Abstract: Techniques are described herein for generating and displaying a confusion matrix wherein a data item belonging to one or more actual classes is predicted into a class. The classes in which the data item may be predicted (the “predicted classes”) are ranked according to a score that in one embodiment indicates the confidence of the prediction. If the data item is predicted into a class that is one of the top K ranked predicted classes, then the prediction is considered accurate and an entry is created in a cell of a confusion matrix indicating the accurate prediction. If the data item is not predicted into a class that is not one of the top K ranked predicted classes, then the prediction is considered inaccurate and an entry is created in a cell of a confusion matrix indicating the inaccurate prediction.

Abstract: Disclosed is a method and system for providing intuitive and efficient representations of positions of interest to a user within encoded media. Various embodiments of the present disclosure provide a heatmap representation that indicates the interestingness of content at different locations within the media. These locations of interest are presented to a user to allow quick jumps to the interesting parts of the content.

Abstract: Disclosed is a method and system for providing attention based advertisement insertion within media content. By placing advertisements within media where there is the most user attention and interest, the system provides a more attractive audience to advertisers. This potentially leads to increased return of interest (ROI) on the advertisement. Additionally, this can enable the media to be comprised of fewer advertisements, and reduce advertisement related abandonment.

Abstract: Embodiments of the invention concern finding correlations between related traded items, such as securities, commodities, currencies, or contracts. The pricing information of multiple traded items from one or more markets is analyzed and a user is provided with a listing of pairs or larger groups of traded items which are most closely correlated. Another embodiment of the invention is a system and method for finding the most related traded item to a particular predefined item. Another embodiment of the invention is a system and method for finding the traded item whose trading history is closest to a predefined curve.

Abstract: The embodiments of the present invention provide methods and systems for automated collection of human-reviewed data. Requesters send data to be reviewed by humans (or data requests) to a data processing system, which is in communication with one or more systems for collecting human-reviewed data (HRD). The methods and systems discussed enables the data processing system to work with one or more of the systems for collecting HRD). In one embodiment, between the data processing system and the systems for collecting HRD are wrappers, which stores parameters specific to the data requests and libraries for transforming the data requests to human intelligent tasks (HITs) specific to each HRD system. The data processing system also includes a number of components that facilitate transforming data requests into HITs, sending the HITs to the HRD collection systems, receiving HRD, and analyzing HRD to improve the quality of collected HRD.

Abstract: A method for providing personal targeted advertising includes receiving, from a user, a submission of a search-ad comprising a personal targeted advertisement associated with at least one rare search term; storing the personal targeted advertisement; detecting a query for the at least one rare search term; and displaying the personal targeted advertisement in response to detecting the query for the at least one rare search term.

Abstract: Techniques are described herein for generating and displaying a confusion matrix wherein a data item belonging to one or more actual classes is predicted into a class. The classes in which the data item may be predicted (the “predicted classes”) are ranked according to a score that in one embodiment indicates the confidence of the prediction. If the data item is predicted into a class that is one of the top K ranked predicted classes, then the prediction is considered accurate and an entry is created in a cell of a confusion matrix indicating the accurate prediction. If the data item is not predicted into a class that is not one of the top K ranked predicted classes, then the prediction is considered inaccurate and an entry is created in a cell of a confusion matrix indicating the inaccurate prediction.

Abstract: A system and method are disclosed for receiving human responses to various requests that may require a human review. A request may be submitted for completion by at least one of a plurality of available users. Upon completion of the request, the user may be eligible for a reward if the requestor approves of the completed request. The reward may include a portion of the revenue generated by the completed request.

Abstract: Embodiments of the invention concern finding correlations between related traded items, such as securities, commodities, currencies, or contracts. The pricing information of multiple traded items from one or more markets is analyzed and a user is provided with a listing of pairs or larger groups of traded items which are most closely correlated. Another embodiment of the invention is a system and method for finding the most related traded item to a particular predefined item. Another embodiment of the invention is a system and method for finding the traded item whose trading history is closest to a predefined curve.

Abstract: A computerized self-service platform assists merchants in creating and managing stores on a computer network. In one aspect, a system for managing a store includes a product database and an ecommerce storefront, with the ecommerce storefront referencing information in the product database. A product administration module (e.g., a catalog manager) allows the merchant to manage the product database on a self-serve basis. Separating the storefront from the product database can yield numerous advantages.

