Amazon Technologies, Inc.

Yahoo! Inc.

Google Inc.

Like.com

Ashwin Devendrappa Tengli

Dragomir Dimitrov Anguelov

Salih Burak Gokturk

Vincent Olivier Vanhoucke

Munjal Shah

Kuang-chih Lee

Hemant Jagadish Kowshik

Diptendu Bhowmick

Rajeev R Rastogi

Rajeev Ramnarain Rastogi

Srinivasan H Sengamedu

Srinivasan Hanumantha Rao Sengamedu

Aravindan Raghuveer

Krishna Prasad Chitrapura

Jeyashankher Ramamirtham

Sandeepkumar Bhuramal Satpal

Pankaj Gulhane

Various approaches enable the detection of 'viewed' content. For example, in digital advertising, when a user visits a web page, the web page can be rendered with an advertisement. Content is considered viewable when a user is given enough opportunity to view the content. Approaches described herein attempt to determine whether a user was given an opportunity to view such of such content by determining and using at least one of a viewability metric, performance activity level metric, snapshot of the content, or combination thereof.

Content viewability detection

Various approaches allow for determining fraudulent activity as it relates to a publisher's reported web address and/or whether content has been viewed, or at least displayed with the opportunity of being viewed. When a user visits a web page, or other such page or application that displays content, the web page can be rendered with advertisements and/or other content. Approaches described herein can verify a publisher's reported web address as well as attempt to determine whether a user is given an opportunity to view such content by determining and utilizing a snapshot of the content.

Publisher identity verification through cross-domain barrier

High conversion rate content can be displayed with primary content from one or more publishers in order to determine whether the content is being displayed to human users or provided to automated processes such as robots. Convertible content such as advertising will generally result in conversions or other actions within an expected range of occurrences. Convertible content performing significantly below the range can be indicative of robotic traffic. Such determinations can be difficult for publishers with low volume traffic, however, as there may not be sufficient data to make an accurate determination. For such publishers, or users viewing content for such publishers, high conversion rate content can be displayed that will allow such determinations to be made with fewer data points. The rates can be used to determine robotic users, which can be blocked, as well as to determine poorly performing placements of the content by the publishers.

Tracking user activity for digital content

A similarity search may be performed on the image of a person, using visual characteristics and information that is known about the person. The search identifies images of other persons that are similar in appearance to the person in the image.

Computer-implemented method for performing similarity searches

Embodiments of methods, systems and/or apparatuses relating to data processing in distributed computing environments are disclosed. In particular, methods, systems, and/or apparatuses for determining information similarly and/or performing related statistical techniques which may be implemented or operated in a distributed computing environment are disclosed.

Information similarity and related statistical techniques for use in distributed computing environments

Web pages are efficiently categorized in a data processor without analyzing the content of the web pages. According to at least one embodiment, data is maintained that represents sample URLs grouped into a plurality of clusters. The sample URLs of a cluster are used to produce a URL regular expression pattern ('URL-regex') that differentiates the sample URLs of the cluster from the sample URLs of other clusters and that covers at least a specified percentage of the sample URLs in the cluster. The process of producing a URL-regex is repeated for each of the clusters producing a URL-regex for each cluster. Web pages are then categorized into one of the clusters by determining which of the URL-regex patterns produced for the clusters match URLs that refer to the web pages. Thus, a web page may be categorized based on a URL that refers to the web page without having to obtain and analyze the content of the web page.

Techniques for categorizing web pages

A method includes generating, electronically, one or more matching patterns for one or more pairs of attribute values. Each pair includes two attribute values. The two attribute values include a first attribute value from a first record and a second attribute value from a second record. The first attribute value and the second attribute value satisfy a first criterion. Further, the method includes identifying, electronically, matching segment between the first attribute value and the second attribute value of a first pair. The method also includes repeating identifying for each pair. Moreover, the method includes computing a similarity score for the first pair using one of the first pair and the matching segment based on the one or more matching patterns and matching segments of the one or more pairs satisfying a second criterion. The method also includes repeating computing for each pair.

Method and system for determining similarity score

Methods and systems that label a web page by collecting a set of inbound labels for the web page, estimating a language model for the web page, computing the likelihood of generating each inbound label given the language model and assigning a score to each inbound label based on this likelihood, and assigning a label to the web page based on the score assigned to each of the set of inbound labels. Inbound labels are preferably collected from the set of web documents linking to the web page. Labels assigned are useful in providing labeled links to web pages from top hosts in search results pages.

Growing community of inventors

Bengaluru, India

Ashwin Devendrappa Tengli