Abstract: Systems and techniques for stream selection from multi-stream storage devices. Notification of a KVS tree write request for a multi-stream storage device is received. The notification includes a KVS tree scope corresponding to data in the write request. A stream identifier (ID) is assigned to the write request based on the KVS tree scope and a stability value of the write request. The stream ID is returned to govern stream assignment to the write request, the stream assignment modifying a write operation of the multi-stream storage device.

Abstract: A system for archiving includes an active memory, an archive memory, an interface, and a processor. The interface is to receive an indication to archive an object in the active memory. The object has a relation to an other object. The processor is to archive the object in the archive memory. The other object maintains the relation with the object that has been archived.

Abstract: Systems and techniques for collecting and using merge tree garbage metrics are described herein. A kvset is created for a node in a KVS tree. Here, a set of kvset metrics for the kvset are computed as part of the node creation. The kvset is added to the node. The node is selected for a compaction operation based on a metric in the set of kvset metrics. The compaction operation is performed on the node.

Abstract: Systems and techniques for merge tree modifications for maintenance operations are described herein. A request for a KVS tree is received. Here, the KVS tree is a data structure including nodes and the nodes include a temporally ordered sequence of kvsets that store keys in sorted order. A parameter set for the KVS tree is received. The request is executed on the KVS tree by modifying operation of the KVS tree in accordance with the parameter.

Abstract: A content-name-resolution (CNR) system can resolve IP addresses for named data objects (NDOs) based on their name. During operation, a CNR server can receive a query from a client device for a source to a named data object. The query can include at least a name prefix for the named data object. The CNR server can identify a cache server that corresponds to the named data object's name prefix, and determines one or more sources associated with the named data object's name prefix. The CNR server then returns, to the client device, a query response that includes a network address for the cache server, and includes a content record specifying the one or more sources.

Abstract: A persistent memory based distributed-journal file system, comprising: 1) Direct mapped persistent memory unit hosting at least a subset of a file system which contains a plurality of files each associated with a metadata record and a self-journal record. 2) A memory storing a program code. 3) One or more processors coupled to the persistent memory unit and the memory for executing the program code which comprises: (a) Creating a resources dataset which maps characteristics of the file system by analyzing the files' metadata records. (b) Receiving an alteration request to alter altered file(s) of the plurality of files. (c) Determining if the alteration is an atomic alteration which is applied to the file system in an atomic operation. (d) Logging an indication of the alteration in the respective self-journal record of the altered file(s) for non-atomic alteration. (e) Applying the alteration in the file system.

Abstract: A system for real-time search, including: a set of partitions, each including a set of segments, each segment corresponding to a time slice of messages posted to the messaging platform, and a real-time search engine configured to receive a search term in parallel with other partitions in the set of partitions, and search at least one of the set of segments in reverse chronological order of the corresponding time slice to identify document identifiers of messages containing the search term; and a search fanout module configured to: receive a search query including the search term; send the search term to each of the set of partitions for parallel searching; and return, in response to the search query, at least one of the identified document identifiers of messages containing the search term.

Abstract: Method and systems for parallel scoring an ensemble model are provided. Aspects include loading data into a first distributed data structure having a plurality of partitions, each partition having loaded data in the form of a set of pairs of data formed of a record to be scored and a partial score for that record. A component model in the ensemble model is selected and processing of the records carried out in parallel across the partitions including updating the partial score for each record. In response to a partial score for a record not meeting an accuracy threshold, the method retains the record in the first distributed data structure to be scored by a subsequent component model. In response to the partial score for a record meeting the accuracy threshold, the method moves the record and updated partial score to an output result data structure to provide a final score.

Abstract: The present disclosure discloses methods and systems for managing data in a database in accordance with a data model, henceforth referred to as a “timeline store” or “timeline model”. The method includes mapping each data reference of one or more data references with a data value of a block of data in a data store. Then, each key reference of one or more key references is mapped with at least one pair of a time reference and the data reference in a timeline store, the time reference determines a point in time at which the key reference is assigned the pair of the time reference and the data reference. Thereafter, the data model is queried to retrieve a set of key references mapped to data references based on a specific ranges of keys and time or time intervals. Finally, the data model is queried to retrieve the raw data bytes associated/mapped with/to any desired data reference.

Abstract: Methods and systems for managing and maintaining a library including organized categories of legal clauses for access by a user in a network computing environment. A clause library manager is provided for identifying a clause for inclusion in a library including multiple clauses. A clause profile including identifying information associated with the clause is generated for the identified clause. A rating is determined for the clause and added to the clause profile. The clause is categorized into a clause category based on the clause profile.

