Abstract: A device may receive text data, audio data, and video data associated with a user, and may process the received data, with a first model, to determine a stress level of the user. The device may process the received data, with second models, to determine depression levels of the user, and may combine the depression levels to identify an overall depression level. The device may process the received data, with a third model, to determine a continuous affect prediction, and may process the received data, with a fourth model, to determine an emotion of the user. The device may process the received data, with a fifth model, to determine a response to the user, and may utilize a sixth model to determine a context for the response. The device may utilize seventh models to generate contextual conversation data, and may perform actions based on the contextual conversational data.

Abstract: The disclosed system and method focus on automatically generating questions from input of written text and/or audio transcripts (e.g., learning materials) to aid in teaching people through testing their knowledge about information they have previously been presented with. These questions may be presented to an end user via a conversational system (e.g., virtual agent or chatbot). The user can iterate through each question, provide feedback for the question, attempt to answer the question, and/or get an answer score for each answer. The disclosed system and method can generate questions tailored to a particular subject by using teaching materials as input. The disclosed system and method can further curate the questions based on various conditions to ensure that the questions are automatically selected and arranged in an order that best suits the subject taught and the learner answering the questions.

Abstract: A device may receive training data that includes datasets associated with natural language processing, and may mask the training data to generate masked training data. The device may train a masked event C-BERT model, with the masked training data, to generate pretrained weights and a trained masked event C-BERT model, and may train an event aware C-BERT model, with the training data and the pretrained weights, to generate a trained event aware C-BERT model. The device may receive natural language text data identifying natural language events, and may process the natural language text data, with the trained masked event C-BERT model, to determine weights. The device may process the natural language text data and the weights, with the trained event aware C-BERT model, to predict causality relationships between the natural language events, and may perform actions, based on the causality relationships.

Abstract: Systems and methods that offer significant improvements to current chatbot conversational experiences are disclosed. The proposed systems and methods are configured to manage conversations in real-time with human customers based on a dynamic and unscripted conversation flow with a virtual assistant. In one embodiment, a knowledge graph or domain model represents the sole or primary source of information for the virtual assistant, thereby removing the reliance on any form of conversational modelling. Based on the information provided by the knowledge graph, the virtual agent chatbot will be equipped to answer customer queries, as well as demonstrate reasoning, offering customers a more natural and efficacious dialogue experience.

Abstract: Automated response generation systems and methods are disclosed. The systems can include a deep learning model specially configured to apply embedding techniques to redesign natural language querying systems for use over knowledge graphs. The disclosed systems and methods employ knowledge graph embedding to reduce knowledge graph sparsity by performing missing link prediction. The systems and methods described generate query graphs with increased flexibility and the ability to handle multi-hope constrained-based queries.

Abstract: A system and method for extracting suggestions from review text is disclosed. The disclosed methods include utilizing natural language processing techniques and knowledge graphs to extract implicit suggestions from review text. In this way, conflicting descriptions can be eliminated and similar descriptions can be consolidated. In later operations, the pruned knowledge graphs may be converted into textual summaries to provide more concise suggestions from the raw review text.

Abstract: A system and method for training and refining a machine learning model is disclosed. The disclosed system and method can further improve the accuracy of trained machine learning models by calculating which threshold values for predictions (e.g., probabilities output by the machine learning model) provide the most accurate results. The system and method may include applying an optimization technique (e.g., multi-objective optimization) to calculate which threshold values result in the best combination of precision and recall. In other words, the system and method adjust threshold values for prediction scores to optimize the objects of precision and recall. A machine learning model trained with these adjusted threshold values can determine when an input belongs to an unknown class because the unknown input has prediction scores below the threshold values for every known class. Embodiments may include refining an intent classifier to better classify unknown intents.

Abstract: The disclosed system and method focus on automatically generating questions from input of written text and/or audio transcripts (e.g., learning materials) to aid in teaching people through testing their knowledge about information they have previously been presented with. These questions may be presented to an end user via a conversational system (e.g., virtual agent or chatbot). The user can iterate through each question, provide feedback for the question, attempt to answer the question, and/or get an answer score for each answer. The disclosed system and method can generate questions tailored to a particular subject by using teaching materials as input. The disclosed system and method can further curate the questions based on various conditions to ensure that the questions are automatically selected and arranged in an order that best suits the subject taught and the learner answering the questions.

Abstract: Systems and methods that offer significant improvements to current virtual agent (VA) conversational experiences are disclosed. The proposed systems and methods are configured to manage conversations in real-time with human customers while accommodating a dynamic goal. The VA includes a goal-driven module with a reinforcement learning-based dialogue manager. The VA is an interactive tool that utilizes both task-specific rewards and sentiment-based rewards to respond to a dynamic goal. The VA is capable of handling dynamic goals with a significantly high success rate. As the system is trained primarily with a user simulator, it can be readily extended for applications across other domains.

Abstract: The present system and method may generally include organizing the task flow of a virtual agent in a way that is controlled by a set of rules and set of conditional probability distributions. The system and method may include receiving a user utterance including a first task, identifying the first task from the user utterance, and obtaining a set of rules related to the plurality of tasks. The set of rules may determine whether pre-tasks and/or pre-conditions are to be executed before executing the first task. The set of rules may also determine whether post-tasks and/or post-conditions are to be executed after executing the first task. The system and method may include executing the task; running a probabilistic graphical model on the plurality of tasks to determine a second task based on the first task; suggesting to the user the second task; and updating the probabilistic graphical model after a threshold number of runs.

