📘 Brigata — DHDK 2025

Scholarly Journal Integration & Query Engine

🎯 Project Goal

This project integrates scholarly journal metadata coming from two heterogeneous sources:

• a graph database (Blazegraph + RDF triples)
• a relational database (SQLite)

and exposes a unified Python query engine that returns results as Python objects (Journal, Category, Area). The architecture is modular, extensible, and designed to support heterogeneous data ingestion and querying.

---

📚 Data Sources

Graph Database (CSV → RDF → Blazegraph)

• Source: Directory of Open Access Journals (DOAJ)
• Contains: ISSN, title, languages, publisher, license, APC, DOAJ Seal
• Loaded via: JournalUploadHandler

Relational Database (JSON → SQLite)

• Source: Scimago Journal Rank (SJR)
• Contains: categories, areas, quartiles
• Loaded via: CategoryUploadHandler

---

🏗️ System Architecture

Overall Workflow

1. CategoryUploadHandler loads JSON into the relational database (SQLite).
2. JournalUploadHandler loads CSV into the graph database (Blazegraph).
3. CategoryQueryHandler and JournalQueryHandler retrieve data as Pandas DataFrames.
4. FullQueryEngine integrates both sources and returns Python objects.

---

🗄️ Relational Database Schema

The schema includes:

• IdentifiableEntity
• HasCategory
• HasArea

This structure supports many-to-many relationships between journals, categories, and areas.

---

🧩 Python Data Model

IdentifiableEntity

Base class for all entities with one or more identifiers.

Journal

• title
• languages
• publisher
• seal
• license
• apc
• hasCategory
• hasArea

Category

• id
• quartile

Area

• id

---

🔍 Query Engine

CategoryQueryHandler

Queries the relational database (SQLite):

• getAllCategories()
• getAllAreas()
• getCategoriesWithQuartile()
• getCategoriesAssignedToAreas()
• getAreasAssignedToCategories()
• getCategoryWithName() — returns all categories whose name partially matches the input string
• getAreaWithName() — returns all areas whose name partially matches the input string

BasicQueryEngine

Provides simple filters over the graph database:

• getAllJournals()
• getJournalsWithTitle()
• getJournalsPublishedBy()
• getJournalsWithLicense()
• getJournalsWithAPC()
• getJournalsWithDOAJSeal()
• getAllCategories()
• getAllAreas()
• getCategoriesWithQuartile()

FullQueryEngine

Provides composite queries combining graph + relational data:

• getJournalsInCategoriesWithQuartile()
• getJournalsInAreasWithLicense()
• getDiamondJournalsInAreasAndCategoriesWithQuartile()
• getJournalByName() — returns journals whose title contains the input string and that have at least one category or area whose name also contains it

---

📦 Project Structure

Brigata/
├── main.py
├── impl.py
├── laura.py
├── Yang.py
├── daniele.py
├── li.py
├── baseHandler.py
├── images/
│   ├── workflow.png
│   ├── relational_database_structure.png
│   ├── datamodel.png
├── data/
│   ├── doaj.csv
│   ├── scimago.json
└── relational.db        # auto-generated

---

⚙️ Installation

Install dependencies:

pip install pandas rdflib SPARQLWrapper sqlalchemy requests

---

🔥 Running Blazegraph

Download Blazegraph:

wget https://github.com/blazegraph/database/releases/download/BLAZEGRAPH_2_1_6_RC/blazegraph.jar

Start the server:

java -server -Xmx1g -jar blazegraph.jar

Verify at http://localhost:9999/blazegraph/.

⚠️ Update the Blazegraph endpoint URL in main.py to match your local setup. The default is http://127.0.0.1:9999/blazegraph/namespace/kb/sparql.

---

🚀 Running the Project

python main.py

This will:

• upload CSV to Blazegraph
• upload JSON to SQLite
• initialize the query engine
• run example queries

---

👥 Team Members

Name	Role	Email
Tianchi Yang	Graph Query Handler (SPARQL)	[email protected]
Yihua Li	Graph Upload Handler (RDF → Blazegraph)	[email protected]
Daniele Camagna	Relational DB + Upload Handler	[email protected]
Laura Bortoli	Query Engine (Basic & Full)	[email protected]