| CPC G10L 15/22 (2013.01) [G10L 15/063 (2013.01); G10L 15/16 (2013.01); G10L 15/30 (2013.01); G10L 2015/223 (2013.01)] | 19 Claims |
|
1. Apparatus for providing pre-processing of a user utterance prior to feeding utterance-related data to a sequential neural network classifier for conversation sentiment scoring, the utterance being expressed, by a user, to an interactive response system during an interaction between the user and the interactive response system, said apparatus comprising a plurality of distributed servers, each of the plurality of distributed servers comprising:
a conversation manager comprising a first processor for receiving a stateless application programming interface (“API”) request, the API request for storing, in a configurable memory, the utterance, previous utterance data and a sequence of labels, each label in the sequence of labels being associated with a previous utterance expressed by a user during the interaction, said previous utterance data being limited to a pre-determined number of utterances occurring immediately prior to the utterance;
a natural language processor, in electronic communication with the first processor, for processing the utterance to output an utterance intent, a semantic meaning of the utterance and an utterance parameter, the utterance parameter comprising one or more words included in the utterance and being associated with the intent, the natural language processor further configured to append the utterance intent, the semantic meaning of the utterance and the utterance parameter to the API request;
a signal extractor for processing the utterance, the utterance intent, the semantic meaning, the utterance parameter, and the previous utterance data extracted from the API request to generate a plurality of utterance signals, wherein the signal extractor is configured to append the plurality of utterance signals to the API request;
an utterance sentiment classifier comprising:
a memory for storing a hierarchy of rules, each rule in the hierarchy of rules being associated with one or more rule signals and a label;
a second processor for, in response to receiving the one or more utterance signals from the signal extractor, iterating through the hierarchy of rules to identify a first rule in the hierarchy for which the one or more utterance signals is a superset of the first rule's one or more rule signals;
wherein the second processor is further configured to append, to the sequence of labels stored in the API request, a label associated with the first rule;
the sequential neural network classifier for:
receiving a data input including the sequence of labels and the label associated with the first rule, the data input not including the utterance;
processing the data input using a trained algorithm; and
based on the processing, appending a sentiment score to the API request, said sentiment score being associated, within the API request, to the utterance;
the conversation manager for:
identifying a response to the user utterance based on the utterance intent, the label and the sentiment score;
appending the response to the API request; and
after the appending, transmitting the API request to the interactive response system; and
the interactive response system for receiving the API request and outputting the response included therein to the user;
wherein:
the pre-processing of the utterance by the natural language processor, the signal extractor and the utterance sentiment classifier reduces the data input to the label and the sequence of labels, thereby increasing a speed at which the sequential neural network classifier returns the sentiment score and decreasing resources consumed by the sequential neural network classifier when processing the data input.
|