Abstract: Computer-implemented methods, computer program products, and computer systems for mitigating frequency loss may include one or more processors configured for receiving first audio data corresponding to unobstructed user utterances, receiving second audio data corresponding to first obstructed user utterances, generating a frequency loss (FL) model representing frequency loss between the first audio data and the second audio data, receiving third audio data corresponding to one or more second obstructed user utterances, processing the third audio data using the FL model to generate fourth audio data corresponding to a frequency loss mitigated version of the second obstructed user utterances, and transmitting the fourth audio data to a recipient computing device. The first obstructed user utterances are obstructed by a facemask and the one or more second obstructed user utterances is obstructed by the facemask. The FL model may be executed as an audio plugin in a web conferencing program.

Abstract: Provided is a method, computer program product, and system for predicting image sharing decisions using machine learning. A processor may receive a set of annotated images and an associated text input from each user of a plurality of users. The processor may train, using the set of annotated images and the associated text input from each user, a neural network model to output an image sharing decision that is specific to a user.

Abstract: Computer-implemented methods, computer program products, and computer systems for mitigating frequency loss may include one or more processors configured for receiving first audio data corresponding to unobstructed user utterances, receiving second audio data corresponding to first obstructed user utterances, generating a frequency loss (FL) model representing frequency loss between the first audio data and the second audio data, receiving third audio data corresponding to one or more second obstructed user utterances, processing the third audio data using the FL model to generate fourth audio data corresponding to a frequency loss mitigated version of the second obstructed user utterances, and transmitting the fourth audio data to a recipient computing device. The first obstructed user utterances are obstructed by a facemask and the one or more second obstructed user utterances is obstructed by the facemask. The FL model may be executed as an audio plugin in a web conferencing program.

Abstract: In response to determining that a native language of a first user is different from a target language of a message to be transmitted by the first user to a second user, a translation model based on a plurality of language efficacies of the first user is created. An optimal action associated with a translation of the message from the native language to the target language is determined based on the created model and a language efficacy of the first user in the native language. The determined optimal action is performed. The message translation comprising the performed optimal action is transmitted to the second user.

Abstract: A method of configuring a widget defining a user interface includes parsing content from a corpus comprising documents describing a plurality of products, selecting a plurality of pillars from the content, wherein the pillars are descriptors describing a product space of the products, determining an affinity of each of the products to each of the pillars, receiving, via the widget, a selection of a first pillar, and displaying a given product in the widget, the given product selected from the products given the selection of the first pillar.

Abstract: A computer receives reviews for an item from a plurality of sources. The computer may identify one or more key features of the item. The computer collects user preferences for the one or more key features of the item. The computer calculates odds ratio for each of the one or more key features of the item. The computer determines an affinity measure for each of the one or more key features based on the calculated odds ratio for each of the one or more key features of the item; and generates a forest plot for the item, where the forest plot comprises a summary measure determined using the affinity measure for each of the one or more key features.

Abstract: In response to determining that a native language of a first user is different from a target language of a message to be transmitted by the first user to a second user, a translation model based on a plurality of language efficacies of the first user is created. An optimal action associated with a translation of the message from the native language to the target language is determined based on the created model and a language efficacy of the first user in the native language. The determined optimal action is performed. The message translation comprising the performed optimal action is transmitted to the second user.

Abstract: A computer-implemented method for building a semantic analysis model. In one embodiment, the computer-implemented method includes creating proxy tags comprising a set of surface form variants. The computer-implemented method creates training examples comprising a combination of terminal tokens and at least one of the proxy tags. The computer-implemented method builds the semantic analysis model using the training examples.

Abstract: A method, computer program product, and computer system, for receiving a first set of ground truth instances from a first source. A second set of ground truth instances may be received from a second source. The first and second sets of ground truth instances may be weighed differently based on a level of trust associated with each of the first and second sources. The weighted first and second sets of ground truth instances may be applied in a machine learning task executed by a computer.

Abstract: A method, computer program product, and computer system for receiving, by a computing device, a question from a user. A first answer provided by a first subject matter expert is identified. A second answer provided by a second subject matter expert is identified. It is determined that a profile of the user matches a first profile of the first subject matter expert more than a second profile of the second subject matter expert. The first answer provided by the first subject matter expert is sent to the user with a preference over the second answer provided by the second subject matter expert based upon, at least in part, determining that the profile of the user matches the first profile of the first subject matter expert more than the second profile of the second subject matter expert.

Abstract: Travel itineraries are automatically prepared based upon user interests and sentiments inferred by deep semantic analysis of user-commented and user-preferred digital works of literature by receiving interests and associated sentiment levels for at least one user according to a deep semantic analysis of a plurality of works of literature, wherein the works of literature have been rated, commented, or both rated and commented by the user; searching repositories of travel items to find one or more matching travel items to the received more interests and associated sentiment levels; and preparing at least one travel itinerary including at least one found matching travel item. Travel itineraries may be prepared responsive to a user-initiated trip planning session, responsive to notification of a new travel item's availability, periodically, or a combination thereof.

