Event-based semantic search and retrieval
A technique for semantic search and retrieval that is event-based, wherein is event is composed of a sequence of observations that are user speech or physical actions. Using a first set of conversations, a machine learning model is trained against groupings of utterances therein to generate a speech act classifier. Observation sequences therein are organized into groupings of events and configured for subsequent event recognition. A set of second (unannotated) conversations are then received. The set of second conversations is evaluated using the speech act classifier and information retrieved from the event recognition to generate event-level metadata that comprises, for each utterance or physical action within an event, one or more associated tags. In response to a query, a search is performed against the metadata. Because the metadata is derived from event recognition, the search is performed against events learned from the set of first conversations. One or more conversation fragments that, from an event-based perspective, are semantically-relevant to the query, are returned.
1. An event-based method of semantic search and retrieval, comprising:
providing a set of first conversations that have been annotated to identify speech acts, physical acts, and events, wherein a speech act is a labeled grouping of semantically-similar utterances, wherein a physical act is a non-linguistic action taken by an actor, and wherein an event is composed of a sequence of observations that are user speech or physical actions;
using the set of first conversations:
training a machine learning model against groupings of utterances to generate a classifier of speech acts; and
organizing a set of inter-related data tables, the set of inter-related data tables including at least a table of events identified, and a table of observations;
receiving a set of second conversations that are unannotated;
evaluating the set of second conversations using the machine learning model and information retrieved from the set of inter-related data tables to generate a set of event-level metadata, wherein the event-level metadata comprises, for a given utterance or physical action within an event, one or more associated tags; and
responsive to receipt of a query, performing a search against the event-level metadata, and returning a response.
2. The method as described in claim 1 wherein the query comprises a filter condition, and the response is based at least in part on the filter condition.
3. The method as described in claim 1 , further including filtering the response according to a filter condition.
4. The method as described in claim 1 wherein the query comprises one of: an utterance, and an ungrammatical collection of words.
5. The method as described in claim 1 wherein the one or more associated tags comprise a moment of interest tag.
6. The method as described in claim 1 wherein the response comprises a list of one or more events as identified from the set of first conversations that are associated to the query.
7. The method as described in claim 6 wherein at least one event in the list is an aliased event that represents an inexact match to an event expressed in the set of first conversations.
8. The method as described in claim 6 wherein the one or more events correspond to conversation fragments retrieved from the set of first conversations.
9. The method as described in claim 1 wherein at least some of the conversations in the set of first and second conversations have at least one or more turns, wherein a turn captures all consecutive utterances from a same conversational entity.
10. The method as described in claim 1 wherein at least some of the conversations in the set of first and second conversations are derived from one of: a human-to-human interaction, and a human-to-conversational bot interaction.
11. The method as described in claim 1 wherein the set of second conversations are received as a data stream in real-time or near real-time.
12. The method as described in claim 1 wherein the set of second conversations comprise an historical corpus of conversational transcripts.
13. The method as described in claim 1 wherein the query represents a conversational moment of interest.
14. The method as described in claim 13 wherein the conversational moment of interest is one of: a speech act label output from the first classifier, and an event label in the table of events.
15. A software-as-a-service computing platform, comprising:
computing hardware;
computer software executing on the computer hardware, the computer software comprising computer program instructions executed on the computing hardware and configured to provide event-based method of semantic search and retrieval with respect to a set of first conversations that have been annotated to identify speech acts, physical acts, and events, wherein a speech act is a labeled grouping of semantically-similar utterances, wherein a physical act is a non-linguistic action taken by an actor, and wherein an event is composed of a sequence of observations that are user speech or physical actions, the computer program instructions configured to:
using the set of first conversations:
train a machine learning model against groupings of utterances to generate a classifier of speech acts; and
organize a set of inter-related data tables, the set of inter-related data tables including at least a table of events identified, and a table of observations;
receive a set of second conversations that are unannotated;
evaluate the set of second conversations using the machine learning model and information retrieved from the set of inter-related data tables to generate a set of event-level metadata, wherein the event-level metadata comprises, for a given utterance or physical action within an event, one or more associated tags; and
responsive to receipt of a query, perform a search against the event-level metadata, and return a response.
16. A method to provide event-based semantic search and retrieval, comprising:
providing a set of first conversations that have been annotated to identify speech acts, physical acts, and events, wherein a speech act is a labeled grouping of semantically-similar utterances, wherein a physical act is a non-linguistic action taken by an actor, and wherein an event is composed of a sequence of observations that are user speech or physical actions;
using the set of first conversations:
training a machine learning model against groupings of utterances to generate a classifier of speech acts; and
organizing observation sequences into groupings of events that are configured for event recognition by one of: event pattern matching, and an event classifier;
receiving a set of second conversations that are unannotated;
evaluating the set of second conversations using the machine learning model and information retrieved from event recognition to generate a set of event-level metadata, wherein the event-level metadata comprises, for a given utterance or physical action within an event, one or more associated tags; and
responsive to receipt of a query, performing a search against the event-level metadata, and returning a response.
17. The method as described in claim 16 further including filtering the response according to at least one filter condition.
18. The method as described in claim 16 wherein the response comprises a list of one or more events as identified from the set of first conversations that are associated to the query.
19. The method as described in claim 18 wherein at least one event in the list is an aliased event that represents an inexact match to an event expressed in the set of first conversations.
20. The method as described in claim 16 wherein the response is a conversational fragment retrieved from the set of first conversations, the conversational fragment having an information structure that is semantically-similar to the query.