Abstract: A method is provided for dictionary expansion. The method acquires an object from a user and adds the object to a set of objects previously acquired from the user that form an expandable dictionary. The method calculates a centroid based on the set. The method calculates a similarity score of each of a plurality of objects relative to the centroid for each of a plurality of object features to calculate a weighted sum of similarity scores for each of the plurality of objects. The method presents candidate objects selected among the plurality of objects based on the weighted sum. The method acquires, from the user, a preferred candidate object among the candidate objects. The method updates weights of the plurality of features to maximize the weighed sum of similarity scores for the preferred candidate object. The method expands the dictionary by adding the preferred candidate object to the expandable dictionary.

Abstract: A method includes computing multiple term distances between pairs of multiple first string terms in a first string and multiple second string terms in a second string, generating a cost matrix based on the term distances, and selecting a set of candidate alignments based on the cost matrix. The method further includes generating multiple alignment scores for the set of candidate alignments, and selecting, from the set of candidate alignments, an alignment between the first string and the second string based on the alignment scores. The method further includes outputting a match identifier based on the alignment.

Abstract: A computer implemented method includes receiving text data, detecting auto-generated text in the received text data to identify tags in the received text to distinguish between the auto-generated text and user generated text, and providing the tagged text data to a machine learning language model.

Abstract: Systems and processes for intelligent device identification are provided. In one example process, audio input may be sampled with a microphone at each of two or more of the plurality of electronic devices. A first electronic device of the plurality of electronic devices for determining a task associated with sampled audio input may be identified. The process may determine the task based on the sampled audio input with the first electronic device and identify identifying a second electronic device of the plurality of electronic devices for performing the task. The task be performed with the second electronic device. The second electronic device is not the first electronic device in some examples.

Abstract: Provided is a method for preventing an accident related to children or pets that may occur by a home appliance using artificial intelligence. According to the present disclosure, the method for preventing the accident comprises comparing a distance between the home appliance and a generation position of the voice signal and a reference distance when the generation position of the voice signal is outside of the home appliance. Then, the present disclosure enables switching a door of the home appliance to a lock state when a distance between the home appliance and the generation position of the voice signal is less than the reference distance. Thus, the present disclosure may enable controlling the home appliance to prevent children or pets from entering an inside of the home appliance.

Abstract: Systems and methods for processing natural language statements. Based on historical records of data associated with an entity, systems and methods provide models for inferring publication of data content associated with the particular entity. The systems and methods may compare newly observed data content to predicted content associated with an entity for evaluating novelty or impact of the newly observed data content.

Abstract: Systems, methods performed by data processing apparatus and computer storage media encoded with computer programs for receiving an utterance from a user in a multi-user environment, each user having an associated set of available resources, determining that the received utterance includes at least one predetermined word, comparing speaker identification features of the uttered predetermined word with speaker identification features of each of a plurality of previous utterances of the predetermined word, the plurality of previous predetermined word utterances corresponding to different known users in the multi-user environment, attempting to identify the user associated with the uttered predetermined word as matching one of the known users in the multi-user environment, and based on a result of the attempt to identify, selectively providing the user with access to one or more resources associated with a corresponding known user.

Abstract: Some aspects disclosed herein are directed to, for example, a system and method comprising a client device receiving an input of at least a portion of a message. The client device may display, on a display of the client device, the at least the portion of the message. The client device may transmit, to a server device, the at least the portion of the message for display via a second client device. The client device may receive, from a user of the client device or a user of the second client device, a request to generate a marker for the at least the portion of the message. In response to the receiving the request to generate the marker, the client device may determine, based on a lexicon, a name for the marker for the at least the portion of the message.

Abstract: Systems and methods for data management using machine learning and artificial intelligence techniques related to topic modeling on text comments are described. The text comments may correspond to a particular transaction conducted by a user. Machine learning text analysis is performed on the text comment to determine one or more topics associated with the text comment. The topic with the highest correlation to the text comment is assigned to the transaction claim. Based on the topic assigned to the transaction claim, various actions may be performed, including remedial actions on a user account. These techniques may be applicable to chargeback fraud, in some embodiments.

Abstract: This disclosure describes systems and techniques receiving a request for information from a user and, in response, outputting the requested information along with unsolicited, interesting content that is related to, yet nonresponsive to, the requested information. In some instances, if the requested information is unknown, the techniques may output an indication that the information is unknown, followed by the additional, unsolicited, interesting content.

Abstract: This application relates to a new paradigm of data reasoning based on natural language inputs. A data structure based on the new paradigm is a canonical of a basic data structure in the shape of a triangle. The basic data structure includes a first node being a primary node, a second node being a context node, a third node being a resultant node. The basic data structure also includes a first link connecting the first and second nodes and configured to assign an attribute of abductive reasoning between the first and second nodes and a second link connecting the second and third nodes and configured to assign an attribute of inductive reasoning between the second and third nodes. The basic data structure further includes a third link connecting the first and third nodes and configured to assign an attribute of deductive reasoning between the first and third nodes.

