The scicloud package provides a word analysis for scientific papers which saved as PDFs. It clusters publication communities based on the words used in the papers. To identify the words relevant to the communities, it uses an indicator species analysis. In the end, it produces a wordcloud and visualizes the communities by year and number of citations with barplots. The package also offers a network analysis approach for which Gephi or a different graphing software is required.
Locate all the PDFs that you want to analyze in a folder called "PDFs" in your working directory, otherwise please provide a specific path to the directory of your PDFs when calling `createScicloudList function.
An API key from Elsevier is also required in order to retrieve additional information needed from Scopus. All papers and their abstracts that match in the search are then used for the analysis.
To acquire your API key, go to https://dev.elsevier.com/. In addition, make sure that you are connected to your institution's network (e.g. via VPN) when executing createScicloudList function.
Refer to the examples below to know how to use the package.
The scicloud functions should be executed in the following order:
createScicloudListrunAnalysisinspectScicloudThe result of createScicloudList and runAnalysis can be saved as a RDS file in your working directory by setting the parameter "saveToWd" to TRUE when executing these functions. To delete the RDS files in the working directory, simply execute deleteRDS and choose one of the delete options. Check the function's documentation for more details.
myAPIKey <- "YOUR_API_KEY"
# retrieving data from PDFs and Scorpus website using API
scicloudList <- createScicloudList(myAPIKey = myAPIKey)
# Run the analysis with a specified no. of cluster
scicloudAnalysis <- runAnalysis(scipusList = scipusList, numberOfClusters = 4)
# Generate a summary of the analysis
scicloudSpecs <- inspectScicloud(scicloudAnalysis)