Abstract: A device receives a query from a user associated with a car dealership and applies the query to a first trained machine learning model configured to predict an intent, and to a second trained machine learning model to predict a set of entities. The device generates a normalized representation of the first query that is database language agnostic based on the predicted intent and the predicted set of entities, and translates the normalized representation into a second query having a format compatible with a language of a database of the car dealership. The device fetches data from the database of the car dealership using the second query, and provides the data for display to the user.

Abstract: In accordance with some embodiments of the disclosed subject matter, mechanisms (which can, for example, include systems, methods, and media) for automated dietary management in healthcare facilities are provided.

Abstract: System and method for interacting with a plurality of data sources are provided. A request may be parsed and an identification parameter identifying a data set may be determined. A field included in the request may be designated as a distribution key. At least one data source may be selected based on a value associated with the distribution key. At least a portion of the request may be sent to a selected data source. Other embodiments are described and claimed.

Abstract: A computer system is described that is configured to make automated repairs to failed electronic transactions. The computer system receives an electronic transaction data file including a plurality of electronic transactions and determines that processing of at least one electronic transactions of the data file has failed due to an error in transaction information. The computer system also identifies at least one financial account in an account database that includes identification information that matches identification information included in the transaction information of the at least one electronic transaction that failed. The computer system may repair the error in the transaction information of the at least one electronic transaction based on information associated with the at least one financial account and send the repaired at least one electronic transaction to an electronic transaction processing queue.

Abstract: Techniques manage log data. Such techniques involve receiving log data describing a state of an application system at a target time point. Such techniques further involve determining whether a level of the log data satisfies a predetermined level. Such techniques further involve, in response to determining that the level of the log data satisfies the predetermined level, storing log data within a predetermined range adjacent to the target time point to a log repository of the application system, the predetermined range indicating the amount of log data expected to be stored.

Abstract: Generating granular clusters for real-time processing is provided. The systems can identify tokens based on aggregating input from computing devices over a time interval. The systems can identify, based on metrics, a subset of tokens for cluster generation. The systems can generate, via a clustering technique, token clusters from the subset of the tokens, each of the token clusters comprising two or more tokens from the subset of the tokens. The systems can apply a de-duplication technique to each of the token clusters. The systems can apply a filtering technique to the token clusters to remove tokens erroneously grouped in a token cluster. The systems can assign, based on a selection process, a label for each of the token clusters. The systems can activate, based on a number of remaining tokens in each of the token clusters, a subset of the token clusters for real-time content selection.

Abstract: The disclosed techniques reduce a responsiveness time for a secondary node state of a database in switching from a second computing node to replace a first computing node acting in a primary node state, with both computing nodes performing the same database queries. The second node receives information regarding queries performed by the first node while in the primary state. In some embodiments, the second node retrieves, from a transaction log, log records detailing operations performed for database transactions. In some embodiments, the second node inserts, based on the log records, data records of the transactions into an in-memory cache of the second node that stores chains of database records from different transactions. Upon receiving sufficient information to switch to the primary state, the second node changes a mode of operation during failover making a committed transaction available for reads by subsequent database queries prior to record reordering.

Abstract: Methods and systems described herein may include processing a change data capture stream in parallel and optimizing the write operations contained in the change data capture stream such that the write operations may be written to a change table in parallel. Processing the change data capture stream may include decoding the transaction data contained in the change data capture stream in parallel. After the transaction data in the change data capture stream has been decoded, the transaction data may be analyzed to identify write operations that write to the same row. Write operations that write to the same row may be merged into a single write operation. The merging of write operations may be performed until each write operation writes to a different row. The write operations may then be written to the change table in parallel.

Abstract: A computer-implemented method to estimate computing cost of database queries. The method includes receiving a query. The method also includes generating one or more access plans for the query. The method further includes dynamically generating a learning model based on each of the generated access plans. The method includes estimating, by the learning model and for each of the one or more access plans, a predicted cost to execute the query. The method includes selecting, based on the predicted cost, a first access path associated with a lowest cost of the predicted costs and executing the query according the selected access path.

Abstract: Provided are methods, systems, devices, apparatuses, and tangible non-transitory computer readable media for navigation and geocoding. The disclosed technology can perform operations including accessing semantic tags associated with images. Each of the semantic tags can be associated with features depicted by one of the images. Further, each of the features can be associated with a geographic location. Based on the semantic tags, landmarks that include the features that satisfy entropic criteria can be identified. The entropic criteria can measure a localized prominence of each of the one or more features. A landmark for use in navigation at the location can be selected based on context data associated with a location on a path that includes a plurality of locations. Furthermore, at least one navigational instruction that references the landmark can be generated.

Abstract: In one embodiment, the present disclosure pertains to dynamically generating normalized master data. In one embodiment, input records comprising string representations of entities are received from multiple sources. The input records may be used as queries to a similarity search data store of master data records. One or more most likely matching master records are returned with corresponding similarity scores. The input record, master record, and a training set are processed using a machine learning algorithm. In one embodiment, one or more similarity scores are incorporated into the machine learning algorithm. The machine learning algorithm produces a final score. Data from input records may be merged into the master records if the final score is greater than a threshold.

