What is LIBSVM label in ARFF term?

Question

I just begin with Data Mining technique and get familiar with some data file format. I have a first look at *.ARFF and now *.LIBSVM.
Can you teel me what is the first number at the begin of line? Can I understand that number in ARFF term

+1 1:70.000000 2:1.000000 3:4.000000 4:130.000000 5:322.000000 6:0.000000 7:2.000000 8:109.000000 9:0.000000 10:2.400000 11:2.000000 12:3.000000 13:3.000000 
-1 1:67.000000 2:0.000000 3:3.000000 4:115.000000 5:564.000000 6:0.000000 7:2.000000 8:160.000000 9:0.000000 10:1.600000 11:2.000000 12:0.000000 13:7.000000 
+1 1:57.000000 2:1.000000 3:2.000000 4:124.000000 5:261.000000 6:0.000000 7:0.000000 8:141.000000 9:0.000000 10:0.300000 11:1.000000 12:0.000000 13:7.000000 
-1 1:64.000000 2:1.000000 3:4.000000 4:128.000000 5:263.000000 6:0.000000 7:0.000000 8:105.000000 9:1.000000 10:0.200000 11:2.000000 12:1.000000 13:7.000000 
-1 1:74.000000 2:0.000000 3:2.000000 4:120.000000 5:269.000000 6:0.000000 7:2.000000 8:121.000000 9:1.000000 10:0.200000 11:1.000000 12:1.000000 13:3.000000 
-1 1:65.000000 2:1.000000 3:4.000000 4:120.000000 5:177.000000 6:0.000000 7:0.000000 8:140.000000 9:0.000000 10:0.400000 11:1.000000 12:0.000000 13:7.000000 

Why I see alot of +1 and -1 repeat in http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/

Answer 1

First number is the class label.

Answer 2

Support Vector Machines use support vectors to represent positive and negative classes.

The first column most likely gives +1 for positive examples, and -1 for negative examples. That is why there is no other value than these two...

You may want to read the documentation of libsvm:

Instructions for using LIBSVM are in the README files in the main directory and some sub-directories. README in the main directory: details all options, data format, and library calls. tools/README: parameter selection and other tools