This R script contains a set of text and data mining algorithms.
Options:
-p INTEGER, --port=INTEGER
bind to port [default = 8080]
--shownodename
display node name in title (helpful to distiguish app running in different docker containers) [default = FALSE]
-s INTEGER, --somdim=INTEGER
som dimension [default = 35]
-i INTEGER, --somiterations=INTEGER
som iterations [default = 100]
-h, --help
Show this help message and exit
For input a text or text file can be used with the following definition
<label1>: <any text>
<label2>: <any text>
...
A working Shiny application can run online within a browser or manually started wih the command source("main.r")
. The application does the following steps:
- cleaning input text, stopword reduction, remove punctation and numbers
- runs the normalized compression distance to analyse dissimilarity of the text input
- runs the multidimensional scaling to create a 2D projection of the dissimilarity data
- runs affinity propagation of 2D projection to build groups of data
Visualisation of Wine Data