Extending automatic transcripts in a unified data representation towards a prosodic-based metadata annotation and evaluation

This paper describes a framework that extends automatic speech transcripts in order to accommodate relevant information coming from manual transcripts, the signal itself, and other resources, like lexica. The proposed automatically collects, relates, computes, stores all together self-contained data source, making it possible easily provide wide range of interconnected suitable for analysis, training, evaluating number processing tasks. main goal this is integrate different linguistic paralinguistic layers knowledge more complete view their representation interactions several domains languages. chain composed two stages, where first consists integrating annotations recognition data, second further enriching previous output prosodic information. described has been used identification analysis structural metadata transcripts. Initially put use detection punctuation marks capitalization recovery also recently studying characterization disfluencies speech. It was already applied Portuguese corpora, English Spanish Broadcast News corpora.