← Back to projects

FindJournal

A tool that helps Algerian researchers find the best-matching journals from the DGRSDT list. Uses a hybrid TF-IDF and semantic embedding algorithm to match research abstracts against 13,600+ indexed journals.

React Typescript Flask Python Tailwind CSS SQLite Docker Text vectorization
Live Demo → GitHub

Overview: FindJournal is a specialized research tool designed to help Algerian researchers identify suitable academic journals from the DGRSDT (Category A) list (~13,600 journals). Researchers can paste their abstracts and receive ranked journal suggestions with similarity scores.

Key Features:

  • Abstract-to-journal matching using hybrid TF-IDF + semantic embeddings
  • Journal browsing with filtering by quartile (Q1-Q4) and minimum SJR
  • Rich metadata per journal: SJR, quartile, H-index, subject categories, research areas
  • Open Access tracking (Diamond OA, Gold OA, subscription-based)

Matching Algorithm: The system combines two complementary approaches:

  1. TF-IDF Matching (60% weight) : keyword-level matching against journal titles, Scimago categories, and OpenAlex topics
  2. Sentence Embeddings (40% weight) : semantic similarity using the all-MiniLM-L6-v2 model to capture conceptual relationships

Data Pipeline: Journals are enriched from multiple sources (DGRSDT official PDF, Scimago quality metrics, and OpenAlex academic topics) then indexed into an SQLite database with pre-computed embeddings.