Abstract: An agent automation system includes a memory configured to store a natural language understanding (NLU) framework and a model, wherein the model includes at least one original meaning representation. The system includes a processor configured to execute instructions of the NLU framework to cause the agent automation system to perform actions including: performing rule-based generalization of the model to generate at least one generalized meaning representation of the model from the at least one original meaning representation of the model; performing rule-based refinement of the model to prune or modify the at least one generalized meaning representation of the model, or the at least one original meaning representation of the model, or a combination thereof; and after performing the rule-based generalization and the rule-based refinement of the model, using the model to extract intents/entities from a received user utterance.

Abstract: The present approaches are generally related to an agent automation framework that is capable of extracting meaning from user utterances, such as requests received by a virtual agent (e.g., a chat agent), and suitably responding to these user utterances. In certain aspects, the agent automation framework includes a NLU framework and an intent-entity model having defined intents and entities that are associated with sample utterances. The NLU framework may include a meaning extraction subsystem designed to generate meaning representations for the sample utterances of the intent-entity model to construct an understanding model, as well as generate meaning representations for a received user utterance to construct an utterance meaning model. The disclosed NLU framework may include a meaning search subsystem that is designed to search the meaning representations of the understanding model to locate matches for meaning representations of the utterance meaning model.

Abstract: An agent automation system implements a virtual agent that is capable of learning new words, or new meanings for known words, based on exchanges between the virtual agent and a user in order to customize the vocabulary of the virtual agent to the needs of the user or users. The agent automation framework has access to a corpus of previous exchanges between the virtual agent and the user, such as one or more chat logs. New words and/or new meanings for known words are identified within the corpus and new word vectors are generated for these new words and/or new meanings for known words and added to refine a word vector distribution model. The refined word vector distribution model is then utilized by the agent automation system to interact with the user.

Abstract: The present disclosure is directed to an agent automation framework that is capable of extracting meaning from user utterances and suitably responding using a search-based natural language understanding (NLU) framework. The NLU framework includes a meaning extraction subsystem capable of detecting multiple alternative meaning representations for a given natural language utterance. Furthermore, the NLU framework includes a meaning search subsystem that enables elastic confidence thresholds (e.g., elastic beam-width meaning searches), forced diversity, and cognitive construction grammar (CCG)-based predictive scoring functions to provide an efficient and effective meaning search. As such, the disclosed meaning extraction subsystem and meaning search subsystem improve the performance, the domain specificity, the inference quality, and/or the efficiency of the NLU framework.

Abstract: An agent automation system includes a memory configured to store a natural language understanding (NLU) framework and a processor configured to execute instructions of the NLU framework to cause the agent automation system to perform actions. These actions comprise: generating an annotated utterance tree of an utterance using a combination of rules-based and machine-learning (ML)-based components, wherein a structure of the annotated utterance tree represents a syntactic structure of the utterance, and wherein nodes of the annotated utterance tree include word vectors that represent semantic meanings of words of the utterance; and using the annotated utterance tree as a basis for intent/entity extraction of the utterance.

Abstract: An agent automation system includes a memory configured to store a corpus of utterances and a semantic mining framework and a processor configured to execute instructions of the semantic mining framework to cause the agent automation system to perform actions, wherein the actions include: detecting intents within the corpus of utterances; producing intent vectors for the intents within the corpus; calculating distances between the intent vectors; generating meaning clusters of intent vectors based on the distances; detecting stable ranges of cluster radius values for the meaning clusters; and generating an intent/entity model from the meaning clusters and the stable ranges of cluster radius values, wherein the agent automation system is configured to use the intent/entity model to classify intents in received natural language requests.

Abstract: Present embodiments include an agent automation framework having a similarity scoring subsystem that performs meaning representation similarity scoring to facilitate extraction of artifacts to address an utterance. The similarity scoring subsystem identifies a CCG form of an utterance-based meaning representation and queries a database to retrieve a comparison function list that enables quantifications of similarities between the meaning representation and candidates within a search space. The comparison functions enable the similarity scoring subsystem to perform computationally-cheapest and/or most efficient comparisons before other comparisons. The similarity scoring subsystem may determine an initial similarity score between the particular meaning representation and the candidates of the search space, then prune non-similar candidates from the search space.

