System and method for discovering and exploring concepts
View Patent ↗A method for identifying concepts in a plurality of interactions includes: filtering, on a processor, the interactions based on intervals; creating, on the processor, a plurality of sentences from the filtered interactions; computing, on the processor, a saliency of each the sentences; pruning away, on the processor, sentences with low saliency for generating a set of informative sentences; clustering, on the processor, the sentences of the set of informative sentences for generating a plurality of sentence clusters, each of the clusters corresponding to a concept of the concepts; computing, on the processor, a saliency of each of the clusters; and naming, on the processor, each of the clusters.
1. A method for automatically performing voice interaction analytics by identifying concepts in a plurality of interactions between customers and agents of a contact center in a plurality of intervals, the method comprising:
identifying, on a processor of a voice interaction analytics system, the interactions occurring in one of the plurality of intervals;
creating, on the processor of the voice interaction analytics system, a plurality of sentences from the identified interactions;
computing, on the processor of the voice interaction analytics system, a saliency of each the sentences;
pruning away, on the processor of the voice interaction analytics system, sentences with low saliency for generating a set of informative sentences;
comparing, on the processor of the voice interaction analytics system, the sentences of the set of informative sentences to identify a plurality of groups of similar sentences;
generating, on the processor of the interaction analytics system, a plurality of sentence clusters, each of the sentence clusters corresponding to a concept of the concepts and comprising a group of similar sentences based on the interaction analytics of the plurality of interactions between customers and agents of the contact center;
computing, on the processor of the voice interaction analytics system, a saliency of each of the sentence clusters;
naming, on the processor of the voice interaction analytics system, each of the sentence clusters;
identifying, on the processor of the voice interaction analytics system, a plurality of sentences of an interaction of the interactions;
comparing, on the processor of the voice interaction analytics system, each of the sentences of the interaction with the plurality of sentence clusters to identify a plurality of matching sentence clusters;
labeling, on the processor of the voice interaction analytics system, the interaction with a plurality of matching concepts, each matching concept corresponding to one of the matching sentence clusters based on the interaction analytics of the plurality of interactions between customers and agents of the contact center in a plurality of intervals; and
displaying, on a display device, the number of interactions labeled with each of the matching concepts during the one of the plurality of intervals based on the interaction analytics of the plurality of interactions between customers and agents of the contact center.
2. The method of claim 1 , wherein the interactions comprise an output of a voice recognition system.
3. The method of claim 2 , further comprising filtering the output of the voice recognition system based on word confidence.
4. The method of claim 2 , wherein the voice recognition system is a large-vocabulary continuous speech recognition system.
5. The method of claim 1 , wherein the intervals are time intervals.
6. The method of claim 1 , wherein the clustering the sentences comprises:
selecting a plurality of template sentences from the set of informative sentences, each of the template sentences corresponding to one of the clusters;
computing a similarity of each sentence in the set of informative sentences to the template sentences;
assigning each of the sentences of the set of informative sentences to a cluster of the clusters in accordance with the computed similarities with the template sentences; and
removing clusters having fewer than a threshold number of assigned sentences.
7. The method of claim 6 , wherein the clustering the sentences further comprises, iteratively, selecting additional template sentences from sentences that were not assigned to a cluster and repeating the selecting the plurality of template sentences, computing the similarity of each sentence to the template sentences; the assigning the each of the sentences to the clusters; and removing clusters having fewer than the threshold number of assigned sentences until all sentences are assigned or until an iteration limit is reached.
8. The method of claim 1 , wherein the naming each of the clusters comprises, for each cluster:
computing a term frequency-inverse document frequency for each sentence in the cluster and
naming the cluster with a sentence of the cluster having the highest term frequency-inverse document frequency.
9. The method of claim 1 , further comprising, in response to generating at least one sentence cluster, outputting an alert, the alert comprising the name of the sentence cluster.
10. An automatic interaction analytics system configured to automatically perform voice interaction analytics by identifying concepts in a plurality of interactions between customers and agents of a contact center, the interaction analytics system comprising:
a processor; and
a memory, wherein the memory has stored thereon instructions that, when executed by the processor, causes the processor to perform voice analytics by identifying concepts in the plurality of voice interactions between the customers and the agents of the contact center in a plurality of intervals by:
identifying a plurality of voice interactions occurring in one of the plurality of intervals;
creating a plurality of sentences from the identified interactions;
computing a saliency of each the sentences;
pruning away sentences with low saliency for generating a set of informative sentences;
comparing the sentences of the set of informative sentences to identify a plurality of groups of similar sentences;
generating a plurality of sentence clusters, each of the sentence clusters corresponding to a concept of the concepts and comprising a group of similar sentences based on the voice interaction analytics of the plurality of interactions between customers and agents of the contact center;
computing a saliency of each of the sentence clusters;
naming each of the sentence clusters;
identifying a plurality of sentences of an interaction of the interactions;
comparing each of the sentences of the interaction with the plurality of sentence clusters to identify a plurality of matching sentence clusters;
labeling the interaction with a plurality of matching concepts, each matching concept corresponding to one of the matching sentence clusters based on the voice interaction analytics of the plurality of interactions between customers and agents of the contact center in a plurality of intervals; and
displaying, on a display device, the number of interactions labeled with each of the matching concepts during the one of the plurality of intervals based on the voice interaction analytics of the plurality of interactions between customers and agents of the contact center.
11. The voice interaction analytics system of claim 10 , wherein the interactions comprise an output of a voice recognition system.
12. The voice interaction analytics system of claim 11 , wherein the instructions further comprise instructions for filtering the output of the voice recognition system based on word confidence.
13. The voice interaction analytics system of claim 11 , wherein the voice recognition system is a large-vocabulary continuous speech recognition system.
14. The voice interaction analytics system of claim 10 , wherein the intervals are time intervals.
15. The voice interaction analytics system of claim 10 , wherein the clustering the sentences comprises:
selecting a plurality of template sentences from the set of informative sentences, each of the template sentences corresponding to one of the clusters;
computing a similarity of each sentence in the set of informative sentences to the template sentences;
assigning each of the sentences of the set of informative sentences to a cluster of the clusters in accordance with the computed similarities with the template sentences; and
removing clusters having fewer than a threshold number of assigned sentences.
16. The voice interaction analytics system of claim 15 , wherein the clustering the sentences further comprises, iteratively, selecting additional template sentences from sentences that were not assigned to a cluster and repeating the selecting the plurality of template sentences, computing the similarity of each sentence to the template sentences; the assigning the each of the sentences to the clusters; and removing clusters having fewer than the threshold number of assigned sentences until all sentences are assigned or until an iteration limit is reached.
17. The voice interaction analytics system of claim 10 , wherein the naming each of the clusters comprises, for each cluster:
computing a term frequency-inverse document frequency for each sentence in the cluster and
naming the cluster with a sentence of the cluster having the highest term frequency-inverse document frequency.