Encoding and using information about distributed group discussions

摘要

Techniques are described for analyzing user-supplied information, including to predict future aspects of additional related information that will be supplied by users. The user-supplied information may include distributed group discussions that involve numerous users and occur via user comments and other content items supplied 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, summarizing and encoding information about multiple selected factors for comments supplied for particular topics or a category during each period of time (such as to quantify an amount of user interest), predicting future values for the selected factors for the topics and category during one or more future period of times, and taking one or more further actions based on the predicted information.

主权项

1. A computer-implemented method comprising:
obtaining, by one or more configured computing systems of a group discussion prediction service, information about a distributed group discussion involving a plurality of users submitting a plurality of content items for the distributed group discussion over multiple time periods, wherein the plurality of content items include multiple content items submitted during each of the multiple time periods; receiving, by the one or more configured computing systems, a selection of one of multiple encoding schemes, each encoding scheme having a distinct group of multiple factors to use in summarizing information about groups of content items; generating, by the one or more configured computing systems and based on the selected encoding scheme, an aggregation of encoded summary information to represent the distributed group discussion for the multiple time periods, the generating of the aggregation including determining, for each of the multiple factors of the selected encoding scheme and each of the multiple time periods, a value that summarizes information for that factor about the multiple content items submitted for the distributed group discussion during that time period; matching, by the one or more configured computing systems, the generated aggregation of encoded summary information to a first portion of a defined prediction template having encoded information representing changes over time related to one or more other distributed group discussions; predicting, by the one or more configured computing systems, future information about the distributed group discussion for multiple future time periods using a second portion of the defined prediction template that is distinct from the first portion, the predicting including determining, for each of the multiple factors of the selected encoding scheme and for each of the multiple future time periods, an expected value that summarizes information for that factor about future content items that are expected to be submitted for the distributed group discussion during that future time period; and providing, by the one or more configured computing systems, at least some of the predicted future information for the multiple future time periods.