This repository contain the process that how I deal with Japanese manifestos and conduct a LDA model on the data.
I use Python as main language in processing the txt files and build the csv file that can be use for further data cleaning and preprocessing. And I use Qnanteda package in R to run NLP processing, including building corpus, defining and delimiting textual features, and conducting topic modeling and LDA.
For more information about Quanteda, users could directly read their Source code from: https://github.com/quanteda/quanteda/ As for the data of Japan House of Representative (HOR), users could access it through: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/QFEPXD