Systems are methods are provided for training an automated agent. The automated agent maintains a reinforcement learning neural network and generates, according to outputs of the reinforcement learning neural network, signals for communicating resource task requests. The system includes a communication interface, a processor, memory, and software code stored in the memory. The software code, when executed, causes the system to: instantiate an automated agent for communicating resource task requests; receive a current feature data structure related to a resource of the resource task requests; maintain a plurality of historical feature data structures related to said resource for a plurality of prior time steps; compute normalized feature data using the current feature data structure and the plurality of historical feature data structures; compute supplemented state data appended with the normalized feature data; and transmit said supplemented state data to the reinforcement learning neural network to train said automated agent.

An online concierge system trains a machine learning conversion model that predicts a probability of receiving an order from a user when the user accesses the online concierge system. The conversion model predicts the probability of receiving the order based on a set of input features that include price and availability information. For each access to the online concierge system, the online concierge system applies the conversion model to a current price and availability and to an optimal price availability. The online concierge system generates a metric as the difference between the two predicted probabilities of receiving an order.

An online system predicts whether a location will experience a threshold increase in traffic over the location's historical average amount of traffic. To predict a future deviation over historical traffic, the online system identifies events within a threshold distance of the location and determines an average number of indications that users will attend events within a threshold radius of the location during a prior time interval. The online system determines a total number of indications that users will attend future events within the threshold distance of the location, disregarding locations associated with less than a threshold number of future events and future events for which the online system received less than a threshold number of indications that users will attend, and determines a ratio of the total number of indications to the average number of indications that users will attend received for the prior events during the time interval.

The present disclosure relates to systems, non-transitory computer-readable media, and methods that utilize a specially trained machine-learning model to generate an emerging user segment based on a target outcome for digital survey responses and respondent attributes of respondents to such digital surveys. In some cases, for instance, the emerging user segment includes a group of users that share the same or similar characteristics as the subset of respondents. By analyzing respondent attributes of digital survey respondents that match a target outcome, the disclosed systems can use the specially trained machine-learning model to dynamically predict users that likely have (or are at risk of having) the same or a similar target outcome—even if such users did not respond to the relevant digital survey.

A system for assessing text content of a product. The system includes a computing device having a processor and a storage device storing computer executable code. The computer executable code, when executed at the processor, is configured to: provide text contents and confounding features of products; train a first regression model using the text content and the confounding features of the products; train the second regression model using the confounding features; operate the first regression model using the text contents and the confounding features to obtain a total loss; operate the second regression model using the confounding features of to obtain a partial loss; subtract the total loss from the partial loss to obtain a residual loss; use the residual loss to evaluate models and parameters for the regression models; and use the first regression model to obtain log odds of the words indicating importance of the words.

A system and method for comparing zones for different versions of a website based on performance metrics are provided. The method includes collecting comparison versions of at least one website specified in a received comparison request; identifying at least one comparison zone in the collected comparison versions, wherein a comparison zone is a zone included in the at least one specified website; collecting at least one comparison zone metric for each of the at least one comparison zones for the collected comparison versions, wherein the at least one comparison zone metric is a numerical value related to a user interaction metric with the at least one comparison zone; analyzing the collected comparison zone metrics to determine zone performance; and returning the analysis results.

A method and a computer system for performing the method of determining an initial value or lifetime value for a domain name is provided. The method for determining an initial value includes obtaining, over a communication network, a domain name from requestor; obtaining, over the communication network, one or more inputs from one or more domain name data sources; applying the one or more inputs and the domain name to an initial lifetime worth computer model, wherein the one or more inputs comprise data related to comparable historical domain names, data from a linguistic model analysis, data from a linguistic frequency list, and data related to a second-level domain to top-level domain relationship analysis; determining, by a hardware processor, an initial lifetime worth for the domain name based on the initial lifetime worth computer model; and providing the initial lifetime worth for the domain name to the requestor.

A tag management system can implement a combined synchronous and asynchronous tag-loading scheme. In an embodiment of this scheme, a synchronous tag may be included at the top of or near the top of a content page. This synchronous tag can hide one or more page elements that are to be modified by a subsequent asynchronous personalization tag. Subsequently, the asynchronous personalization tag can update the page element and then cause the page element to be displayed. As a result, flicker between the old and new page elements can be reduced or avoided.

A tag management system can implement a combined synchronous and asynchronous tag-loading scheme. In an embodiment of this scheme, a synchronous tag may be included at the top of or near the top of a content page. This synchronous tag can hide one or more page elements that are to be modified by a subsequent asynchronous personalization tag. Subsequently, the asynchronous personalization tag can update the page element and then cause the page element to be displayed. As a result, flicker between the old and new page elements can be reduced or avoided.

Method for monitoring the presence of a point-of-sale display in a shop, at the sight of consumers, the method comprising: acquiring signals from mobile devices by at least one signal sensor, sending one output from the signal sensor to an analyzing device, wherein the analyzing device: calculates an audience rate of the point-of-sale display on the basis of the output, and determines, on the basis of the audience rate, if the point-of-sale display is in the shop at the sight of consumers or not.