Automatically augmenting message exchange threads based on tone of message

1. A method implemented using one or more processors, comprising:

determining, from a message exchange thread involving one or more message exchange clients, a message contributed to the message exchange thread by a participant, wherein the message is directed by the participant to a personal assistant module participating in the message exchange thread as part of a conversation between the participant and the personal assistant module;

processing the message using one or more machine learning models to calculate a chatter score associated with the message and a search query suitability score associated with the message, wherein the chatter score represents a similarity between the message and previous conversational content known to be idle chatter, and wherein the search query suitability score represents a similarity between the message and previous messages that were directed to personal assistant modules in order to cause the personal assistant modules to conduct searches;

comparing the chatter score with the search query suitability score;

based on the comparing, formulating new content to be incorporated into the message exchange thread by selectively performing one or the other of the following:

formulating a search query based on the message, obtaining search results that are responsive to the search query, and formulating the new content using at least some of the responsive search results; or

formulating, as the new content, a conversational response to the message that does not include search results that are responsive to the search query; and

incorporating, as a message from the personal assistant module, the new content into the message exchange thread.

2. The method of claim 1 , wherein the incorporating comprises inserting the response into a transcript of the message exchange thread that is displayed in a graphical user interface of a message exchange client operating on a given client computing device.

3. The method of claim 1 , wherein calculating the chatter score comprises applying the message as input across a first machine learning model, wherein the first machine learning model provides, as output, the chatter score.

4. The method of claim 3 , wherein calculating the search query suitability score comprises applying the message as input across the first machine learning model, wherein the first machine learning model also provides, as output, the search query suitability score.

5. The method of claim 3 , wherein calculating the search query suitability score comprises applying the message as input across the second machine learning model, wherein the second machine learning model provides, as output, the search query suitability score.

6. The method of claim 3 , wherein the machine learning model is trained on at least one positive training example, wherein the at least one positive training example includes an instance in which one or more participants of a prior message exchange thread responded positively to incorporation of a conversational response to a prior message of the prior message exchange thread or incorporation of content responsive to a prior search query generated based on the prior message.

7. The method of claim 1 , wherein the search query suitability score is further calculated based at least in part on one or more known entities or entity types mentioned in the message.

8. The method of claim 1 , wherein the new content comprises a graphical element that is operable by a second participant to incorporate an automatically-generated conversational response to the message.

9. A system comprising one or more processors and memory storing instructions that, in response to execution of the instructions, cause the one or more processors to:

determine, from a message exchange thread involving one or more message exchange clients, a message contributed to the message exchange thread by a participant, wherein the message is directed by the participant to a personal assistant module participating in the message exchange thread as part of a conversation between the participant and the personal assistant module;

process the message using one or more machine learning models to calculate a chatter score associated with the message and a search query suitability score associated with the message, wherein the chatter score represents a similarity between the message and previous conversational content known to be chatter, and wherein the search query suitability score represents a similarity between the message and previous messages that were directed to personal assistant modules in order to cause the personal assistant modules to conduct searches;

based on a comparison of the chatter score with the search query suitability score, formulate new content to be incorporated into the message exchange thread, wherein the instructions to formulate include instructions to selectively perform one or the other of the following:

formulate a search query based on the message, obtaining search results that are responsive to the search query, and formulating the new content using at least some of the responsive search results; or

formulate, as the new content, a conversational response to the message that does not include search results that are responsive to the search query; and

incorporate, as a message from the personal assistant module, the new content into the message exchange thread.

10. The system of claim 9 , wherein the instructions to incorporate include instructions to insert the response into a transcript of the message exchange thread that is displayed in a graphical user interface of a message exchange client operating on a given client computing device.

11. The system of claim 9 , wherein the instructions to calculate the chatter score including instructions to apply the message as input across a first machine learning model, wherein the first machine learning model provides, as output, the chatter score.

12. The system of claim 11 , wherein the instructions to calculate the search query suitability score instructions to apply the message as input across the first machine learning model, wherein the first machine learning model also provides, as output, the search query suitability score.

13. The system of claim 11 , wherein the instructions to calculate the search query suitability score include instructions to apply the message as input across the second machine learning model, wherein the second machine learning model provides, as output, the search query suitability score.

14. The system of claim 11 , wherein the machine learning model is trained on at least one positive training example, wherein the at least one positive training example includes an instance in which one or more participants of a prior message exchange thread responded positively to incorporation of a conversational response to a prior message of the prior message exchange thread or incorporation of content responsive to a prior search query generated based on the prior message.

15. The system of claim 9 , wherein the search query suitability score is further calculated based at least in part on one or more known entities or entity types mentioned in the message.

16. The system of claim 9 , wherein the new content comprises a graphical element that is operable by a second participant to incorporate an automatically-generated conversational response to the message.

17. At least one non-transitory computer-readable medium comprising instructions that, in response to execution by one or more processors, cause the one or more processors to:

determine, from a message exchange thread involving one or more message exchange clients, a message contributed to the message exchange thread by a participant, wherein the message is directed by the participant to a personal assistant module participating in the message exchange thread as part of a conversation between the participant and the personal assistant module;

process the message using one or more machine learning models to calculate a chatter score associated with the message and a search query suitability score associated with the message, wherein the chatter score represents a similarity between the message and previous conversational content known to be chatter, and wherein the search query suitability score represents a similarity between the message and previous messages that were directed to personal assistant modules in order to cause the personal assistant modules to conduct searches;

based on a comparison of the chatter score with the search query suitability score, formulate new content to be incorporated into the message exchange thread, wherein the

instructions to formulate include instructions to selectively perform one or the other of the following:

formulate a search query based on the message, obtaining search results that are responsive to the search query, and formulating the new content using at least some of the responsive search results; or

formulate, as the new content, a conversational response to the message that does not include search results that are responsive to the search query; and

incorporate, as a message from the personal assistant module, the new content into the message exchange thread.

18. The at least one non-transitory computer-readable medium of claim 17 , wherein the instructions to incorporate include instructions to insert the response into a transcript of the message exchange thread that is displayed in a graphical user interface of a message exchange client operating on a given client computing device.

19. The at least one non-transitory computer-readable medium of claim 17 , wherein the instructions to calculate the chatter score including instructions to apply the message as input across a first machine learning model, wherein the first machine learning model provides, as output, the chatter score.

20. The at least one non-transitory computer-readable medium of claim 19 , wherein the instructions to calculate the search query suitability score instructions to apply the message as input across the first machine learning model, wherein the first machine learning model also provides, as output, the search query suitability score.