helpContent <- function() { ## biblioAI ---- biblioAI <- "

🧠 Biblio AI: AI-Powered Bibliometric Analysis

Starting from version 5.0, Biblioshiny introduces Biblio AI, a new suite of features powered by Google's Gemini models. This integration allows users to receive automatic interpretations, critical insights, and narrative summaries of their bibliometric results – directly within the platform.

Note: Biblio AI requires a Chrome-based browser (such as Google Chrome or Microsoft Edge) installed on your computer to work correctly.

✨ What does Biblio AI do?

Biblio AI enhances the core analytical modules of Biblioshiny by providing contextual, AI-generated commentary on several results, such as:

In each of these sections, users can activate the Biblio AI panel to access dynamic text explanations, perfect for use in scientific writing, presentations, or reporting.

🔧 How to enable Biblio AI?

To enable Biblio AI, follow these simple steps:

  1. Register at Google AI Studio (free access available).
  2. Generate an API Key enabled for Gemini model access (Free Tier supported).
  3. Enter your API Key in the Settings section of Biblioshiny.

The interface will guide you through the secure and local setup. Your API key is used only on your device to interact with the AI model.

🎯 Why use Biblio AI?

" ## Info ---- info <- "

📚 Supported Bibliographic Databases and Suggested File Formats

Biblioshiny imports and analyzes collections exported from the following bibliographic databases:

Web of Science, Scopus, and OpenAlex allow users to export the complete set of metadata, making it possible to perform all analyses implemented in Biblioshiny.

Some other databases, such as Dimensions, PubMed, and Cochrane Library, provide only a limited set of metadata. This may impose restrictions on the range of analyses that can be conducted using those datasets.

The following table reports, for each supported database:

Databases, available metadata and suggested file formats

Source Format Exported metadata Recommended Format
Web of Science
  • 'BibTeX'
  • 'Plaintext'
  • 'EndNote Desktop'
  • All except Cited references
  • All
  • All
 Plaintext
Scopus
  • 'BibTeX'
  • 'CSV'
  • All except Cited references
  • All
 CSV
OpenAlex
  • 'CSV / Excel' (web export)
  • 'API'
  • User-selected columns; no cited references (see OpenAlex section below)
  • All metadata, including cited references
 API
Dimensions
  • 'Excel'
  • 'API' (no longer supported — Dimensions has moved to a closed commercial access model)
  • All but with limited reference info (only Ids)
  • —
 Excel
The Lens
  • 'CSV export'
  • All but with limited reference info (only Ids)
 CSV
PubMed
  • 'PubMed export'
  • 'API'
  • Document, Authors, Content info
  • Document, Authors, Content info
 PubMed export
Cochrane Library
  • 'Plaintext'
  • Document, Authors, Content info
 Plaintext

🅾️ Importing from OpenAlex

OpenAlex can be imported into Biblioshiny in two ways:

⚠️ Which columns to select in the web export?

When you export a CSV/Excel file from the OpenAlex website, make sure to include at least the required fields. If any of them is missing, Biblioshiny cannot build the bibliographic collection and the import is stopped with an explanatory message.

Level OpenAlex export columns Why they are needed
Required
(minimum set)		 Work ID, Title, Author, Source, Year Unique identifier (de-duplication, short reference), document, author and source identification, and publication year. Without these the collection cannot be created.
Recommended Citation count, DOI, Author IDs, ORCID, Institution & Institution IDs, Country, Language, Abstract, Corresponding author & Corresponding institution, Keywords / Concepts Enable impact analysis, author disambiguation, affiliation and country analysis, language filtering, text/content analysis (abstract, keywords) and corresponding-author analysis. The import still works without them, but the related analyses are limited or unavailable. A warning lists the missing fields.

🔗 Cited references: enrich your collection via the API

The OpenAlex web CSV/Excel export does not include the cited references (and does not include the document type). Therefore citation-based analyses — such as Historical Direct Citation Network, Co-citation and Bibliographic Coupling — are not available from a plain web export.

To obtain them you can:

Note: reference enrichment via the API requires an internet connection and may take some time for large collections.

" ## Team Publications ---- publications <- "

📖 Main Authors' References (Bibliometrics)

" ## SAAS ---- saas <- "

SAAS Workflow

Search - Appraisal - Analysis - Synthesis

SEARCH Data Collection • API Integration • Database Export • Query Design • PubMed, OpenAlex • Web of Science APPRAISAL Quality Assessment • Data Filtering • Quality Control • Duplicate Removal • Field Validation • Data Cleaning ANALYSIS Deep Investigation • Bibliometric Analysis • Network Analysis • Citation Analysis • Conceptual Structure • Content Analysis ∑ SYNTHESIS Results & Reporting • Report Generation • Key Insights • Visualization • AI Summary • Publication Output Integrated Tools & Features Science Mapping ✓ Co-citation Analysis ✓ Bibliographic Coupling ✓ Co-word Analysis ✓ Thematic Evolution ✓ Trend Topics ✓ Three Fields Plot ✓ Historiograph ✓ Factorial Analysis Network Analysis ✓ Collaboration Networks ✓ Country Collaboration ✓ Institution Networks ✓ Author Networks ✓ Citation Networks ✓ Co-occurrence Networks ✓ Community Detection ✓ Centrality Measures AI-Enhanced Analysis ✓ Biblio AI Assistant ✓ Document Summarization ✓ PDF Content Analysis ✓ Citation Context Extraction ✓ Reference Matching ✓ Readability Metrics ✓ Topic Modeling ✓ Natural Language Processing Visualization ✓ Interactive Plots ✓ Network Graphs ✓ Thematic Maps ✓ Word Clouds ✓ Sankey Diagrams ✓ Heatmaps ✓ Time Series Plots ✓ Export Options Iterative Refinement

About the SAAS Workflow

The SAAS workflow represents the comprehensive process of bibliometrix and biblioshiny for conducting scientific bibliometric analysis. Each phase is designed to ensure methodological rigor and reliable results:

The iterative cycle allows continuous refinement of the analysis by returning to previous phases based on obtained results.

SAAS Workflow developed by:
Massimo Aria & Corrado Cuccurullo
University of Naples Federico II, Italy

Aria, M., & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959-975.

Aria, M., Cuccurullo, C., D'Aniello, L., & Spano, M. (2026). Biblioshiny and the SAAS Workflow: An integrated framework for transparent and reproducible science mapping. Journal of Informetrics, 20(3), 101837.

" ## Filters ---- filters <- "

🔍 Filters: Refining Your Bibliometric Collection

The Filters module provides a comprehensive set of tools to refine and subset your bibliographic collection based on multiple metadata criteria. By applying filters, you can focus your analysis on specific document types, time periods, geographic regions, journals, or citation thresholds—enabling more targeted and meaningful bibliometric insights.

Filters are organized into four thematic panels, each addressing different aspects of bibliographic metadata. At the top of the page, a real-time summary displays how many documents, sources, and authors remain after applying your filter selections.


📊 Real-Time Filter Summary

Located at the top of the Filters page, this summary updates dynamically as you adjust filter settings:

These indicators help you assess the impact of your filters before applying them, ensuring your subset maintains sufficient size for robust analysis.


