JToBI

JToBI, short for Java ToBI, is a Java-based toolkit for applying and developing ToBI-style prosodic annotation on speech corpora. It implements core elements of the ToBI (Tone and Break Indices) framework, providing a structured way to label tonal targets, boundary tones, and break indices that organize utterances into prosodic phrases. The system is designed to be extensible, enabling multiple languages and ToBI variants to be represented within a common architecture.

Key concepts in JToBI include tones (such as pitch accents and boundary tones), break indices that mark junctures between prosodic units, and hierarchical tier structures that map tonal events to time-aligned portions of speech. The toolkit supports language-specific adaptations, allowing researchers to implement English ToBI conventions as well as other language inventories and their particular phonological realizations.

Functionality and workflow typically involve preparing audio and transcripts, annotating utterances with ToBI-like tiers, and aligning annotations with phonetic or audio data. JToBI offers a data model and tooling for creating, editing, and validating annotations, along with import and export capabilities. It can export to common phonetic formats such as TextGrid and ToBI-style text formats, facilitating interoperability with other analysis tools and corpora.

Development and usage context: JToBI emerged from the ToBI community to provide a platform-independent, extensible Java implementation of ToBI. It is used in academic research and teaching to study prosodic patterns, compare language-specific intonation systems, and support large-scale annotation projects. As with many academic tools, users should verify compatibility with current workflows and related annotation standards.

See also: ToBI, PToBI, TToBI, IToBI.