Frequently Asked Questions
Q1 - I have installed Bibliometrix using the code install.packages(“bibliometrix”) but when I digit Biblioshiny I get an error:
biblioshiny()
> Error in biblioshiny() : could not find function "biblioshiny"
Before using a package, you have to load it through the function library. So, digit:
library(bibliometrix)
and then start using bibliometrix/biblioshiny
Q2 - I downloaded the .bib file from Scopus and I am trying to import it into Biblioshiny. It shows an error "replacement has length zero" or an error “undefined columns selected”.
The error usually appears when:
- The export file does not contain all mandatory metadata that bibliometrix needs to work correctly;
- The database/format you selected in bibliometrix/biblioshiny does not fit the format of the export file you are trying to import (i.e. your bib file was exported from Scopus but you are trying to import it as WoS bib file)
- The export file format is not-supported by bibliometrix/biblioshiny (i.e. Mendeley, Zotero, etc.)
Please, read carefully the bibliometrix tutorial about how to download data from a bibliographic database.
Q3 - If I have a database made in excel and I want to do an analysis in bibliometrix/biblioshiny, how can I load the database for the analysis?
Bibliometrix/Biblioshiny supports collection coming from Web of Science (plaintext, bibtex or EndNote formats), Scopus (bibtex or CSV formats), Pubmed (Pubmed format or API call), Cochrane Library (plaintext) and Dimensions (API call, csv or xlsx formats).
It is not possible to import a homemade excel database.
Please, read carefully the bibliometrix tutorial about how to download data from a bibliographic database.
Q4 - I am starting to use Bibliometrix/Biblioshiny as a tool to perform some scientometrics studies. I have access to both Scopus and WoS at my institution. Therefore, I was wondering if you would advise me on which one is better to use? I am aware of the differences in terms of document coverage and the collection of data (number of entries, etc), but I am not totally sure if one of the two gives more reliable or better quality results when processed by your tool.
WoS is, in general, preferable for data quality (reference items are standardized, the availability of Keywords Plus, very few missing data, etc.) but if you need to analyze Art & Humanities publications or conference proceedings, Scopus is the better choice.
Q5 - I want to load a bibtex file created by a reference manager software but when I run the function convert2df it returns an error:
Error in seq.default(iStart, iStop) : 'from' must be a finite number
The problem is the data format. Unfortunately, bibtex is not a proper standardized format, so it could change when created from different sources.
Bibliometrix/biblioshiny need a bibtex file formatted exactly as exported by WoS or Scopus and with the full set of mandatory metadata (authors' name, affiliations, references, etc.). If not, it will not work.
Q6 - How can I upload multiple export files in Biblioshiny?
You have just to put your export files into a single zip-compressed file and then import it with Biblioshiny.
Q7 - When reading your publications, you tend to only use WoS rather than merging WoS and Scopus. What are the reasons for this?
WoS and Scopus databases use a very different approach to codify the bibliographic metadata.
I.e. WoS applies a pre-processing analysis on reference lists rewriting references as FIRST AUTHOR, YEAR, JOURNAL, ISSUE, DOI while Scopus stores the full APA record as an author included it in his manuscript.
This means that, at the moment, it is not possible to do citation and co-citation analyses on merged databases. Moreover, in this specific example about references, Scopus data needs a huge cleaning phase to match reference items (often the same reference is written in several different ways!).
When you try to merge databases from different DBs you have also to answer some critical questions:
- When a manuscript is included in both DBs, what citation metric will you consider? (Total citation number are quite different among different sources)
- Which content metadata will you consider to perform a content analysis? WoS provides “Keyword Plus” (ID tag) that many publications advice as the best content field to perform topic analysis. But Scopus does not provide this metadata.
- If you want to perform a collaboration analysis among affiliations, how to match affiliation names? Affiliation info is stored without any pre-processing phase.
- etc.
In general, at the moment, WoS metadata is characterized by a higher quality than Scopus and, if it is possible, I prefer to use WoS. On the other hand, Scopus is a better choice when your analysis regards Arts & Humanities. In this case, Scopus DB is extremely richer than WoS.
