Back to results
US 11,900,943 B2Granted
System and method of text zoning
AssigneeVerint Systems Ltd.
InventorsRoni Romano, Yair Horesh, Jeremie Dreyfuss
Life of this patent
Estimating the term…
Abstract
A method of zoning a transcription of audio data includes separating the transcription of audio data into a plurality of utterances. A that each word in an utterances is a meaning unit boundary is calculated. The utterance is split into two new utterances at a work with a maximum calculated probability. At least one of the two new utterances that is shorter than a maximum utterance threshold is identified as a meaning unit.
Claims24
A method of performing speech analytics, the method comprising:
- producing a set of meaning units by zoning a digital transcript file of audio data, wherein the zoning comprises:
- identifying a plurality of utterances within the digital transcript file representing audio data, the identifying performed by applying an acoustic model to the audio data or a linguistic model to the digital transcript file;
- for each utterance having fewer than a predetermined number of words, defining the utterance as a meaning unit for the set of meaning units;
- for each utterance having more than a predetermined number of words, defining the utterance as comprising more than one meaning unit and parsing data associated with the utterance into data windows of predetermined length of N words for determination of a meaning unit boundary within the utterance;
- and
- for each data window, for each word W1 . . . N within each data window, calculating a probability and entropy that the each word W1 . . . N is the meaning unit boundary by applying another n-gram linguistic model to the data window, the probability and the entropy indicative of words indicative of a change in speaker, words appearing before or after a particular word, word doublets or triplets are spoken together, or a likelihood of a short combination or words;
- splitting each data window into meaning units at a word Wk having a maximum calculated probability of being the meaning unit boundary, such that each meaning unit includes the word Wk having the highest probability indicating a new meaning unit;
- and
- applying automated speech analytics to the set of meaning units.
6 more elements
A non-transitory computer readable medium comprising instructions that, when executed by a processor of a processing system, cause the processing system to perform a method, the method comprising:
- produce a set of meaning units by zoning a digital transcript file of audio data, wherein the instructions to produce the set of meaning units further cause the processing system to:
- identify a plurality of utterances within the digital transcript file representing audio data, performed by applying an acoustic model to the audio data or a linguistic model to the digital transcript file;
- for each utterance having fewer than a predetermined number of words, define the utterance as a meaning unit for the set of meaning units;
- for each utterance having more than a predetermined number of words, define the utterance as comprising more than one meaning unit and parse data associated with the utterance into data windows of predetermined length of N words for determination of a meaning unit boundary within the utterance;
- and
- for each data window, for each word W1 . . . N within each data window, calculate a probability and entropy that the each word W1 . . . N is the meaning unit boundary by applying another n-gram linguistic model to the data window, the probability and the entropy indicative of words indicative of a change in speaker, words appearing before or after a particular word, word doublets or triplets are spoken together, or a likelihood of a short combination or words;
- split each data window into meaning units at a word Wk having a maximum calculated probability of being the meaning unit boundary, such that each meaning unit includes the word Wk having the highest probability indicating a new meaning unit;
- and
- apply automated speech analytics to the set of meaning units.
6 more elements
A system for performing speech analytics, comprising:
- a memory comprising executable instructions;
- a processor configured to execute the executable instructions and cause the system to:
- produce a set of meaning units by zoning a digital transcript file of audio data, wherein the instructions to produce the set of meaning units further cause the system to:
- identify a plurality of utterances within the digital transcript file representing audio data, performed by applying an acoustic model to the audio data or a linguistic model to the digital transcript file;
- for each utterance having fewer than a predetermined number of words, define the utterance as a meaning unit for the set of meaning units;
- for each utterance having more than a predetermined number of words, define the utterance as comprising more than one meaning unit and parse data associated with the utterance into data windows of predetermined length of N words for determination of a meaning unit boundary within the utterance;
- and
- for each data window, for each word W1 . . . N within each data window, calculate a probability and entropy that the each word W1 . . . N is the meaning unit boundary by applying another n-gram linguistic model to the data window, the probability and the entropy indicative of words indicative of a change in speaker, words appearing before or after a particular word, word doublets or triplets are spoken together, or a likelihood of a short combination or words;
- split each data window into meaning units at a word Wk having a maximum calculated probability of being the meaning unit boundary, such that each meaning unit includes the word Wk having the highest probability indicating a new meaning unit;
- and
- apply automated speech analytics to the set of meaning units.
8 more elements
Dependencies are inferred from the claim text.