A speech corpus (or spoken corpus) is a database of speech audio files and text transcriptions. In speech technology, speech corpora are used, among other things, to create acoustic models (which can then be used with a speech recognition or speaker identification engine). In linguistics, spoken corpora are used to do research into phonetic, conversation analysis, dialectology and other fields.
A corpus is one such database. Corpora is the plural of corpus (i.e. it is many such databases).
There are two types of speech corpora:
- Read Speech, which includes:
- * Book excerpts
- * Broadcast news
- * Lists of words
- * Sequences of numbers
- Spontaneous Speech, which includes:
- * Dialogs â between two or more people (includes meetings; one such corpus is the KEC);
- * Narratives â a person telling a story (one such corpus is the Buckeye Corpus);
- * Map-tasks â one person explains a route on a map to another;
- * Appointment-tasks â two people try to find a common meeting time based on individual schedules.
A special kind of speech corpora are non-native speech databases that contain speech with a foreign accent.
See also
References
- Edwards, Jane / Lampert, Martin (eds.) (1992): Talking Data â Transcription and Coding in Discourse Research. Hillsdale: Erlbaum.
- Leech, Geoffrey / Myers, Greg / Thomas, Jenny (eds.) (1995): Spoken English on Computer: Transcription, Markup and Application. Harlow: Longman.
External links