Q8 - I would like to understand the difference between the concepts “Global Citation” and “Local Citation”.
Global Citations (TC) means the Total Citations that an article, included in your collection, has received from documents indexed on a bibliographic database (WoS, Scopus, etc.). So, TC counts citations received by a selected article "all over the world".
In Most Cited References, Bibliometrix counts the local citations, in other words, the citations that a reference (a document included in, at least, one of the bibliographies of the articles in your collection) has received from documents included in your collection.
The reference lists include all documents from bibliographies as then many of the references could not be part of your collection!
When a reference is also part of your collection, it is called “Cited document”.
So, Local citations are citations received by a reference article "internally to your collection".
Please, take a look at biblioshiny slides where we explain this concept more in deep (https://bibliometrix.org/biblioshiny/assets/player/KeynoteDHTMLPlayer.html).
Q9 - Before analyzing my database in Biblioshiny, I want to pass the raw data downloaded from WOS through a title and abstract screening process. How to do that?
Biblioshiny reads Excel files that were previously converted from one of the supported export formats and exported using Biblioshiny.
If you need to filter some documents by title/abstract/keyword, this is my advice:
1) load raw files in Biblioshiny (putting them in one zip file). Biblioshiny will convert your collection in a data frame
2) export this collection in Excel using the export button
3) open the file Excel and remove rows (documents) that you do not need.
Please, pay attention to:
-do not remove any column
-do not change any column name
-do not change the data format (all characters are uppercase, semicolons separate different items in the same column, etc.)
4) load your filtered Excel file in Biblioshiny using the load button
Everything should be fine.
Q10 - In Bibliometrix, Can I work citation analysis from Pubmed data?
No, It isn't possible because PubMed does not store citation metadata (no Cited References, no total citations, etc.)
Q11 - A question about the indicator “Average Years from Publication”. Does it refer to the average years to an article to be cited?
Yes, it does.
Q12 - I have a question about its filter function: I imported 4687 documents from WoS. But 10 of them were removed after filtering, despite I did not apply any filter (see the screenshot attached). Do you know why these 10 documents were removed?
Maybe one of the fields used in the filter menu could contain missing values.
Usually, it depends on publication years. Some recent articles in press could not yet have a PY value.
Try to export your data in excel and inspect the column PY.
Q13 - I have difficulties in interpreting the graphic about "Authors' Production Over Time". What does the color of the circles mean? Does the darker color refer to works with the greatest impact? Does circle size refers to more posts and circle color to greater impact?
Given an Author “XY”:
- the line represents his timeline
- a bubble at a given year means that “XY” published at least a document in that year
- the bubble size is proportional to the n. of documents XY published in that year
- the color intensity is proportional to the total citations per year of the document published in that year.
Q14 - I was unable to find inside documentation on how to make a "Trend Topic" like graph but outside the Biblioshiny environment. What the command?
You can use the function fieldByYear.
Here an example:
library(bibliometrix)
data(management)
res <- fieldByYear(management, field = "ID", timespan = c(2005,2015),
min.freq = 5, n.items = 5, graph = TRUE)
Q15 - I am reaching out to get your help on the problem I am facing when running Biblioshiny for Bibliometrix. Basically, all is good except that I am getting a co-citation network in black color which confuses me to identify. I've provided the network below:
You are using a non-compatible browser (i.e. Microsoft Edge or Apple Safari).
Please, use Mozilla Firefox or Google Chrome as default browser.
Q16 - How can I perform a conceptual analysis?
Conceptual analysis can be done in various ways:
- the first one is the network analysis, the most common among bibliometric analyses
- the second one is the factorial analysis. It is a data reduction technique
- the third is a mixed approach between the two previous approaches that we have called thematic maps and evolution.
Topic modeling is a fourth way. But it is not yet included in bibliometrix/biblioshiny.
All these techniques are well explained in the slides you find on www.bibliometrix.org
Q17 - It was not clear to me how to interpret the clustering system on the networks.
