Abstract: A computer-implemented process for training a natural language processing (NLP) agent having a reinforced learning model includes the following operations. A type of document from a document corpus is identified using metadata particularly associated with the document. The NLP agent tokenizes the document to generate a plurality of tokens. Using a schema identified from the type of the document, one of the plurality of tokens is compared to a system of record (SOR) field from the schema. A similarity score between the one of the plurality of tokens with a correct value and a reward based upon the similarity score are generated. A determination is made that an optimum minimum average similarity rate has not been obtained. Based upon the determination, the reinforced learning model is trained using a loss function that includes the reward.

Abstract: Aspects of the invention include techniques for decomposing trend and seasonality components in a forecast of parametric time series data. A non-limiting example method includes receiving time series data that includes a plurality of values taken over a first period of time. A forecast is generated using the time series data. The forecast can include one or more predicted values over a second period of time. The forecast is decomposed into N components and 2N coalitions are defined for the N components. A coalition value is determined for each coalition of the 2N coalitions.

Abstract: An embodiment for learning-based automatic selection of artificial intelligence applications. The embodiment may receive a user request for an exam, the user request including exam information. The embodiment may automatically identify an exam type cluster corresponding to the received exam information. The embodiment may automatically detect applicable AI applications corresponding to the identified exam type cluster. The embodiment may automatically run each applicable AI application on a series of relevant test sets to generate a score for each applicable AI application. The embodiment may automatically recommend to a user a highest-scoring applicable AI application.

Abstract: A processor may identify one or more data fields of interest. The processor may receive data for each of the data fields of interest. The processor may compute one or more bias scores for the data fields of interest based on the data. The processor may develop a bias matrix with the bias scores. The processor may display the bias matrix to a user.

Abstract: Embodiments include a hierarchical metadata prediction system that includes a first line-based predictor having a first line for storage of metadata entries, and a second line-based predictor configured to store metadata entries from the first line-based predictor. The second line-based predictor has a second line, the second line including a plurality of containers, the plurality of containers including at least a first set of containers having a first size and a second set of containers having a second size. The system also includes a processing device configured to transfer one or more metadata entries between the first line-based predictor and the second-line based predictor. Embodiments also include a computer-implemented method and a computer program product.

Abstract: Participant inclusion determination can include generating a textual analysis of a draft of an electronic communication in response to a sender preparing the draft for conveyance over an electronic communications network. A word relevance score can be computed for each word of the draft based on the textual analysis. Probabilistically weighted distances between the sender, an initial recipient of the electronic communication, and more additional recipient candidates for additionally receiving the electronic communication can be determined. The probabilistically weighted distances can correspond to hierarchical distances within a hierarchical structure corresponding to an organization in which the sender, the initial recipient, and one or more additional recipient candidates are members.

Abstract: A system may include a memory and a processor in communication with the memory. The processor may be configured to perform operations. The operations may include extracting a first multiple dimensional dataset and a second multiple dimensional dataset and encoding the first multiple dimensional dataset into a first encoded dataset and the second multiple dimensional dataset into a second encoded dataset. The operations may include combining the first encoded dataset and the second encoded dataset into a combined encoded dataset and conveying the combined encoded dataset to a user.

Abstract: Synchronizing requests with a respective context includes, responsive to a determination that there are more pages to explore, performing regular crawling operations for a current page, recording a current page in a list of explored pages and extracting links from the current page. Responsive to a determination that there are more links to extract, a next link to analyze is selected to form a selected link and responsive to a determination that there is a new request associated with the selected link, a new request identifier is created and saved as an entry in a hashmap. Responsive to a determination that there is not a new request associated with selected link, a request associated with the selected link is updated with a new link value when the link value differs.

Abstract: Method and system are provided for running a smaller memory-address width program in a larger memory-address width address space. The method includes: dividing a smaller memory-address width program executable code into a set of portions; reserving a first virtual storage area in a part of an address space accessed using a smaller memory-address width address; and reserving a set of second virtual storage areas in a part of the address space accessed using a larger memory-address width address to accommodate the program executable code. The method provides a relocation mechanism to relocate a processor thread by translating using a relocation factor from an address in the reserved first virtual storage area to the one of the reserved second virtual storage areas containing the executable code.

Abstract: A modular two-stage neural architecture is used in translating a natural language question into a logic form such as a SPARQL Protocol and RDF Query Language (SPARQL) query. In a first stage, a neural machine translation (NMT)-based sequence-to-sequence (Seq2Seq) model translates a question into a sketch of the desired SPARQL query called a SPARQL silhouette. In a second stage a neural graph search module predicts the correct relations in the underlying knowledge graph.