Abstract: A method comprises generating a first classification model, e.g., metadata-based, for determining whether a video belongs to a category; generating a second classification model, e.g., content-based, for determining whether the video belongs to a category, the first classification model and second classification model being based on different modalities; and generating a fusion model that blends the categorization results of the models. Each classification model may classify the video to multiple categories. During operation, a method obtains a video; uses the first classification model, the second classification model and the fusion model to determine whether the video belongs to a category; and indexes the video in a video index. The method may enable selection of a category corresponding to the video search results. The category may be identified based on a query profile, which may be learned from users' query logs or popular queries and click history.

Abstract: A technique is provided for generating one or more XML documents from a relational database using the XPath data model. Data stored on a data storage device that is connected to a computer is transformed. Initially, a document object model tee is generated using a document access definition, which defines the mapping between an XML tree structure and relational tables. The document object model tree is traversed to obtain information to retrieve relational data. The relational data is mapped to one or more XML documents.

Abstract: A method, apparatus, and article of manufacture for a computer implemented modification system for supporting localized data changes in a computer. A change is executed to data in a computer having a local database. The change is performed by the computer to modify data in the local database stored within a data storage device at the computer. The change is indicated as being global. When the local database on the computer is synchronized with a main database on another computer, only global changes are transmitted from the local database to the main database.

Abstract: A computerized self-service platform assists merchants in creating and managing stores on a computer network. In one aspect, a system for managing a store includes a product database and an ecommerce storefront, with the ecommerce storefront referencing information in the product database. A product administration module (e.g., a catalog manager) allows the merchant to manage the product database on a self-serve basis. Separating the storefront from the product database can yield numerous advantages.

Abstract: Provided is a method, system, and program for accessing rows of data from a base table satisfying a search criteria. An initial request to access rows in the base table is received, including the search criteria specifying predicates for one or more columns in the base table. A data structure is generated to include information on qualifying base table rows that satisfy the search criteria. Notification is provided that the qualifying base table rows are accessible before information on all the qualifying base table rows is included in the data structure. A request is then received for one qualifying base table row. A determination is made as to whether the data structure includes information on the requested qualifying base table row. If so, the information in the data structure is used to provide access to the requested qualifying base table row.

Abstract: A technique is provided for creating metadata for fast search of XML documents stored as column data. Data is stored in a data store connected to a computer. A main table is created in a relational database management system, wherein the main table has a column for storing an extensible markup language (XML) document in a native XML format One or more side tables are created in the relational database management system, wherein each side table has one or more columns that store one or more attributes extracted from die XML document for fast searching of the XML document. Thereafter, the side tables are used to locate data in the main table.

Abstract: The present invention discloses a method, apparatus, and article of manufacture for generating a database application. In accordance with the present invention, a features list is built for the database application. The features list contains user-selected functions. The database application is dynamically configured based on the built features list.

Abstract: A technique is provided to store fragmented XML data into a relational database by decomposing XML documents with application specific mappings. Data stored on a data store that is connected to a computer is transformed. Initially, an XML document containing XML data is received. A document access definition that identifies one or more relational tables and columns is received. The XML data is mapped from the application DTD to the relational tables and columns using the document access definition by generating a first document object model tree using the XML document, generating a second document object model tree using the document access definition, and mapping the data from the first document object model tree into columns in one or more relational rabies using the second document object model tree.

Abstract: A technique is provided for generating one or more XML documents from a single SQL query. Data stored on a data storage device that is connected to a computer is transformed. A query is received that selects data stored in a relational database management system on the data storage device, wherein a data access definition defines: (1) a collection of one or more tables in the relational database management system for storing attributes from an XML document, (2) how data stored in the tables maps to the XML document, (3) a query for mapping the data stored in the tables to the XML document, and (4) a table that will contain the XML document after the XML document is generated. Then, one or more XML documents are generated from the selected data using the data access definition.

Abstract: An extender for a computer-implemented relational database system is disclosed for storing, querying, and retrieving structured documents. The extender provides a new abstract data type and includes a plurality of user defined functions for storing, querying, and retrieving structured documents internally, as character-based large objects (CLOB), or externally, in flat files or URLs, for example. A method and apparatus using an extender having a “conditional select” functionality is further disclosed for querying structured documents. The “conditional select” functionality is based on user defined functions residing within the extender being applied to a projection in a structured query in the database system. A method and apparatus using an extender having a set of user defined functions is further disclosed for creating one or more indexes for structured documents stored in the relational database system.