Problems analysing opinion in social networks

Analysis of sentiment of community by the name of the community

To judge a factor so subjective like a person's opinion on something is a tough task and must go beyond a simple title. A robust analysis must be made, using several factors such as the content of the forums in the community.

Discussions on a brand/product always occur in communities of the same name brand/product

To get things worse, I believe in this case what's a big issue is noisy data like SPAM and topics without any relevance in forums (ex:"Would you hookup with the person above?").

To think on the analysis of the social media as a clipping

On the opinion retrieval task, it is common to exclude page blocks judged without importance, like a blogroll and the trackbacks, and sometimes they are interpreted as part of the Blog's template, though blocks like these could give auxiliary information to the opinions, if they are treated in a different way.

Who executes the maketing strategies also measures the results

Doing this is more an attempt to lower the spends, because they don't understand the complexity of the problem of Buzz analysis. Or simply, it could be a marketing over a marketing, "not telling all the truth in statistics" for example, it happens...

The monitoring task on the Web is a complex challenge

I believe it is already possible to build Search Engines for that specific purpose using techniques available today. And the analysis of feeling can be obtained with the classifier SVM, or to use fuzzy logic, and even a combination of these or other techniques, finally: is possible and I believe that gives good results.

It will be possible to analyse socio-demographic data with accuracy

I believe that this type of error is close to an end, because the Web is increasingly moving to something more personal, that is, take Google Friend Connect for example, from the time that the person logs into his Google account and let comments and ratings, he already has left somewhere in its footsteps into Google domain valuable information that help in creating a profile. And with that profile much can be done, especially recommendations of any kind.

Is it possible to analyze networks with metrics for popularity?

In the social media, audience is just one of the factors, as well as the analysis of opinions. All this is part of a complex network that will help create the Web 3.0, as personal services and ease of use in the end, fold up a Web omnipresent, and then everyone should and will need to use.

[E.LIFE]