Code for our subtitle extraction algorithm used in our CHI 2014 paper: "Smart Subtitles for Vocabulary Learning"

Paper: CHI 2014

ACM DL

Video

Web app system code

Subtitle extraction system code

LaTeX sources for paper

Language learners often use subtitled videos to help them learn. However, standard subtitles are geared more towards comprehension than vocabulary learning, as translations are nonliteral and are provided only for phrases, not vocabulary. This paper presents Smart Subtitles, which are interactive subtitles tailored towards vocabulary learning. Smart Subtitles can be automatically generated from common video sources such as subtitled DVDs. They provide features such as vocabulary definitions on hover, and dialog-based video navigation. In our pilot study with intermediate learners studying Chinese, participants correctly defined over twice as many new words in a post-viewing vocabulary test when they used Smart Subtitles, compared to dual Chinese-English subtitles. Learners spent the same amount of time watching clips with each tool, and enjoyed viewing videos with Smart Subtitles as much as with dual subtitles. Learners understood videos equally well using either tool, as indicated by selfassessments and independent evaluations of their summaries.

MIT

Geza Kovacs