System and method for using network data to improve event predictions
First Claim
1. A system for using internet social network data for event predictions, the system comprising:
- one or more processors and a memory, the memory being a non-transitory computer-readable medium having executable instructions encoded thereon, such that upon execution of the instructions, the one or more processors perform operations of;
receiving an internet social media post stream from an internet social network;
generating and maintaining a social network database and an audit trail database;
filtering new social media posts on the internet social network to extract new social media posts related to possible future civil unrest by searching text in each new social media post for keywords related to social unrest and a future date, passing the new social media posts through a text-oriented logistic regression classifier, and geo-coding the new social media posts, wherein the new social media posts are geo-coded based on an l1-multivariate median of locations of all users who have shared the new social media posts;
comparing the new social media posts against the social network database and audit trail database to provide a quality score, the quality score representing a likelihood that at least one of the new social media posts is a true prediction of future civil unrest; and
wherein if the quality score is determined to be true, publishing the detection as a true prediction of future civil unrest.
1 Assignment
0 Petitions
Accused Products
Abstract
Described is a system for event prediction on microblogs. In operation, the system receives a social media post stream from a social network to generate and maintain a social network database and an audit trail database. New social media posts are filtered to determine if the new social media posts are related to a possible future event. The new social media posts are compared against the social network database and audit trail database to provide a quality score, the quality score representing a likelihood that at least one of the new social media posts is a true prediction of a future event. If the quality score is determined to be true, the detection is published as a true prediction of a future event.
-
Citations
18 Claims
-
1. A system for using internet social network data for event predictions, the system comprising:
one or more processors and a memory, the memory being a non-transitory computer-readable medium having executable instructions encoded thereon, such that upon execution of the instructions, the one or more processors perform operations of; receiving an internet social media post stream from an internet social network; generating and maintaining a social network database and an audit trail database; filtering new social media posts on the internet social network to extract new social media posts related to possible future civil unrest by searching text in each new social media post for keywords related to social unrest and a future date, passing the new social media posts through a text-oriented logistic regression classifier, and geo-coding the new social media posts, wherein the new social media posts are geo-coded based on an l1-multivariate median of locations of all users who have shared the new social media posts; comparing the new social media posts against the social network database and audit trail database to provide a quality score, the quality score representing a likelihood that at least one of the new social media posts is a true prediction of future civil unrest; and wherein if the quality score is determined to be true, publishing the detection as a true prediction of future civil unrest. - View Dependent Claims (2, 3, 4, 5, 6)
-
7. A computer program product for social unrest prediction on microblogs using internet social network data, the computer program product comprising:
a non-transitory computer-readable medium having executable instructions encoded thereon, such that upon execution of the instructions by one or more processors, the one or more processors perform operations of; receiving an internet social media post stream from an internet social network; generating and maintaining a social network database and an audit trail database; filtering new social media posts on the internet social network to extract new social media posts related to possible future civil unrest by searching text in each new social media post for keywords related to social unrest and a future date, passing the new social media posts through a text-oriented logistic regression classifier, and geo-coding the new social media posts, wherein the new social media posts are geo-coded based on an l1-multivariate median of locations of all users who have shared the new social media posts; comparing the new social media posts against the social network database and audit trail database to provide a quality score, the quality score representing a likelihood that at least one of the new social media posts is a true prediction of future civil unrest; and wherein if the quality score is determined to be true, publishing the detection as a true prediction of future civil unrest. - View Dependent Claims (8, 9, 10, 11, 12)
-
13. A computer implemented method for social unrest prediction on microblogs using internet social network data, the method comprising an act of:
-
causing one or more processers to execute instructions encoded on a non-transitory computer-readable medium, such that upon execution, the one or more processors perform operations of; receiving an internet social media post stream from an internet social network; generating and maintaining a social network database and an audit trail database; filtering new social media posts on the internet social network to extract new social media posts related to possible future civil unrest by searching text in each new social media post for keywords related to social unrest and a future date, passing the new social media posts through a text-oriented logistic regression classifier, and geo-coding the new social media posts, wherein the new social media posts are geo-coded based on an l1-multivariate median of locations of all users who have shared the new social media posts; comparing the new social media posts against the social network database and audit trail database to provide a quality score, the quality score representing a likelihood that at least one of the new social media posts is a true prediction of future civil unrest; and wherein if the quality score is determined to be true, publishing the detection as a true prediction of future civil unrest. - View Dependent Claims (14, 15, 16, 17, 18)
-
Specification