Abstract: A device may receive text data, audio data, and video data associated with a user, and may process the received data, with a first model, to determine a stress level of the user. The device may process the received data, with second models, to determine depression levels of the user, and may combine the depression levels to identify an overall depression level. The device may process the received data, with a third model, to determine a continuous affect prediction, and may process the received data, with a fourth model, to determine an emotion of the user. The device may process the received data, with a fifth model, to determine a response to the user, and may utilize a sixth model to determine a context for the response. The device may utilize seventh models to generate contextual conversation data, and may perform actions based on the contextual conversational data.

Abstract: Techniques for building knowledge graphs from conversational data are disclosed. The systems include a high-performance relation classifier developed with active learning and requiring minimal supervision. The classifier is used to classify relation triples extracted from conversational text, which are then used to populate the knowledge graph. A heuristic for constructing the knowledge graph is also disclosed. The proposed embodiments provide a way to efficiently build and/or augment knowledge graphs and improve the quality of the generated responses by a dialogue agent despite a sparsity of data.

Abstract: A device may receive training data that includes datasets associated with natural language processing, and may mask the training data to generate masked training data. The device may train a masked event C-BERT model, with the masked training data, to generate pretrained weights and a trained masked event C-BERT model, and may train an event aware C-BERT model, with the training data and the pretrained weights, to generate a trained event aware C-BERT model. The device may receive natural language text data identifying natural language events, and may process the natural language text data, with the trained masked event C-BERT model, to determine weights. The device may process the natural language text data and the weights, with the trained event aware C-BERT model, to predict causality relationships between the natural language events, and may perform actions, based on the causality relationships.

Abstract: A device may receive first information that identifies an input associated with a virtual agent application executing on a user device. The virtual agent application may provide an interface for a project involving a plurality of user devices. The device may determine, based on the first information that identifies the input, a first response based on second information. The device may determine, based on at least one of the first information that identifies the input or the first response and without user input, a second response. The device may provide, to the virtual agent application of the user device, fourth information that identifies at least one of the first response or the second response.

Abstract: Systems and methods that offer significant improvements to current chatbot conversational experiences are disclosed. The proposed systems and methods are configured to manage conversations in real-time with human customers based on a dynamic and unscripted conversation flow with a virtual assistant. In one embodiment, a knowledge graph or domain model represents the sole or primary source of information for the virtual assistant, thereby removing the reliance on any form of conversational modelling. Based on the information provided by the knowledge graph, the virtual agent chatbot will be equipped to answer customer queries, as well as demonstrate reasoning, offering customers a more natural and efficacious dialogue experience.

Abstract: The present system and method may generally include organizing the task flow of a virtual agent in a way that is controlled by a set of rules and set of conditional probability distributions. The system and method may include receiving a user utterance including a first task, identifying the first task from the user utterance, and obtaining a set of rules related to the plurality of tasks. The set of rules may determine whether pre-tasks and/or pre-conditions are to be executed before executing the first task. The set of rules may also determine whether post-tasks and/or post-conditions are to be executed after executing the first task. The system and method may include executing the task; running a probabilistic graphical model on the plurality of tasks to determine a second task based on the first task; suggesting to the user the second task; and updating the probabilistic graphical model after a threshold number of runs.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for creating structured data using data received from unstructured textual data sources. One of the methods includes receiving unstructured textual data, identifying one or more keywords in the unstructured textual data, determining one or more patterns included in the unstructured textual data using the identified keywords, identifying one or more intelligence types that correspond with the unstructured textual data using the determined patterns, and associating, for each of the identified intelligence types, a data subset from the unstructured textual data with the respective intelligence type.

Abstract: Methods and systems including computer programs encoded on a computer storage medium, for interactive content recommendation. In one aspect, a method includes receiving a request for content by a user, determining a user intent based on the received request, providing to the user a first attribute responsive to the user intent, receiving a first attribute value responsive to the first attribute, providing a second attribute, and receiving a second attribute value responsive to the second attribute. A particular content vector including a first content attribute and a second content attribute for a particular content item is identified where the first content attribute and the second content attribute sufficiently match the first attribute value and the second attribute value. The particular content item is provided as a suggested content item, and, responsive to a user selection of the particular content item, provided for presentation on the user device.

Abstract: A system and method for spoken language understanding using recurrent neural networks (“RNNs”) is disclosed. The system and method jointly performs the following three functions when processing the word sequence of a user utterance: (1) classify a user's speech act into a dialogue act category, (2) identify a user's intent, and (3) extract semantic constituents from the word sequence. The system and method includes using a bidirectional RNN to convert a word sequence into a hidden state representation. By providing two different orderings of the word sequence, the bidirectional nature of the RNN improves the accuracy of performing the above-mentioned three functions. The system and method includes performing the three functions jointly. The system and method uses attention, which improves the efficiency and accuracy of the spoken language understanding system by focusing on certain parts of a word sequence. The three functions can be jointly trained, which increases efficiency.

Abstract: In the pharmaceutical research and development process, it may be necessary to process large amounts of medical records or clinical literature, to ensure safety of patients consuming a drug. A pharmacovigilance system may assist in this process by efficiently and automatically processing medical records to extract information and relationships contained therein and may also form a preliminary assessment regarding a medical or clinical judgment. The pharmacovigilance system may automatically generate reports based on this information, which may be validated by trained clinicians and medical experts.