1️⃣ General Filters

The General panel provides fundamental filters applicable to most bibliometric collections:

Document Type

Language

Publication Year


2️⃣ Journal (J) Filters

The Journal panel enables filtering based on publication venues, journal rankings, or Bradford's Law zones:

Upload a List of Journals

Journal of Informetrics
Scientometrics
Journal of the Association for Information Science and Technology
Research Policy

Upload a Journal Ranking List

Journal,Quartile
Journal of Informetrics,Q1
Scientometrics,Q1
Library Quarterly,Q2
Online Information Review,Q3

with ISSN columns:

Journal,ISSN,eISSN,Quartile
Journal of Informetrics,1751-1577,1875-5879,Q1
Scientometrics,0138-9130,1588-2861,Q1

Source by Bradford Law Zones


3️⃣ Author's Country (AU) Filters

The Author's Country panel enables geographic filtering based on author affiliations:

Region

Country


4️⃣ Documents (DOC) Filters

The Documents panel provides citation-based filters with interactive histograms:

Total Citations

Total Citations per Year


🎛️ Filter Workflow

Follow these steps to apply filters effectively:

  1. Review the initial collection: Check the summary counts (Documents, Sources, Authors) before applying filters.
  2. Select filter criteria: Adjust settings across the four panels based on your research objectives.
  3. Monitor real-time updates: The summary at the top updates dynamically as you change selections, showing how many documents remain.
  4. Click 'Apply': Once satisfied with your selections, click the blue Apply button to activate the filters.
  5. Verify results: Check the updated summary to ensure your filters produced the expected subset size.
  6. Proceed to analysis: Navigate to other modules (Overview, Sources, Authors, etc.) to analyze the filtered collection.
  7. Reset if needed: Click the Reset button to clear all filters and restore the original dataset.

💡 Best Practices


⚠️ Important Considerations


🔍 Use Case Examples

Example 1: Analyzing Top-Tier Recent Research

Example 2: National Research Assessment

Example 3: Historical Foundational Literature


📚 References

Bradford, S. C. (1934). Sources of information on specific subjects. Engineering, 137, 85–86.

Aria, M., & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Garfield, E. (2009). From the science of science to Scientometrics: visualizing the history of science with HistCite software. Journal of Informetrics, 3(3), 173–179. https://doi.org/10.1016/j.joi.2009.03.009

" ## Author Profile ---- authorProfile <- "

👤 Author Profile Overview

The Author Profile page provides a dual-perspective bibliometric overview of each author included in the collection:

🔹 Global Profile

The Global Profile presents the author's complete scientific output, based on metadata retrieved from OpenAlex via the openalexR R package. This profile includes all publications authored by the researcher, regardless of whether they are part of the current collection.

Main features of the Global Profile include:

Data Source: OpenAlex API (via openalexR)
Unique Identifier: OpenAlex Author ID (e.g., A5014455237)

🔸 Local Profile

The Local Profile focuses exclusively on the subset of the author's publications that are included in the user-defined collection currently under analysis in the project.

Main features of the Local Profile include:

This local profile helps contextualize the author's role and impact within the specific research topic or dataset under investigation.

🔄 Interpretation and Use

The Global Profile offers a broad, external view of the author's overall scholarly influence, while the Local Profile highlights their specific relevance within the current study.

This dual visualization is particularly useful for:

📚 References

Priem, J. et al. (2022). OpenAlex: A fully-open index of scholarly works, authors, venues, institutions, and concepts. Retrieved from https://openalex.org

Aria, M., Le, T., Cuccurullo, C., Belfiore, A., & Choe, J. (2024). openalexR: An R-Tool for Collecting Bibliometric Data from OpenAlex. R Journal, 15(4), 167–180. https://doi.org/10.32614/RJ-2023-089

Aria, M. et al. (2023). openalexR: An R package for programmatic access to OpenAlex metadata. CRAN. Retrieved from https://cran.r-project.org/package=openalexR

Hirsch, J.E. (2005). An index to quantify an individual's scientific research output. Proceedings of the National Academy of Sciences, 102(46), 16569–16572. https://doi.org/10.1073/pnas.0507655102

" # referenceMatching <- "

🔗 Reference Matching: Algorithm and Usage

The Reference Matching module implements an advanced algorithm to identify and link cited references within a bibliometric collection to the actual documents present in the dataset. This process enables accurate citation network analysis, co-citation studies, and identification of highly-cited works within the collection.

🔬 Algorithm Overview

The reference matching algorithm follows a multi-step procedure designed to maximize accuracy while handling noisy and incomplete bibliographic data:

  1. Reference Extraction: Cited references are parsed from the reference list (CR field) of each document in the collection. Each reference string is decomposed into structured components: first author surname, publication year, journal/source, volume, page, and DOI (when available).
  2. Data Normalization: Both references and documents undergo extensive normalization to reduce variability:
    • Author names are standardized (e.g., removing accents, abbreviations, and middle initials)
    • Journal titles are normalized using abbreviation lookup tables and string similarity methods
    • Years, volumes, and page numbers are cleaned and formatted uniformly
  3. Blocking Strategy: To improve computational efficiency, references are grouped into blocks based on first author surname and publication year. Only references and documents within the same block are compared, reducing the search space significantly.
  4. Similarity Computation: For each reference-document pair within a block, a matching score is calculated using a weighted combination of similarity measures:
    • DOI matching (if available): exact match = 100% confidence
    • First author similarity: string distance (Jaro-Winkler or Levenshtein)
    • Year match: exact or within ±1 year tolerance
    • Journal/source similarity: string distance between normalized titles
    • Volume and page matching: exact or fuzzy comparison
  5. Threshold-Based Assignment: A reference is matched to a document if the combined similarity score exceeds a predefined threshold (typically 0.85–0.95). The threshold can be adjusted by the user to balance precision and recall.
  6. Ambiguity Resolution: In cases where a reference matches multiple documents, the algorithm selects the candidate with the highest similarity score. If scores are nearly identical, the match is flagged for manual review.

💡 Usage in Biblioshiny

To perform reference matching in Biblioshiny, follow these steps:

  1. Load your bibliographic collection: Ensure that your dataset includes the CR (Cited References) field, which is available in full exports from Web of Science, Scopus, and OpenAlex.
  2. Navigate to the Reference Matching module: Access the module from the Analysis menu or the Citation Network section.
  3. Configure matching parameters:
    • Similarity threshold: Adjust the matching threshold to control precision (higher values = stricter matching, fewer false positives).
    • Normalization options: Enable or disable specific normalization rules (e.g., journal abbreviation matching, fuzzy year tolerance).
  4. Run the algorithm: Click Start Matching to initiate the process. Depending on the collection size, this may take several minutes.
  5. Review results: The output includes:
    • A summary table of matched and unmatched references
    • A list of ambiguous matches for manual inspection
    • Network visualization options (e.g., co-citation network, historiograph)
  6. Export matched data: The matched citation network can be exported for further analysis in external tools (e.g., Gephi, Pajek) or used directly in Biblioshiny for advanced network analysis.

⚙️ Key Parameters and Options

📊 Applications

Reference matching is essential for several bibliometric analyses:

⚠️ Important Notes

📚 References

Aria, M., D'Aniello, L., & Spano, M. (2026). A multi-phase reference matching algorithm for bibliometric analysis: design, implementation, and evaluation. Scientometrics. https://doi.org/10.1007/s11192-026-05763-2

Aria, M., & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Garfield, E. (1979). Citation indexing: Its theory and application in science, technology, and humanities. New York: Wiley.

Small, H. (1973). Co-citation in the scientific literature: A new measure of the relationship between two documents. Journal of the American Society for Information Science, 24(4), 265–269. https://doi.org/10.1002/asi.4630240406

" # importOrLoad <- "

📥 Import or Load: Building Your Bibliometric Collection

The Import or Load module is the starting point for any bibliometric analysis in Biblioshiny. This section allows users to build their bibliographic collection by either importing raw files from supported databases or loading pre-processed bibliometrix files saved in previous sessions.


📂 Three Import Options

Biblioshiny offers three flexible ways to create or load a bibliographic collection:

1. Import Raw File(s)

Import bibliographic data directly from supported databases in their native export formats.

Supported Databases:

Import Process:

  1. Click Browse to select one or more raw export files from your computer
  2. Biblioshiny automatically detects the database format and parses the metadata
  3. The system converts the raw data into a standardized bibliometrix data frame
  4. A Conversion Results summary displays the number of documents successfully imported
  5. View a preview table showing key metadata fields (DOI, Authors, Title, Journal, etc.)

Important Notes:

2. Load Bibliometrix File(s)

Resume work on a previously processed collection by loading .rdata or .xlsx files generated by Biblioshiny or the bibliometrix R package.