Bibliometrix/Biblioshiny implements different community detection algorithms to identify clusters into a network (i.e. Louvain, Walktap, MDS, etc.). You can choose which one to use using the argument "cluster".
For more information about the community detection algorithm, you can read:
Lancichinetti, A., & Fortunato, S. (2009). Community detection algorithms: a comparative analysis. Physical review E, 80(5), 056117.
Q18 - Is it possible to retrieve the code behind some of your graphs in the biblioshiny? I would like to be able to reproduce these graphs inside my code with all my analysis.
bibliometrix sources are published on GitHub.
Moreover, the Biblioshiny code is in the folder "bibliometrix/inst" in your R package library.
Q19 - I have a question regarding exporting an organization network map for VOSViewer. In Bibliometrix, it is possible to export a map directly to VOSViewer by using the function net2vosviewer. Is it possible to perform the same export via Biblioshiny?
In Biblioshiny, you can export maps in pajek format and then read them using Vosviewer.
Q20 - Where can I find a glossary of the column names of bibliographic the data frame?
Data frame columns are named using the standard Clarivate Analytics WoS Field Tag codify.
The main field tags are
For a complete list of field tags see https://www.bibliometrix.org/documents/Field_Tags_bibliometrix.pdf
Q21 - I have installed Bibliometrix using the code install.packages(“bibliometrix”) but when I digit library(bibliometrix) I get an error:
Error: package or namespace load failed for ‘bibliometrix’ in loadNamespace(i, c(lib.loc, .libPaths()), versionCheck = vI[[i]]):
there is no package called ‘XML’
Would you mind help me fix this problem?
This is a well-known bug in R. Sometimes R is unable to install all dependencies of a certain package.
In your case, bibliometrix cannot work because XML package is missing.
Please, manually install XML and all other packages that should be missing using the command:
install.packages(“package name”)
i.e.
install.packages(“XML”)
Q22 - I have read your paper "bibliometrix: An R-tool for comprehensive science mapping analysis" published in the Journal of Informetrics in 2017. Trying to replicate your code using a new dataset from Dimensions (but also from Scopus, WoS, Lens, and Pubmed), the function convert2df() returns many errors.
During the last three years, we released many updates of the package including new features and larger support to new databases such as Dimensions, PubMed, and Lens. Consequently, also the function convert2df changed and the example code published in our paper is now out of date.
You can read more about the new bibliometrix 4.0 (and its web interface Biblioshiny) on the website www.bibliometrix.org. We shared many documents:
- a tutorial about the main functions implemented in bibliometrix
https://www.bibliometrix.org/vignettes/Introduction_to_bibliometrix.html
- a tutorial about “how to import data from the main databases”
https://www.bibliometrix.org/vignettes/Data-Importing-and-Converting.html
- a tutorial about “how to use Biblioshiny”
https://bibliometrix.org/biblioshiny/assets/player/KeynoteDHTMLPlayer.html
Moreover, YouTube offers several video tutorials (many in Spanish) created by scholars all over the world:
https://www.youtube.com/results?search_query=bibliometrix+biblioshiny
Q23 - Why is the sum of the frequencies greater than the number of items in the collection in Country Scientific Production?
1) Country Scientific Production measures the number of "authors appearances by country affiliations". This means that if in an article there are three authors working respectively in the USA, SPAIN and ITALY then the counter of appearances for each of these three countries will be increased by 1. In other words each article is attributed to the countries of all its co-authors and will therefore be counted as many times as there are authors (in the example above, three times). It follows that the sum of the production indicator necessarily exceeds the total number of articles (at least if all articles are single-signed).
Q24 - I installed bibliometrix version 4.1.2 and when I try to view the dendrogram in Factorial Analysis, Biblioshiny returns an error.
What can I do?
The error is probably due to a problem with the visNetwork library not importing all the necessary packages correctly.
To solve the problem, run the following command line in R or RStudio:
install.packages(c("ggraph", "tidygraph", "sparkline"))