Skip to content

CONFIGURATION.md

Latest commit

 

History

History
161 lines (124 loc) · 3.59 KB

CONFIGURATION.md

File metadata and controls

161 lines (124 loc) · 3.59 KB
Raw

Configuration Guide

This document explains how to configure the PRISMA workflow automation project using the config.json file.

Overview

The config.json file is the central configuration for the PRISMA workflow automation project. It defines the search strategy, database settings, and output preferences.

Configuration Structure

The configuration file has the following main sections:

{
  "project": { ... },
  "search_terms": { ... },
  "boolean_operators": { ... },
  "database_defaults": { ... },
  "databases": {
    "pubmed": { ... },
    "scopus": { ... },
    "embase": { ... },
    "cochrane": { ... }
  },
  "screening": { ... },
  "data_extraction": { ... },
  "output": { ... }
}

Project Section

The project section contains metadata about the systematic review:

"project": {
  "title": "Example Systematic Review",
  "description": "A systematic review of lung cancer treatments",
  "authors": ["Researcher A", "Researcher B"],
  "date": "2025-04-01"
}

Search Terms

The search_terms section follows the PICO framework:

"search_terms": {
  "population": ["lung cancer", "NSCLC", "non-small cell lung cancer"],
  "intervention": ["immunotherapy", "pembrolizumab", "nivolumab"],
  "comparison": ["chemotherapy", "radiation therapy"],
  "outcome": ["survival", "progression-free survival"],
  "study_design": ["randomized controlled trial", "systematic review"]
}

Boolean Operators

The boolean_operators section defines how search terms are combined:

"boolean_operators": {
  "within_category": "OR",
  "between_categories": "AND"
}

Database Defaults

The database_defaults section contains settings that apply to all databases by default:

"database_defaults": {
  "enabled": true,
  "fields": ["title", "abstract", "keywords"],
  "date_range": {
    "start_year": 2000,
    "end_year": 2025
  },
  "languages": ["english"],
  "max_results": 1000
}

Database-Specific Settings

Each database has its own section with specific settings that override the defaults:

"databases": {
  "scopus": {
    "enabled": true,
    "fields": ["title", "abstract", "keywords"],
    "subject_areas": ["medicine", "health sciences"],
    "document_types": ["article", "review"]
  }
}

Screening Criteria

The screening section defines inclusion and exclusion criteria:

"screening": {
  "title_abstract": {
    "inclusion": ["human studies", "adult patients"],
    "exclusion": ["case reports", "animal studies"]
  },
  "full_text": {
    "inclusion": ["sample size > 100", "follow-up > 1 year"],
    "exclusion": ["non-English full text", "conference abstracts"]
  }
}

Data Extraction

The data_extraction section defines fields to extract from included studies:

"data_extraction": {
  "fields": [
    "study_design",
    "sample_size",
    "intervention_details",
    "outcome_measures",
    "results",
    "conclusions"
  ]
}

Output Settings

The output section defines how results are exported:

"output": {
  "format": "bibtex",
  "deduplicate": true,
  "export_path": "output/results"
}

Example Configuration

See the project's config.json file for a complete example configuration.

Modifying the Configuration

You can modify the configuration in several ways:

  1. Directly edit the config.json file
  2. Use the CLI command: python -m src.python.cli config edit
  3. Programmatically via the ConfigLoader class

Configuration Validation

The ConfigLoader class validates the configuration when loading it. If there are errors, it will raise exceptions with helpful error messages.