topic similarity, co-author distance, institution proximity, and recency alignment, along with evidence like overlapping concepts and shared co-authors.

How we built it

The backend is built with Python and Flask, exposing REST endpoints that wrap the OpenAlex API for topics, authors, institutions, and co-author networks. We also added custom endpoints for trending topics/scientists and match-scoring, with synthetic defaults that can later be replaced by real algorithms.

On the frontend, we used React with Tailwind CSS and Chart.js to create a responsive and accessible UI. React Router handles multi-page navigation; react-leaflet powers the institution map, and react-force-graph-2d draws the co-author networks. We implemented a dynamic overlay for trending suggestions, multi-state modals for author profiles, and a progress indicator for match evaluation. The app falls back to an offline data set when network requests fail, ensuring demos always work.

Challenges we ran into

Data sparsity and API limits: Building dynamic visualizations required careful handling of the OpenAlex API rate limits and dealing with incomplete or missing data.
Interactive graph rendering: Tuning force-directed graphs to display complex co-author networks in a performant and readable way took experimentation with physics parameters and drawing optimizations.
Responsive design: Adapting large data visualizations to small screens without overwhelming the user required custom layout logic and thoughtful use of Tailwind’s utility classes.
Match-scoring logic: Designing a compatibility model that could be computed quickly but still feel meaningful was non-trivial; we implemented a placeholder function and drafted plans for a more sophisticated algorithm.

Accomplishments that we're proud of

A polished, multi-page frontend with intuitive navigation and clean visuals.
Real-time integration with the OpenAlex knowledge graph, including fallback behavior for offline scenarios.
An interactive co-author network visualization that scales to dozens of nodes without sacrificing usability.
A modular backend with endpoints ready to be extended by ML or rule-based scoring for compatibility.
A user experience that makes exploring research connections engaging and accessible.

What we learned

Building CollabNet taught us about:

Working with open academic data sources and understanding their structure and limitations.
Designing data-driven UI components in React, and using charts, maps, and graphs effectively.
Creating a consistent design system with Tailwind CSS and customizing it to match a professional aesthetic.
Structuring a Flask API to support both straightforward data retrieval and more complex analytics endpoints.

What's next for CollabNet

There’s a lot of room to grow:

Enhanced scoring algorithms: Integrate machine learning models for compatibility, leveraging users’ research histories and preferences.
Personalized dashboards: Allow users to build profiles, save searches, and receive recommendations.
Real-time collaboration tools: Add chat, calendar integration, and document sharing to support matched researchers.
Expanded data sources: Pull in conference proceedings, funding opportunities, and more to enrich the collaboration discovery process.

CollabNet lays the foundation for a smarter, more connected research community, and we’re excited to continue developing it.