Hello everyone, very glad to share my
thoughts here!
After three courses taken, I get two
essential points about Social Media Analytics and want to share my opinions
with you.
Social circle exists in many websites, such
as Facebook, Twitter, Google+, Renren, Weibo, Qzone, Wechat, and even Baidu
Post Bar which can also be identified as a kind of social network.
fig.1 major social media
Based on the complex relationship in these
social circles, a website can form a social graph and use it as a reference to
recommend new friends to a certain person so as to maintain or expand the
volume of its users. This is the first aspect and the other one is as what I
have mentioned in class, big data.
Analyzing deeply into the specific messages one
posts or sends to another person, a website can use the technique of NLP to get
learned about what a certain person or a circle of friends mostly concerned
about. Using the data, ads can be pushed, resulting in boosting consumption. In
this way, business value can be realized.
fig.2
using the cosine of two item of news to identify similarity
I ever read a book by Dr. Jun Wu. The book “The Beauty of Mathematics” mentioned the
basic techniques used by Google, including NLP and its effect on search and ads
in several chapters. The book introduced the use of cosine theorem and Bayesian
network in the classification of news and key words. My knowledge about these
aspects are enlightened by this interesting book, so I recommend it to you
here.
