This is a research project to automatically segment speech transcriptions into segments, conforming to the TED subtitling guidelines. This is an important step to simplify the process of video subtitling.