A computer-implemented method comprising: receiving a job from a plurality of priority queues, the job including a deliverable and a plurality of commands; performing a correction procedure on the deliverable, wherein the correction procedure comprises, for each of the plurality of commands: transforming a respective command into a structure of keywords, the structure associated with a flow of execution, the structure including a plurality of nodes, the plurality of nodes including a root node and plurality of parent nodes, each parent node of the plurality of parent nodes having at least one child node, each parent node of the plurality of parent nodes including a keyword in the respective command; traversing the structure according to the flow of execution, executing one or more keywords at one or more parent nodes of the plurality of parent nodes; determining an output of the respective command based on the execution of the one or more keywords at the one or more parent nodes of the plurality of parent nodes.

Systems and methods for accelerated stabilization of data packet metadata are disclosed herein. The system can include a memory having a content database and a user profile database. The system can include a user device having a first network interface and a first I/O subsystem. The system can include one or more servers. The one or more servers can: retrieve data packet metadata for a data packet; determine that the data packet metadata is unstable; identify a set of potential recipients of the data packet; select one of the set of potential recipients as the recipient of the data packet; provide the data packet to the recipient of the data packet; receive a response from the recipient to the provided data packet; and automatically update the data packet metadata based on the response received from the recipient.

A system and process for dynamic security monitoring of test taking. The system combines historical information about the test taker or relevant conditions, and includes sensors which objectively monitor test taker's actions, inactions, and involuntary response to the test given before and during the test, and at a specific test location. The information collected from the sensors is compared to a plurality of predetermined individual risk factors, which indicate a possibility of test fraud by the individual test taker to create a test event risk profile. The individual risk profile is combined with non-individual specific risk factors to create a holistic test event security profile. Security resources are then dynamically assigned to or removed from the test event based on the unique test event security profile.

A system and process for dynamic security monitoring of test taking. The system combines historical information about the test taker or relevant conditions, and includes sensors which objectively monitor test taker's actions, inactions, and involuntary response to the test given before and during the test, and at a specific test location. The information collected from the sensors is compared to a plurality of predetermined individual risk factors, which indicate a possibility of test fraud by the individual test taker to create a test event risk profile. The individual risk profile is combined with non-individual specific risk factors to create a holistic test event security profile. Security resources are then dynamically assigned to or removed from the test event based on the unique test event security profile.

An augmented reality system for developing and providing augmented reality learning experiences is disclosed. The augmented reality system advantageously combines augmented reality with the capabilities of cloud technology to provide a pull-based collaborative model, in which students and instructors collaborate by uploading, sharing, and downloading augmented reality learning content. The augmented reality system enables students to improve the augmented reality learning content by adding contributions to the original augmented reality learning content that was created by an instructor.

An augmented reality system for developing and providing augmented reality learning experiences is disclosed. The augmented reality system advantageously combines augmented reality with the capabilities of cloud technology to provide a pull-based collaborative model, in which students and instructors collaborate by uploading, sharing, and downloading augmented reality learning content. The augmented reality system enables students to improve the augmented reality learning content by adding contributions to the original augmented reality learning content that was created by an instructor.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for developing cognitive and behavioral metrics associated with a user. In some implementations, a system obtains data from a writing implement, the data indicative of a user performing a task by writing with the writing implement against a receiving device. The system extracts features from the obtained data. The system determines metrics that reflect characteristics of the user, the metrics that reflect cognitive characteristics of the user and metrics that reflect behavioral characteristics of the user. Based on the extracted features and the determined metrics, the system generates a user profile for the user. Based on the features indicating characteristics of the writing behavior of the user, the system modifies the generated user profile as a performance of the user changes over time. The system provides recommendations to improve the performance of the user.

The present disclosure relates to a method of reducing a size of an artificial intelligence model by an electronic device, including: inputting an input value for training to a first model; training the first model for performing a specific task based on the input value; inputting the input value to a second model; and training the second model based on an output value of the first model, in which the first model may be an artificial intelligence model larger in size than the second model.

The present disclosure relates to a method of reducing a size of an artificial intelligence model by an electronic device, including: inputting an input value for training to a first model; training the first model for performing a specific task based on the input value; inputting the input value to a second model; and training the second model based on an output value of the first model, in which the first model may be an artificial intelligence model larger in size than the second model.

A method and system for online proctoring of tests while preserving privacy of test-taker is disclosed. Proctoring data, which include video and audio data from at least one camera and a microphone monitoring the test-taker and the test environment, is chopped up into data fragments. Each fragment is altered to replace personally identifiable information, and the altered fragment is encrypted using a cryptographic key. The chronological order of fragments is also scrambled. Encrypted and altered data fragments are distributed to a pool of proctors who review the encrypted fragment for suspicious behavior. Suspicious fragments are further compared with original, unaltered versions of the fragments to confirm suspicious behavior, and render a verdict. The test-taker is aware of, and explicitly consents to the processing of a fragment by a proctor. A secure, custom viewer for the fragments also allows the test-taker to control the number of times a proctoring data segment can be viewed. Our method and system ensure the privacy of the proctoring data by explicitly authorizing every entity that processes a proctoring data fragment, and limiting number of views of the fragment, while allowing independent evaluation of proctoring data for different forms of cheating.