Abstract: Techniques for generating recommendation cluster within a social network service are described. Consistent with some embodiments, sample members in a social network service are identified. The sample members may be associated with prior member activity involving a source member. A cluster category this then selected based on a member attribute shared by a plurality of the sample members. In turn, a recommendation cluster is generated based on the selected cluster category. Generating the recommendation duster may involve selecting member profiles that match the cluster category. The member profiles selected in this way form the recommendation cluster. One or more of the member profiles of the recommendation cluster are then surfaced to a client device operated by the source member.

Abstract: A method includes identifying a buddy database projection that corresponds to a recovering database projection that is to be recovered to a recovering node of a database system. A source container containing data of the buddy database projection is identified on a source node of the database system, the source container being characterized by a range of epochs that includes an epoch for which data is to be recovered to the recovering node. Data is copied from the identified source container to the recovering node.

Abstract: The disclosed system provides a computerized map system with map elements represented by markers. The system arranges map elements by category and subcategory, with each type of marker corresponding to a category. The system may divide the geographic area based on neighborhood and city area. The system also provides the user with controls to search the objects, to display results in layers, to navigate the objects, and to quickly compare information about the objects using standardized positions for similar kinds of information. The system also enables users to memorialize searches and browsing by pinning map objects that the user is interested in. In an embodiment, a number of tiers of advertising are also provided, such that business owners as clients can pay for advertising content and customizability appropriate for their businesses.

Abstract: A data-processing system facilitates processing a data stream to assist other devices to pre-fetch elements of the data stream, out-of-sequence, for uninterrupted playback. During operation, the system can receive a data file, and segments the data file into a sequence of content objects. The system then determines a target content object that is to be processed out-of-sequence, and a corresponding source content object. The system then inserts a reference to the target content object, into the source content object, and stores the sequence of content objects. A client device can disseminate interest to obtain the data stream's content objects. Upon receiving a content object, the client device can determine whether the content object includes a reference to other content objects. If so, the client device can disseminate interests for these referenced content objects.

Abstract: A data storage system may implement rate-limiting secondary index creation for an online table. A secondary index may be generated for a table stored in a data store. The table may be incrementally indexed, maintaining the updates determined according to indexing different portions of the table in a queue of pending updates that are subsequently applied at the secondary index. Prior to indexing a portion of the table, an evaluation of a current number of pending updates in the queue of pending updates may be performed with respect to a throttle threshold. If the current number of pending updates exceeds the throttle threshold, then indexing the portion of the table may be throttled. Received updates to previously indexed portions of the table, may be applied to the table and placed in the queue of pending updates without an evaluation of the current number of pending requests.

Abstract: A service metadata replication system includes an ingester that scrapes or receives service data including metadata values for service objects from various services. The ingester formats the metadata values for storage in a service store. The service store can asynchronously replicate a portion of the metadata values through a write intake to a storage cluster. In addition, an inter-regional replicator asynchronously replicates one or more additional metadata values to the storage cluster from another service instance that is remotely located. An authorization runtime scrapes or receives security information regarding the service instance. A query processor can issue a query request for a portion of the metadata values and a portion of the additional metadata values to a read intake. The read intake can then satisfy the query request based on the contents of the authorization runtime and storage cluster.

Abstract: Methods and systems for creating a compact tree node representation of an XML document. One implementation commences by allocating memory for storing an XML tree index data structure, then allocating another separate portion of memory to store a hash table. Then, traversing an XML document to process the traversed nodes as follows: (a) when the traversed node is an element node, then adding the element node to the XML tree index data structure (b) when the traversed node is a text node, then populating a text node index into the XML tree index data structure and copying the text node values to the hash table, and (c) when the traversed node is an attribute node, then populating an attribute node index into the XML tree index data structure. Such a structure supports fast index-based tree restructuring, and permits very large XML document to be accessed within tight memory size constraints.

Abstract: A processing device identifies a file (or data block) and selects a storage location for the file based on a plurality of data distribution policies. To select the storage location, the processing device applies, at a top distribution layer, a first data distribution policy to select a first storage pool from a plurality of storage pools based on a first storage criterion. The processing device then applies, at a bottom distribution layer, a second data distribution policy associated with the first storage pool to select a first storage node from a plurality of storage nodes in the first storage pool based on a second storage criterion. The processing device stores at least a portion of the file in the first storage node of the first storage pool.

Abstract: An approach for identifying missing content is provided. An approach includes: receiving a search string; determining a skill rating associated with a searcher that generated the search string; and searching a database using the search string. The approach also includes: updating at least one of a first score associated with the search string and a second score associated with the database based on the skill rating; and generating an alert based on one of the first score exceeding a first threshold and the second score exceeding a second threshold.

