Scimeetr

• Install
• Introduction
• From data to reading list
  • loading and exploring bibliometric data
  • Mapping scientific community
  • Automatically generating a reading list of seminal papers
• In depth description of each steps
  • How to get bibliometric data?
    • Retrieving data from Scopus
    • Retrieving data from Web of Science
  • How to upload bibliometric data into R
  • Exploring scimeetr data
    • Printing and summary
    • Characterizing the corpus of papers contained by a scimeetr object
      • Characterize the corpus with keywords
      • Characterize the corpus with journals
      • Characterize the corpus with authors
    • Scimeetr object structure and navigation
  • Making reading lists
    • Core
      • Most cited per year
      • More cited than expected
    • Experts
      • Recent paper of a few experts
      • Paper of experts group
    • Centrality
      • Betweeness
      • Closeness
      • Connectness
      • Page rank
      • Cite most others
  • Finding the main communities of research
  • Focusing on a sub-community
  • Dive to a sub-community

Install

scimeetr can be installed directly from the R console using the following lines :

if (!require("devtools")) install.packages("devtools")

devtools::install_github("MaximeRivest/scimeetr")

Introduction

Scimeetr helps explore the scholarly literature. It contains a suit of function that let someone:

• load bibliometric data into R
• make a map of peer reviewed papers by creating various networks
• find research community
• characterise the research communities
• generate reading list

This tutorial is composed of two self-contained section. The first section show case the whole process with all the default parameters. The second section describes each function in more detail by presenting the rational for the function, the algorithms used and the options.

<a href="#top">Back to top</a>

From data to reading list

You can automatically generate a reading list of seminal papers in a research litterature by using only those three functions: ìmport_wos_files, scimap, and scilist. This first section describes this process in more details.

<a href="#top">Back to top</a>

loading and exploring bibliometric data

The first step in exploring the literature is to retrieve bibliometric data from the Web of Science or Scopus. In this first tutorial I use a dataset from the Web of Science about ecological networks.

library(scimeetr)
scimeetr_list <- import_wos_files("path/to/folder/")

Then,summary can be used to get a quick characterisation of the data.

summary(scimeetr_list)

## 
##  # Summary of Scimeetr #
## -----------------------
##     Number of papers:  742
##     Number of different reference:  28526
## 
##     Average number of reference per paper:  51
## 
##     Quantiles of total citation per paper: 
## 
##      0%     25%     50%     75%    100% 
##    0.00    2.00    7.00   19.75 1333.00 
## 
##     Mean number of citation per paper:  19.81536
## 
##     Average number of citation per paper per year:  1.2
## 
## 
##   Table of the 10 most mentionned keywords 
## 
##                       Keyword    Frequency
## 1                BIODIVERSITY           57
## 2                 AGRICULTURE           46
## 3  COMMON AGRICULTURAL POLICY           32
## 4          ECOSYSTEM SERVICES           31
## 5                CONSERVATION           28
## 6    AGRI-ENVIRONMENT SCHEMES           27
## 7     AGRI-ENVIRONMENT SCHEME           20
## 8  AGRI-ENVIRONMENTAL SCHEMES           19
## 9         AGRICULTURAL POLICY           18
## 10              WATER QUALITY           18
## 
## 
## 
##   Table of the 10 most productive journal 
## 
##                                             Journal    Frequency
## 1                                   LAND USE POLICY           84
## 2              AGRICULTURE ECOSYSTEMS & ENVIRONMENT           37
## 3               JOURNAL OF ENVIRONMENTAL MANAGEMENT           33
## 4                           BIOLOGICAL CONSERVATION           24
## 5                        JOURNAL OF APPLIED ECOLOGY           21
## 6                              ECOLOGICAL ECONOMICS           17
## 7                          JOURNAL OF RURAL STUDIES           17
## 8                              AGRICULTURAL SYSTEMS           14
## 9  JOURNAL OF ENVIRONMENTAL PLANNING AND MANAGEMENT           14
## 10                     LANDSCAPE AND URBAN PLANNING           14
## 
## 
## 
##   Table of the most descriminant keywords 
## 
##        comID                      tag
## 1 com1 (742)                         
## 2                        BIODIVERSITY
## 3                        CONSERVATION
## 4                          MANAGEMENT
## 5                         AGRICULTURE
## 6            AGRI-ENVIRONMENT SCHEMES
## 7                  ECOSYSTEM SERVICES

From this summary, we see that there is 396 papers in my data set which overal cites 16567 different elements. On average, each paper cites 53 elements.

Than we learn that, in this research community, 25% of the papers are cited less than 2 times, 50% are cited less than 9 times and 75% are cited less than ~23 times. There are papers that are cited up to 1333 times. The average citation per paper is ~25. This is much higher than the median (9), thus most paper are cited only a few times and a few papers are profusely cited. When correcting for the age of the paper, we learn that papers are cited 2 times per year on average.

By looking at the most frequent keyword and journals, we learn that this community of research is about biodiversity, agriculture, ecosystem services and policy. Keyword and journal frequency tables efficiently reveal the theme of a scientific community.

<a href="#top">Back to top</a>

Mapping scientific community

The previous characterisation is great, but it is limited if your dataset contains many different scientific communities. By detecting the scientific communities present within a dataset a map of science can be drawn and each cluster can be characterised on its own. The function scimap can be used for this task.

scimap_result <- scimap(scimeetr_list)

The function returns all the data that scimeetr_list contained and more. For example communities have been identified and now if the function summary is used on scim_result. In addition of the previous information. The descriminant keywords of each communities constituating the main community are listed.

summary(scimap_result)

## 
##  # Summary of Scimeetr #
## -----------------------
##     Number of papers:  742
##     Number of different reference:  28526
## 
##     Average number of reference per paper:  51
## 
##     Quantiles of total citation per paper: 
## 
##      0%     25%     50%     75%    100% 
##    0.00    2.00    7.00   19.75 1333.00 
## 
##     Mean number of citation per paper:  19.81536
## 
##     Average number of citation per paper per year:  1.2
## 
## 
##   Table of the 10 most mentionned keywords 
## 
##                       Keyword    Frequency
## 1                BIODIVERSITY           57
## 2                 AGRICULTURE           46
## 3  COMMON AGRICULTURAL POLICY           32
## 4          ECOSYSTEM SERVICES           31
## 5                CONSERVATION           28
## 6    AGRI-ENVIRONMENT SCHEMES           27
## 7     AGRI-ENVIRONMENT SCHEME           20
## 8  AGRI-ENVIRONMENTAL SCHEMES           19
## 9         AGRICULTURAL POLICY           18
## 10              WATER QUALITY           18
## 
## 
## 
##   Table of the 10 most productive journal 
## 
##                                             Journal    Frequency
## 1                                   LAND USE POLICY           84
## 2              AGRICULTURE ECOSYSTEMS & ENVIRONMENT           37
## 3               JOURNAL OF ENVIRONMENTAL MANAGEMENT           33
## 4                           BIOLOGICAL CONSERVATION           24
## 5                        JOURNAL OF APPLIED ECOLOGY           21
## 6                              ECOLOGICAL ECONOMICS           17
## 7                          JOURNAL OF RURAL STUDIES           17
## 8                              AGRICULTURAL SYSTE