Skip to content

cicirello/bibliometrics

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

bibliometrics

Packages and Releases PyPI GitHub release (latest by date)
PyPI Downloads Downloads Downloads/month Downloads/week
Build Status build CodeQL
Security Snyk security score
Other Info Python Version from PEP 621 TOML License GitHub top language
Support GitHub Sponsors Liberapay Ko-Fi

This command line utility does the following:

The intention of this utility is as a tool for a researcher to generate an SVG of their own bibliometrics only. For example, I am using it to generate and update such an SVG for my own profile twice monthly. It is not intended for analyzing multiple researchers' bibliometrics, and requests for such functionality will be rejected (there are other tools for that).

Blog Post: Your Citation Metrics in an SVG for Your Website (Posted on DEV.to on July 22, 2022)

Table of Contents

This README is organized as follows:

  • Supported Bibliometrics: list of the bibliometrics with brief descriptions of what they are.
  • Samples: provides examples of the output of this utility.
  • Configuration: explains how to configure the utility, such as colors for the SVG, file locations, etc.
  • Configuring the Scholar ID: explains the two ways of providing your Google Scholar ID to the utility.
  • Usage: how to install and run.
  • Respect Google Scholar's robots.txt: explains the relevant portions of Google Scholar's robots.txt as it relates to this, or any tool, designed to gather information from Scholar. Note that most other tools that provide more functionality (e.g., all of the ones I looked at before implementing this) do not respect that robots.txt. If you wish to submit an issue or pull request requesting additional functionality, please know that any such request must be possible to implement without violating Scholar's robots.txt. Otherwise, the issue or pull request will be closed.
  • Support the Project: different ways that you can support the project.
  • Contribute: contribution guidelines.
  • License.

Supported Bibliometrics

The bibliometrics utility computes the following bibliometrics:

  • Total citations: total number of citations to the researcher's publications.
  • Five-year citations: total citations within past 5 years.
  • Most-cited paper: number of citations to the researcher's most-cited paper.
  • h-index: the maximum h such that the researcher's h most-cited papers have been cited at least h times each.
  • m-quotient: h-index / n, the number of years since first publication, which was introduced in the same article as the h-index itself as a way of adjusting for length of publication history.
  • g-index: the maximum g such that the researcher's g most-cited papers have been cited an average of g times each.
  • i10-index: number of papers cited at least 10 times each.
  • i100-index: number of papers cited at least 100 times each.
  • i1000-index: number of papers cited at least 1000 times each.
  • i10000-index: number of papers cited at least 10000 times each.
  • w-index: number of papers cited at least 10w times each.
  • o-index: the geometric mean of the researcher's h-index and number of citations to the most-cited paper.
  • Various bibliometrics intended to complement the h-index, providing additional information about the researcher's h-core, where the h-core refers to the h most-cited papers. These complementary bibliometrics include the following:
    • h-median: the median citations of the papers in the researcher's h-core.
    • e-index: the square root of the excess citations to the papers in the h-core, where excess citations is defined as the total citations to the papers in the h-core minus $h^2$.
    • r-index: the square root of the total citations to the papers in the researcher's h-core.
    • a-index: the average number of citations to the papers in the researcher's h-core.

Samples

Here are a couple sample SVGs.

Example 1

Example 2

Here is a sample of the JSON summary also generated by the utility:

{
    "a-index": 71.56,
    "e-index": 34.12,
    "five-year-cites": 364,
    "g-index": 44,
    "h-index": 25,
    "h-median": 48,
    "i10-index": 33,
    "i100-index": 3,
    "m-quotient": 1.0,
    "most-cited": 228,
    "o-index": 75,
    "r-index": 42.3,
    "total-cites": 2052,
    "w-index": 8
}

The keys in this JSON are all lowercased, and correspond to the keys expected if you customize the order in your configuration file. However, in your configuration file they are case-insensitive. See the next section for how to configure.

The above sample is also found in this repository: bibliometrics.json, and the sample SVGs are found in the images directory.

Configuration

The utility looks for a configuration file .bibliometrics.config.json in your current working directory (please note the . at the start of the filename). A sample is found at the root of this repository: .bibliometrics.config.json.

To generate the JSON summary of your bibliometrics, specify the filename (optionally with path) via the "jsonOutputFile" field. If this field is not present, then no JSON file will be generated.

To compute the m-quotient, you must provide the year of your first publication in the "firstPubYear" field. The bibliometrics utility does not attempt to scrape this from your Scholar profile.

To change the order that the bibliometrics appear in the SVG, or to explicitly exclude one or more bibliometrics, you can use the "include" field. This field is an array of keys associated with the various bibliometrics. If this field is not present, then the following default order is used: [ "total-cites", "five-year-cites", "most-cited", "h-index", "g-index", "i10-index", "i100-index", "i1000-index", "i10000-index", "w-index", "o-index", "h-median", "m-quotient", "e-index", "r-index", "a-index" ]. There is no reason to use this field if the only thing you want to do is to exclude bibliometrics that have the value 0. Such bibliometrics will be excluded by default. The list of keys for the bibliometrics to include is case-insensitive.

The "svgConfig" field is an array of JSON objects, such that each object configures one SVG. Each of the JSON objects in this array includes the following fields:

  • "filename" is the filename (optionally with path) to the target SVG file.
  • "background" is the background color.
  • "border" is the border color.
  • "title" is the title color.
  • "text" is the color of the rest of the text.
  • "include" is similar to the top-level field of the same name, but applies only to one SVG, whereas the top-level field applies to all. If both the top-level "include" field and the more specific field by the same name are used, then the top-level "include" overrides the default, and the individual SVG's "include" in turn overrides the top-level "include".