Abstract: Techniques are provided for improving performance of spatial queries by defining a grid that divides the domain space into cells, and then using a cell-to-item mapping to determine which items do not have to be individually evaluated against the location criteria of the spatial queries. Based on the cell to which an item belongs, the item may automatically qualify as a match, be automatically disqualified, or require item-specific evaluation. To account for items with size, the query window of a spatial query may be expanded. To limit the degree to which the query window is expanded, a plurality of grids may be established for the domain space, where each grid has differently sized cells, and items are assigned to grids based on the size of the items.

Abstract: Provided are techniques for a Data-Partitioned Secondary Index (DPSI) partition level join. While using a Data-Partitioned Secondary Index (DPSI) to perform a join of an outer table and an inner table, a different task from multiple tasks is assigned to each partition of the inner table. With each task, a join is performed of the outer table and the assigned partition of the inner table using the DPSI to generate results. The results from each different task are merged.

Abstract: A method and system for fuzzy full text search is disclosed. The system includes an inverted index where tokens are organized in a Next Valid Character (NVC) tree. Each path through the tree defines a token. The document lists for the tokens are stored at leaf nodes. When performing a fuzzy full text search, the system uses an edit distance greater than zero. After receiving search tokens, the system traverses the NVC tree to generate document error lists. The system then compares the error lists to identify relevant document identifiers and provides a result set.

Abstract: Techniques are described for performing synthesis of relationships between a plurality of concept definitions automatically derived from a faceted domain of information. Some embodiments involve identifying at least one facet attribute in an active concept definition specified by user input. In response to determining that at least one explicit relationship and/or at least one implicit relationship exist(s) between the active concept definition and a first concept definition of the plurality of concept definitions, a relationship is synthesized between the active concept definition and the first concept definition.

Abstract: A method and apparatus are provided for navigating through a media object collection using a mobile device. Users navigate to the target media objects by repeatedly specifying filters. After a user specifies a filter, the filter is applied in combination with all previously specified filters to produce a filtered set of media objects. After a filter is applied, a display level is selected for each hierarchical dimension for which the media objects have metadata. The selection of the display level may be based on the number of distinct values that the media objects in the filtered set have for each level of the hierarchical dimension. Single-action controls are then displayed for each of the values that the media objects in the filtered set have for the selected display level.

Abstract: A method for detecting solutions to a problem using content in online discussion sources. The method includes receiving a request, such request identifying a problem, and searching multiple online discussion sources for content related to the problem. Responsive to finding content related to the problem, the method searches the multiple online discussion sources for a plurality of solutions to the problem. Responsive to finding a plurality of solutions to the problem, the method forms groups containing the solutions from each of the multiple online discussion sources. The method then determines a likeliness to solve the problem for each of the groups and ranks the groups based on the determined likeliness to solve the problem. The method then determines that the rank of at least one group meets a threshold value, wherein the threshold value is based on a confidence in the likeliness to solve the problem.

Abstract: The invention relates to a method of synchronizing a plurality of clients accessing a database, each client executing a plurality of tasks on the database, wherein the method comprises for each of the clients the steps of accumulating the time of one or more tasks performed by the client after the issuance of a synchronization request and rejecting a request for the opening of a new task of the client, if the accumulated task time exceeds a maximum accumulated task time.

Abstract: A system and method is disclosed for presenting relevant advertising to web pages, including web logs. The ads can be selected by reference to an underlying content presented in web logs, based on extracting topics from the pages, and using these topics as the basis for a search query. Ads can then be matched against another set of topics derived from web pages forming the search results to the query.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for distributing content are disclosed. In one aspect, a method includes accessing data specifying a plurality of search queries. Content distribution campaigns (“campaigns”) in which distribution of at least one content item is conditioned on a distribution parameter matching one of the search queries are identified. Two or more similar campaigns are identified, and a search query that matches a distribution parameter in at least one of the similar campaigns is identified as a candidate content distribution parameter. A trend score for the candidate content distribution parameter is determined based on a change in a submission rate of search queries that match the candidate distribution parameter. Suggestion data suggesting the candidate content distribution parameter as an additional content distribution parameter for at least one of the similar campaigns is provided based on the trend score.

Abstract: The application describes systems and methods for preventing the distribution of large volumes of electronic data by loading selected sensitive files with pad data to increase the size of the files such that file transfer, distribution, or downloading to removable media storage devices is more cumbersome. In one aspect, a file loading system comprises a datastore for storing a plurality of data files where each of the plurality of data files includes information and a processor arranged to: access the plurality of data files in the datastore, and load a data pad into one or more of the plurality of data files to increase the size of the one or more of the plurality of data files.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for query categorization based on image results. In one aspect, a method includes receiving images from image results responsive to a query, wherein each of the images is associated with an order in the image results and respective user behavior data for the image as a search result for the first query, and associating one or more of the first images with a plurality of annotations based on analysis of the selected first images' content.