Use Cases:

Supported Formats:

3. Use a Sample Collection

Perfect for testing and learning Biblioshiny's features without preparing your own data.


🔍 Post-Import Features

After successfully importing or loading a collection, you can:


💾 Exporting Collections

Once your collection is loaded, you can export it in multiple formats:


⚠️ Best Practices


📚 References

Aria, M., & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

" ## API ---- api <- "

🌐 API: Gathering Data from Bibliographic Databases

The API module enables users to retrieve bibliographic data directly from supported databases using their Application Programming Interfaces (APIs). This feature eliminates the need for manual file downloads and allows for dynamic, reproducible data collection based on custom search queries.


🔌 Supported APIs

Biblioshiny currently integrates with the following bibliographic database APIs:

Each database API has specific query capabilities, rate limits, and metadata coverage. Biblioshiny automatically handles API authentication, pagination, and data formatting.


🔧 How to Use the API Module

The API workflow consists of three main steps:

Step 1: Configure API Query

Define your search parameters using the interactive query builder:

Example PubMed Query:

Search Terms: bibliometrics[Title/Abstract] AND scientometrics[Title/Abstract]
Start Year: 2010
End Year: 2025

Step 2: Test and Validate Query

Before downloading the full dataset, it's essential to validate your query:

  1. Click Try the Query to send a test request to the API
  2. The Query Translation panel shows the exact API call being executed
  3. The Documents Returned section displays the number of results matching your query (e.g., 'PubMed API returns 898 documents')
  4. If the result count is too high or too low, refine your search terms and test again

Important: Some APIs have result limits (e.g., PubMed allows up to 10,000 records per query). If your query exceeds the limit, consider narrowing the time range or adding more specific keywords.

Step 3: Download Data

Once satisfied with the query validation:

  1. Adjust the Total Documents to Download slider (range: 1 to the maximum returned by the query)
  2. Click 2. Download to retrieve the bibliographic metadata
  3. The API will fetch records in batches, respecting rate limits automatically
  4. Progress is displayed in real-time (e.g., 'Fetching records 1-500...')
  5. Downloaded data is automatically converted into the bibliometrix format

Step 4: Export and Analyze

After successful data retrieval:


🔑 API Authentication

Some databases require API keys or institutional access:

To configure API keys in Biblioshiny, navigate to Settings and enter your credentials in the API Configuration section.

📊 Advantages of API-Based Data Collection


⚠️ Important Considerations


📚 References

Aria, M., Le, T., Cuccurullo, C., Belfiore, A., & Choe, J. (2023). openalexR: An R-Tool for Collecting Bibliometric Data from OpenAlex. R Journal, 15(4). https://doi.org/10.32614/rj-2023-089

NCBI Resource Coordinators. (2018). Database resources of the National Center for Biotechnology Information. Nucleic Acids Research, 46(D1), D8–D13. https://doi.org/10.1093/nar/gkx1095

" ## Merge Collections ---- mergeCollections <- "

🔀 Merge Collections: Combining Data from Multiple Sources

The Merge Collections module allows users to combine bibliographic datasets from different databases (Web of Science, Scopus, OpenAlex, PubMed, etc.) into a single unified collection. This functionality is essential for comprehensive literature reviews, cross-database validation, and maximizing metadata coverage by leveraging the strengths of multiple sources.


🎯 Why Merge Collections?


🔧 How to Merge Collections

The merge process in Biblioshiny is straightforward:

  1. Navigate to Merge Collections: Select Data > Merge Collections from the main menu
  2. Select Collection Files: Click Browse and select two or more bibliometrix files to merge:
    • Supported formats: .rdata, .xlsx
    • Files can originate from different databases (e.g., wos_collection.rdata + scopus_collection.xlsx)
    • Files must be valid bibliometrix data frames (created via Import or Load, or R package)
  3. Configure Merge Options:
    • Remove Duplicates: Enable (recommended) to automatically detect and remove duplicate records
    • Verbose Output: Enable to display detailed information about the merge process and duplicates removed
  4. Click Start: The merge algorithm combines the collections, standardizes metadata fields, and removes duplicates
  5. Review Results: A summary displays the total number of documents and how many duplicates were removed
  6. Export Merged Collection: Save the unified dataset for future analysis

🔬 Merge Algorithm Overview

The merge process follows a sophisticated multi-stage algorithm implemented by the mergeDbSources() function:

Stage 1: Database Identification and Ordering

Stage 2: Field Alignment

Stage 3: Duplicate Detection

Duplicates are identified using a two-step matching strategy:

Step 3.1: DOI-Based Matching

Step 3.2: Title-Year Matching

Stage 4: Author Name Standardization

When merging collections from multiple databases, author name formats are standardized:

Stage 5: Metadata Integration


📊 Merge Statistics and Validation

After merging, the system provides detailed statistics:

Example Output:

Merging 3 collections:
- WoS: 1,500 documents
- Scopus: 1,800 documents
- OpenAlex: 2,000 documents
Total: 5,300 documents

Removing duplicates...
- 450 duplicates removed by DOI
- 320 duplicates removed by title-year match

Final collection: 4,530 documents

📌 Best Practices


⚠️ Important Considerations


🔍 Example Use Cases


📚 References

Aria, M., & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Visser, M., van Eck, N. J., & Waltman, L. (2021). Large-scale comparison of bibliographic data sources: Scopus, Web of Science, Dimensions, Crossref, and Microsoft Academic. Quantitative Science Studies, 2(1), 20–41. https://doi.org/10.1162/qss_a_00112

Martín-Martín, A., Thelwall, M., Orduna-Malea, E., & Delgado López-Cózar, E. (2021). Google Scholar, Microsoft Academic, Scopus, Dimensions, Web of Science, and OpenCitations' COCI: a multidisciplinary comparison of coverage via citations. Scientometrics, 126(1), 871–906. https://doi.org/10.1007/s11192-020-03690-4

" ## Main Information ---- mainInformation <- "

📊 Main Information: Overview of Your Bibliometric Collection

The Main Information page provides a comprehensive, at-a-glance summary of the key bibliometric indicators for your collection. This dashboard-style interface displays 12 core metrics organized into visual cards, allowing you to quickly assess the scope, composition, and characteristics of your dataset.

This section is the ideal starting point for understanding your collection before diving into more detailed analyses. It answers fundamental questions such as: How large is my dataset? What is the temporal coverage? How collaborative is the research? How impactful are the documents?


📈 Core Metrics Explained

The Main Information dashboard displays the following indicators:

1. Timespan

2. Sources

3. Documents

4. Annual Growth Rate

5. Authors

6. Authors of Single-Authored Docs

7. International Co-Authorship

8. Co-Authors per Document

9. Author's Keywords (DE)

10. References

11. Document Average Age

12. Average Citations per Document


🧠 Biblio AI Integration

If Biblio AI is enabled, you can click the Biblio AI tab to receive an automated narrative summary of these indicators. The AI-generated text provides contextualized interpretations, highlights notable patterns, and offers insights suitable for inclusion in research reports or presentations.

Example AI-generated insights:


📋 Viewing Options

The Main Information page offers three viewing modes via tabs at the top:


💡 How to Use Main Information

This section is designed for multiple purposes:


📌 Best Practices


⚠️ Important Considerations


🔍 Next Steps

After reviewing the Main Information dashboard, proceed to more detailed analyses:


📚 References

Aria, M., & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Zupic, I., & Čater, T. (2015). Bibliometric methods in management and organization. Organizational Research Methods, 18(3), 429–472. https://doi.org/10.1177/1094428114562629

" ## Life Cycle ---- lifeCycle <- "

📈 Life Cycle of Scientific Production: Modeling Research Topic Evolution

The Life Cycle of Scientific Production module implements a logistic growth model to analyze the temporal dynamics of research topics. This approach, grounded in the theory of scientific paradigms and innovation diffusion, allows researchers to identify the current developmental stage of a field, predict future trends, and estimate when a topic will reach maturity or saturation.

