A method for generating human-machine hybrid predictions of answers to forecasting problems includes: parsing text of an individual forecasting problem to identify keywords; generating machine models based on the keywords; scraping data sources based on the keywords to collect scraped data relevant to the individual forecasting problem; providing the scraped data to the machine models; receiving machine predictions of answers to the individual forecasting problem from the machine models based on the scraped data; providing, by the computer system via a user interface, the scraped data to human participants; receiving, by the computer system via the user interface, human predictions of answers to the individual forecasting problem from the human participants; aggregating the machine predictions with the human predictions to generate aggregated predictions; and generating and outputting a hybrid prediction based on the aggregated predictions.

Interaction data sets of consumers of an offering set with an application associated with an offering set are obtained. At least some of the interactions are associated with respective content presentation operations. A neural network model is trained, using the interaction data sets, to generate predictions of actions of a consumer at whom a set of content presentation operations is directed, and to provide action-attribution scores for the content presentation operations. The model includes convolutional layers and an attention module. Weights learned in the attention module are used to estimate the action-attribution scores. A trained version of the model is stored.

A system, computer-readable storage medium storing at least one program, and computer-implemented method for providing recommendations based on social network sharing activity. Sharing activity relating to the sharing of the content item on a social network by a first user is accessed. Consumption information related to the consumption of the content item. A correlation between the sharing activity and the consumption information is determined. A recommendation is then generated based on the correlation.

A machine is configured to generate in real time personalized online banner images for users based on data pertaining to user behavior in relation to an image of a product. For example, the machine receives a user selection indicating one or more data features associated with the user. The one or more data features include a data feature pertaining to user behavior in relation to an image of a product. The machine generates, using a machine learning algorithm, a data representation of the machine learning algorithm based on the one or more data features including the data feature pertaining to user behavior in relation to the image of the product. The data representation includes one or more data features pertaining to one or more characteristics of online banner images. The machine generates an online banner image for the user based on the data representation.

Disclosed examples determine a first level of loyalty to a brand for a household based on first data records of consumer data corresponding to a first period of time; determine a second level of loyalty to the brand for the household based on second data records of consumer data corresponding to a second period of time after the first time period; determine consumer metrics based on the first level of loyalty and the second level of loyalty; and generate a report based on the consumer metrics.

Systems and methods for optimizing pricing of products within a retailer are provided. Such systems and methods include determining a set of products to be included in an elasticity computation. Next, the number of days to collect transaction logs for a given product is determined, responsive to sales volumes for each given product. These transaction logs are then collected and used to compute the elasticities for these products. Products that were not included for calculation of elasticities have elasticities imputed for them. A set of constraints are received. Optimal prices are then generated based upon the objectives, rules and price elasticities.

Aspects of the subject disclosure may include, for example, obtaining contextual information relating to a user, where the contextual information comprises location data that identifies a location of the user, identifying media content that relates to the contextual information and to profile data associated with the user, deriving, from the media content, personalized media content based on the profile data associated with the user, causing a target device to provide an immersion environment that includes the personalized media content, detecting user interaction data relating to the immersion environment, and performing an action relating to the personalized media content based on the detecting the user interaction data. Other embodiments are disclosed.

Various systems and methods for providing a tool to entities that determines the optimal usage of data items are disclosed. For example, the tool can generate a model that uses various characteristics to predict how likely it is that a viewer will watch (or listen to) the media program being promoted. The model can then determine an increase in revenue that would result from the assignment of a media promo using the predicted likelihood and subtracting a known opportunity cost from this determined revenue increase to determine a net revenue value. The model can repeat this determination for any number of viewers and aggregate the determined net revenue values to generate an aggregated net revenue value. The tool may include a user interface in which a content provider can adjust various variables to see how adjusting one or more variables affects the aggregated net revenue value.

Methods, systems, and computer programs are presented for the determination of optimal communication scheduling. Send Time Optimization (STO) uses machine learning (ML) to recommend a personalized send time based on a recipient's past engagement patterns. The purpose of the ML model is to learn patterns in the data automatically and use the patterns to make personalized predictions for each recipient. The send time recommended by the model is the time at which the model believes the recipient will be most likely to engage with the message, such as clicking or opening, and use of the send time mode is expected to increase engagement from recipients. Additional customizations include communication-frequency optimization, communication-channel selection, and engagement-scoring model.

A system, method, and computer program product for implementing electronically generated promotion deployment is provided. The method includes receiving electronic data comprising shipment identification for a shipment of an item from a shipping client to a recipient. A profile is generated for the recipient and a further client associated with a freight carrier is authorized to generate visible promotional structures associated with packages for delivery by the freight carrier. Promotional attributes from the further client are cross referenced with respect to a package comprising the item for the recipient. The cross referencing may include matching demographic profiles, of recipients of pending delivery packages, to additional profiles selected by a promotional entity for each pending delivery package. A visible promotional structure associated with the promotional attributes is generated and presented during delivery of the package. The visible promotional structure may include a physical structure or a digital structure.