Abstract: A natural language understanding (NLU) framework includes a lookup source framework that enables the lookup sources to be created and applied to understanding utterances. Each lookup source is associated with a respective lookup source template that defines the compile-time and inference-time behavior of the lookup source. For example, a lookup source template indicates which plugins are used by the lookup source, and may define property values that determine the operational behavior of each of these plugins during compile-time and/or inference-time operation of the lookup source. The lookup source framework includes a template manager that manages lookup source templates and determines a suitable lookup source template for each lookup source.

Abstract: A natural language understanding (NLU) framework includes an ensemble scoring system designed to receive indicators determined by various systems of the NLU framework when inferencing a user utterance. The ensemble scoring system uses the received indicators, along with a set of ensemble scoring weights, to determine a respective ensemble score for each artifact of the utterance identified during inference. For example, segmentations provided by a lookup source system may be used to boost scores of intent and/or entities identified during a meaning search operation of a NLU system. The NLU framework may also include an ensemble scoring weight optimization subsystem that automatically determines optimized ensemble scoring weight values from labeled training data using an optimization plugin. Accordingly, the NLU framework enables these indicators to be suitably weighted and combined to provide a desired level of performance (e.g.

Abstract: A natural language understanding (NLU) framework includes an ensemble scoring system that uses received indicators, along with a set of ensemble scoring weights and ensemble scoring rules, to determine a respective ensemble score for each artifact of the utterance identified during inference. The ensemble scoring rules enable boosting of the respective ensemble score of an extracted intent of an utterance in response to a sufficient or important entity associated with the intent also being extracted from the utterance. Based on one or more ensemble scoring rules, the ensemble scoring system may refer to an intent-entity model to determine sufficient or important entities associated with an extracted intent, and boost the respective ensemble artifact score of the intent when the ensemble scoring system determines, with a suitable confidence, that a sufficient entity or important entity of the intent was extracted by the NLU framework during inference of the user utterance.

Abstract: A natural language understanding (NLU) framework includes an a concept system that performs concept matching of user utterances. The concept system generates a concept cluster model from sample utterances of an intent-entity model, and then trains a machine learning (ML) concept model based on the concept cluster model. Once trained, the concept model receives semantic vectors representing potential concepts extracted from utterances, and provides concept indicators to an ensemble scoring system. These concept indicators include indications of which concepts of the concept model that matched to the potential concepts, which intents of the intent-entity model are related to these concepts, and concept-relationship scores indicating a strength and/or uniqueness of the relationship between each concept-intent combination.

Abstract: A natural language understanding (NLU) framework includes a domain-aware vector encoding (DAVE) framework. The DAVE framework enables a designer to create a DAVE system having a domain-agnostic semantic (DAS) model and a corresponding trained vector translator (VT) model. The DAVE system uses the DAS model to generate domain-agnostic semantic vectors for portions of a user utterance, and then uses the VT model to translate the domain-agnostic semantic vectors into a domain-aware semantic vectors to be used by a NLU system of the NLU framework during a meaning search operation. The VT model is also designed to provide predicted intent classifications for the portions the user utterance. Both the NLU system and the DAVE system of the NLU framework are highly configurable and refer to various NLU constraints during operation, including performance constraints and resource constraints provided by a designer or user of the NLU framework.

Abstract: A natural language understanding (NLU) framework includes a lookup source framework, which enables a lookup source system to be defined having one or more lookup sources. The lookup source system can operate in a number of different manners to facilitate repository-aware inference of user utterances, for example, by facilitating vocabulary injection during compilation of an utterance meaning model and/or an understanding model. Additionally, the lookup source system can be leveraged to cleanse client-specific training data of sensitive values to generate generic training data that can be used to train the NLU framework of other clients. The lookup sources can be compiled in a synchronous or asynchronous manner, which enables lookup sources to be compiled in an on-demand basis from test source data. Additionally, understanding models that reference lookup sources can be periodically recompiled while leveraging the latest versions of the lookup sources for vocabulary injection.

Abstract: A natural language understanding (NLU) framework includes a lookup source framework, which enables a lookup source system to be defined having one or more lookup sources. In particular, taxonomy lookup sources can be compiled from suitable taxonomy source data that represents relationships between various entities within a domain of a client. These taxonomy lookup sources can extract taxonomy segmentations from utterances, such as received user utterances and sample utterances of an intent-entity model. The taxonomy segmentations can then be leveraged by the NLU system to perform vocabulary injection to expand the number of meaning representations in the utterance meaning model and/or the understanding model, increasing the likelihood of matches being located during a meaning search operation. Additionally, the taxonomy lookup sources can be leveraged by the NLU system to enable validation of sample utterances submitted for inclusion in the intent-entity model.