By fitting a logistic curve to the annual publication counts in your collection, this analysis reveals whether a research area is in its emergence phase, rapid growth phase, maturity phase, or decline phase.


📐 The Logistic Growth Model

The life cycle analysis is based on the logistic growth function, which models how the cumulative number of publications evolves over time:

Formula:

P(t) = K / (1 + exp(-b(t - t₀)))

Where:

The annual publication rate is derived as the first derivative of P(t), producing a bell-shaped curve that peaks at the inflection point and gradually declines as the topic approaches saturation.


🔬 Model Overview: Key Parameters

The Model Overview section displays four fundamental indicators derived from the fitted logistic model:

1. Saturation (K)

2. Peak Year (Tm)

3. Peak Annual

4. Growth Duration (Δt)


✅ Model Fit Quality

The Model Fit Quality section assesses how well the logistic curve fits the observed publication data using four statistical metrics:

1. R² (Coefficient of Determination)

2. RMSE (Root Mean Squared Error)

3. AIC (Akaike Information Criterion)

4. BIC (Bayesian Information Criterion)

Overall Assessment: Biblioshiny automatically classifies model fit as Excellent, Good, or Poor based primarily on R² values. An 'Excellent' fit (R² > 0.90) validates the use of logistic growth assumptions for forecasting.


📍 Current Status

This section provides a snapshot of the topic's present state relative to its life cycle trajectory:

Example Interpretation: If Progress to Saturation = 10.0%, the topic is in the rapid growth phase, with 90% of its publication potential still ahead. This signals a promising emerging field attracting increasing scholarly attention.


🏁 Milestone Years

The Milestone Years section predicts when the topic will reach specific saturation thresholds:

Example:

10% of K: 2021.0
50% of K: 2029.3 (+9 years)
90% of K: 2037.6 (+18 years)
99% of K: 2046.7 (+27 years)

This indicates the topic emerged around 2021, will peak in 2029, and approach saturation by 2038, with a full life cycle spanning approximately 25 years.

The system also classifies the topic's current phase (e.g., 'rapid growth phase' if between 10-50% of K) to aid interpretation.


🚀 Forecast

The Forecast section projects future publication output based on the fitted logistic model:

Example:

Projection for 2025: 2183 cumulative (436 annual)
Projection for 2030: 4898 cumulative (587 annual)

This suggests the topic will grow from ~900 publications (current) to over 4,800 by 2030, with annual output peaking around 587 publications per year.

Important: Forecasts assume the logistic model remains valid (no disruptive events, paradigm shifts, or external shocks). Long-term forecasts (>10 years) should be interpreted with caution.


📊 Visualizations

The Plot tab provides two complementary graphs:

1. Life Cycle - Annual Publications

Interpretation: This bell-shaped curve shows how publication activity rises, peaks, and eventually declines. The shape reveals the topic's maturity:

2. Cumulative Growth Curve

Interpretation: This S-shaped curve illustrates the topic's total knowledge accumulation over time. The curve's position and steepness reveal:


🧠 Biblio AI Integration

The Biblio AI tab allows you to generate AI-powered narrative interpretations of the life cycle analysis. Key features include:

Example Prompt Enhancement:

The analysis was performed on a collection downloaded from WOS focusing on machine learning applications in healthcare from 1990-2020.

This contextual information helps Biblio AI produce more accurate and domain-relevant interpretations.

💡 Use Cases


📌 Best Practices


⚠️ Important Considerations


🔍 Interpreting Fit Quality Issues

If your model shows poor fit (R² < 0.70), consider these potential causes:

Solution: Try narrowing your collection (e.g., focusing on a specific subtopic or time range) or exploring alternative growth models.


📚 References

Aria, M., Misuraca, M., & Spano, M. (2020). Mapping the evolution of social research and data science on 30 years of Social Indicators Research. Social Indicators Research, 149, 803–831. https://doi.org/10.1007/s11205-020-02281-3

Bettencourt, L. M., Kaiser, D. I., & Kaur, J. (2009). Scientific discovery and topological transitions in collaboration networks. Journal of Informetrics, 3(3), 210–221. https://doi.org/10.1016/j.joi.2009.03.001

Rogers, E. M. (2003). Diffusion of Innovations (5th ed.). New York: Free Press.

Small, H., & Upham, S. P. (2009). Citation structure of an emerging research area on the verge of application. Scientometrics, 79(2), 365–375. https://doi.org/10.1007/s11192-009-0424-0

Wang, Q. (2018). A bibliometric model for identifying emerging research topics. Journal of the Association for Information Science and Technology, 69(2), 290–304. https://doi.org/10.1002/asi.23930

" ## Three Field Plot ---- threeFieldPlot <- "

🔀 Three-Field Plot

The Three-Field Plot is an advanced visualization tool that reveals the relationships among three distinct bibliographic dimensions through an interactive Sankey diagram. This plot enables researchers to explore the complex connections between different metadata fields, making it particularly useful for understanding how research topics, authors, sources, and references are interconnected within a scientific domain.

🎯 Purpose and Application

The Three-Field Plot serves multiple analytical purposes:

📊 How It Works

The visualization consists of three vertical columns representing different bibliographic fields:

The width of each flow (colored band) is proportional to the frequency of co-occurrence between elements. Thicker flows indicate stronger associations, while thinner ones represent weaker connections.

⚙️ Configuration Options

The Options panel allows you to customize the plot:

💡 Common Field Combinations

Some particularly insightful field combinations include:

🔍 Interpretation Guidelines

📌 Best Practices

⚠️ Limitations

🤖 Biblio AI Integration

When Biblio AI is enabled, you can generate automatic interpretations of the Three-Field Plot. The AI will:

📚 Key References

Aria, M., & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Chen, C. (2017). Science Mapping: A Systematic Review of the Literature. Journal of Data and Information Science, 2(2), 1–40. https://doi.org/10.1515/jdis-2017-0006

" ## Content Analysis ---- contentAnalysis <- "

📄 Scientific Article Content Analysis

Content Analysis is a specialized feature in Biblioshiny that enables researchers to perform deep, AI-enhanced analysis of individual scientific articles in PDF format. This tool goes beyond traditional bibliometric analysis by examining the full text of documents, extracting citations with their surrounding context, and revealing patterns in how research is cited and discussed within scholarly narratives.

This menu integrates in Biblioshiny the functions included in the R library contentanalysis by Aria and Cuccurullo (https://cran.r-project.org/package=contentanalysis). The module is built on the bibliometrix ecosystem and integrates advanced text mining capabilities to support:


🎯 Purpose and Applications

Content Analysis is particularly valuable for:


📥 Step 1: Import PDF File

The analysis begins by uploading a scientific article in PDF format. The system supports both single-column and multi-column layouts (specify the number of columns for accurate text extraction).

Citation Format Detection: The tool uses AI-enhanced extraction to identify citations in multiple formats:

AI-Enhanced Extraction: When enabled, this feature uses advanced AI models to improve citation detection accuracy, particularly useful for:

Note: AI-enhanced extraction requires a configured Google Gemini API key in Settings. See the Biblio AI help section for setup instructions.


⚙️ Step 2: Analysis Parameters

Users can customize the extraction and analysis through several key parameters:

Context Window Size (words)

Defines the number of words to extract before and after each citation. Default is 20 words.

Max Distance for Network (characters)

Defines the maximum character distance between two citations to be considered 'co-occurring' in the network analysis. Default is 800 characters (roughly 120-150 words).

Advanced Options

📊 Analysis Results and Tabs


1️⃣ Descriptive Statistics

Provides an overview of the document's structural and linguistic characteristics:

Document Metrics:

Readability Indices:

Text Statistics:

N-grams Analysis: Displays the most frequent unigrams (single words), bigrams (two-word phrases), and trigrams (three-word phrases) in the document. This reveals key concepts and repeated terminology.


2️⃣ Word Trends

Visualizes how selected terms are distributed across the document's sections. This analysis helps understand the thematic flow and identify where specific concepts are emphasized.

