• Add more than 200 popular javascript projects from github to '/ dataset / words' for the quality of TF-IDF scores.
• Well-made open source is popular (assume)
  • However, popular open source is not always well-crafted. :)

purpose

• Extract Topic Word (using TF-IDF text mining) , a.k.a analyzeTopicWords

tasks

• parse source code @ref https://github.com/fkling/astexplorer
  • parse vue component
  • parse flow script
  • parse typescript
  • parse jsx
• Analyze 13 javascript-based projects that have received many star of github
  • Analyze name extraction / frequency
  • Separate names by word (by frequency)
  • Filter stopword from word list
    • Do I have to filter out customary words?
  • Saving structured data of this information as a text-based file
  • Automatically create final word list file by typing github address
  • filter only noun word
• Analyze 100 javascript-based projects that have received many star of [github]
  • download most star js github project 100 by script
• Make topic word cloud image
  • extract topic word list by specific top 100 project (using TF-IDF text mining)
  • make guide for word cloud image file(png)
  • support cli
    • commander.js
    • Refactoring scripts for publish.
• Prepare publish
  • Modifying README.md
  • versioning v0.2.0
  • publish npm

• Bitbucket
• Google Code
• Codeplex
• Sourceforge
• Fedora Project
• GitLab

ref

• https://github.com/wahengchang/js-meter

• support tsx file

How about supporting color set options..?

reference : https://github.com/vuejs/vue-cli source code