Configuration Guide

This guide explains how to configure the Literature Review Tool for your research needs.

Configuration File Format

The tool uses JSON configuration files. Here’s a complete example:

{
  "research_topic": "fundus image dataset",
  "keywords": "medical-imaging, machine-learning, fundus, retinal, ophthalmology",
  "initial_prisma_values": {
    "inclusion_criteria": ["review", "thesis", "journal", "book"],
    "exclusion_criteria": ["non-english", "conference"],
    "databases": ["PubMed", "CrossRef", "arXiv", "CORE", "SemanticScholar", "IEEE", "Springer", "DBLP", "Scopus"],
    "date_range": "2015-2025"
  },
  "api_keys": {
    "CORE": "your_core_api_key_here",
    "IEEE": "your_ieee_api_key_here",
    "Springer": "your_springer_api_key_here",
    "Scopus": "your_scopus_api_key_here"
  }
}

Configuration Fields

Research Topic (Optional)

"research_topic": "machine learning in healthcare"

Keywords

Keywords can be specified in two formats:

List format:

"keywords": ["machine learning", "deep learning", "neural networks"]

Comma-separated string:

"keywords": "machine learning, deep learning, neural networks"

Auto-generated keywords: If you don’t provide keywords, they’ll be generated from the research topic with a warning:

{
  "research_topic": "machine learning applications",
  "initial_prisma_values": { ... }
}

PRISMA Values

The initial_prisma_values object contains PRISMA methodology parameters:

Inclusion Criteria

Publications matching ANY inclusion criterion will be included:

"inclusion_criteria": ["review", "thesis", "journal", "book"]

Exclusion Criteria

Publications matching ANY exclusion criterion will be excluded:

"exclusion_criteria": ["non-english", "conference", "preprint"]

Databases

List of databases to search:

"databases": ["PubMed", "CrossRef", "arXiv", "CORE", "SemanticScholar", "IEEE", "Springer", "DBLP", "Scopus"]

Available databases:

Date Range

Filter publications by year:

"date_range": "2015-2025"

Format: "START_YEAR-END_YEAR"

Field-Specific Criteria (Advanced)

You can specify database-specific fields using the field:value format for more precise filtering:

{
  "research_topic": "machine learning",
  "initial_prisma_values": {
    "inclusion_criteria": [
      "type:journal-article",
      "language:english",
      "journal:nature"
    ],
    "exclusion_criteria": [
      "type:conference-paper",
      "source:arxiv",
      "language:non-english"
    ],
    "databases": ["PubMed", "CrossRef", "arXiv"],
    "date_range": "2020-2025"
  }
}

Supported Fields

Field Description Example Values
type Publication type journal-article, conference-paper, book-chapter
publication_type Alternate for type Same as type
pubtype Alternate for type Same as type
language Publication language english, spanish, french
source Database source pubmed, arxiv, crossref
journal Journal name nature, science, plos one
venue Venue name (conferences) Same as journal
authors Author names Author name to search
document_type Document type (CORE) Document classification

Database-Specific Mappings

Different databases use different field names. The tool automatically maps your criteria to the appropriate field:

PubMed/Europe PMC

CrossRef

arXiv

CORE

SemanticScholar

IEEE, Springer, DBLP, Scopus

Backward Compatibility

Criteria without field specifications default to the type field:

"inclusion_criteria": ["journal", "review"]

This is equivalent to:

"inclusion_criteria": ["type:journal", "type:review"]

API Keys

Some databases require API keys. Add them to the api_keys section:

"api_keys": {
  "CORE": "your_core_api_key",
  "IEEE": "your_ieee_api_key",
  "Springer": "your_springer_api_key",
  "Scopus": "your_scopus_api_key"
}

See Databases for information on obtaining API keys.

Example Configurations

Minimal Configuration

{
  "research_topic": "artificial intelligence",
  "initial_prisma_values": {
    "inclusion_criteria": ["journal"],
    "exclusion_criteria": ["conference"],
    "databases": ["PubMed", "CrossRef"],
    "date_range": "2020-2024"
  }
}

Comprehensive Configuration

{
  "research_topic": "deep learning for medical imaging",
  "keywords": ["deep learning", "medical imaging", "neural networks", "radiology", "diagnosis"],
  "initial_prisma_values": {
    "inclusion_criteria": [
      "type:journal-article",
      "type:review",
      "language:english"
    ],
    "exclusion_criteria": [
      "type:conference-paper",
      "type:preprint",
      "language:non-english"
    ],
    "databases": ["PubMed", "CrossRef", "IEEE", "Springer", "Scopus"],
    "date_range": "2018-2024"
  },
  "api_keys": {
    "IEEE": "your_ieee_api_key",
    "Springer": "your_springer_api_key",
    "Scopus": "your_scopus_api_key"
  }
}

Computer Science Research

{
  "research_topic": "blockchain security",
  "keywords": ["blockchain", "security", "cryptography", "distributed systems"],
  "initial_prisma_values": {
    "inclusion_criteria": [
      "type:journal-article",
      "language:english"
    ],
    "exclusion_criteria": [
      "type:conference-paper"
    ],
    "databases": ["arXiv", "DBLP", "IEEE", "SemanticScholar"],
    "date_range": "2019-2024"
  },
  "api_keys": {
    "IEEE": "your_ieee_api_key"
  }
}

Configuration Validation

The tool validates your configuration file and provides helpful error messages:

Best Practices

  1. Start simple: Use a basic configuration and refine iteratively
  2. Test with small page sizes: Verify your configuration works before large searches
  3. Version control: Keep your configurations in version control
  4. Document your choices: Add comments in a separate README
  5. Review exclusions: Check exclusion reasons to refine criteria
  6. Use meaningful names: Save configs with descriptive filenames (e.g., ml_healthcare_2024.json)

Next Steps