Analyzing conversations to automatically identify action items

• US 10,181,326 B2
• Filed: 12/26/2017
• Issued: 01/15/2019
• Est. Priority Date: 06/01/2015
• Status: Active Grant

• Current Assignee: Zoominfo Converse LLC
• Original Assignee: AffectLayer, Inc. (ZoomInfo Technologies, Inc.)
• Inventors: Raanani, Roy, Levy, Russell, Breakstone, Micha Yochanan, Facher, Dominik
• Primary Examiner(s): Addy, Thjuan K
• Application Number:
  US15/854,642
• Publication Number:
  US 20180122383A1
• Time in Patent Office: 385 Days
• Field of Search: 37926507, 37926505, 37926506, 3792661, 37926503
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 17/26 Recognition of special voice characteristics, e.g. for use in lie detectors; Recognition of animal voices
• CPC Class Codes: G10L 2015/228 of application context
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 2203/357 Autocues for dialog assistance
• CPC Class Codes: H04M 3/5175 Call or contact centers supervision arrangements
• CPC Class Codes: H04M 3/5183 Call or contact centers with computer-telephony arrangements

Abstract

An action item identification system automatically determines action items by analyzing conversations of representatives with customers. The action item identification system retrieves recordings of various conversations, extracts features of each of the conversations, and analyzes the features to determine a set of features that is indicative of an action item associated with the corresponding conversation. The set of features is further analyzed to generate the action item in an action item manifest (a) as a summary of what is discussed in the conversations or (b) verbatim from the conversations.

Claims

1. A computer-implemented method, comprising:

retrieving multiple recordings of conversations associated with multiple representatives, wherein each of the conversations is between at least one of the representatives and at least one of multiple customers;

extracting multiple features from each of the recordings, wherein the multiple features indicate characteristics of any of (a) a customer of multiple customers in the corresponding conversation, (b) a representative of multiple representatives in the corresponding conversation, (c) the corresponding conversation;

analyzing features associated with a first conversation of the multiple conversations to determine a set of features that is indicative of an action item, wherein the action item is a task to be performed by a first customer of the multiple customers or a first representative of the multiple representatives associated with the first conversation; and

generating, based on the set of features, information regarding the action item.

2. The computer-implemented method of claim 1, wherein analyzing the features includes determining the set of features based on a usage of one or more words or phrases in the first conversation that are indicative of the action item.

3. The computer-implemented method of claim 2, wherein determining the set of features includes determining a specified word or a specified phrase in the first conversation which are explicitly indicative of the action item.

4. The computer-implemented method of claim 2, wherein determining the set of features includes determining a set of words or phrases in the first conversation based on which the action item is implicit.

5. The computer-implemented method of claim 4, wherein determining the set of words or phrases in the first conversation based on which the action item is implicit includes analyzing a portion of the first conversation prior to or after the set of words or phrases to determine the action item that is implied.

6. The computer-implemented method of claim 1, wherein generating the information regarding the action item includes generating the set of features, the set of features including one or more words or phrases in the first conversation that are indicative of the action item.

7. The computer-implemented method of claim 1, wherein generating the information regarding the action item includes generating a summary of the action item.

8. The computer-implemented method of claim 7, wherein generating the summary of the action item includes generating the summary by performing a semantic analysis of the set of features, the set of features including one or more words or phrases in the first conversation that are indicative of the action item.

9. The computer-implemented method of claim 7, wherein generating the summary of the action item includes generating the summary to include a context of the first conversation, the context including identification of the first customer of the first representative.

10. The computer-implemented method of claim 7, wherein the summary includes any information obtained using metadata associated with the first conversation.

11. The computer-implemented method of claim 1, wherein generating the summary includes generating the summary using any of a rule-based, an artificial intelligence, a machine learning, or natural language processing technique.

12. The computer-implemented method of claim 1 further comprising:

determining one or more features that are indicative of a next step, the next step being the action item or any activity other than the action item that is expected out of the first customer or the first representative subsequent to the first conversation.

13. The computer-implemented method of claim 12 further comprising:

generating information regarding the next step based on the one or more features, the one or more features including one or more words or phrases in the first conversation that are indicative of the next step.

14. The computer-implemented method of claim 1, wherein retrieving the multiple recordings include retrieving a data stream associated with a real-time conversation between the first representative and the first customer.

15. The computer-implemented method of claim 14 further comprising:

extracting one or more features from the real-time conversation that are indicative of the action item.

16. The computer-implemented method of claim 1, wherein extracting the features includes:

generating features that include a transcription, vocabulary and a language model of the conversations as a first output.

17. The computer-implemented method of claim 16, wherein extracting the features includes:

generating, using the first output, features that include semantic information from the conversations.

18. The computer-implemented method of claim 1, wherein extracting the features includes generating features that include data regarding conversation flow.

19. The computer-implemented method of claim 1, wherein extracting the features includes generating features related to a representative-customer pair in a conversation of the conversations.

20. The computer-implemented method of claim 1, wherein extracting the features includes extracting a visual feature associated with a conversation of the conversations.

21. The computer-implemented method of claim 1, wherein extracting the features includes extracting the features using any of an artificial intelligence, a machine learning, or natural language processing technique.

22. The computer-implemented method of claim 1, wherein analyzing the features of each of the conversations includes analyzing the features using any of an artificial intelligence, a machine learning, or natural language processing technique.

23. The computer-implemented method of claim 1, wherein at least one of the recordings includes a recording of a video call between one of the customers and one of the representatives.

24. The computer-implemented method of claim 1, wherein at least one of the recordings includes an online meeting between one of the customers and one of the representatives.

25. The computer-implemented method of claim 1, wherein at least one of the recordings includes a recording of a virtual reality-based conversation between one of the customers and one of the representatives.

26. The computer-implemented method of claim 1, wherein at least one of the recordings includes a recording of an augmented reality-based conversation between one of the customers and one of the representatives.

27. The computer-implemented method of claim 1, wherein at least one of the multiple recordings includes an e-mail conversation between one of the customers and one of the multiple representatives.

28. A non-transitory computer-readable storage medium storing computer-readable instructions, comprising:

instructions for retrieving multiple recordings of conversations associated with multiple representatives, wherein each of the conversations is between at least one of the representatives and at least one of multiple customers;

instructions for extracting multiple features from each of the recordings, wherein the multiple features indicate characteristics of any of (a) a customer of multiple customers in the corresponding conversation, (b) a representative of multiple representatives in the corresponding conversation, (c) the corresponding conversation;

instructions for analyzing features associated with a first conversation of the multiple conversations to determine a set of features that is indicative of an action item, wherein the action item is a task to be performed by a first customer of the multiple customers or a first representative of the multiple representatives associated with the first conversation; and

instructions for generating, based on the set of features, information regarding the action item, the information including (a) one or more words or phrases in the first conversation that are indicative of the action item or (b) a summary of the action item, the summary generated based on the one or more words or phrases in the first conversation that are indicative of the action item.

29. The non-transitory computer-readable storage medium of claim 28, wherein the instructions for determining the set of features include:

instructions for determining a set of words or phrases in the first conversation based on which the action item is implicit determining the set of features.

30. A system, comprising:

a first component that is configured to extract multiple features from a recording of a conversation, wherein the conversation is associated with at least one of multiple customers and at least one of multiple representatives;

a second component that is configured to analyze the features of the conversation to determine a set of features that is indicative of an action item, wherein the action item is a task to be performed by a first customer of the multiple customers or a first representative of the multiple representatives; and

a third component that is configured to generate information regarding the action item based on the set of features.