2. Why we need data mining in social
media
People share more data
Products are more customized
More users produce more data
Market research
Product development and design
3. Graph mining
Graphs appear in all form of information
In this approach communities corresponding to the same, are clustered
together
These clustered data tend to have more similarities
4. Graph mining usage
It is used by google to group links on a result page based on meta data
Graph mining also sorts the results based on the hit count of a link.
Each time an user clicks a links, the corresponding page rank is incremented
The links with top counts are grouped together
5. Graph mining on facebook
Searching people: Friends of friends who are single in Pollachi, here the
people with the higher mutual friends are grouped and sorted accordingly
Searching places: Restaurants in Chennai liked by friends
Searching interests: Movie most of my friend likes
Adds are tailored for user
6. Text mining
Text mining is an emerging technology that attempts to extract meaningful
information from unstructured textual data.
Social networks contain a lot of texts and posts made daily
7. Usage of text mining
Automatic processing of emails:
Texts are classified automatically
Junk mails are separated
We could potentially extract a complete information from a website
automatically
8.
9. Text mining methods
Cluster analysis:
Automatic or semi automatic analysis of large data and extract previously unknown
patterns
Anomaly detection:
It is the search of items or events that do not fit the expected pattern
10. Conclusion
Valuable information is hidden in vast amount of social media
Presenting suitable opportunities on extracting data form social media, we
could learn a lot
A great example for knowledge extraction on social media is that, Twitter
found that BJP is gonna win the election before the voting day itself