A system for detection of email risk automatically determines that a first party is considered by the system to be trusted by a second party, based on at least one of determining that the first party is on a whitelist and that the first party is in an address book associated with the second party. A message addressed to the second party from a third party is received. A risk determination of the message is performed by determining whether the message comprises a hyperlink and by determining whether a display name of the first party and a display name of third party are the same or that a domain name of the first party and a domain name of the third party are similar, wherein similarity is determined based on having a string distance below a first threshold or being conceptually similar based on a list of conceptually similar character strings. Responsive to determining that the message poses a risk, a security action is automatically performed comprising at least one of marking the message up with a warning, quarantining the message, performing a report generating action comprising including information about the message in a report accessible to an admin of the system, and replacing the hyperlink in the message with a proxy hyperlink.

Systems, methods, and computer program products for adaptively splitting electronic chats are provided. An e-discovery system comprises a computer processor and a non-transitory, computer-readable medium embodying thereon a set of computer instructions executable by the computer processor. The set of computer instructions includes instructions for: sending a chat query to a remote electronic chat service; receiving an electronic chat responsive to the chat query, the electronic chat embodying a set of electronic chat messages; adaptively splitting the set of electronic chat messages into a set of conversations, each conversation in the set of conversations comprising a subset of electronic chat messages from the set of electronic chat messages; and storing each conversation from the set of conversations as a separate document.

Methods and systems for a cascading model architecture for providing information on a reply email. Training sample data can be created using a user's incoming reply emails from external computer devices to a user's computer device. A receptivity neural network model can be trained using the training sample data of the reply emails, and a trained receptivity neural network model can be used to determine a receptivity classification for whether new reply emails are positive reply emails or non-positive reply emails. Sample data of non-positive reply emails can be augmented, and an objection identification neural network model can be trained on the augmented sample data of the non-positive reply emails. A trained objection identification neural network model can be used to determine a objection classification for new non-positive reply emails. Explainability information for a classified reply email can be determined to provide information key words and/or key phrases that were used by the trained receptivity neural network model and the trained objection identification neural network model to classify new positive reply emails and new non-positive reply emails.

Methods and systems for a cascading model architecture for providing information on a reply email. Training sample data can be created using a user's incoming reply emails from external computer devices to a user's computer device. A receptivity neural network model can be trained using the training sample data of the reply emails, and a trained receptivity neural network model can be used to determine a receptivity classification for whether new reply emails are positive reply emails or non-positive reply emails. Sample data of non-positive reply emails can be augmented, and an objection identification neural network model can be trained on the augmented sample data of the non-positive reply emails. A trained objection identification neural network model can be used to determine a objection classification for new non-positive reply emails. Explainability information for a classified reply email can be determined to provide information key words and/or key phrases that were used by the trained receptivity neural network model and the trained objection identification neural network model to classify new positive reply emails and new non-positive reply emails.

In one embodiment, a method for generating a message to a friend of a user is provided, comprising: processing activity data of a first user measured by an activity monitoring device to update a value of an activity metric for the first user; identifying a change in an inequality relationship between the value of the activity metric for the first user and a value of the activity metric for a second user; in response to identifying the change in the inequality relationship, prompting the first user to generate a message to the second user.

In one embodiment, a method for generating a message to a friend of a user is provided, comprising: processing activity data of a first user measured by an activity monitoring device to update a value of an activity metric for the first user; identifying a change in an inequality relationship between the value of the activity metric for the first user and a value of the activity metric for a second user; in response to identifying the change in the inequality relationship, prompting the first user to generate a message to the second user.

A system includes a first computing device client associated with a first user in a community of users operable to send content to publish to a data aggregation server. The data aggregation server is operable to receive the content to publish from the first computing device client, host a first user profile associated with the first user of the first computing device client, the first user profile identifying a targeted recipient in the community of users, and disseminate automatically the content received from the first computing device client to a second computing device client associated with the targeted recipient, without receiving input from the first computing device client explicitly specifying the targeted recipient to whom the content is disseminated.

A technique for message history display includes combining message histories for multiple different messaging services. A system constructed according to the technique may include, for example, a message history database; a history aggregation engine that aggregates message logs for storage in the message history database; and a history provisioning engine that provides an aggregated message log associated with the user from the message history database to a requesting device. A method according to the technique may include, for example, identifying a device in association with a user profile; providing an online platform that receives messages from and sends messages to the device; and creating an aggregated log from messages sent to and from the device.

A technique for message history display includes combining message histories for multiple different messaging services. A system constructed according to the technique may include, for example, a message history database; a history aggregation engine that aggregates message logs for storage in the message history database; and a history provisioning engine that provides an aggregated message log associated with the user from the message history database to a requesting device. A method according to the technique may include, for example, identifying a device in association with a user profile; providing an online platform that receives messages from and sends messages to the device; and creating an aggregated log from messages sent to and from the device.

The present disclosure relates to systems for identifying instances of natural language input, determining intent classifications associated with instances of natural language input, and generating responses based on the determined intent classifications. In particular, the disclosed systems intelligently identify and group instances of natural language input based on characteristics of the user input. Additionally, the disclosed systems determine intent classifications for the instances of natural language input based message queuing in order to delay responses to the user input in ways that increase accuracy of the responses, while retaining a conversational aspect of the ongoing chat. Moreover, in one or more embodiments, the disclosed systems generate responses utilizing natural language.