Cbs Interactive, Inc.

Stubhub, Inc.

Clifford Lyon

Ron H Rothman

Ned Rhinelander

Girish Vasvani

Mats Nilsson

Geoff Lester

Camila Franco

Vikram Chalasani

Operations to blend electronic inventory are provided. The operations may include obtain a request for multiple seats at a venue and determine a combination of data objects from a data storage that represent seats at the venue that meet the request. The combination of the data objects may include at least one of multiple first data objects and at least one of multiple second data objects. The first data objects may be tagged with a first tag and are associated with first seats in the venue controlled by a first entity and the second data objects may be tagged with a second tag and are associated with second seats in the venue controlled by a second entity, and the combination of the data objects is determined irrespective of whether the data objects include the first tag or the second tag.

Blending electronic inventory

Various embodiments are generally directed to an automated A/B testing system using a combination of classical and alternative statistical analysis to control the performance A/B tests. An apparatus includes a processor circuit; and logic to commence collection of a specified number of samples of user responses to multiple versions of a user interface in a test, employ an alternative statistical analysis to analyze the samples as they are collected, determine whether a proposed version elicits a statistically significant improvement in user response over a control version, terminate the test before the specified number of samples are collected in response to a statistically significant improvement over the control version elicited by the proposed version, and complete collection of the specified number of samples and employ a classical statistical analysis to analyze the samples in response to the proposed version not eliciting the statistically significant improvement. Other embodiments are described and claimed.

Automated experiment scheduling

Assigning tags to content on a page stored within a digital document environment is described. An activity log is received for a document including data indicating search queries resulting in the location of the document by a search engine. One or more keywords are determined from the activity log, and the document is tagged with a tag associated with the keywords.

Assigning tags to digital content

Various embodiments are generally directed to a visual configuration environment for configuring A/B tests of variations of aspects of content presentation in a user interface of a portal of a server to support remotely accessing services that it provides. A method includes identifying multiple visual areas of a visible portion of a user interface in a first data script; visually presenting on a display the multiple visual areas in at least one menu in a first table to enable selection of first and second visual areas of the multiple visual areas for inclusion in an A/B test of multiple versions of the user interface; and visually presenting a list of all possible combinations of variations in content presentation selected in response to the visual presentation of the first table for the first and second visual areas in the form of a second table. Other embodiments are described and claimed herein.

A/B test configuration environment

Systems and methods for clustering content according to similarity are provided that identify and group similar content using a set of tags associated with the content. A topic model of a group of content is built, producing a probability distribution of topic membership for the content. Individual items of content are then clustered using a clustering algorithm, and a distance matrix from the probability distribution is built. Based on the distance matrix, individual items of content are labeled as 'must-link' or 'cannot-link' pairs with the group of content. The topic model is then embedded into successively smaller dimensions using a kernel method, until the clustering is stable with respect to both the behavioral and content domains.

System and method for clustering content according to similarity

Disclosed herein are systems, methods, and computer-readable storage media for consistently associating a user ID with a user. In one aspect, a method may comprise receiving a data request at a transceiver accessible by a server computing device, wherein, in response to receiving the data request, the server computing device is configured to: access, by one or more processors coupled to the transceiver, a first value embedded in a first cookie associated with a first expiration time; set a second cookie embedded with a second value; and reset the first cookie responsive to receiving an indication that the first cookie is expiring, wherein to reset the first cookie comprises replacing the first cookie with a third cookie set with a value equal to the first value and a third expiration time later than the first expiration time. Other embodiments are disclosed and claimed.

Growing community of inventors

Melrose, MA, United States of America

Clifford Lyon