Abstract: Machine learning techniques are leveraged to provide personalized assistance on a computing device. In some configurations a timeline of a user's interactions with the computing device is generated. For example, screenshots and audio streams may be saved as entries in the timeline. Context—the state of the computing device when the entry is created, such as which documents and websites are open—is also stored. Entries in the timeline are processed by a model to generate embedding vectors. The timeline may be searched by finding the embedding vector that is closest to an embedding vector derived from a search query. The user may select a query result, causing the associated context to be restored. For example, if the query is “show me all documents related to my upcoming trip to Japan”, the query result may open documents and websites that were open when booking a flight to Japan.

Abstract: A database system is operable to determine a query for execution. A disjunction of a plurality of conjunctive normal form (CNF) filtering predicates is determined based on filtering predicates of the query. An IO pipeline is generated based on the disjunction of the plurality of CNF filtering predicates that includes, based on applying at least one type of optimization, at least one shared IO element common to multiple ones of plurality of the CNF filtering predicates. The IO pipeline is executed in accordance with execution of the query to generate a filtered subset of a plurality of rows that each meet at least one of the plurality of CNF filtering predicates.

Abstract: Disclosed techniques and solutions can provide improved snapshot replication. Typically, an initial replica obtained using snapshot replication is periodically updated. However, the update process can unnecessarily consume computing resources if data in a source data object has not changed with respect to data in a replica data object. Disclosed techniques check to determine whether a snapshot replica is out of date before obtaining a new snapshot. The checks can be performed on manual request or on the occurrence of triggers, such as receiving a query that accesses the replica data object or according to a schedule. Information for current and prior versions of the remote data object can be compared to determine whether a replica is out of date, such as digest values of contents of the remote data object or timestamps associated with the remote data object.

Abstract: A document classification apparatus includes a processing circuit. The processing circuit configured to: divide a sentence included in a target document into word units; divide a sentence included in a supplementary text into word units; calculate an appearance statistic of a word obtained from the target document and the supplementary-text; provide, to the word, an attribute representing specialty, specificity, or category independence based on the appearance statistic; extract an anchor word from words included in the supplementary text based on the attribute; perform target learning in accordance with an attribute of the anchor word and generate a word embedding vector; and estimate a classification category of a document based on the word embedding vector.

Abstract: A computing device may be configured to continuously, repeatedly, or recursively generate, train, improve, focus, or refine the machine learning classifier models that are used data anomalies. The computing device may create a corpus of data based on architecture or standards documents, generate classifier models based on the corpus of data, collect information from one or more data sources, generate feature vectors based on the collected information, apply the feature vectors to the classifier models to generate an analysis result, and identify a data anomaly based on the generated analysis result.

Abstract: Method and system of sentiment-based selective engagement of users in a network system. The method includes transmitting a request, over a network system, for engaging users associated with content data authored at least partly in accordance with sentiment expressions, the request including a set of engagement criteria, discovering, responsive to the request, a sentiment block digital asset in accordance with the set of engagement criteria, the sentiment block digital asset being based on the content authored in accordance with one or more sentiment expressions, and engaging the users based on the set of engagement criteria in conjunction with a set of consent rules associated with the at least one sentiment block digital asset.

Abstract: Methods and apparatuses relate to transferring user-specific data that can be validated via a hash tree. If not all of the underlying data of the hash tree is transferred, then in place of the non-transferred data, a suitable hash value from the hash tree can be transferred. The data to be transferred includes at least one reference data element which is suitable for checking the authenticity of the sender.

Abstract: Method and system of generating and deploying a tokenized digital asset. The method includes monitoring generation of a sentiment expression in accordance with social media content data, the sentiment expression rendered in association with a sentiment community of interest, generating a sentiment block digital asset in accordance with the sentiment expression and rendering, as a tokenized digital asset, a digital token data file that is a unique representation of the sentiment block digital asset.

Abstract: Example embodiments of the present disclosure provide a method, a device, and a computer program product for processing a workflow chart. The method includes encoding structural information of the workflow chart including a plurality of nodes and a plurality of edges by using a graph neural network to acquire a vector representation of the structural information; acquiring textual description data about the workflow chart at the nodes; training a language model based on the acquired textual description data and the acquired vector representation to acquire a pretrained language model; and fine-tuning the pretrained language model through training data of a specific task to acquire a fine-tuned language model. Through the method for processing the workflow chart of the present disclosure, the combination of the graph neural network and the language model not only can process a large number of complex workflow charts, but also can generate effective natural language outputs.

Abstract: Computer implemented systems and methods are disclosed for importing data from electronic data files. In accordance with some embodiments, a file format is assigned to a source electronic data files by a data importation system. The data importation system may further identify a file type identifier associated with the source electronic data file and map the source electronic data file to a transformation template. The data importation system may further store the file format, file type identifier, and an indication of the transformation template as a file type profile associated with the source electronic data file in a database.