Features:

Interpretation Examples:

Distribution Statistics: Provides detailed frequency counts and statistical measures (mean, standard deviation, skewness) for each tracked term across segments.


3️⃣ In-Context Citations

Displays each extracted citation with its surrounding context window, enabling qualitative citation analysis. This is one of the most powerful features for understanding why and how sources are cited.

Features:

Analytical Uses:

Export Contexts: Use the 'Export Contexts' button to download all citation contexts as a structured dataset (CSV format) for external qualitative coding or further analysis in tools like NVivo, ATLAS.ti, or custom R scripts.


4️⃣ Network Analysis

Generates a citation co-occurrence network that visualizes which references are cited near each other in the text. This reveals the intellectual structure and conceptual clusters within the paper.

Network Construction:

Interpretation:

Network Metrics: While not explicitly displayed, users can infer centrality and clustering qualitatively:

Export Network: Download the network as a graph file (GraphML or edge list format) for advanced analysis in specialized network software like Gephi, Cytoscape, or igraph in R.

Legend: The right panel displays a legend showing the color coding for different document sections. This helps quickly identify which parts of the paper contribute most to the citation network.


5️⃣ References

Displays the complete bibliography extracted from the document. References are automatically parsed and enriched with metadata from Crossref and OpenAlex databases when DOIs are available.

Features:

Reference Matching Quality:

Total References: The header shows the breakdown:


6️⃣ BiblioAI Summary

Uses Google Gemini AI models to generate intelligent, context-aware summaries of the analyzed document. This feature transforms raw PDF content into structured, actionable insights.

Summary Types:

Customization:

Requirements:

Best Practices:


🧠 Integration with Biblio AI

Content Analysis is fully integrated with the Biblio AI ecosystem. Throughout the analysis tabs, users can activate AI-assisted interpretation panels that:

These dynamic interpretations adapt to your specific document and parameters, providing contextualized guidance rather than generic explanations.


💡 Advanced Use Cases

1. Citation Context Mining for Meta-Research

Researchers studying citation practices can use Content Analysis to:

2. Quality and Transparency Assessment

Evaluate methodological transparency by:

3. Comparative Literature Review

Analyze multiple papers on the same topic by:

4. Pedagogical Applications

Use Content Analysis to teach:

5. Editorial and Peer Review Support

Editors and reviewers can use Content Analysis to:


📌 Best Practices


⚠️ Limitations and Considerations


🔄 Integration with Biblioshiny Workflow

Content Analysis complements other Biblioshiny modules:


📚 Key References

Main References on Bibliometrix and Content Analysis Tools:

Aria, M., & Cuccurullo, C. (2025). contentanalysis: Scientific Content and Citation Analysis from PDF Documents. [R package]. https://doi.org/10.32614/CRAN.package.contentanalysis

Aria, M., & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Aria, M., Cuccurullo, C., D'Aniello, L., Misuraca, M., & Spano, M. (2024). Comparative science mapping: a novel conceptual structure analysis with metadata. Scientometrics. https://doi.org/10.1007/s11192-024-05161-6

Aria, M., Cuccurullo, C., D'Aniello, L., Misuraca, M., & Spano, M. (2022). Thematic Analysis as a New Culturomic Tool: The Social Media Coverage on COVID-19 Pandemic in Italy. Sustainability, 14(6), 3643. https://doi.org/10.3390/su14063643

Aria, M., Misuraca, M., & Spano, M. (2020). Mapping the evolution of social research and data science on 30 years of Social Indicators Research. Social Indicators Research, 149, 803–831. https://doi.org/10.1007/s11205-020-02281-3


Quantitative Content Analysis - Foundational Works:

Berelson, B. (1952). Content Analysis in Communication Research. Glencoe, IL: Free Press.

Krippendorff, K. (2018). Content Analysis: An Introduction to Its Methodology (4th ed.). Thousand Oaks, CA: SAGE Publications.

Neuendorf, K. A. (2017). The Content Analysis Guidebook (2nd ed.). Thousand Oaks, CA: SAGE Publications.

Weber, R. P. (1990). Basic Content Analysis (2nd ed.). Newbury Park, CA: SAGE Publications.


🎓 Further Reading

For more information on using Content Analysis and related techniques, see:

" ## Co-occurrence Network ---- coOccurrenceNetwork <- "

🔗 Co-occurrence Network

The Co-occurrence Network is a powerful tool for mapping the conceptual structure of a research field. It builds a network where nodes represent terms (keywords, title words, or abstract terms) and edges connect terms that appear together in the same document. The strength of an edge reflects how frequently two terms co-occur, revealing the thematic relationships within a scientific domain.

🎓 Theoretical Foundations

Co-occurrence analysis is rooted in the idea that terms appearing together in scientific documents are semantically related. When two keywords frequently co-occur across multiple publications, they are likely associated with the same research theme or concept. This principle has been widely used in scientometrics since the pioneering work of Callon et al. (1983) on co-word analysis.

The resulting network provides a map of the knowledge structure of a field, highlighting:

⚙️ Parameters and Options

The analysis can be configured through several parameters:

🔎 Interpreting Results


📚 Key References

Callon, M., Courtial, J.-P., Turner, W. A., & Bauin, S. (1983). From translations to problematic networks: An introduction to co-word analysis. Social Science Information, 22(2), 191-235. https://doi.org/10.1177/053901883022002003

Aria, M. & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959-975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Van Eck, N. J., & Waltman, L. (2010). Software survey: VOSviewer, a computer program for bibliometric mapping. Scientometrics, 84(2), 523-538. https://doi.org/10.1007/s11192-009-0146-3

Cobo, M. J., Lopez-Herrera, A. G., Herrera-Viedma, E., & Herrera, F. (2011). An approach for detecting, quantifying, and visualizing the evolution of a research field. Journal of Informetrics, 5(1), 146-166. https://doi.org/10.1016/j.joi.2010.10.002

" ## Thematic Map ---- thematicMap <- "

🗺 Thematic Map

The Thematic Map (also known as the Strategic Diagram) is a two-dimensional visualization that classifies research themes according to their centrality (importance within the field) and density (internal coherence of the theme). Originally proposed by Callon et al. (1991) and further developed by Cobo et al. (2011), it provides a comprehensive snapshot of the thematic landscape of a scientific domain.

🎓 Theoretical Foundations

The Thematic Map is built from a co-occurrence network of terms, normalised by association strength (van Eck & Waltman, 2009): the co-occurrence of two terms is divided by the product of their frequencies, so that frequent terms do not dominate the network. A community detection algorithm (Louvain by default) is applied to this normalised network to identify the thematic groups, as in Aria et al. (2026), and each cluster is characterized by two metrics:

📋 Fuzzy Publication-to-Cluster Assignment

Following the framework of Aria et al. (2026), publications are not assigned to a single cluster through hard partitioning. Instead, each publication receives a graded membership degree across all thematic clusters, reflecting the multithematic nature of scientific documents. The PageRank centrality of each term is computed within its cluster, on the co-occurrences among the cluster's terms (a term with no link inside its cluster gets the baseline (1 − d) / n, with d = 0.85 and n the number of terms of the cluster). The similarity of a publication to a cluster is the sum, over the terms they share, of the term's PageRank in the cluster divided by its frequency in the collection: a publication's affiliation with a cluster is stronger when its terms are central (high PageRank) and distinctive (low frequency) within that cluster. The similarities are then normalised to a distribution over clusters summing to one. A publication whose terms all lie outside the clusters (left out by the term filters) belongs to every cluster with the same degree 1 / (number of clusters); a publication with no term in the chosen field has no membership.

The Documents table reports the membership degrees of each publication, and the cluster it is assigned to when its membership there is at least 0.5.

