Method and apparatus for automatic correlation of multi-channel interactions

1. A method for categorizing interactions in a call center of an organization, comprising:

capturing in the call center at least one vocal interaction and at least one non-vocal interaction, using logging or capturing devices, wherein the at least one vocal interaction and the at least one non-vocal interaction are captured in accordance with pre-defined rules that regulate which interaction is to be captured, and wherein the at least one vocal interaction and the at least one non-vocal interaction having dissimilar contents of common semantic context;

retrieving at least one first word from the at least one vocal interaction;

retrieving at least one second word from the at least one non-vocal interaction;

assigning the at least one vocal interaction into a first category, wherein the first category is based on technical data of the at least one vocal interaction, using the at least one first word;

assigning the at least one non-vocal interaction into a second category, wherein the second category is based on technical data of the at least one non-vocal interaction, using the at least one second word; and

associating the first category and the second category into a multi-channel category based on the common semantic context thereof, thus aggregating the at least one vocal interaction and the at least one non-vocal interaction.

2. The method of claim 1 wherein the first category and the second category are pre-determined sub-categories, and wherein associating the first category and the second category comprises unifying the pre-determined sub-categories.

3. The method of claim 1 wherein the first category and the second category are clusters, and wherein associating the first category and the second category comprises unifying the clusters.

4. The method of claim 1 further comprising performing initial filtering for selecting interactions to be processed.

5. The method of claim 4 wherein the initial filtering relates to Computer-Telephony-Integration data or to Customer-Relationship-Management data or to meta data associated with the at least one vocal interaction or the at least one non-vocal interaction.

6. The method of claim 1 further comprising filtering the least one vocal interaction.

7. The method of claim 6 wherein filtering the at least one vocal interaction relates to emotion detection, talk analysis, accent identification or language identification.

8. The method of claim 1 further comprising filtering the least one non-vocal interaction.

9. The method of claim 8 wherein filtering the at least one non-vocal interaction relates to sentiment analysis.

10. The method of claim 1 further comprising normalization and injection of the at least one vocal interaction or the least one non-vocal interaction into a unified format.

11. The method of claim 1 further comprising analyzing the multi-channel category.

12. An apparatus for multi-channel categorizing interactions in a call center of an organization, comprising:

a logging or capturing component for capturing in the call center at least one vocal interaction and at least one non-vocal interaction, wherein the at least one vocal interaction and the at least one non-vocal interaction are captured in accordance with pre-defined rules that regulate which interaction is to be captured, and wherein the at least one vocal interaction and the at least one non-vocal interaction having dissimilar contents of common semantic context;

an audio analysis engine for retrieving at least one first word from the at least one vocal interaction;

a text analysis engine for retrieving at least one second word from the at least one non-vocal interaction;

a grouping component for assigning the at least one vocal interaction into a first category using the at least one first word and assigning the at least one non-vocal interaction into a second category using the at least one second word, wherein the first category is based on technical data of the at least one vocal interaction and the second category is based on the at least one non-vocal interaction; and

an aggregation component for associating the first category and the second category based on the common semantic context thereof into a multi-channel category, thus aggregating the at least one vocal interaction and the at least one non-vocal interaction.

13. The apparatus of claim 12 wherein the first category and the second category are pre-determined sub-categories, and wherein the aggregation component unifies the first category and the second category into one or more categories.

14. The apparatus of claim 12 wherein the first category and the second category are clusters, and wherein the aggregation component unifies the first category and the second category into one or more clusters.

15. The apparatus of claim 12 wherein the audio analysis engine is a word spotting engine or a speech to text engine.

16. The apparatus of claim 12 further comprising one or more of the engines selected from the group consisting of: emotion detection engine; talk analysis engine; call flow analysis engine; accent identification engine; and language identification engine.

17. The apparatus of claim 12 further comprising a normalization and injection component for normalizing and injecting the at least one vocal interaction or the least one non-vocal interaction into a unified format.

18. The apparatus of claim 12 further comprising an advanced analysis engine for analyzing the multi-channel category.

19. The apparatus of claim 12 further comprising a filtering component for filtering out interactions unsuitable for further analysis.

20. A non-transitory computer readable storage medium containing a set of instructions for a general purpose computer, the set of instructions comprising:

capturing in the call center at least one vocal interaction and at least one non-vocal interaction, using logging or capturing devices, wherein the at least one vocal interaction and the at least one non-vocal interaction are captured in accordance with pre-defined rules that regulate which interaction is to be captured, and wherein the at least one vocal interaction and the at least one non-vocal interaction having dissimilar contents of common semantic context;

retrieving at least one first word from the at least one vocal interaction;

retrieving at least one second word from the at least one non-vocal interaction;

assigning the at least one vocal interaction into a first category, wherein the first category is based on technical data of the at least one vocal interaction, using the at least one first word;

assigning the at least one non-vocal interaction into a second category, wherein the second category is based on technical data of the at least one non-vocal interaction, using the at least one second word; and

associating the first category and the second category into a multi-channel category based on the common semantic context thereof, thus aggregating the at least one vocal interaction and the at least one non-vocal interaction.