Machine-learning based multi-step engagement strategy modification is described. Rather than rely heavily on human involvement to manage content delivery over the course of a campaign, the described learning-based engagement system modifies a multi-step engagement strategy, originally created by an engagement-system user, by leveraging machine-learning models. In particular, these leveraged machine-learning models are trained using data describing user interactions with delivered content as those interactions occur over the course of the campaign. Initially, the learning-based engagement system obtains a multi-step engagement strategy created by an engagement-system user. As the multi-step engagement strategy is deployed, the learning-based engagement system randomly adjusts aspects of the sequence of deliveries for some users. Based on data describing the interactions of recipients with deliveries served according to both the user-created and random multi-step engagement strategies, the machine-learning models generate a modified multi-step engagement strategy.

Systems and methods are disclosed for executing the electronic distribution of electronic content to a display associated with a transit-oriented vehicle. The method includes receiving, from an advertiser or content provider, a request to transmit electronic content to the display, identifying a geographical zone associated with the transit-oriented vehicle; quantifying the relative position of any one of the plurality of electronic devices to the geographical zone as the transit-oriented vehicle moves along a predetermined path; identifying a displacement pattern generated by the plurality of directional vectors for the plurality of electronic devices as the geographical zone moves along the predetermined path; tailoring the electronic content based on the displacement pattern generated by the plurality of directional vectors; and transmitting the tailored electronic content to the display.

Systems and methods are disclosed for executing the electronic distribution of electronic content to a display associated with a transit-oriented vehicle. The method includes receiving, from an advertiser or content provider, a request to transmit electronic content to the display, identifying a geographical zone associated with the transit-oriented vehicle; quantifying the relative position of any one of the plurality of electronic devices to the geographical zone as the transit-oriented vehicle moves along a predetermined path; identifying a displacement pattern generated by the plurality of directional vectors for the plurality of electronic devices as the geographical zone moves along the predetermined path; tailoring the electronic content based on the displacement pattern generated by the plurality of directional vectors; and transmitting the tailored electronic content to the display.

Aspects of the present disclosure provide techniques for intelligently presenting targeted offers to a user of a software application. Embodiments include receiving received request from a user of the software application to access a portion of the software application. Using a predictive model, a predictive score is generated for the user. The predictive model generally may be a model trained using user account data reduced into n-tuples of predictive attributes representative of a plurality of users, correlated with clickstream data associated with the plurality of users indicating whether a user interacted with a targeted offer. The predictive score generally represents a likelihood that the user will interact with the targeted offer based on data stored by the software application for the user. The targeted offer is presented to the user based, at least in part, on a determination that the predictive score for the user exceeds a threshold score.

An online system receives information describing items for sale posted by non-merchant users of the online system and retrieves a first set of attributes associated with each non-merchant user. The online system also receives information describing a first set of user interactions by potential purchasing users with each item and then retrieves a machine-learning model trained to infer a second set of attributes associated with a non-merchant user of the online system, in which the model is trained based on the first set of attributes and the information describing the first set of user interactions. The online system then retrieves information describing a second set of user interactions by potential purchasing users with an item for sale posted by the non-merchant user and uses the model to infer the second set of attributes associated with the non-merchant user based on the information describing the second set of user interactions.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for posting messages on a social messaging platform. One of the methods includes providing a message composition interface; receiving a user request to include a promotion with a composed message; in response to the request, processing the content of the composed message and the user's profile on the social messaging platform to determine one or more social signals; computing an engagement score for each user in a pool of users, wherein the engagement score for each user indicates the likelihood that the user will engage with the message; selecting candidate users from the pool of users; selecting a plurality of target users from the candidate users based on a promotion value and the respective engagement scores; and posting the message on the platform including adding the message to a message stream associated with the selected target users.

Various embodiments of a framework which allow dynamic testing of many creative content and other messages simultaneously using metrics-based optimization. A “multi-armed bandit” algorithmic approach employed, as an alternative to limited AB-type testing, to automatically select a set of content parameters based on the content parameters' respective probabilities, render the selected parameters to generate content sent to a user, and, after obtaining feedback in the form of user interaction data, update the parameters for future, iterative selection of content parameters. This framework can be used in essentially any setting to allow for the provision of feedback, including user interaction data.

A method for implementing an enterprise's outbound campaign in which an offer is communicated to target customers in a manner personalized for each. The method includes: providing a customer profile database; providing a personalization platform; updating the customer profile of a first target customer according to data received from a personal bot running on a device of the first target customer; deriving interaction predictors for the first target customer relating to behavioral tendency for a interaction type; receiving an enterprise campaign dataset from the enterprise related to the outbound campaign, the enterprise campaign dataset including: information describing the offer, context; and list of the target customers; augmenting, at the personalization platform, the enterprise campaign dataset with the interaction predictors related to the first target customer to produce an enriched campaign dataset; and transmitting the enriched campaign dataset to the enterprise.

Provided are methods and systems for optimization and personalization of existing communication channels. An example method commences with aggregating user data received from a plurality of data sources and creating a user profile for a user associated with the user data. The method includes creating a business logic for user interactions of the user via the existing communication channels. The business logic includes trigger conditions and actions corresponding to the trigger conditions. The method continues with mapping, using a recommendation algorithm, content to the user according to the business logic. The content is templatized to create personalized communication messages for the existing communication channels. The method includes receiving feedback data in response to the user interactions with the personalized communication messages. Based on the feedback data, the user profile is updated, and the recommendation algorithm and the next suggestion communication action are updated based on the updated user profile.

Techniques for prefetching operation cost based digital content and digital content with emphasis that overcome the challenges of conventional systems are described. In one example, a computing device may receive digital content representations of digital content from a service provider system, which are displayed on a user interface of the computing device. Thereafter, the computing device may also receive digital content as prefetches having a changed display characteristic as emphasizing a portion of the digital content based on a model trained using machine learning. Alternatively, the computing device may receive digital content as a prefetch based on a model trained using machine learning in which the model addresses a likelihood of conversion of a good or service and an operation cost of providing the digital content. Upon receiving a user input selecting one of the digital content representations, digital content is rendered in the user interface of the computing device.