Methods for analyzing large time series data set?

I have a large data set (separated in to smaller data sets) containing data in the following format:

# Start 'Data Entry'
:: 09:53:06.488000        # Time of data logged
>> MessageName|mouse move # Type of measurement, in this case a mouse movement
>> Position|(465, 454)    # Position of mouse after movement
>> BPM|98.34              # Some measurement described as a float
# End 'Data Entry'

:: 09:53:21.267000 
>> MessageName|key down
>> Ascii|d
>> BPM|98.21

For each individual data set (data taken from a specific person), I have tens of thousands of these data entries.

I need to identify relationships between between BPM and the other data over time.

For example, I could look at the relationship between the mouse positions over time and the respective BPM, and find whether or not say, the angle between mouse positions has a relationship to BPM.

What I need is a way to analyze the data to find relationships that I can't derive myself, and patterns within the data that aren't obvious to the human eye.

I have looked into 'MINE' but I'm not sure if it's suited to my needs.

Are there specific algorithms, or techniques, dedicated to this type of data analysis.