Abstract: Mechanisms for training a Question and Answer (QA) system are provided. The QA system receives a training question for processing by the QA system and processes the training question to generate an answer to the training question, from a portion of content. The QA system identifies a repeatable pattern of content present in the portion of content in association with the answer to the training question. The QA system applies the repeatable pattern of content to other portions of content to generate at least one additional training question and at least one additional entry in a ground truth data structure to thereby expand a set of training questions and expand the ground truth data structure. The QA system is then trained using the expanded set of training questions and expanded ground truth data structure.

Abstract: A method, system or computer usable program product for utilizing a dialectical model for providing an answer to a user query including applying NLP to the query to generate a first set of NLP structures; generating a contrary set of NLP structures contrary to the user query; evaluating the first set of NLP structures to identify a first set of evidence; evaluating the contrary set of NLP structures to identify a second set of evidence; evaluating a first set of hypotheses from the first set of NLP structures and a contrary set of hypotheses from the contrary set of NLP structures to determine a set of answers to the user query; summarizing the set of answers including indicating derivation thereof; and providing the summarized set of answers to the user.

Abstract: Systems and methods for cognitive emergency task coordination are disclosed. In embodiments, computer-implemented method comprises: receiving, by a computing device, participant data from one or more participant devices; inferring, by the computing device, skills or expertise of one or more available participants based on the participant data; generating, by the computing device, a task list including tasks associated with an emergency; assigning, by the computing device, select participants of the one or more available participants to one or more of the tasks in the task list based on the inferred skills or expertise; and sending, by the computing device, coordination data to the select participants, the coordination data including information regarding the one or more of the tasks assigned to the respective select participants.

Abstract: An approach is provided for automatically generating user-specific interaction modes for processing question and answers at the information handling system by receiving a question from a user, extracting user context parameters identifying a usage scenario for the user, identifying first input and output presentation modes for the user based on the extracted user context parameters, monitoring user interaction with the system in relation to the question, and adjusting the first input and output presentation modes based on the extracted user context parameters and detected user interaction with the system.

Abstract: Described herein are systems, devices, methods, computer-readable media, techniques, and methodologies for cloud-to-edge data flow migration. More specifically, example embodiments relate to migrating at least a portion of cloud-based data processing functionality from a cloud environment to network edge environment. The functionality can be migrated to anywhere along the network edge. Input data received or generated at a network edge device may be classified according to one or more classification criteria. At least a portion of cloud-based data processing functionality may then be selected for migration to the network edge based at least in part on the classified input data and/or one or more network characteristics. After migration, at least a portion of the input data may be processed by the migrated functionality. The input data processed at the network edge may then be sent to the cloud environment for additional processing by cloud-based data processing functionality.

Abstract: An interleave module interleaves training and task support to assist a user to complete a task while simultaneously providing training to the user. The interleave module helps the user work towards a desired state using a task knowledge base while interjecting appropriate training based on need, the user's mental state and timing constraints. The interleave module may use a variety of input sensors for a subject and the agent to help determine actions to be taken to achieve the task goals and the user's mental state for training.

Abstract: A method, computer program product, and a system where a processor(s) obtains data related to medical treatment and a current physical state of a given patient. The processor(s) cognitively analyzes the data to identify potential patient behavioral patterns of the given patient. The processor(s) obtains general patient data and correlates the general patient data with a portion of the data to identify elements in the general patient data with a potential impact on the identified potential patient behavioral patterns. The processor(s) determines impacts of the elements on the identified potential behavioral patterns and applies the impacts to generate a data structure comprising baseline behavioral patterns, i.e., a predictive model to utilize in determining probabilities that the given patient will exhibit behaviors comprising the baseline behavioral patterns during a defined future time interval.

Abstract: Described herein are systems, devices, methods, computer-readable media, techniques, and methodologies for cloud-to-edge data flow migration. More specifically, example embodiments relate to migrating at least a portion of cloud-based data processing functionality from a cloud environment to network edge environment. The functionality can be migrated to anywhere along the network edge. Input data received or generated at a network edge device may be classified according to one or more classification criteria. At least a portion of cloud-based data processing functionality may then be selected for migration to the network edge based at least in part on the classified input data and/or one or more network characteristics. After migration, at least a portion of the input data may be processed by the migrated functionality. The input data processed at the network edge may then be sent to the cloud environment for additional processing by cloud-based data processing functionality.

Abstract: A method of configuring a widget defining a user interface includes parsing content from a corpus comprising documents describing a plurality of products, selecting a plurality of pillars from the content, wherein the pillars are descriptors describing a product space of the products, determining an affinity of each of the products to each of the pillars, receiving, via the widget, a selection of a first pillar, and displaying a given product in the widget, the given product selected from the products given the selection of the first pillar.