Speech Analytics Guide
The system offers speech transcription and phonetic search to allow searching in voice conversations. Refer to the corresponding guides to understand the capabilities, language support, limitations and license requirements for each.
The following table provides a brief overview of the options:
| Speech Transcription | Scribe / Transcription for Trader Voice | Phonetic Search |
|---|---|---|---|
Technology | Large Vocabulary Continuous Speech Recognition (LVCSR) | Large Vocabulary Continuous Speech Recognition (LVCSR) | Phonetic Indexing |
Languages | 48 languages | English | English-GB, English-US |
Recognize persons, product, and company names with good accuracy | Yes, with limitations* | Yes, with limitations* | Yes |
Trained for financial services | No | Yes | No |
Storage requirement | Small, compressed text file | Small, compressed text file | Index files are about the same size as media files |
Recognition/Indexing time | Moderate | Moderate | Fast |
Search | Very Fast | Very Fast | Moderate** |
View context | Yes | Yes | No |
Text analytics | No | Yes | No |
Requires 3rd party servers | No | Yes | No |
* Might require phonetic boosting or language customization
** The system can search for pre-defined phrases automatically during indexing and the system is caching search results to make searching faster