# Research Networks

<details>

<summary>Definition</summary>

Research Networks illustrate the collaborative relationships between researchers, departments, and research groups within an institution. Connections are mapped to show how expertise flows across disciplines, highlighting hubs of innovation and areas of concentrated productivity. Metrics such as the number of publications provide insight into research activity and output. These networks reveal the internal dynamics of collaboration and their potential to drive advancements in knowledge and create opportunities for broader application and growth.

</details>

<details>

<summary>Navigation</summary>

This view is rendered as a force-directed graph (FDD) where:

* Nodes (spheres) = Individual researchers
* Edges (lines) = Collaboration relationships (e.g., co-authorships or joint project participation)
* Node color = Knowledge Area classification
* Node size = Number of publications

Interactions:

* Hover reveals researcher name, knowledge area, and total publications
* Click isolates that researcher’s network
* Filters allow exploration by academic institution

</details>

<details>

<summary>Methodology</summary>

The Research Network is generated from academic metadata scraped and structured.

1. Data Collection (Scraper Phase) - Scrape author profiles, project data, and publication records from institutional databases; clean and format the data.
2. Entity Assembly - Each researcher is identified as a node. Connections (edges) are established based on shared publication authorship, co-participation in the same research group, project collaboration
3. Knowledge Area Classification - Researchers are categorized into Knowledge Areas based on Natural Language Processing classification of their publication metadata. Color coding corresponds to these classifications
4. Network Visualization - A force-directed layout is used, which simulates physics: more tightly connected academic researchers are drawn closer together, while loosely related ones drift apart. This spatialization reveals natural clusters, central sectors, and structural gaps

</details>

<details>

<summary>Calculation</summary>

This is not a numerical KPI, but rather a structural and relational insight into the collaboration patterns within research institutions. However, each node in the network includes computed attributes such as:

* Publication Count – Number of indexed outputs (e.g., papers, articles, patents)
* Degree Centrality – Number of direct collaboration links (co-authorships) per researcher
* (Planned) Eigenvector or Betweenness Centrality – Measures of influence or network bridging roles
* Knowledge Area – Thematic classification used to group researchers based on the nature of their work

</details>

<details>

<summary>Interpretation</summary>

The Research Networks view enables institutional leaders and planners to:

* Spot highly productive researchers or underutilized talent
* Identify isolated research groups in need of greater integration
* Analyze the interdisciplinary strength of departments
* Understand how collaboration patterns align with innovation output

Over time, this view can be extended to track longitudinal changes, show the evolution of collaboration, or integrate with Innovation Pipelines to trace the pathway from research to startup activity.

</details>