Abstract: Aspects of the present disclosure relate to neural-based ontology generation and refinement. A set of input data can be received. A set of entities can be extracted from the set of input data using a named-entity recognition (NER) process, each entity having a corresponding label, the corresponding labels making up a label set. The label set can be compared to concepts in a set of reference ontologies. Labels that match to concepts in the set of reference ontologies can be selected as a candidate concept set. Relations associated with the candidate concepts within the set of reference ontologies can be identified as a candidate relation set. An ontology can then be generated using the candidate concept set and candidate relation set.

Abstract: A text-to-speech synthesis method using machine learning, the text-to-speech synthesis method is disclosed. The method includes generating a single artificial neural network text-to-speech synthesis model by performing machine learning based on a plurality of learning texts and speech data corresponding to the plurality of learning texts, receiving an input text, receiving an articulatory feature of a speaker, generating output speech data for the input text reflecting the articulatory feature of the speaker by inputting the articulatory feature of the speaker to the single artificial neural network text-to-speech synthesis model.

Abstract: A method and system for voice commands recognition. The system comprises a video camera and a microphone producing an audio/video recording of a user issuing vocal commands and at least one processor connected to the video camera and the microphone. The at least one processor has an associated memory having stored therein processor executable code causing the processor to perform the steps of: obtain the audio/video recording from the video camera and the microphone; extract video features from the audio/video recording and store the result in a first matrix; extract audio features from the audio/video recording and store the result in a second matrix; apply a speech-to-text engine to the audio portion of the audio/video recording and store the resulting syllables in a text file; and identify via a neural network the vocal commands of the user based on the first matrix, the second matrix and the text file.

Abstract: Disclosed herein is an artificial intelligence apparatus including an input interface configured to receive speech data, and a processor configured to detect a non-utterance interval included in the speech data and determine presence/absence of a second utterance after the non-utterance interval according to characteristics of a first utterance before the non-utterance interval, when the non-utterance interval exceeds a set time.

Abstract: Systems and methods for extracting data from unstructured data sources based on proximity co-reference resolution model. The method includes receiving an electronic document from an unstructured data source and extracting entities from the electronic document. The method also includes receiving fields to be extracted from the electronic document and generating keywords based on the fields. Each of the entities is associated with at least one of the fields. The method further includes identifying keywords in the electronic document based on the generated keywords and calculating, for each of the fields, proximity scores based on a proximity co-reference resolution model. The method also includes, for each of the fields, identifying a field-entity pair based on the calculated proximity scores and generating for display on a user device the field-entity pair.

Abstract: Techniques for entity and relationship detect from unstructured text as a service are described. A service may receive a request to identify entities within a provided unstructured text element, and the service may segment and tokenize the unstructured text and send the result to multiple services implementing multiple deep machine learning models trained to identify particular entities. The service may send additional requests to an additional service or services implementing additional deep machine learning models to identify relationships between detected attributes and ones of the detected entities. The outputs from all services can be analyzed and consolidated into a single result that identifies the entities, any attributes of the entities, and confidence scores indicating the confidence in each detected entity.

Abstract: Guided exploration of data is provided. A semantic graph corresponding to a dataset is generated using identified relations among columns of the dataset that are identified based on mapping the columns to main concepts in a generic ontology. A subgraph of the semantic graph is formed based on identification of nodes corresponding to relevant central concepts within the semantic graph using graph centrality metrics. A plurality of paths is identified in the subgraph using a bi-directional multiple hop search from the nodes corresponding to the relevant central concepts in the subgraph. A relevance score is assigned to each path in the plurality of paths using a graph-theoretic metric and the graph centrality metrics. A set of natural language questions based on relevant central concepts and concept relations corresponding to nodes in each respective path with an assigned relevance score greater than a minimum threshold score is output to the user.

Abstract: For a passage text and a corresponding answer text, perform a word-level soft alignment to obtain contextualized passage embeddings and contextualized answer embeddings, and a hidden level soft alignment on the contextualized passage embeddings and the contextualized answer embeddings to obtain a passage embedding matrix. Construct a passage graph of the passage text based on the passage embedding matrix, and apply a bidirectional gated graph neural network to the passage graph until a final state embedding is determined, during which intermediate node embeddings are fused from both incoming and outgoing edges. Obtain a graph-level embedding from the final state embedding, and decode the final state embedding to generate an output sequence word-by-word. Train a machine learning model to generate at least one question corresponding to the passage text and the answer text, by evaluating the output sequence with a hybrid evaluator combining cross-entropy evaluation and reinforcement learning evaluation.

Abstract: The present disclosure is generally directed to increasing the scalability of onboarding network resources, such as a digital component, to a voice-based network. The system enables the navigating of and interaction with digital components using voice or speech input and output interfaces on a computing device. The system can receive and processes an input audio signal to identify a digital component. The system enables voice-based interaction with the previously unregistered digital component via the input and output interfaces.