Abstract: A natural language understanding (NLU) framework includes a lookup source framework, which enables a lookup source system to be defined having one or more lookup sources. Each lookup source of the lookup source system includes a respective source data representation that is compiled from respective source data. For example, a source data representation may include source data arranged in a finite state transducer (IFST) structure as a set of finite-state automata (FSA) states, wherein each state is associated with a token that represents underlying source data. Different producers can be applied during compilation of a source data representation to derive additional states within the source data representation from the source data. Certain states of the source data representation that contain sensitive data can be selectively protected through encryption and/or obfuscation, while other portions of the source data representation that are not sensitive may remain in clear-text form.

Abstract: A natural language understanding (NLU) framework includes a lookup source system having one or more lookup sources. Each lookup source includes a respective source data representation that is compiled from respective source data. Once compiled, a user utterance can be submitted to the lookup source system, which generates segmentations of the user utterance. Each segmentation generally includes a collection of non-overlapping segments, and each segment generally describes how tokens of the user utterance can be grouped together and matched to the states of the source data representations. During lookup source inference, matches can be made to produced states or using fuzzy matchers that have corresponding of scoring adjustments. These scoring adjustments may be used by a segmentation scoring subsystem, potentially in combination with one or more additional segmentation scoring plugins, to score and rank the segmentations determined by the lookup source system for the user utterance.

Abstract: A natural language understanding (NLU) framework includes a modeling and optimization system that enables enhanced understanding and explainability to the operation of the NLU framework. The NLU framework includes a configuration vector storing settings of various components that may be applied during NLU inference of an utterance, such as which components should be activated or deactivated, as well as which numerical values (e.g., threshold values, coefficients, weight values) that are used by these components during operation. By using this configuration vector to systematically disable and adjust numerical parameters of the components of the NLU framework, and then determining the performance of the NLU framework in these configurations, the modeling and optimization system determines relationships between, as well as the relative importance of, the components of the NLU framework.

Abstract: Present embodiment include a prosody subsystem of a natural language understanding (NLU) framework that is designed to analyze collections of written messages for various prosodic cues to break down the collection into a suitable level of granularity (e.g., into episodes, sessions, segments, utterances, and/or intent segments) for consumption by other components of the NLU framework, enabling operation of the NLU framework. These prosodic cues may include, for example, source prosodic cues that are based on the author and the conversation channel associated with each message, temporal prosodic cues that are based on a respective time associated with each message, and/or written prosodic cues that are based on the content of each message. For example, to improve the domain specificity of the agent automation system, intent segments extracted by the prosody subsystem may be consumed by a training process for a ML-based structure subsystem of the NLU framework.

Abstract: The present approaches are generally related to an agent automation framework that is capable of extracting meaning from user utterances, such as requests received by a virtual agent (e.g., a chat agent), and suitably responding to these user utterances. In certain aspects, the agent automation framework includes a NLU framework and an intent-entity model having defined intents and entities that are associated with sample utterances. The NLU framework may include a meaning extraction subsystem designed to generate meaning representations for the sample utterances of the intent-entity model to construct an understanding model, as well as generate meaning representations for a received user utterance to construct an utterance meaning model. The disclosed NLU framework may include a meaning search subsystem that is designed to search the meaning representations of the understanding model to locate matches for meaning representations of the utterance meaning model.

Abstract: Present embodiment include a prosody subsystem of a natural language understanding (NLU) framework that is designed to analyze collections of written messages for various prosodic cues to break down the collection into a suitable level of granularity (e.g., into episodes, sessions, segments, utterances, and/or intent segments) for consumption by other components of the NLU framework, enabling operation of the NLU framework. These prosodic cues may include, for example, source prosodic cues that are based on the author and the conversation channel associated with each message, temporal prosodic cues that are based on a respective time associated with each message, and/or written prosodic cues that are based on the content of each message. For example, to improve the domain specificity of the agent automation system, intent segments extracted by the prosody subsystem may be consumed by a training process for a ML-based structure subsystem of the NLU framework.