Academic World Multi Database Dashboard

A CSS-styled Dash (Plotly) dashboard for CS students to discover potential faculty collaborators through overlap in research topics, co-authorship links, and university partnerships.

Purpose

• Scenario:

  • Explore and manage the Academic World dataset across three databases to understand faculty profiles, keyword trends, and collaborations.

• Target users:

  • Prospective graduate students exploring programs and potential advisors
  • Academic administrators or database managers reviewing, updating, or enhancing faculty profiles

• Objectives:

  • Query insights (keywords, publications) with interactive filters.
  • Edit data (faculty info, collaborations) to see immediate effects.
  • Demonstrate use of three different database systems in a single dashboard.

Demo

• Video demo: Illinois Media Space Project Demo Fahad & Usman

Installation

1. Create and activate a Python 3.10+ virtual environment.
2. Install dependencies:

   pip install -r requirements.txt

3. Create a .env file and set credentials:

   MYSQL_HOST=localhost
   MYSQL_PORT=3306
   MYSQL_USER=root
   MYSQL_PASSWORD=your_mysql_password
   MYSQL_DATABASE=academicworld
   
   MONGODB_URI=mongodb://localhost:27017/
   MONGODB_DATABASE=academicworld
   
   NEO4J_URI=bolt://localhost:7687
   NEO4J_USER=neo4j
   NEO4J_PASSWORD=your_neo4j_password
   NEO4J_DATABASE=neo4j

4. (Optional) Verify connectivity:

   python3 test_connections.py

Usage

Run the dashboard:

python3 dashboard.py

Open your browser at http://127.0.0.1:8050.

Inputs are unified at the top (University → Faculty). Widgets auto-refresh every 30 seconds.

Widget 1: Faculty Profile (MongoDB)

Select University → Faculty to view profile from MongoDB.

• What: Detailed profile card for the selected faculty (name, email, position, affiliation, keywords).
• Inputs: University → Faculty.
• Database: MongoDB
• Purpose: Immediate context for the selected person; keywords also populate Widget 4.

Widget 2: Faculty Keywords by University (MySQL)

Select a University to see top keywords by number of faculty.

• What: Bar chart of top keywords among faculty at the selected university (faculty count per keyword).
• Inputs: University.
• Database: MySQL with prepared statements.
• Purpose: Snapshot of research themes at a school; helps choose a direction to drill down.

Widget 3: Top Keywords by Year (MySQL View)

Select a Year to see top keywords by paper count (powered by view vw_keywords_by_year).

• What: Bar chart of the year’s most frequent keywords by publication count.
• Inputs: Year.
• Database: MySQL
• Purpose: Fast analytics for yearly trends via a reusable view; complements Widgets 2 and 4.

Widget 4: Keyword Popularity Over Years (MySQL)

Select a keyword to view trend across years.

• What: Line chart of publication frequency for a chosen keyword across years.
• Inputs: Keyword (from selected faculty’s profile in Widget 1).
• Database: MySQL
• Purpose: Visualize topic momentum over time; identify emerging or declining areas.

Widget 5: Collaborations (View & Edit) (Neo4j)

View top collaborators by joint publications.

• Add collaboration between the selected faculty and another faculty.
• What: Bar chart of top collaborators (joint publication counts) with controls to Add/Remove collaborations.
• Inputs: Faculty (via unified filters), collaborator target (dropdown), Add/Remove buttons.
• Database: Neo4j
• Purpose: Explore and manage collaboration networks; demonstrates graph reads/writes.

Widget 6: Edit Faculty Information (MySQL, MongoDB, Neo4j)

Edit name, email, position, research interest, affiliation, keywords; updates all three DBs.

• What: Form to edit faculty profile fields; persists to all three databases.
• Inputs: Editable text fields + Update button.
• Databases: MySQL (transactional update), MongoDB (document update), Neo4j (node/relations update).
• Purpose: Keep profiles consistent across stores; demonstrates cross-DB updates and a MySQL transaction.

Design

Project Structure

├── assets/
│   ├── style.css
├── env
├── config.py
├── dashboard.py
├── mongodb_utils.py
├── mysql_utils.py
├── neo4j_utils.py
├── README.md
├── requirements.txt
├── test_connections.py

Architecture:

• Frontend: Dash/Plotly web app (dashboard.py, assets/style.css).
• Data Access:
  • mysql_utils.py for MySQL queries and updates.
  • mongodb_utils.py for MongoDB profile reads/updates.
  • neo4j_utils.py for Neo4j graph queries and collaboration edits.
• Config via environment variables in config.py.

Data flow:

• User inputs → callbacks → DB utilities → results plotted or persisted.
• Collaboration edits trigger immediate chart refresh via a hidden store.

Implementation

• Frameworks/Libraries:
  • Dash
  • Plotly
  • mysql-connector-python
  • pymongo
  • Neo4j
  • Python driver
  • python-dotenv
• Key files:
  • dashboard.py: layout + callbacks for 6 widgets.
  • mysql_utils.py, mongodb_utils.py, neo4j_utils.py: database logic.
  • assets/style.css: UI styles (centered buttons, smaller tiles).

Database Techniques

• View (MySQL):
  • vw_keywords_by_year used by Widget 3 to compute keyword counts per year.
• Transaction (MySQL):
  • update_faculty performs atomic updates to faculty and keywords.
• Indexing (MongoDB):
  • Indexes on faculty.id (unique) and keywords.name to speed lookups/updates.
• Constraint (Neo4j):
  • Unique constraint on (:Faculty {id}) for consistent identity.
• Prepared Statements (MySQL):
  • Parameterized queries used across utilities to avoid injection.

Contributions

• Usman Asghar

  • Developed database access and integration for MySQL, MongoDB, and Neo4j.
  • Completed the functionality for Widgets 1,2 and 4
  • Added styling and fixed the layout of how the dashboard is presented
  • Added the filters on top to select university first then proffesor from that university
  • Keyword Popularity widget 4 is filtered to select only from the keywords of that proffesor
  • Made the ReadME file
  • Time Spent: 22 Hours

• Fahad Khan

  • Completed the functionality for Widgets 3,5 and 6
  • Fixed the stlying to get design of each card that holds each widget
  • Added the callback functionality to retrieve the information from each databases
  • Recorded, edited, and narrated the demo video and uploaded it to Illinois Media
  • Added the backend logic for the widgets 5 and 6 to updated that database with the changes in the data
  • Time Spent: 20 Hours

• Combined Time Spent: 42 hours