Early stage concept of how to learn Japanese using customised Twitter feeds.

This app consists of multiple steps:

• TWITTER: Connect and parse Twitter feed into a NoSQL database
  • Extract username, timestamp and tweet
  • Mecab (Japanese morphological analyser) to extract vocab/kanji compounds
  • Import into BlitzDB
• DICTIONARY: Convert JMDICT and KANJIDIC2 into a graph store for quick lookup (and just for fun)
  • Both JMDICT and KANJIDIC2 are in one database, which allows cross referencing without joins
• ANKI: Lookup definitions of each entry in BlitzDB in CayleyDB
  • Update BlitzDB entry with English meanings
  • Parse each document in the NoSQL database to be conducive to study
  • Import each document from BlitzDB into Anki
• Actually learn some Japanese instead of stuffing around with programming

• Python3.4
• Mecab
• CayleyDB

• blitzdb
• Mecab
• pprint
• requests
• tweepy
• lxml

• JMDICT
• KANJIDIC2

• Twitter import into BlitzDB
• JMDICT && KANJIDIC2 import into CayleyDB

• Command line arguments
• Anki Stuff
  • Anki plugin UI
  • Schema for Anki cards
  • Anki data import