Systems, devices, and techniques are disclosed for recommendations using a transformer neural network. User activity data including items and actions associated with users and a catalog including descriptions of the items may be received. User vectors for the users, item vectors for the items and action vectors the actions may be generated by applying singular vector decomposition to the user activity data. Sequence vectors may be generated based on item vectors and the action vectors. Transformer vectors may be generated by applying a text-to-text transferring transformer to descriptions of the items. Similarity vectors may be generated based on the transformer vectors. Merged vectors may be generated by merging the sequence vector, transformer vector, and similarity vector for items. A set of probabilities may be determined by inputting the user vector for the user, merged vectors for the items, and sequence vectors for the actions to a transformer neural network.

A system and method for providing contextual recommendations in gaming media streams. One or more users request a gaming media stream broadcast corresponding to game play. A broadcast service obtains the gaming media stream that includes a gaming stream identifier and user identifier. The broadcast service can generate a set of recommendations based on at least one of social network information, historical information, and identification of events and items in the gaming media stream. The broadcast service provides the gaming media stream with one or more recommended items.

A system for providing product recommendations to online visitors to an e-commerce website is provided. The system may include program comprising instructions that, when executed by a processor, cause the processor to sort a list of products based on a comparison of a user's interactions with the e-commerce website and previous user interactions with the same e-commerce website.

A method includes obtaining (i) device data associated with multiple electronic devices and (ii) advertisement data associated with one or more advertisement campaigns. The method also includes identifying first features and second features, where the first features correspond to the device data and the second features correspond to the advertisement data. The method further includes generating a graph relating usage history of the multiple electronic devices and one or more advertisement genres. The method also includes identifying a specified electronic device from among the multiple electronic devices and an advertisement segment from among the one or more advertisement campaigns using the first features, the second features, and the graph. In addition, the method includes providing the advertisement segment to the specified electronic device.

Methods, systems, and storage media for assigning multiple identifiers to user accounts are disclosed. Exemplary implementations may: receive an indication of an event; associate the event with a user; associate the user with at least a first account identifier of a plurality of accounts of the user; associate the user with at least a second account identifier of the plurality of accounts of the user, the second account identifier different than the first account identifier; add at least one of the first account identifier or the second account identifier to a custom audience relating to the event; target the custom audience through directed advertising; and deliver advertisements to the custom audience.

A tracking system is provided that tracks user browsing activity while maintaining user privacy. It is determined whether a cookie is stored in local storage of a client device. If the cookie is not stored: a tracking identification (ID) is generated; the tracking ID, a cookie rotation timestamp, and a failed request counter in a generated cookie is stored; and the generated cookie is copied and attached to a transmitted event. If the cookie is stored: it is determined whether a time indicated by a stored cookie rotation timestamp is in the past; if the time is in the past, a flag requesting a new cookie value is set for a transmitted event; and if the time is not in the past, the stored cookie is copied and attached to a transmitted event.

An automated, dynamic digital financial management tool is described herein. The financial management tool enables a user to access all forms of payments, debt and transactions to be deducted automatically depending on which form of payment is chosen. The financial management tool documents every item/service purchased. The documentation enables a company/manufacturer/store to provide marketing and recall information to the user. The financial management tool is able to include any and/or all financial aspects of a user's life. The financial management tool is able to be implemented using a universal card. The financial management tool is able to include a pay station to receive payments. The financial management tool is able to include automated services.

Systems and methods are disclosed configured to identify a location where current or predicted ride requests exceed a threshold within a time period. Driver location information for one or more drivers is used to identify a driver to satisfy the identified current or predicted ride requests at the location that exceed the threshold within the time period. In response to determining that a driver has at least a first likelihood of accepting a ride request at the location within the time period, item content is selected using a history of establishments frequented by the driver. The selected item of content, associated with a token, is provided for presentation on a display to the driver. In response to determining that the identified driver accepted the ride request at the location within the time period, the token is enabled to be utilized by the identified driver at an entity.

A system and methods for an offer placement system for delivering targeted marketing offers to consumers during a session with an online (web-based) Internet portal, such as an online banking portal of a financial institution. A component receives advertising campaign data from an advertiser including a plurality of targeted marketing offers for delivery to consumers in response to a particular condition (e.g. an offer-triggering event), and stores the campaign data in a database. A component determines the occurrence of an offer-triggering event caused by a consumer during an online session with the online portal, for example, viewing a list of banking transactions. A component identifies a predetermined targeted marketing offer stored in the database that corresponds to the offer-triggering event. A component delivers information corresponding to the targeted marketing offer to the consumer during the online session with the portal.

An online photo management service that stores a collection of photos belonging to a user can send re-engagement messages to the user that can include photos automatically selected from the collection. The selection can be based on a scoring algorithm that rates the photos according to a set of attributes and computes a score based on the attributes and a set of weights. Based on user responses to re-engagement messages, the weights can be tuned to more reliably select photos likely to result in user re-engagement with the stored collection of photos.