Abstract: A behavioral targeting system determines user profiles from online activity. The system includes a plurality of models that define parameters for determining a user profile score. Event information, which comprises on-line activity of the user, is received at an entity. To generate a user profile score, a model is selected. The model comprises recency, intensity and frequency dimension parameters. The behavioral targeting system generates a user profile score for a target objective, such as brand advertising or direct response advertising. The parameters from the model are applied to generate the user profile score in a category. The behavioral targeting system has application for use in ad serving to on-line users.

Abstract: Data quality measurement is provided for use in a data processing stream, which comprises at least one upstream data processing system and at least one downstream data processing system. An input alert component can be used to provide a measurement of data prior to its input to a data processing system (e.g., a downstream data processing system or an upstream data processing system). An output alert component can be used to provide a measurement on data output by a data processing system. A self-consistency component can be used to measure consistency between items of input, or output data. An end-to-end component can be used to measure data quality using data items from both input data and output data. These components can be used in some combination, or independent of the other, and in any order. In addition, the data quality measurements can be performed separate from that processing performed by either the upstream or downstream processing system.

Abstract: A method attributes credit for conversions to a plurality of event types. A plurality of probabilities that a single impression of each of a plurality of respective event types causes a conversion by one of a plurality of users are defined. A processor is used to select values for the plurality of probabilities so as to minimize a total squared error based on differences between an observed outcome for each user and a respective expected outcome for each user calculated from the plurality of probabilities.

Abstract: Systems and methods are provided for predicting visitor traffic to a network of web site pages. The systems and methods are used, as an example, to predict the inventory of total available online advertisements available within the network for a forthcoming period. The visitor traffic includes page viewing, listening or transacting on web pages within a web site, wherein the web pages are categorized by subject, interest areas or specific user queries such as word or phrase searches. For each page whose traffic is being predicted, the system takes into account annual seasonality, day-of-week, holidays, special events, short histories, user demographics, user web behavior (viewing, listening and transacting) and parent and child web page characteristics.

Abstract: Systems and methods are provided for predicting visitor traffic to a network of web site pages. The systems and methods are used, as an example, to predict the inventory of total available online advertisements available within the network for a forthcoming period. The visitor traffic includes page viewing, listening or transacting on web pages within a web site, wherein the web pages are categorized by subject, interest areas or specific user queries such as word or phrase searches. For each page whose traffic is being predicted, the system takes into account annual seasonality, day-of-week, holidays, special events, short histories, user demographics, user web behavior (viewing, listening and transacting) and parent and child web page characteristics.

Abstract: Systems and methods are provided for predicting visitor traffic to a network of web site pages. The systems and methods are used, as an example, to predict the inventory of total available online advertisements available within the network for a forthcoming period. The visitor traffic includes page viewing, listening or transacting on web pages within a web site, wherein the web pages are categorized by subject, interest areas or specific user queries such as word or phrase searches. For each page whose traffic is being predicted, the system takes into account annual seasonality, day-of-week, holidays, special events, short histories, user demographics, user web behavior (viewing, listening and transacting) and parent and child web page characteristics.

Abstract: A system for automatically tracking the location of a vehicle includes a visual image detector mounted on the vehicle for producing as the vehicle moves along a route digitized strips of image data representing successive panoramic views of scenery about the vehicle at respective locations along the route. A sparse tracking subsystem processes and stores only selected ones of the image data strips representing substantially spaced apart successive locations along the route, for use as a sparse database. A dense tracking subsystem processes and stores as a dense database every successive one of the image data strips representing location along the route, whereby the dense tracking subsystem provides more accurate location of the vehicle when it retraces some portion of the route than the sparse tracking subsystem.

Abstract: A system for automatically detecting and recognizing the identity of a deformable object such as a human face, within an arbitrary image scene. The system comprises an object detector implemented as a probabilistic DBNN, for determining whether the object is within the arbitrary image scene and a feature localizer also implemented as a probabilistic DBNN, for determining the position of an identifying feature on the object such as the eyes. A feature extractor is coupled to the feature localizer and receives coordinates sent from the feature localizer which are indicative of the position of the identifying feature and also extracts from the coordinates information relating to other features of the object such as the eyebrows and nose, which are used to create a low resolution image of the object. A probabilistic DBNN based object recognizer for determining the identity of the object receives the low resolution image of the object inputted from the feature extractor to identify the object.