LATEST PROJECTS

ML Vocabulary Builder

For the sake of practicing what I have learned from Machine Learning and other data mining courses, I built my own application:

 

http://mlvocabularybuilder.elasticbeanstalk.com/

 

Motivation:

Typically we memorize vocabulary in any language by two ways: repeating or associating with known words.  However, most of the Language study focuses on the repeating aspect.   Studying the association behavior is difficult since each person has his or her own subtle way of establishing links between words that look similar(have same foreign language root), have similar pronounciation and have similar meaning.

 

Goal:

  • By collecting subject's familiarity on words, predict which of the remaining unexplored words that the subjects are most or least familiar with

  • By collecting subject's familiarity on words, predict which words the subject might be mixed up with others words.

  • Classify the word bank into different pools based on features that subjects are sensitive about for subject to study separately.

 

Features involved in the ML model

  • Foreign Language Root

  • Similarity with other words that have different meanings

  • Frequency of words(Difficulty)

  • Length of words

  • Part of Speech