This approach has two key advantages:

📈 The Four Quadrants

These two dimensions define four quadrants, each with a distinct strategic interpretation:

⚙️ Parameters and Options

🔎 Interpreting Results


📚 Key References

Aria, M., D'Aniello, L., Misuraca, M., & Spano, M. (2026). Rethinking thematic evolution in science mapping: An integrated framework for longitudinal analysis. Journal of Informetrics, 20, 101877. https://doi.org/10.1016/j.joi.2026.101877

Callon, M., Courtial, J.-P., & Laville, F. (1991). Co-word analysis as a tool for describing the network of interactions between basic and technological research: The case of polymer chemistry. Scientometrics, 22(1), 155-205. https://doi.org/10.1007/BF02019280

Cobo, M. J., Lopez-Herrera, A. G., Herrera-Viedma, E., & Herrera, F. (2011). An approach for detecting, quantifying, and visualizing the evolution of a research field. Journal of Informetrics, 5(1), 146-166. https://doi.org/10.1016/j.joi.2010.10.002

Aria, M. & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959-975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Cobo, M. J., Lopez-Herrera, A. G., Herrera-Viedma, E., & Herrera, F. (2012). SciMAT: A new science mapping analysis software tool. Journal of the American Society for Information Science and Technology, 63(8), 1609-1630. https://doi.org/10.1002/asi.22688

" ## Thematic Evolution ---- thematicEvolution <- "

🕐 Thematic Evolution

The Thematic Evolution analysis tracks how research themes emerge, grow, merge, split, and decline over time. By dividing the publication timeline into consecutive time slices and performing co-occurrence analysis within each slice, this method reveals the dynamic evolution of a field's conceptual structure. The implementation in bibliometrix adopts the integrated framework proposed by Aria et al. (2026), which reconceptualises thematic evolution as a fully network-based process where cross-sectional thematic detection, graded document affiliation, and inter-temporal linkage are jointly modelled within a coherent relational architecture.

🎓 Theoretical Foundations

This approach extends static thematic mapping into a longitudinal framework. It was originally formalised by Cobo et al. (2011), who introduced overlap-based similarity measures to identify continuities, splits, and mergers between themes across consecutive periods. Aria et al. (2026) further developed the methodology by addressing a key structural asymmetry in existing approaches: while cross-sectional theme detection is explicitly relational (based on weighted co-occurrence networks), lineage construction was traditionally reduced to set-theoretic comparisons of term lists, discarding the relational structure that defines a theme's semantic core.

The integrated framework resolves this asymmetry through three key innovations:

📈 How It Works

  1. Time slicing: The publication period is divided into consecutive time windows (e.g., 2000-2005, 2006-2010, 2011-2015)
  2. Thematic identification: Within each slice, a co-occurrence matrix is built and normalised using association strength. Community detection is applied to the normalised network to identify thematic clusters, characterised by centrality and density measures
  3. Fuzzy document assignment: Each publication is assigned graded membership degrees to all clusters, based on the PageRank centrality, within each cluster, of the terms it shares with it (see the Thematic Map help). The fuzzy cardinality of a cluster is the sum of the membership degrees of the publications
  4. Lineage computation: For a theme h of a period and a theme j of the next one, let PR be the PageRank of a term within its theme and PRtot the total PageRank of a theme. The weighted inclusion index Iw(h, j) is the share of PRtot(h) carried by the terms that h shares with j; it is directional: it measures how much of the source theme is carried over to the successor. The importance index Ω(h, j) is the square root of the sum, over the shared terms, of PR(h) × PR(j), divided by PRtot(h) × PRtot(j); it is high when the shared terms are central in both themes. The lineage strength is L = α Iw + (1 − α) Ω, a number between 0 and 1
  5. Automatic lineage detection: A link from a theme to a theme of the next period enters the evolutionary graph when its lineage strength is at least θ (absolute criterion), or when it is among the k strongest links leaving that theme (relative criterion). The relative criterion keeps the continuations that are weak in absolute terms but are the main successors of their theme
  6. Evolution mapping: The result is a temporally ordered directed graph, visualised as a Sankey diagram where the thickness of a flow is its lineage strength

⚙️ Parameters and Options

The default values (α = 0.5, θ = 0.10, k = 2) are those of the application in Aria et al. (2026) and should be regarded as a starting point: collections with a broader scope, larger vocabularies or shorter periods may require recalibration. θ and k together set how dense the evolutionary graph is.

🔎 Interpreting Results

The Table tab reports, for every pair of themes of consecutive periods that share at least one term, the shared words, the weighted inclusion, the inclusion index, the occurrences, the stability, the importance index, the lineage strength, and whether the link is kept in the evolutionary graph. In R, thematicEvolution() also returns the pattern of each theme (emergent, disappearing, continuation, split, merge) in Nodes and the pathways in pathways.

📑 Layers: the thematic maps stacked

The Layers tab draws the thematic map of each period as a plane, the oldest at the bottom, as in Angelelli et al. (2025). Each theme sits on its plane at its relevance degree (centrality) and development degree (density), and the themes of adjacent periods are linked as in the evolutionary graph. The Sankey diagram shows the flows but not the positions, the thematic maps the positions but not the flows: the layers show both, so you can follow a theme as it moves across the quadrants while it merges with or splits from other themes.


📚 Key References

Aria, M., D'Aniello, L., Misuraca, M., & Spano, M. (2026). Rethinking thematic evolution in science mapping: An integrated framework for longitudinal analysis. Journal of Informetrics, 20, 101877. https://doi.org/10.1016/j.joi.2026.101877

Angelelli, M., Ciavolino, E., Ringle, C. M., Sarstedt, M., & Aria, M. (2025). Conceptual structure and thematic evolution in partial least squares structural equation modeling research. Quality & Quantity, 59(3), 2753–2798. https://doi.org/10.1007/s11135-025-02071-4

Cobo, M. J., Lopez-Herrera, A. G., Herrera-Viedma, E., & Herrera, F. (2011). An approach for detecting, quantifying, and visualizing the evolution of a research field. Journal of Informetrics, 5(1), 146-166. https://doi.org/10.1016/j.joi.2010.10.002

Aria, M. & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959-975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Callon, M., Courtial, J.-P., & Laville, F. (1991). Co-word analysis as a tool for describing the network of interactions between basic and technological research: The case of polymer chemistry. Scientometrics, 22(1), 155-205. https://doi.org/10.1007/BF02019280

Morris, S. A., & Van der Veer Martens, B. (2008). The cognitive structure of scientific revolutions. Scientometrics, 75(3), 423-442. https://doi.org/10.1007/s11192-007-1804-x

" ## Factorial Analysis ---- factorialAnalysis <- "

📊 Factorial Analysis

Factorial Analysis applies dimensionality reduction techniques to bibliometric data, projecting high-dimensional document-term relationships onto a low-dimensional space for visualization and interpretation. The primary method used is Correspondence Analysis (CA), a multivariate statistical technique particularly suited to analyzing contingency tables and categorical data.

🎓 Theoretical Foundations

Correspondence Analysis (CA) decomposes a document-term matrix to identify the principal dimensions of variation in the data. It simultaneously represents both documents (rows) and terms (columns) in the same factorial space, revealing the associations between them.

Key concepts include:

When applied to bibliometric data, CA reveals the intellectual structure of a field by identifying the main dimensions along which research topics and authors are organized.

📈 How It Works

  1. Matrix construction: A document-term (or author-term) co-occurrence matrix is built from the bibliographic data
  2. Decomposition: Correspondence Analysis decomposes this matrix into factorial axes that capture the most important dimensions of variation
  3. Projection: Documents and terms are projected onto the first two or three factorial axes
  4. Clustering: Hierarchical clustering can be applied to the factorial coordinates to group similar documents or terms

⚙️ Parameters and Options

🔎 Interpreting Results


📚 Key References

Greenacre, M. J. (2007). Correspondence Analysis in Practice (2nd ed.). London: Chapman and Hall/CRC. https://doi.org/10.1201/9781420011234

Benzecri, J.-P. (1973). L'Analyse des Donnees. Volume II: L'Analyse des Correspondances. Paris: Dunod.

