Enhanced customer experience means a higher measure of satisfaction, customer retention and increased selling effectiveness essential for any company to prosper – because, without happy customers, there is no business. To create the best customer experiences, especially for call centers, organizations must utilize the tools available on the market, just like CallMiner (News - Alert) has successfully done with its speech analytics tool.
Recent, CallMiner, a provider of speech and voice analytics and creator of Eureka, has addressed the long term issue of Phonetics vs. LVCSR in the call center. Based on its research, it found the pros and cons of each system:
Phonetic Speech Analytics
Pros:
- This approach generally has higher recall rate.
- Words that are not in a predefined vocabulary can still be found.
- Take into account what the possible sequences of sounds are and their frequency.
- Typically faster at processing (turning the phone calls into data).
Cons:
- Since there are many possible sequences in the lattice, the term may be found in many places where it was never actually said, for instance if the actual words were “see a list.”
- Requires a larger footprint for storage.
- Size of the “vocabulary” being used is very small as phonetics relies only on the sounds of the language and there are only few tens of unique phonemes in most languages.
- Search process itself is much slower since phonemes cannot be as efficiently indexed the way whole words can.
LVCSR Speech Analytics
Pros:
- Accuracy is much higher than just the single word lookup of a phonetic approach (accuracy is much higher than just the single word lookup of a phonetic approach).
- Search time is actually much faster than phonetic.
- Provides a transcript of the words around the key term.
- The larger contexts that the sounds occur in are taken into account as well.
Cons:
- Lower recall due to unusual words or recognition errors.
- Initial processing of the audio also takes a bit longer than with a phonetic approach.
- The words in the search terms need to be in the dictionary in order to be found by the transcription search engine.
With all these pros and cons, CallMiner’s conclusion on which system to choose "comes down to looking at what is 'best fit' is for your business problem in terms of cost, value, and manual effort. In general, phonetics is an appropriate fit for search-seldom applications. For larger enterprise applications like speech analytics and business intelligence mining, the LVCSR approach to speech analytics is better suited. In the grand scheme of things, using phonetics is a bit like using cassette tapes in an era of MP3s. It still works and has niche benefits, but for the most part the technology is outdated and there are more powerful options available to contact centers."
So, all in all, it is up to the organization to take a look at the above fact list and decide which speech recognition system fits it best. For more information about Phonetics vs. LVCSR, click here.
Edited by Alisen Downey