Analyzing distributed group discussions
Inventors
Browning, Randolph • Snelling, David Aaron • Papp, Stefan Nicholas • Zimmermann, Bernhard G. • Downs, Oliver B.
Assignees
Publication Number
US-9386107-B1
Publication Date
2016-07-05
Expiration Date
2033-03-06
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
Techniques are described for analyzing user-supplied information, including in at least some situations to predict future aspects of additional related information that will be supplied by users. The user-supplied information that is analyzed may, for example, include distributed group discussions that involve numerous users and occur via user comments made to one or more social networking sites and/or other computer-accessible sites. The analysis of user-supplied information may, for example, include determining particular topics that are of interest for a specified category during one or more periods of time, quantifying an amount of user interest in particular topics and the category during the period of time, predicting future amounts of user interest in the particular topics and the category during one or more future period of times, and taking one or more further actions based on the predicted information.
Core Innovation
The invention provides techniques for analyzing user-supplied information, especially distributed group discussions occurring on social networking sites and other computer-accessible platforms. The system obtains textual comments from multiple users over different time periods, analyzes these comments to determine topics of discussion and content categories, quantifies user interest in these topics, and predicts expected future comments and user engagement for specific topics and categories.
A core aspect of the solution is the ability to automate the collection and analysis of distributed user comments across various sources and timeframes. The invention addresses challenges inherent to the distributed and dynamic nature of such discussions by grouping comments, associating them with relevant topics and user attributes, and then leveraging historical patterns and prediction templates to forecast future discussion activity, topic relevance, and related metrics.
The problem being solved, as described in the background, is the difficulty in obtaining and disseminating timely information about ongoing and past distributed discussions, due to their occurrence across multiple platforms and among numerous, often uncoordinated, users. The present invention overcomes this by providing systems and methods that aggregate, analyze, and project trends in distributed group discussions, enabling more informed, responsive actions based on these insights.
Claims Coverage
The patent includes several independent claims covering methods, systems, and computer-readable media for the analysis and prediction of distributed group discussions based on user-supplied comments.
Analyzing distributed user comments to determine topics and predict future discussion
The invention obtains information from social networking sites or other network-accessible sites, including textual comments from users over multiple prior time periods. The contents of these comments are analyzed to identify topics and match them with specified content categories based on keywords. A subset of topics is determined based on thresholds for the quantity of comments mentioning those topics within categories, excluding overly rare or common topics. The system then identifies the number of comments mentioning a topic in each time period, matches these quantities to a portion of a prediction template, and predicts expected future quantities of comments associated with the topic using information from another portion of the prediction template. The predicted information is then provided to recipients via electronic interactions.
Generating and using prediction templates and trend information for forecasting
Trend prediction or template matching is employed for forecasting expected future comment quantities. A first histogram is generated to represent prior comment quantities, and this is matched to a subset of a prediction template, itself a longer-term histogram based on prior actual data. Prediction templates are generated or updated based on whether previous predictions matched actual future comment quantities, allowing the system to dynamically improve forecasts and choose appropriate templates for current and future situations.
Providing actionable output and automated responses based on predictions
The system provides the forecasted information to one or more recipients, such as clients or other users, through electronic interactions. Further, based on predicted expected quantities, the system may determine specific types of information to make available to users and can automatically supply additional textual comments to social networking sites or data sources that correspond to the predicted discussion topics and categories.
Cross-sectional analysis by user, source, or location attributes
The system can determine increases or decreases in the number of comments associated with specific users, geographical locations, or information sources, relative to others. It can also perform influence analysis to identify users who have greater effect on discussions, and sentiment analysis to track increases or decreases in positive or negative sentiment over time. This information is provided as actionable indications to recipients.
Computer-implemented systems and storage media for executing the above methods
The claimed invention includes systems comprising one or more processors and memories, executing software instructions to perform the methods of obtaining, analyzing, predicting, and providing information about distributed group discussions, and generating new prediction templates based on observed data. Non-transitory computer-readable media are also claimed, storing instructions for performing these operations.
In summary, the claims broadly cover methods, systems, and media for automated analysis of distributed group discussions through obtaining, categorizing, quantifying, and predicting comment trends using templates and trend lines, with actionable output and dynamic adaptation based on ongoing data.
Stated Advantages
The analysis can be performed in real-time or near-real-time, enabling timely insights into ongoing distributed discussions.
The system provides automated predictions about future discussion activity, allowing entities to plan or respond proactively.
By grouping and analyzing comments from multiple sources, topics, and time periods, the invention allows for a comprehensive, cross-sectional understanding of distributed group discussions.
Predicted information can trigger automated actions, including providing supplemental or corrective information to influence discussion directions.
The creation and selection of dynamic prediction templates enhance predictive accuracy over time.
The system supports the quantification of user influence and sentiment shifts within discussions, delivering actionable insights for clients and system operators.
Documented Applications
Providing real-time or near-real-time analysis and prediction of interest in specific topics or categories on social networking or other computer-accessible sites.
Supplying analyzed and predicted information about topics or categories to clients, such as news organizations, companies, or entities interested in user engagement for planning or targeted actions.
Automatically providing supplemental, supportive, or refutational information or comments to influence the course of discussions based on predicted trends.
Analyzing distributed discussions for sentiment trends and user influence to inform advertising, targeting, or other automated actions.
Interested in licensing this patent?