Abstract: Disclosed embodiments include apparatuses, methods and storage media associated with modifying a food record database. The method comprises receiving a plurality of food records from a plurality of sources, each of the plurality of food records comprising at least a food record description, the plurality of sources including (i) at least one non-user entity that is an owner of a third party food database and (ii) users of the food record database. The method further comprises receiving search requests from users, and returning one or more top search results from the food record database in response. The method also comprises determining a score for a particular food record identified by the top search results, wherein the score is calculated based at least in part on one of: a number of times the particular food record has been included in the top search results of the search requests or a number of times the particular food record has been logged.

Abstract: A system including one or more processors and one or more non-transitory computer-readable media storing computing instructions that, when executed on the one or more processors, cause the one or more processors to perform functions including: receiving a respective item description and at least one respective attribute value for each item of a set of items; generating at least one respective text embedding; generating a graph of the set of items based on at least co-view data to create pairs of items that are co-viewed by joining respective pairs of items; training the text embedding model and a machine learning model using a neural loss function based on the graph; and automatically determining, using the machine learning model, as trained, a label for each item of the set of items. Other embodiments are disclosed.

Abstract: The present invention is directed to generating an n-dimensional map using the results of a query for compounds enumerated within a collection of documents describing a particular biological target of interest and a curated set of compounds not enumerated in the collection of documents. Both sets of compounds (document coded and curated coded) are converted into coded forms and placed in the n-dimensional map. One or more processors are configured to evaluate the distance between the curated coded forms and the closest cluster of document coded forms. Based on the distance between a coded form and the document coded forms, the curated coded forms can be ranked regarding the likelihood of interacting with the particular biological target.

Abstract: A computer-implemented method includes: obtaining input data comprising a plurality of data values, wherein each data value is associated with a time value; obtaining a plurality of threshold criteria; and (a) selecting, or generating, a dataset or a plurality of datasets that are different to one another from the input data, and determining whether each selected, or generated dataset meets a threshold criterion of the plurality of threshold criteria; and (b) causing each selected, or generated, dataset that meets the threshold criterion to be stored in memory in association with the threshold criterion, repeating (a) and (b) for each threshold criterion of the plurality of threshold criteria, thus causing a plurality of datasets to be stored in the memory, wherein each stored dataset meets a threshold criterion of the plurality of threshold criteria.

Abstract: Systems and methods are described for identifying and resolving performance issues of automated components. The automated components are segmented into groups by applying a K-means clustering algorithm thereto based on segmentation feature values respectively associated therewith, wherein an initial set of centroids for the K-means clustering algorithm is selected by applying a set of context rules to the automated components. Then, for each group, a performance ranking is generated based at least on a set of performance feature values associated with each of the automated components in the group and a feature importance value for each of the performance features. The feature importance values are determined by training a machine learning based classification model to classify automated components into each of the groups, wherein the training is performed based on the respective performance feature values of the automated components and the respective groups to which they were assigned.

Abstract: Systems and methods are provided to implement a fast recovery process in a partitioned replicated data store. In some embodiments, the data store is configured to store an object in a plurality of partitions and replicate data in each partition in a group of replica nodes to satisfy a durability model. In response to a replica failure, the data store performs a split operation to create a plurality of new partitions. The partition's data is split into subsets corresponding to the new partitions. The subsets are transmitted, in parallel, from the surviving replica nodes of the partition to new replica nodes in the new partitions. The new partitions then replicate respective subsets of data in their respective replication groups using a chained replication technique. The recovery process allows the data store to return into compliance with the durability model more quickly, by parallelizing the copying of data.

Abstract: Methods, systems, apparatus, and computer-readable media for deriving data elements from queries. In some implementations, a query is received, and one or more computers determine that the query involves an operation that satisfies one or more criteria. In response data is saved that indicates a derived data element corresponding to the operation. The one or more computers provide data causing a representation of the derived data element to be presented, such as data causing an interactive control representing the derived data element to be presented on a user interface.

Abstract: Methods, systems, and media for personalizing computerized services based on mood and/or behavior information from multiple data sources are provided.

Abstract: Implementations described and claimed herein provide systems and methods for electronic discovery management. In one implementation, a staging path from source(s) to a staging area in a target location is generated automatically in connection with a collection request. Criteria for the collection request is obtained automatically and an encrypted export key for the collection request is captured using first robot(s). An export of a responsive data collection is obtained automatically from the source(s) using the first robot(s). The responsive data collection is exported along the staging path based on the criteria and the encrypted export key. An image of the responsive data collection is generated in the target location by sending parameter(s) to second robot(s), and the collection request is fulfilled by triggering a compression of the image of the responsive data collection into forensic container(s) using the second robot(s).