Aki Ariga

Field Data Scientist at Cloudera.

I’m interested in natural language processing and machine learning, and applying those technologies in production. I love to think how leverage data with technology for bussiness. I also love to write Ruby, Python and Julia for ML.

I am a podcaster of technical podcast rubyist.club.

Blogs

• Japanese Blog
• English Blog

Codes & Notebooks

Codes

• tabula-py
  • Simple wrapper of tabula-java: extract table from PDF into pandas DataFrame
• mecab-on-pyspark
  • Example code for distributed Python package with conda. This code enables to run C++ and Python code without install dependent packages
• TinySegmenter.jl
  • Julia implimentation of compact Japanese tokenizer
• Mykytea-python, Mykytea-ruby
  • Python and Ruby wrapper of KyTea, Japanese morophological analyzer
• sparkavro
  • Load Avro data into Spark with sparklyr
• docker-sphinx-recommonmark
  • Docker image for sphinx with recommonmark
• cloudera-sparklyr
  • Build script and Demo for Cloudera Director with Sparklyr
• cdsw-simple-serving-python
  • Example project for Cloudera Data Science Workbench
• cdsw-serve-docker
  • Dockernized for cdsw-simple-serving-python. An API server example for Cloudera Data Science Workbench
• homebrew-cloudera
  • Homebrew Formulas for cloudera tools
• cloudera-parcel
  • Cloudera parcel for R
• spacyr-sparklyr
  • Demo for spaCy on R with Sparklyr

Full code list, see GitHub

Notebooks

• ibis-demo
  • Demo notebooks of ibis, Python productivity framework for the Apache Hadoop ecosystem.
• chezou/notebooks
  • tutorial machine learning or data science, written in Japanese

Publications

• 「仕事ではじめる機械学習」オライリー・ジャパン ("Machine Learning at work" from O'reilly Japan. Japanese)

Presentations

• 2017-12-06: Starata Data Conference Singapore (Singapore)
  • Train, predict, and serve: How to put your machine learning model into production
  • Conference page
  • Slide
• 2017-11-07: Cloudera World Tokyo 2017 (Tokyo, Japan)
  • 機械学習システムのデプロイパターン (in Japanese)
  • Slide
• 2017-06-27: JSAI SIG-SWO #42 (Tokyo, Japan)
  • Invited talk: データサイエンティストからみた統合されたデータ分析基盤の恩恵 (in Japanese)
  • Slide
• 2017-06-27: Data Engineering and Data Analysis Workshop #1 (Tokyo, Japan)
  • Cloudera Data Science WorkbenchとPySparkを使って好きなPythonライブラリを分散で使う (in Japanese)
  • Slide
• 2017-02-07: Big Data Analytics Tokyo (Tokyo, Japan)
  • A data enginnering and data science platform based on Hadoop/Spark (in Japanese)
  • Slide
• 2016-11-08: Cloudera World Tokyo 2016 (Tokyo, Japan)
  • 大規模データに対するデータサイエンスの進め方 (in Japanese)
  • Slide

See more in SlideShare

Communities

• Founder of kawasaki.rb, regional Ruby community
  • Organized Kanagawa Ruby Kaigi 01 and Kawasaki Ruby Kaigi 01
• Founder of Machine Learning Casual Talks
• Co-founder of Julia Tokyo

Papers

See Google scholar

Patents

See Linkedin