what kind of help? please specify it.

I am currently working for a  software company.
I need to do the data mining for their new project, and i am the only one that have the statistics background. The new project is like .... searching something from the web sites as well as discussion forums, then search the keywords. However, the problem is we have a huge amount of data; and I think the most tough problem is how to select the data. Take an example, we need to exact the keyword " what is the best sushi in Beijing " from a discussion forum,and what we use Java to do it .
there are many ppl talking about sushi in Beijing on the forum....so hard to figure out what they are talking about because someone says " good" but actually means bad.
So.... my duty is data mining, so I find it really hard to figure out what kind of data we need ( like some data are just outliers ... etc.)

that s why i am here to try to get help from some body who has great experience on this area, coz i just graduate from the University, my background is statistical related, but i never do those data mining stuffs when i am in the university.

Thanks

没有高人指点下么？
本人菜鸟一个。。。

First, you need to find the data for you analysis. Maybe you can get some data from some internet, such as dianping.com. Second you need to choose the right tools for your analysis. You need to know it's supervised learning or not.

这个我也知道阿哥。。。 问题啥tools阿？
数据现在非常零乱。。。我用Java介入twitter 和facebook之类的地方搞出来的
要用啥样的模型去fit我也不知道呢
初步想还是regression....我sas用的不好。。。基本打算用R