Given enough data, this can be done automatically. A simple way to do it self-organizing maps (SOMs). On the top of this wiki page: http://en.wikipedia.org/wiki/Self-organizing_map there is an example where members of congress are grouped based on voting patterns.
Superficially it might seem that the SOM has created a left-right spectrum of sorts. Actually this is not the case, as the clustering distances show. Instead, what has been done is just that members have been grouped on a 2-d plane such that members with similar views are closer to each other.
Given enough data, this can be done automatically. A simple way to do it self-organizing maps (SOMs). On the top of this wiki page: http://en.wikipedia.org/wiki/Self-organizing_map there is an example where members of congress are grouped based on voting patterns.
Superficially it might seem that the SOM has created a left-right spectrum of sorts. Actually this is not the case, as the clustering distances show. Instead, what has been done is just that members have been grouped on a 2-d plane such that members with similar views are closer to each other.