Document clustering with basic machine learning
This is a pretty naive "script" in Go, that looks for *.md files in a directory or its subdirectories, reads them, parses them into words, stems and eliminates stopwords, and then does K-Means Clustering to categorize the texts.
It's a Sunday project to see if I could get auto-categorization for my blog posts.
It's not super-fast, and my guess is because it uses a bunch of maps of strings instead of arrays:
cluster-text $ time go run main.go ~/repos/xaprb/xaprb-src/content/blog > output.txt
real 26m18.683s
user 26m13.341s
sys 0m8.029s
Clustered 1158 docs with 9877 words into 50 clusters