Back to results
US 11,599,731 B2Granted
Generating recommendations by using communicative discourse trees of conversations
AssigneeOracle International Corporation
InventorBoris Galitsky
Life of this patent
Estimating the term…
Loading patent…
Loading patent…
AssigneeOracle International Corporation
InventorBoris Galitsky
Estimating the term…
Techniques are disclosed for improved autonomous agents that can provide a recommendation in a non-intrusive, conversational manner. In an aspect, a method determines a first sentiment score for a first utterance and a second sentiment score for a second utterance, each sentiment score indicating an emotion indicated by the respective utterance. The method further identifies that a difference between the first sentiment score and the second sentiment score is greater than a threshold. The method further extracts a noun phrase from the second utterance. The method identifies a text fragment that includes an entity that corresponds to the noun phrase. The method identifies that the text fragment addresses a claim of the second utterance. The method forms a third utterance that includes the a recommendation related to the second utterance and adds the third utterance to the sequence of utterances after the second utterance.
A method for providing a recommendation in conversational form, the method comprising:
A non-transitory computer-readable storage medium storing computer-executable program instructions, wherein when executed by a processing device, the program instructions cause the processing device to perform operations comprising:
A system comprising:
Dependencies are inferred from the claim text.
Drawings are temporarily unavailable. Open the PDF at USPTO
From the US filing history. Foreign family members aren’t listed.
Semantic analysis: discourse or dialogue representation
Querying: query formulation
Natural language analysis: grammatical analysis; Style critique
Phrasal analysis, e.g. finite state techniques or chunking: named entity recognition
About 14 years 8 months of term left · ends 14 Jun 2041 (estimated)
This disclosure is generally concerned with linguistics. More specifically, this disclosure relates to improved autonomous agents that use communicative discourse trees and other discourse techniques to provide a recommendation and identify an appropriate place at which to insert the recommendations into a conversation.
Computer-implemented applications of linguistics are increasing due to the greatly increased speed of processors and capacity of memory. For example, computer-based analysis of language discourse facilitates numerous applications, such as automated agents that can answer questions received from user devices.
But such solutions are unable to fully analyze utterances in a conversation between a user device and an autonomous agent. Consequently, these systems are unable to develop appropriate recommendations. Moreover, these systems are often overly intrusive from a user perspective.
FIG. 1 shows an exemplary rhetoric classification environment, in accordance with an aspect.
FIG. 2 depicts an example of a discourse tree in accordance with an aspect.
FIG. 3 depicts a further example of a discourse tree in accordance with an aspect.
Disclosed solutions provide technical improvements to the area of computer-implemented linguistics. In particular, disclosed solutions provide improved autonomous agents (“chatbots”) that can analyze textual information present in utterances of a dialogue (e.g., between a user device and an autonomous agent) to determine a changing sentiment in the dialogue. Based on the change in sentiment, disclosed solutions generate an appropriate recommendation, and insert the recommendation at an appropriate position the dialogue.
A goal of recommendation systems is to help obtain items of interest. To do so, recommendation systems determine interests of a user and present items that satisfy those interests. But in contrast to disclosed solutions, existing systems attempt to determine user interests by using various different techniques, such as explicitly asking questions or guiding a user through a sequence of iterations, suggesting specific items, and using feedback from users to refine suggestions in subsequent iterations. Such approaches are not only intrusive, but are also only functional in a limited set of domains and often cannot consistently provide relevant and timely recommendations. For instance, these approaches fail to take into consideration the huge amount of knowledge, both structured and non-structured, found in user utterances during a dialogue.
In contrast, using preference and intent information explicitly provided by a user, certain aspects use techniques such as rhetorical structure theory and communicative discourse trees (“CDTs”) to provide improved recommendations. In particular, certain aspects determine emotions expressed in text and an intent of a user and to match the intent with a recommendation. CDTs are discourse trees that are supplemented with one or more communicative actions. A communicative action is a cooperative action undertaken by individuals based on mutual deliberation and argumentation. Using communicative discourse trees and other techniques, improved performance over traditional statistical-based approaches can be accomplished, such as improved sentiment detection and logical analysis of sentences. Further, once a recommendation is constructed, certain aspects join a conversation when a recommendation is appropriate, thereby maximizing an effectiveness of the recommendation.