Abstract: A system ranks results. The system may receive a list of links. The system may identify a source with which each of the links is associated and rank the list of links based at least in part on a quality of the identified sources.

Abstract: Techniques are described for revising data partition size for use in generating predictive models. In one example, a method includes determining an initial number of base model partitions of data from a plurality of data sources; determining an initial base model partition size based at least in part on the initial number of base model partitions; and evaluating the initial base model partition size at least in part with reference to at least one base model partition size reference. The method further includes determining a finalized number of base model partitions based at least in part on the initial base model partition size; determining a revised base model partition size; and generating revised base models based at least in part on the revised base model partition size, including using a predictive modeling framework to randomly assign input data records from the plurality of data sources into the base model partitions.

Abstract: Performing snapshot conscious internal file modification for network-attached storage is presented herein. A file system can comprise a first component configured to modify, during a service request, storage for a subset of data blocks of a file—the service request not being recognized by an external entity as a change of content of the file. Further, the file system can comprise a second component configured to prevent, based on the service request, a copy of the storage from being created for servicing of a snapshot—the snapshot comprising a point-in-time copy of the file system.

Abstract: An approach is described for sorting and displaying documents according to sentiment level in an online community. An associated system may include a processor and a memory storing an application program, which, when executed on the processor, performs an operation that may include selecting a review topic in an online community and identifying a plurality of documents contributed for the review topic. The plurality of documents may include at least one of a product review submission, a marketing survey submission, a social network activity stream post, a discussion forum post, a weblog post, and an audiovisual sample. The operation further may include obtaining sentiment data associated with each of the plurality of documents developing a sentiment model based on the obtained sentiment data. Additionally, the operation may include organizing and presenting the plurality of documents in an online community interface based on the sentiment model.

Abstract: A system for model validation comprises an input interface, a model builder, a model validator, a model committer, and an attachment point. The input interface is for receiving a set of input data, wherein a first input data of the set of input data is associated with a first model level validation and a first attachment point. The model builder is for determining a model that is used to update a database based at least in part on the set of input data. The model validator is for determining whether the model is valid using model validations, wherein the model validations include the first model level validation. The model committer is for committing the model in the event the model is valid. The attachment point determiner is for determining a failure associated attachment point in the event the model is not valid, wherein the failure associated attachment point comprises the first attachment point in the event that the first model level validation failed.

Abstract: The invention provides methods and apparatuses for opening a webpage, invoking a client, and creating a light application. The method for opening a webpage comprises: receiving a webpage opening request containing a link; according to the link, searching a preset application list for a client application and/or a light application corresponding to the link; and when the client application and/or the light application exists in the application list, opening the client application and/or invoking the light application to open a webpage corresponding to the link. Thereby, a webpage corresponding to a link can be opened in a more friendly manner, the use of a user is facilitated, the efficiency of acquiring information by the user is improved, and the user experience is enhanced.

Abstract: The present invention extends to methods, systems, and computing system program products for iteratively calculating a Z-score for Big Data. Embodiments of the invention include iteratively calculating one or more components of a Z-score for a modified computation subset based on the one or more components of a Z-score calculated for a previous computation subset and then calculating a Z-score for a selected data element in the modified computation subset based on one or more of the iteratively calculated components. Iteratively calculating a Z-score avoids visiting all data elements in the modified computation subset and performing redundant computations thereby increasing calculation efficiency, saving computing resources and reducing computing system's power consumption.

Abstract: The present invention extends to methods, systems, and computer program products for training a classification model to predict categories. In one implementation, a method identifies category mappings generated for dominant queries associated with a query log. The method identifies mappings between a first set of queries and categories shown for the first set of queries, and identifies mappings between a second set of queries and clicked products for the second set of queries. A classification model is trained based on the mappings generated for dominant queries, the mappings between queries and the shown categories, and the mappings between queries and the clicked products.

Abstract: Embodiments of the invention are directed to systems, methods, and computer program products for automating data validation. The system is configured to receive the initial data set in the first data structure from a network of distributed servers; transform the initial data set into an expected data set having the second data structure using a first mapping logic; use the expected data set to validate a converted data set that was generated by converting the initial data set into the second data structure using a second mapping logic; and transmit control signals configured to cause a user device associated with the user to receive a notification indicating that the match is within the pre-determined tolerance level.

Abstract: Systems and methods may provide for implementing a dynamic search service. In one example, the method may include receiving information relating to a topic of interest to a user, analyzing the information to determine a search parameter, conducting a coordinated search of content utilizing a plurality of device, generating and gathering results of a coordinated search from the plurality of devices, and generating a result communication including at least one result of the coordinated search.