project_description.md

Project Statement

A web application enabling users to search in scientific articles from a list of key words

Scientific Article

• Title
• Abstract
• Authors (1,*)
• Institutions (1,*)
• Keywords (1,*)
• References (1,*)
• PDF URL
• Publication date
• Textual content

Features

Authentication

• Client types: Admin, Moderator, User
• Email/Password Authentication
• Google Authentication (Optional)

User Functionalities

• Search by [title, keywords, authors, content] (One Query / Advanced search)
• Display search results sorted by publication date (recent first)
  • Filter search results by [keywords, authors, institutions, publication dates interval]
  • Display article details [text format, PDF]
  • Save article in favorite list
• Consult favorite lists

Admin Functionalities

• CRUD on moderators by Email/Username (First Name, Last Name)
• Upload scientific articles through Local files
• Upload scientific articles through URL (Single / Drive Directory)
• Handle file duplicates

PDF Extractor

• Handle [1 column, 2 columns] PDF articles
• Extract scientific articles characteristics from PDF text (Hard code / Regex / AI)
• Extracted article sent in a search index to ElasticSearch

Moderator Functionalities

• Consult uploaded scientific articles by the admin
• Correct eventiual errors on PDF extraction
• Update scientific article index in ElasticSearch
• Delete scientific article in case to many errors

Elastic Search

• Index a scientific article
• Search in indexed articles from user query

Technical Specifications

User Experience

• User-friendly responsive interface

Technologies

• Backend with python (FastAPI)
• Frontend with ReactJS
• ElasticSearch for scientific articles search and indexing
• SQL SGBD for non-article data

Design Pattern

• MVC design

Code documentation

• Auto-generated documentation

Tests

• 3 Unit tests
• 1 Automated functional test (Selenium) for "Articles Search" functionality

Deployment

• Docker deployment