Abstract: A data analysis apparatus performs to acquire series data having character strings related to an information system to be analyzed as elements and series data having numerical values indicating a state of a device constituting the information system as elements, each of the series data having an index that enables comparison of element order relations within series and between series, classify the elements of each piece of the series data into classification classes, and output series data having classification values indicating classification classes as elements, integrate series data having classification values of the character strings as elements and series data having classification values of the numerical values as elements into one piece of series data, and detect an occurrence of a frequent pattern which is a combination of frequently occurring elements, using the one piece of series data obtained through integration.

Abstract: A document management system trains a machine learned model configured to facilitate contract generation. The document management system generates a training set for the machine learned model using user input. From a first set of agreement documents, the document management system identifies a first set of sentences relevant to clause terms input by the user. The document management system identifies, from a second set of agreement documents, a second set of sentences similar to the first set of sentences. In response to user feedback about the second set of sentences, the document management system updates the training set. The document management system retrains the machine learned model using the updated training set.

Abstract: Systems and methods for managing a business intelligence database are described, for example, by receiving ETL event data associated with an execution of a set of tasks of the ETL process; processing the ETL event data to identify at least one of a resource and a first task limiting performance of the ETL process, wherein the resource is memory, storage, processor bandwidth, or network bandwidth; and automatically modifying the at least one of the resource and the first task to improve performance of the ETL process.

Abstract: Embodiments of the disclosed technologies include generating a search prompt based on an input portion of an online dialog involving a user of a computing device. The search prompt includes a dialog summarization instruction configured to instruct a generative artificial intelligence model to generate and output a dialog summary. The search prompt is sent to a first generative model. In response to the search prompt, a search query is generated and output by the first generative model based on the dialog summary. The search query is sent to a search system. Search result data is determined based on an execution of the search query by the search system. At least some of the search result data is included in an output portion of the online dialog. The output portion is configured to be displayed at the computing device in response to the input portion of the online dialog.

Abstract: A system for grouping architectural information within a plurality of layers for common formatting in a database. The system may include at least one processor; and at least one memory configured to execute the instructions to perform operations. The operations may include group according to the meta-model, architectural datasets together to structure database architectural information into a common format; establish an architectural domain for grouping the database architectural information in a plurality of layers; organize the meta-model to establish possession of information in the plurality of layers; and provide architectural data corresponding to each data domain for each of the plurality of layers from one or more data sources.

Abstract: Implementations of a system for downscaling commodity flow data may include a first commodity flow database at a first level of geographic/temporal granularity, a second commodity flow database including commodity data for a plurality of commodities at a second level of geographic/temporal granularity and a group of high speed multiple core processor clusters where each core of may be assigned to process data associated with a commodity from the second commodity flow database and data of the first commodity flow database. The system may also include a downscaling module to generate downscaled and a quality check module to generate quality checked data; and a merging module configured for merging downscaled data and quality checked data in parallel. The system may also include a database module configured for writing data asynchronously according to a first-in first-out queue system in a commodity flow database within a period of hours.

Abstract: An apparatus, computer-readable medium, and computer-implemented method for extracting a subset, including receiving a request comprising one or more criteria indicating one or more criteria tables in a plurality of tables of the database, storing an entity graph corresponding to a schema of the database and the one or more criteria, determining one or more top-most criteria entities in the entity graph, and generating the subset from the plurality of tables by traversing the entity graph starting at the one or more top-most criteria entities without traversing any directed edge more than once.

Abstract: An improved query system in which registered sex offender information published in state and territory databases is obtained using an improved data scraper, merged, cleaned, and/or enriched to remove inaccuracies is described herein. Users can then perform a single search or multiple searches simultaneously (e.g., perform batch searches) and view the results in a single content page. The improved query system can use the merged and/or cleaned data to generate a more limited set of search results, reducing the amount of mismatched or inaccurate information that a user may have to navigate through. Thus, users can perform one or more searches and view accurate search results without having to navigate through many different windows or content pages or perform additional navigational steps to identify relevant information.

Abstract: A computer-implemented method for performing natural language-based data integration includes causing execution of a data integration application on a remote device via a network and causing surfacing of a GUI corresponding to the data integration application on a display of the remote device. The method includes receiving, via the GUI, a natural language input representing a data integration task, generating, via an LLM, a set of ordered activities corresponding to the data integration task represented by the natural language input, and selecting, via the LLM, one or more APIs for performing each activity within the set of ordered activities. The method also includes generating a data pipeline based on the set of ordered activities and the API(s) for performing each activity, as well as back-translating the data pipeline to a desired data format for execution by the data integration application.

Abstract: A data processing architecture controls data processing arbitration in a high performance computing system that includes one or more premises. Individual premises can include one or more server computers executing an instance of a local file system and including one or more temporary data storage devices. Individual instances of the local file system can access files stored in objects of a primary data store. Individual objects of the primary data store can be accessed using a common identifier indicating a storage location of the individual objects in the primary data store.