Yahoo! Inc.

Yahoo Assets LLC

Verizon Media Inc.

Oath Inc.

Excalibur Ip, LLC

Jean-Marc Langlois

Scott J Gaffney

Xing Yi

Arkady Borkovsky

Douglas Wesley Cook

Tomi Johan Poutanen

Hongyuan Zha

Nathan Liu

Hongwei Shang

Changsung Kang

Prabhakar Krishnamurthy

Neeti Narayan

Yi Chang

Dawei Yin

Bruno Fernandez Ruiz

Aharon Lavi

Belle L Tseng

Abhay K Gupta

Kostas Tsioutsiouliklis

Hongbo Deng

Xin Fan

Bruno Fernandez-Ruiz

Alyssa Glass

David Mark Ciemiewicz

Choon Hui Teo

Jiliang Tang

Ya Zhang

Pengyuan Wang

Tim Daly

Hua Ouyang

Yiping Han

Yan Jiao

Chao Tan

Jin Lim

Todd Beaupre

Disclosed are systems and methods for improving interactions with and between computers in content providing, searching and/or hosting systems supported by or configured with devices, servers and/or platforms. The disclosed systems and methods provide a novel messaging framework that automatically applies a multi-factor analysis technique to incoming and received messages in order to properly identify a message's type and category, which dictates the manner in which the message is displayed within a recipient's inbox. The disclosed framework operates on two levels: i) it determines whether a message is from a human or machine sender (H/M classification), and ii) it determines the messages category (MAGMA categorization).

Computerized system and method for multi-factor message classification and delivery

One or more computing devices, systems, and/or methods for oversampling for imbalanced test data are provided. A classifier is configured to classify data points as either belonging to a first class or a second class. A determination may be made that the first class and the second class are imbalanced where a first number of data points estimated to be part of the first class is a threshold amount less than a second number of data points estimated to be part of the second class. An oversampling ratio is determined for the first class. The oversampling ratio is used to select a sample set of data points for editorial labeling, where the sampling set of data points comprises a total number of data points below a threshold amount.

Oversampling for imbalanced test data

The present teaching relates to ranking search content. In one example, a plurality of documents is received to be ranked with respect to a query. Features are extracted from the query and the plurality of documents. The plurality of documents is ranked based on a ranking model and the extracted features. The ranking model is derived to remove one or more documents from the plurality of documents that are less relevant to the query and order remaining documents based on their relevance to the query. The ordered remaining documents are provided as a search result with respect to the query.

Method and system for ranking search content

Method, system, and programs for measuring user engagement. In one example, a model generated based on user activities with respect to a plurality pieces of content is obtained. One or more actual occurrences of the user activities with respect to one piece of the plurality pieces of content are identified. One or more future occurrences of the user activities with respect to the piece of content are estimated based on the model. A user engagement score with respect to the piece of content is calculated based on the one or more actual occurrences of the user activities and the one or more future occurrences of the user activities.

Method and system for measuring user engagement using click/skip in content stream

The present teaching relates to rewriting a query and providing search results. In one example, a plurality of queries is obtained. For each of the plurality of queries, one or more search results are identified. The one or more search results have been obtained in response to the query and have been previously selected by a user submitting the query. A plurality of titles is obtained. Each of the titles corresponds to one of the one or more search results with respect to one of the plurality of queries. A model is generated based on the plurality of queries and the plurality of titles. The model is to be used for rewriting a query.

Method and system for rewriting a query

Method and system for measuring user engagement using click/skip in content stream using a probability model

Methods, systems, and computer programs are presented for selecting content items for presentation to a user device of a user. One method includes an operation for measuring dwelltimes for a plurality of content items, where the dwelltime for a content item is based on the amount of time that the content item is displayed to a viewer. Further, the method includes an operation for training a classifier of content items based on the measured dwelltimes and based on features associated with the first set of content items, where the training includes calculating a relative importance of respective features for increasing measured dwelltimes. Additionally, the method includes an operation for sending additional content items to be presented on the user device based on the one or more ranking parameters, where the additional content items are associated with one or more ranking parameters that are expected to increase dwelltimes for the user of the user device.

Customized content stream utilizing dwelltime-based machine learning

A method for adjusting pricing for advertisements of non-guaranteed delivery ('NDG') advertising auctions includes retrieving dwell time information for users that engage advertisements streamed to the users in a display content stream. The method aggregates the advertisements into different groups according to display context (such as advertisement category, viewing device, product in which the advertisement is streamed, etc.) and computes short-click ratios of advertisements in each group based on a short-click threshold assigned to the group and the user dwell time information for the group. The method further determines average dwell time by users on advertisements in respective groups. The method adjusts pricing of the advertisements of a group for NGD bidding based on the short-click ratio and average dwell time of the group, to favor groups having higher dwell times and lower short-click ratios. Prices may be dynamically adjusted across different contexts based on user engagement information.

User engagement-based contextually-dependent automated pricing for non-guaranteed delivery

Embodiments of the present teachings disclose method, system, and programs for a multi-phase ranking system for implementation with a personalized content system. The disclosed method, system, and programs utilize a weighted AND system to compute a dot product of the user profile and a content profile in a first phase, a content quality indicator in the second phase and a rules filter in a third phase.

Method and system for multi-phase ranking for content personalization

Methods, systems, and computer programs are presented for selecting news articles for presentation to a user. One method includes an operation for measuring dwelltimes for a first set of news items, where the dwelltime for a news item is based on the amount of time that the news item is displayed to a viewer. Further, the method includes an operation for training a classifier of news items based on the measured dwelltimes and based on features associated with the first set of news items. Additionally, the method includes an operation for ranking with the classifier a second set of news items for presentation to the user, the ranking also using the profile of the user for delivery of customized news to the user. The ranked second set of news item is then presented to the user.

Customized news stream utilizing dwelltime-based machine learning

A server system of an online information system displays advertising items and content items retrieved from storage devices as a stream viewable by a user on a user device. The advertisement items and the content items are ordered in the stream by a ranking score computed for each of the advertisement items and each of the content items. A quality scoring system determines an affinity score between a user and a present content item based on features of the present content item matching user profile parameters associated with the user and identifies post-interaction satisfaction with a prior content item. The quality scoring system determines a quality score based on the affinity score and the post-interaction satisfaction. The quality score is used for ordering items in the stream. The server system transmits a web page including the stream to a user device over a network. In this manner, advertising items and content items compete in a unified marketplace for inclusion in the stream for viewing by the end user.

Quality scoring system for advertisements and content in an online system

A method for selecting a social media network user. The method comprises obtaining one or more parameters indicative of quality of social media network content from the social media network user, ranking the social media network user based on the one or more parameters, and determining whether the social media network user is selected based on the ranking.

Social media based content selection system

Growing community of inventors

Menlo Park, CA, United States of America

Jean-Marc Langlois