Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for AI based privacy amplification. A data security system allows for data transmitted between devices to be secured using varying levels of data security that are adjusted dynamically based on the determined sensitivity level of the data. The data security system uses AI (e.g., machine learning models) to predict the sensitivity level of data being transmitted between the devices in real-time and applies an appropriate level of data security based on the predicted sensitivity level. Dynamically adjusting the level of data security that is used based on the sensitivity level of the data provides for heightened levels of data security to protect highly sensitive data, and lower levels of data security to conserve computing resources when protecting less sensitive data.

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for AI based privacy amplification. A data security system allows for data transmitted between devices to be secured using varying levels of data security that are adjusted dynamically based on the determined sensitivity level of the data. The data security system uses AI (e.g., machine learning models) to predict the sensitivity level of data being transmitted between the devices in real-time and applies an appropriate level of data security based on the predicted sensitivity level. Dynamically adjusting the level of data security that is used based on the sensitivity level of the data provides for heightened levels of data security to protect highly sensitive data, and lower levels of data security to conserve computing resources when protecting less sensitive data.

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for differential bid generation using machine learning. A bid management system generates an input based on features of a content item and. The bid management system uses the input to determine acceptance probability score distributions for a set of candidate users. The bid management system uses the acceptance probability score distributions along with user attributes describing the set of candidate users along with campaign parameters to determine a subset of the candidate users to receive an offer to view the content item along with bid values to offer to each candidate user.

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for determining user engagement with a content item. A computing device accesses at least one image of eyes of a user that is captured while a client device is presenting a first content item on a display of the client device. The computing device determines, based on using the at least one image as input in a neural network, a gaze of the user. The gaze including coordinates at which the user is looking in relation to the client device. The neural network was trained based on machine generated images of a modeled human user looking at various coordinates. The computing device determines, based on the gaze of the user, an engagement score for the user. The engagement score indicates a level of engagement of the user with the first content item.

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for differential bid generation using machine learning. A bid management system generates an input based on features of a content item and. The bid management system uses the input to determine acceptance probability score distributions for a set of candidate users. The bid management system uses the acceptance probability score distributions along with user attributes describing the set of candidate users along with campaign parameters to determine a subset of the candidate users to receive an offer to view the content item along with bid values to offer to each candidate user.

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for determining user engagement with a content item. A computing device accesses at least one image of eyes of a user that is captured while a client device is presenting a first content item on a display of the client device. The computing device determines, based on using the at least one image as input in a neural network, a gaze of the user. The gaze including coordinates at which the user is looking in relation to the client device. The neural network was trained based on machine generated images of a modeled human user looking at various coordinates. The computing device determines, based on the gaze of the user, an engagement score for the user. The engagement score indicates a level of engagement of the user with the first content item.

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for determining user engagement with a content item. A computing device accesses at least one image of eyes of a user that is captured while a client device is presenting a first content item on a display of the client device. The computing device determines, based on using the at least one image as input in a neural network, a gaze of the user. The gaze including coordinates at which the user is looking in relation to the client device. The neural network was trained based on machine generated images of a modeled human user looking at various coordinates. The computing device determines, based on the gaze of the user, an engagement score for the user. The engagement score indicates a level of engagement of the user with the first content item.

Abstract: An ontology-driven information system includes a plurality of models, each of which expresses an aspect of a business domain using concepts and relationships between concepts. An ontology, which is in communication with each of the plurality of models, provides uniform definitions for the concepts and relationships between concepts used in the plurality of models. A method for executing an interaction flow model includes receiving an event and categorizing the received event. Once the event is categorized, a situation that matches the categorized received event is identified. One or more tasks are then executed for the situation. The execution of the one or more tasks can include either an interpretation of a model or the execution of a method of an object. The information system also includes a user and application interface and a reasoning engine that is in communication with the user and application interface.

Abstract: An ontology-driven information system includes a plurality of models, each of which expresses an aspect of a business domain using concepts and relationships between concepts. An ontology, which is in communication with each of the plurality of models, provides uniform definitions for the concepts and relationships between concepts used in the plurality of models. A method for executing an interaction flow model includes receiving an event and categorizing the received event. Once the event is categorized, a situation that matches the categorized received event is identified. One or more tasks are then executed for the situation. The execution of the one or more tasks can include either an interpretation of a model or the execution of a method of an object. The information system also includes a user and application interface and a reasoning engine that is in communication with the user and application interface.