The colors can be defined in any format that is valid within an SVG. For example, you can specify RGB, two hex digits for each color channel, with #010409; or RGB with one hex digit for each color channel, #123. You can also use SVG named colors, such as white, as well as RGBA such as rgba(56,139,253,0.4). If it is valid as a color in SVG, then it should work. The utility simply inserts it for the relevant color within the SVG without validation.

Here is a sample .bibliometrics.config.json (using the default order of the bibliometrics, and providing the year of first publication):

{
    "jsonOutputFile": "bibliometrics.json",
    "firstPubYear": 1999,
    "svgConfig": [
        {
            "background": "#010409",
            "border": "rgba(56,139,253,0.4)",
            "filename": "images/bibliometrics2.svg",
            "text": "#c9d1d9",
            "title": "#58a6ff"
        },
        {
            "background": "#f6f8fa",
            "border": "rgba(84,174,255,0.4)",
            "filename": "images/bibliometrics.svg",
            "text": "#24292f",
            "title": "#0969da"
        }
    ]
}

Here is another sample that generates three SVGs, overriding the default order at the top-level to exclude the i10-index (and related indexes), and then overriding it again for one of the three SVGs to additionally exclude the g-index, w-index, o-index, h-median, e-index, r-index, and a-index. This example also does not specify "firstPubYear", which means that the m-quotient won't be calculated:

{
    "jsonOutputFile": "bibliometrics.json",
    "include": [
        "total-cites", 
        "five-year-cites", 
        "most-cited", 
        "h-index", 
        "g-index", 
        "w-index",
        "o-index",
        "h-median",
        "e-index", 
        "r-index", 
        "a-index"
    ],
    "svgConfig": [
        {
            "background": "#010409",
            "border": "rgba(56,139,253,0.4)",
            "filename": "images/bibliometrics3.svg",
            "text": "#c9d1d9",
            "title": "#58a6ff"
        },
        {
            "background": "#010409",
            "border": "rgba(56,139,253,0.4)",
            "filename": "images/bibliometrics2.svg",
            "text": "#c9d1d9",
            "title": "#58a6ff",
            "include": ["total-cites", "five-year-cites", "most-cited", "h-index"]
        },
        {
            "background": "#f6f8fa",
            "border": "rgba(84,174,255,0.4)",
            "filename": "images/bibliometrics.svg",
            "text": "#24292f",
            "title": "#0969da"
        }
    ]
}

Configuring the Scholar ID

There are two ways to provide your Google Scholar ID to the utility:

  • in the configuration file (see above section) via a field "scholarID" (not shown in the example in the repository); or
  • via an environment variable SCHOLAR_ID.

Usage

Installing

To install from PyPi (Unix and MacOS):

python3 -m pip install bibliometrics

To install from PyPI (Windows):

py -m pip install bibliometrics

To upgrade to latest version from PyPi (Unix and MacOS):

python3 -m pip install --upgrade bibliometrics

To upgrade to latest version from PyPI (Windows):

py -m pip install --upgrade bibliometrics

Running

To use this utility, first ensure that you configure it as specified above. Then execute the following (Unix and MacOS):

python3 -m bibliometrics

Or on Windows:

py -m bibliometrics

Respect Google Scholar's robots.txt

If you use this utility, please respect Google Scholar's robots.txt. The reason that the g-index (as well as i100-index, i1000-index, i10000-index, w-index) is only computed by this utility if it is less than 100 derives from Scholar's robots.txt. Likewise, the reason that the e-index, r-index, and a-index are only computed if your h-index is at most 100 relates to this as well. In the case of the h-median, we can compute it as long as your h-index is less than 200. Here is the relevant excerpt:

Allow: /citations?user=
Disallow: /citations?*cstart=

The first line above allows getting a user's profile page. But the second line above disallows querying a specific starting reference. This means if you are respecting Google Scholar's robots.txt then you are limited to the first page of the profile, which can include up to 100 publications. Therefore, we can compute g-index as long as it is not greater than 100. However, if we compute 100, we cannot be certain if it is correct or if it is actually higher, so this utility excludes a g-index of 100 as well. The same is true for i100-index, i1000-index, and i10000-index. There is a similar issue for e-index, r-index, and a-index. In those cases, to compute the relevant bibliometric, we need the number of citations of your h most-cited articles. Therefore, we only compute these if your h-index is at most 100.

Support the Project

You can support the project in a number of ways:

  • Starring: If you find this utility useful, consider starring the GitHub repository.
  • Sharing with Others: Consider sharing it with others who you feel might find it useful.
  • Reporting Issues: If you find a bug or have a suggestion for a new feature, please report it via the Issue tracker. Please note, however, that feature requests will be evaluated for whether they can be implemented while respecting Google Scholar's robots.txt, which means limited to what can be computed from the first page of your Google Scholar profile.
  • Contributing Code: If there is an open issue that you think you can help with, submit a pull request.
  • Sponsoring: You can also consider sponsoring via one of the following:
GitHub Sponsors Liberapay
Ko-Fi

Contribute

If you would like to contribute start by reading the contribution guidelines. This project has adopted the Contributor Covenant Code of Conduct.

The intention of this utility is as a tool for a researcher to generate an SVG of their own bibliometrics, so issues requesting, or pull requests implementing, support for multiple scholar profiles will be rejected.

License

The code in this repository is released under the MIT License.