Conversational bot interaction with utterance ranking

Abstract

A conversational bot system uses a set of conversations that have been annotated to identify speech acts, wherein a speech act is a labeled grouping of utterances. To facilitate processing, a data model associated with a multi-turn conversation is received. The data model comprises an observation history. Upon receipt of query that includes a sequence of at least two or more utterances, an utterance ranking algorithm is applied. The algorithm selectively reorders the utterances in the sequence into a ranked order of importance that reflects a lowest to highest priority of response. In response to applying the utterance ranking algorithm, the data model is then updated to reflect the ranked order. In one embodiment, updating the data model positions the highest priority utterance as a most recent utterance in the observation history. The updated data model is then used to attempt to generate a coherent response to the query.

What is claimed is as follows: 
     
         1 . A method for imitating a human conversational response using a set of conversations that have been annotated to identify speech acts, and physical acts, wherein a speech act is a labeled grouping of utterances, comprising:
 in association with an automated conversational bot executing in a computing system:
 receiving a data model associated with a multi-turn conversation, the data model comprising an observation history; 
 upon receipt of a query that includes a sequence of utterances, applying an utterance ranking algorithm that outputs a ranked order of importance of the utterances in the sequence, the ranked order of importance reflecting a lowest to highest priority of response, and wherein a position of a given utterance in the ranked order of importance also reflects an importance of the given utterance within a context of the multi-turn conversation; 
 in response to applying the utterance ranking algorithm, updating the data model to reflect the ranked order; and 
 using the updated data model to attempt to generate a coherent response to the query for the automated conversational bot; and 
   the automated conversational bot returning the coherent response to the query.   
     
     
         2 . The method as described in  claim 1  wherein updating the data model positions a highest priority utterance as a most recent utterance in the observation history. 
     
     
         3 . The method as described in  claim 2  wherein using the data model to attempt to generate a coherent response searches for a correct action to a next highest priority utterance when a response to the highest priority utterance cannot be found. 
     
     
         4 . The method as described in  claim 1  wherein the given utterance is an utterance that has a low frequency of language use.