International Business Machines Corporation

Sharon Krisher

Leonid Bolshinsky

Vladimir Gamaley

Eitan Shapiro

Examples of techniques for performing multilingual queries are described herein. A computer-implemented method includes monitoring, via a processor, a client for language settings during user-generated content creation and detect one or more language settings. The method includes analyzing user-created content to detect a language of a text. The method includes compiling a list of scored preferred languages for the client based on the detected language settings and the detected language of the text. The method includes intercepting a query from the client. The method includes analyzing a text of the intercepted query in a plurality of selected languages based on a language setting of a user interface application, a detected language of the query, and a predetermined number of preferred languages of the client to produce results of analysis for each selected language. The method includes generating a multilingual query based on results of analysis for the selected languages.

Generating multilingual queries

Examples of techniques for sharing content based on topics are described herein. A method includes extracting a content topic from a piece of content. The method includes extracting a user topic based on a user activity. The method also includes matching the content topic with the user topic and generating a user score based on a detected number activities a user performs that include a matched user topic. The method further includes generating a sorted list of users with matched user topics, the list to be sorted by the user score.

Sharing content based on extracted topics

A method, computer program product, and computer system is provided. A processor identifies a shared property between a user of a social network and a group of the social network. A processor identifies an unshared property between the user of the social network and the group of the social network. A processor determines a first percentage of the one or more members of the group with the shared property. A processor determines a second percentage of the one or more members of the group with the unshared property. A processor generates a recommendation score for the group based, at least in part, on the first and second percentage.

Determining a group recommendation

Embodiments of the present invention provide methods, computer program products, and systems for web link quality analysis. Embodiments of the present invention can be used to receive a first indication that a web link has been accessed by a first device and record one or more resources of the web link and actions taken on the one or more resources of the web link and generate a score for the web link based on one or more similar web links to the web link and then displaying the score associated with the web link. Embodiments of the present invention can be used to generate a second score for the web link based on actions taken on the web link and an amount of accessed resources of the one or more resources of the web link and generate a display of the second score associated with the web link.

Web link quality analysis and prediction in social networks

A method, executed by a computer, for optimizing searches includes receiving a boolean search query comprising a plurality of operators and operands and usage information corresponding to a user, determining modifications to be made to the boolean search query according to the usage information, and modifying the boolean search query according to the modifications. A computer program product and computer system corresponding to the above method are also disclosed herein.

Optimizing searches

Techniques are described for automatic suggestion for socially connected experts in electronic discussions. In one example, techniques include extracting topics from an electronic discussion, generating a list of experts in the topics, and editing the list according to social closeness to participants in the electronic discussion.

Automatic suggestion of experts for electronic discussions

The method includes monitoring a computing device for language settings during user-generated content creation and detect one or more language settings. The method further includes analyzing user-created content to detect a language from a text of the user-generated content. The method further includes compiling a list of scored preferred languages for the computing device based on the detected language settings and the detected language of the text. The method further includes intercepting a query from the computing device. The method further includes analyzing a text of the intercepted query in a plurality of selected languages based on a language setting of a user interface application, a detected language of the query, and a predetermined number of preferred languages of the computing device to produce results of analysis for each selected language. The method further includes generating a multilingual query based on the results of analysis for the selected languages.

A computer system with instructions for providing quality of service includes identifying a high network traffic event at a node, providing a user score to each user account of a plurality of user accounts, wherein the plurality of user accounts corresponds to a plurality of user devices connected to the node, and responsive to identifying the high network traffic event, providing instructions to a subset of user devices of the plurality of user devices, wherein the instructions instruct the subset of devices to enter a low traffic mode, and wherein each device of the subset of devices corresponds to a user account with a user score below a threshold.

Quality of service for high network traffic events

A method for providing quality of service includes identifying a high network traffic event at a node, providing a user score to each user account of a plurality of user accounts, wherein the plurality of user accounts corresponds to a plurality of user devices connected to the node, and responsive to identifying the high network traffic event, providing instructions to a subset of user devices of the plurality of user devices, wherein the instructions instruct the subset of devices to enter a low traffic mode, and wherein each device of the subset of devices corresponds to a user account with a user score below a threshold. A computer program product and computer system corresponding to the above method are also disclosed herein.

Some users (also called sharers) desire to share content with recipients (also called recipients) over a virtual community or other computerized means. Some of the content to be shared (first-order content) may include links, hyperlinks, or other references (links) to additional content (each individually a second-order content). These content layers may continue indefinitely and the content layers may contain links both within and without the domain or virtual community of the first-order content. A sharer may be unaware of whether various recipients have access to any of the various layers of content (first-order or otherwise). Embodiments of the present invention transmit the likelihood for individual recipients to access various layers of content to the sharer.

Access predictions for determining whether to share content

A computer program product with instructions for providing quality of service includes identifying a high network traffic event at a node, providing a user score to each user account of a plurality of user accounts, wherein the plurality of user accounts corresponds to a plurality of user devices connected to the node, and responsive to identifying the high network traffic event, providing instructions to a subset of user devices of the plurality of user devices, wherein the instructions instruct the subset of devices to enter a low traffic mode, and wherein each device of the subset of devices corresponds to a user account with a user score below a threshold.

In an approach to automating multilingual indexing, a computer receives text of a conversation between at least two users. The computer detects at least one language associated with the text. The computer determines whether the language associated with the text is detected with a confidence level that exceeds a threshold. The computer retrieves text from one or more previous conversations between the two users. The computer detects at least one language associated with the text. The computer determines whether the at least one language associated with the text is detected with a confidence level that exceeds a pre-defined threshold. The computer analyzes the text using at least one of the detected languages to create one or more terms. The computer indexes the one or more terms and stores a boost value associated with each of the one or more indexed terms corresponding to confidence level of the detected language.

Growing community of inventors

Haifa, Israel

Sharon Krisher