Aria, M. & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959-975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Cuccurullo, C., Aria, M., & Sarto, F. (2016). Foundations and trends in performance management. A twenty-five years bibliometric analysis in business and public administration domains. Scientometrics, 108(2), 595-611. https://doi.org/10.1007/s11192-016-1948-8

" ## Co-Citation Network ---- coCitationNetwork <- "

📑 Co-Citation Network

The Co-Citation Network maps the intellectual structure of a research field by analyzing patterns in how documents cite the same references. Two references are co-cited when they are both cited by the same document. The more frequently two references are co-cited, the stronger their intellectual relationship. This technique, introduced by Small (1973), is one of the foundational methods of bibliometric analysis.

🎓 Theoretical Foundations

Co-citation analysis rests on two complementary approaches:

Both approaches build networks where:

⚙️ Parameters and Options

🔎 Interpreting Results


📚 Key References

Small, H. (1973). Co-citation in the scientific literature: A new measure of the relationship between two documents. Journal of the American Society for Information Science, 24(4), 265-269. https://doi.org/10.1002/asi.4630240406

Kessler, M. M. (1963). Bibliographic coupling between scientific papers. American Documentation, 14(1), 10-25. https://doi.org/10.1002/asi.5090140103

Aria, M. & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959-975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., D'Aniello, L., & Spano, M. (2026). A multi-phase reference matching algorithm for bibliometric analysis: design, implementation, and evaluation. Scientometrics. https://doi.org/10.1007/s11192-026-05763-2

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Small, H. (1999). Visualizing science by citation mapping. Journal of the American Society for Information Science, 50(9), 799-813. DOI link

" ## Historiograph ---- historiograph <- "

📚 Historiograph

The Historiograph (or Historiographic Map) is a chronological visualization of a research field's evolution based on direct citation relationships. Introduced by Garfield (2004) and further developed by Garfield, Pudovkin, and Istomin (2003), this method constructs a time-based network where documents are positioned along a timeline and connected by citation links, identifying the most influential works and the intellectual pathways of a field.

🎓 Theoretical Foundations

Unlike co-citation analysis, which measures indirect relationships, the historiograph uses direct citations to map the intellectual development of a field. The underlying idea is that the citation chain from a foundational paper through successive works traces the evolution of ideas and methodologies over time.

Key principles:

📈 How It Works

  1. Local citation analysis: The algorithm identifies citation links among the documents in your collection
  2. Network construction: A directed citation network is built where edges represent citation relationships
  3. Chronological layout: Documents are arranged on a timeline based on their publication year
  4. Path identification: The main citation paths through the network reveal the intellectual evolution of the field

⚙️ Parameters and Options

🔎 Interpreting Results


📚 Key References

Garfield, E. (2004). Historiographic mapping of knowledge domains literature. Journal of Information Science, 30(2), 119-145. https://doi.org/10.1177/0165551504042802

Garfield, E., Pudovkin, A. I., & Istomin, V. S. (2003). Why do we need algorithmic historiography? Journal of the American Society for Information Science and Technology, 54(5), 400-412. https://doi.org/10.1002/asi.10226

Aria, M. & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959-975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., D'Aniello, L., & Spano, M. (2026). A multi-phase reference matching algorithm for bibliometric analysis: design, implementation, and evaluation. Scientometrics. https://doi.org/10.1007/s11192-026-05763-2

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Lucio-Arias, D., & Leydesdorff, L. (2008). Main-path analysis and path-dependent transitions in HistCite-based historiograms. Journal of the American Society for Information Science and Technology, 59(12), 1948-1962. https://doi.org/10.1002/asi.20903

" ## Trend Topics ---- trendTopics <- "

📈 Trend Topics

The Trend Topics analysis identifies and visualises the temporal dynamics of key terms within a bibliometric collection. By tracking the frequency of keywords (or other terms) across publication years, this tool reveals which topics are gaining momentum, which have reached maturity, and which are declining, providing a clear picture of the evolution of research interests over time.

🎓 Theoretical Foundations

Trend analysis in bibliometrics is based on the observation that the vocabulary used in scientific publications reflects the intellectual priorities of a research community at a given point in time. By monitoring the rise and fall of specific terms across successive years, it is possible to detect shifts in research focus, the emergence of new paradigms, and the obsolescence of older topics.

This approach complements other temporal analyses (such as thematic evolution) by focusing on individual terms rather than thematic clusters, offering a more granular view of how specific concepts gain or lose traction within a field.

📈 How It Works

  1. Term extraction: Keywords or other terms (Author's Keywords, Keywords Plus, title words, etc.) are extracted from each document in the collection
  2. Temporal aggregation: For each term, the frequency of occurrence is computed for each publication year
  3. Term selection: The top terms are selected based on their overall frequency or their trend characteristics
  4. Visualisation: The selected terms are plotted along a timeline, showing for each term the year range in which it appears and its frequency peak

⚙️ Parameters and Options

🔎 Interpreting Results

💡 Tips for Analysis


📚 Key References

Aria, M. & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959-975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

Cobo, M. J., Lopez-Herrera, A. G., Herrera-Viedma, E., & Herrera, F. (2011). An approach for detecting, quantifying, and visualizing the evolution of a research field. Journal of Informetrics, 5(1), 146-166. https://doi.org/10.1016/j.joi.2010.10.002

Noyons, E. C. M., Moed, H. F., & van Raan, A. F. J. (1999). Integrating research performance analysis and science mapping. Scientometrics, 46(3), 591-604. https://doi.org/10.1007/BF02459614

" bradfordLaw <- "

Bradford's Law of Scattering

Bradford's Law is one of the foundational laws of bibliometrics, first formulated by Samuel C. Bradford in 1934. The law describes the phenomenon of concentration and dispersion in scientific publishing: a small number of core journals account for a disproportionately large share of the literature on a given topic, while the remaining literature is scattered across an increasingly large number of peripheral journals.


The Original Formulation

Bradford observed that if scientific journals are ranked in order of decreasing productivity on a given subject and then divided into groups (zones) that each contain approximately the same number of articles, the number of journals in successive zones increases geometrically. Specifically, if the journals are partitioned into three zones such that each zone contains roughly one-third of the total articles, the number of journals in each zone follows the ratio:

1 : k : k²

where k is the Bradford multiplier. The first zone (Zone 1, or the “core”) contains a small number of highly productive journals; the second zone (Zone 2) contains a larger number of moderately productive journals; and the third zone (Zone 3, or the “periphery”) contains a very large number of journals that each contribute only one or a few articles.


The Bradford Distribution

More formally, the law can be expressed through the Bradford distribution. If sources are ranked in decreasing order of productivity (n1 ≥ n2 ≥ … ≥ nS) and the cumulative number of articles is plotted against the logarithm of the source rank, the resulting curve — known as the Bradford bibliograph — shows a characteristic pattern: an initial concave region (corresponding to the core journals) followed by a linear region (corresponding to the intermediate and peripheral journals).

The linear portion of the curve indicates that:

C(r) = a + b · log(r)

where C(r) is the cumulative number of articles contributed by the top r sources, and a and b are constants estimated via linear regression.

This logarithmic relationship implies that each successive “doubling” of the number of journals yields a roughly constant increment in the total number of articles — a fundamental feature of the skewed, heavy-tailed distributions that pervade bibliometric data.


Practical Interpretation of Bradford Zones

Biblioshiny implements Bradford's Law by partitioning the sources in the collection into zones. The partition serves multiple analytical and practical purposes:


Diagnostic Information from the Bradford Curve

The shape of the Bradford curve provides diagnostic information about the field:


Goodness-of-Fit Test

To assess whether the empirical source distribution conforms to the theoretical Bradford distribution, Biblioshiny performs a Kolmogorov-Smirnov (KS) goodness-of-fit test. This test compares the empirical cumulative distribution of articles across sources with the theoretical distribution predicted by the fitted Bradford model.


References

Bradford, S. C. (1934). Sources of information on specific subjects. Engineering, 137, 85–86.