Abstract: A system and related method score a fairness of an outcome model. The method comprises receiving a set of original transaction records (OTRs), and selecting an OTR subset of the OTRs according to a subset selection criteria in order to reduce a number of OTRs to send to outcome model. For each OTR in the subset a perturbed transaction record (PTR) is created based on the OTR that includes changing at least one attribute in the PTR from the OTR, sending the OTR and the PTR to the outcome model, receiving an OTR outcome and a PTR outcome from the outcome model, and determining a record bias score for the OTR outcome and the PTR outcome respectively that indicates bias in the respective outcome. The OTR and the PTR bias score are stored in a bias determination system (BDS) database.

Abstract: A method, according to one approach, includes forming an underlayer of a magnetic recording medium. The underlayer includes first encapsulated nanoparticles each comprising a first magnetic nanoparticle encapsulated by a first aromatic polymer, and a first polymeric binder binding the first encapsulated nanoparticles. A magnetic recording layer is formed above the underlayer. The magnetic recording layer includes second encapsulated nanoparticles each comprising a second magnetic nanoparticle encapsulated by an encapsulating layer, and a second polymeric binder binding the second encapsulated nanoparticles.

Abstract: Analysis of edge closures of metal surface particles based on a graph structure.

Abstract: In one aspect of the present disclosure, the degree of side erasure occurring in neighboring tracks is controlled by controlling the skew angle of the tape head, relative to the longitudinal direction of motion of the tape past the head as a track is written. In one embodiment, the tape head is rotated relative to the longitudinal direction of tape travel, and the trailing edge of the tape head which forms an acute angle with the edge of a track being written, is used for writing tracks at a desired track pitch. Other aspects and advantages may be provided, depending upon the particular application.

Abstract: Embodiments of the present invention provide a computer system a computer program product, and a method that comprises determining a task that is being performed based on a set of observed micromovements; in response to determining the task that is being performed, detecting an anomaly in at least one micromovement in the set of observed micromovements based on a comparison of the at least one micromovement to one or more micromovements associated with an established baseline of micromovements; and generating a notification on a user interface that details the detected anomaly.

Abstract: Detecting system log anomalies by receiving multivariate time-series system log data an multivariate metric data, inferring system metrics from the system log data and metric data, receiving a metric causal graph including causal relationships between system metrics, determining a univariate variation score for the system metrics, determining a causal variation score for the multivariate time series system metric data, according to the causal graph, determining an overall activity score according to the univariate variation score, and causal variation score, and altering a review window duration according to the activity score.

Abstract: The present disclosure relates to a method, system and computer program product for electronic health record (EHR) data synthetization. According to the method, an original EHR dataset X is captured. A latent space Z is generated from the original EHR dataset X, wherein dimensionality of Z is lower than that of X. A stochastic process prior module is applied to the latent space Z. Synthetic EHR dataset X? is reconstructed from the latent space Z after being applied with the stochastic process prior.

Abstract: An approach for generating a compiled version of a program source code. At least two different executable files may be generated applying at least two different compiler optimization settings for compiling a first part and applying at least two different compiler optimization settings for compiling a second part of the source code. A main performance part of the source code may be determined dependent on values of a target quantity of the at least two different executable files. The main performance part is the part of the first part and the second part that has a greater influence on the target quantity. The compiled version of the program source code may be generated by compiling the source code applying a higher optimization level of the compiler for compiling the main performance part than for compiling the remaining part of the source code.

Abstract: A mechanism is provided in a data processing system to implement a multi-sensor health monitoring platform. The mechanism applies a machine learning model to predict patient needs and patient activity trends based on physiological features and activity features of the patient. The mechanism applies the machine learning model to predict energy requirements for a plurality of medical sensors based on the predicted patient needs and patient activity trends. The mechanism schedules recharging of the plurality of medical sensors based on the predicted energy requirements and identifying one or more sensors to set to an activate state based on the predicted patient needs and patient activity trends. The mechanism collecting sensor data from the one or more sensors and applies the machine learning model to generate a point-of-care recommendation based on the collected sensor data.

Abstract: A method and resulting structures for a semiconductor device includes forming a source terminal of a semiconductor fin on a substrate. An energy barrier is formed on a surface of the source terminal. A channel is formed on a surface of the energy barrier, and a drain terminal is formed on a surface of the channel. The drain terminal and the channel are recessed on either sides of the channel, and the energy barrier is etched in recesses formed by the recessing. The source terminal is recessed using timed etching to remove a portion of the source terminal in the recesses formed by etching the energy barrier. A first bottom spacer is formed on a surface of the source terminal and a sidewall of the semiconductor fin, and a gate stack is formed on the surface of the first bottom spacer.