A R shiny application that generates a network of word cooccurrences given a sample text. The user can visualize the network and download a simple edge list of this network for further analysis. This application was created as a teaching tool for PL4246, Networks in Psychology.
Application flow:
- Upload a .txt file containing some text. Please keep this file size small-ish as the application/server is not optimized for large files.
- If your text data does not work, make sure that (i) it is a .txt file, (ii) there is no rich text formatting and (iii) add an extra hard return on the last line
- Wait a bit for the network to generate.
- Download the edge list if needed.
The application does not retain any user data.
Assumes you have R and RStudio installed and know your way around it.
The following packages are needed: shiny, igraph, tidyverse, tokenizers - install via install.packages('PackageName')
Download the files in this repository by clicking on 'Code', 'Download ZIP'.
Open the app.R file in Rstudio, make sure your working directory is pointed to wherever the app files are at, and then click on 'Run App'.
A demo of this application can be found here: http://r-server.csqsiew.xyz/cooccurNet/
- Words that co-occur within a window size of 2 are connected (i.e., bigrams)
- Words that straddle different sentences are not considered as co-occurring.
output.csvis a directed edge list (from -> to) with frequency of occurrence (counts) in the 3rd column.- See
lorem.txtandoutput.csvfor examples of input and output files.
last updated: 12th April 2024
Please feel free to adapt for your own use.
If you found this useful, feel free to tip me here: https://ko-fi.com/csqsiew or check out my work here: http://hello.csqsiew.xyz/
Thanks! :)