METHOD AND APPARATUS FOR TARGETING MESSAGES TO USERS IN A SOCIAL NETWORK
A method and apparatus for targeting messages to users in a social network, for example by first identifying topics in the social network is provided. One embodiment of a method for discovering topics in a social network includes collecting information from the social network, the information including at least one of: interactions between users of the social network or profile information for the users, determining a global topic model including at least one topic, based on the collected information, and locally refining the global topic model in accordance with the collected information.
1 . A method for grouping users of a computer network, the method comprising:
algorithmically analyzing content of a first set of observed interactions among the users;
automatically inferring a topic, based at least in part on the algorithmically analyzing;
associating a subset of the users with the topic; and
determining a relative strength of a relationship between a first user and a second user in the subset of the users, wherein the relative strength is based at least in part on a quantitative measure of a second set of observed interactions involving the first user and the second user.
2 . The method of claim 1 , wherein the first set of observed interactions includes electronic mail.
3 . The method of claim 1 , wherein the first set of observed interactions includes an instant message.
4 . The method of claim 1 , wherein the first set of observed interactions includes a posting to a web site, blog, or online forum.
5 . The method of claim 1 , wherein the first set of observed interactions includes a comment or tag made on a web site or blog.
6 . The method of claim 1 , wherein the algorithmically analyzing comprises:
extracting a first word from the first set of observed interactions;
generating an initial topic by associating the first word with a second word;
associating at least some of the users with the initial topic, based at least in part on a similarity measure that relates the initial topic to profile information for the at least some of the users.
7 . The method of claim 6 , wherein the automatically inferring comprises:
calculating a centroid of the profile information, wherein the centroid is the topic.
8 . The method of claim 6 , wherein the automatically inferring comprises:
merging the initial topic with another topic.
9 . The method of claim 1 , further comprising:
refining the topic based, at least in part, on the first set of observed interactions.
10 . The method of claim 9 , wherein the refining comprises:
filtering the first set of observed interactions according to a criterion related to the topic;
adjusting a weight of a link relating a pair of interactions in the first set of observed interactions, in accordance with a result of the filtering; and
indicating the weight in a model of the topic.
11 . The method of claim 10 , wherein the criterion is a frequency of appearance in the first set of observed interactions of a word related to the topic.
12 . The method of claim 11 , wherein the adjusting comprises:
setting the weight to a value that is proportional to the frequency.
13 . The method of claim 10 , further comprising:
dividing the topic into a plurality of sub-topics, based at least in part on the link.
14 . The method of claim 1 , wherein the automatically inferring comprises:
mapping the word into an ontology to produce a normalized set of concepts that defines the topic.
15 . The method of claim 14 , wherein the ontology is community-generated.
16 . The method of claim 1 , further comprising:
obtaining a second third set of observed interactions among the users; and
dynamically updating the topic, based at least in part on the third set of observed interactions.
17 . The method of claim 16 , wherein the dynamically updating is performed incrementally.
18 . The method of claim 17 , wherein the dynamically updating comprises:
filtering the third set of observed interactions according to a criterion related to the topic;
adjusting a weight of a link relating a pair of interactions in a cumulative set of interactions that includes the first set of interactions, the second set of interactions, and the third set of interactions, in accordance with a result of the filtering; and
indicating the weight in a model of the topic.
19 . The method of claim 16 , wherein the dynamically updating comprises adjusting the relative strength.
20 . A computer-readable storage device having stored thereon a plurality of instructions, the plurality of instructions including instructions which, when executed by a processor, cause the processor to perform a method for grouping users of a computer network, comprising:
algorithmically analyzing content of a first set of observed interactions among the users;
automatically inferring a topic, based at least in part on the algorithmically analyzing;
associating a subset of the users with the topic; and
determining a relative strength of a relationship between a first user and a second user in the subset of the users, wherein the relative strength is based at least in part on a quantitative measure of a second set of observed interactions involving the first user and the second user.