Brookes, B. C. (1969). Bradford's law and the bibliography of science. Nature, 224(5223), 953–956. https://doi.org/10.1038/224953a0

Egghe, L. (1986). The dual of Bradford's law. Journal of the American Society for Information Science, 37(4), 246–255. https://doi.org/10.1002/(SICI)1097-4571(198607)37:4<246::AID-ASI10>3.0.CO;2-F

Aria, M. & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

" authorsDominance <- "

Authors' Dominance

The Dominance Factor (DF), proposed by Kumar & Kumar (2008), measures how often an author leads the collaborations they take part in. It is the share of an author's multi-authored articles in which they are the first author:

DF = FAA / MAA

where FAA is the number of multi-authored articles the author signs as first author, and MAA is the number of multi-authored articles the author signs in any position. Single-authored articles are left out: an author with no multi-authored article has no Dominance Factor and is not shown.


How to Read the Plot

Biblioshiny selects the most productive authors of the collection (the Number of Authors option) and places each of them by:

Two dashed lines split the authors into four groups: the vertical one is the median productivity of the plotted authors, the horizontal one is DF = 0.5, above which an author is first author in at least half of their multi-authored articles.

Each bubble is named above. When two names would overlap, or a name would cover another bubble, only the author with the higher DF is named; hover over a bubble to see the names of all its authors. Authors with the same number of articles, multi-authored articles and DF fall on the same point, which is drawn once with their names stacked. A low DF does not mean low quality: in fields where authors are listed alphabetically, or where the senior author signs last, the first position says little about who led the work.


References

Kumar, S. & Kumar, S. (2008). Collaboration in research productivity in oil seed research institutes of India. Proceedings of the Fourth International Conference on Webometrics, Informetrics and Scientometrics & Ninth COLLNET Meeting, Berlin, Germany.

Aria, M. & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

" lotkaLaw <- "

Lotka's Law of Scientific Productivity

Lotka's Law is one of the foundational laws of bibliometrics, first formulated by Alfred J. Lotka in 1926. The law describes the frequency distribution of scientific productivity among authors: a large proportion of authors publish only one paper, while a small proportion are highly prolific.


The Inverse Power Law

Lotka observed that the number of authors who publish exactly n articles follows an inverse power law:

f(n) = C / nβ

where:

In Lotka's original formulation, β = 2, meaning that approximately 1/n² of the authors who publish one paper will publish n papers. For example, about 1/4 of authors with one paper will have two papers, about 1/9 will have three papers, and so on.


Estimation Method

Biblioshiny estimates the parameters of Lotka's Law using linear regression in log-log space:

log10(f) = log10(C) − β · log10(n)

This provides the estimated β and C coefficients, along with the R² goodness-of-fit measure. The plot displays three curves:


Kolmogorov-Smirnov Goodness-of-Fit Tests

Two Kolmogorov-Smirnov (KS) tests are performed to assess how well the data conform to Lotka's Law:


Interpreting the Results


References

Lotka, A. J. (1926). The frequency distribution of scientific productivity. Journal of the Washington Academy of Sciences, 16(12), 317–323.

Pao, M. L. (1985). Lotka's law: A testing procedure. Information Processing & Management, 21(4), 305–320. https://doi.org/10.1016/0306-4573(85)90055-X

Nicholls, P. T. (1986). Empirical validation of Lotka's law. Information Processing & Management, 22(5), 417–419. https://doi.org/10.1016/0306-4573(86)90007-0

Aria, M. & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

" rpys <- "

Reference Publication Year Spectroscopy (RPYS)

Reference Publication Year Spectroscopy (RPYS) is a method for detecting the historical roots of research fields, topics, or researchers by analysing the publication years of the cited references in a body of literature. The method was introduced by Marx et al. (2014) and later extended by Thor et al. (2018).


The Spectrogram

The RPYS spectrogram plots two curves:

The top 10 peaks of the deviation curve are automatically identified and labelled in the spectrogram. These peaks correspond to the most influential reference publication years.


Types of Citation Sequences

Following the methodology of Thor et al. (2018), Biblioshiny applies Configural Frequency Analysis (CFA) to classify the citation dynamics of individual cited references over time. For each cited reference, CRExplorer computes standardised residuals (z-scores) that compare the observed citation counts with expected values under a model of independence. Based on the z-scores, each citing year is classified as:

The resulting sequence of symbols across citing years characterises the citation trajectory of each reference. Four types of sequences are identified:

Type Definition
Sleeping Beauty Publication cited below average (“−”; z < −1) in at least two of the first three citing years, and above average (“+”; z > 1) in at least one of the following citing years.
Constant Performer Publication cited in more than 80% of the citing years at least once. In more than 80% of the citing years it has been cited at least on the average level (“o”; −1 ≤ z ≤ 1) or above (“+”; z > 1).
Hot Paper Publication cited above average (“+”; z > 1) in at least two of the first three citing years after publication.
Life Cycle Publication cited in at least two of the first four years on the average level (“o”) or lower (“−”), in at least two of the following years above average (“+”; z > 1), and in the last three years on the average level (“o”) or lower (“−”).

A single cited reference may belong to more than one type. When this occurs, all applicable types are reported.


References

Marx, W., Bornmann, L., Barth, A., & Leydesdorff, L. (2014). Detecting the historical roots of research fields by reference publication year spectroscopy (RPYS). Journal of the Association for Information Science and Technology, 65(4), 751–764. https://doi.org/10.1002/asi.23089

Thor, A., Bornmann, L., Marx, W., & Mutz, R. (2018). Identifying single influential publications in a research field: New analysis opportunities of the CRExplorer. Scientometrics, 116, 591–608. https://doi.org/10.1007/s11192-018-2733-7

Barth, A., Marx, W., Bornmann, L., & Mutz, R. (2014). On the origins and the historical roots of the Higgs boson research from a bibliometric perspective. The European Physical Journal Plus, 129, 111. https://doi.org/10.1140/epjp/i2014-14111-6

Aria, M. & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

" ## PRISMA Diagram ---- prisma <- "

PRISMA Flow Diagram for Bibliometric Studies

The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram documents the process of selecting studies for analysis. In bibliometric research, the PRISMA diagram tracks how the initial set of records retrieved from databases is progressively refined through filtering criteria.

Phases

How to use

  1. In the Identification box, enter the database name, your search query, and the total number of records retrieved.
  2. Optionally, add a secondary source (e.g. records added via DOI or from another database).
  3. Click Add Step to define each filtering stage. For each step, select the criterion type, optionally describe the filter values (e.g. “Article, Review” or “1985–2025”), and enter the number of records remaining after applying that filter.
  4. The PRISMA phase (Screening or Eligibility) and the number of excluded records are computed automatically.
  5. Click Generate PRISMA Diagram to produce the flow chart.
  6. Use the download button to export the diagram as a high-resolution PNG, or the report button to add it to the workbook.

References

Page, M. J., et al. (2021). The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ, 372, n71. https://doi.org/10.1136/bmj.n71

Aria, M., & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007

Aria, M., & Cuccurullo, C. (2026). Science Mapping Analysis: A Primer with Biblioshiny. McGraw-Hill, New York, NY, USA. ISBN: 978-88-386-2297-7.

" return(list( biblioAI = biblioAI, saas = saas, info = info, publications = publications, filters = filters, authorProfile = authorProfile, referenceMatching = referenceMatching, importOrLoad = importOrLoad, api = api, mergeCollections = mergeCollections, mainInformation = mainInformation, lifeCycle = lifeCycle, threeFieldPlot = threeFieldPlot, contentAnalysis = contentAnalysis, coOccurrenceNetwork = coOccurrenceNetwork, thematicMap = thematicMap, thematicEvolution = thematicEvolution, factorialAnalysis = factorialAnalysis, coCitationNetwork = coCitationNetwork, historiograph = historiograph, trendTopics = trendTopics, bradfordLaw = bradfordLaw, lotkaLaw = lotkaLaw, authorsDominance = authorsDominance, rpys = rpys, prisma = prisma )) }