2008. 2. 12. 11:01
force-directed placement netflix prize/시각화2008. 2. 12. 11:01
Re: A question about your visualization of Netflix movie data set.
The movies can be thought of as nodes in a graph, and the similarities can be thought of as weighted edges. Then a force directed layout is used to layout the graph. Take a look at:
On 2/10/08, 전광성 <deltakam@naver.com> wrote:
The movies can be thought of as nodes in a graph, and the similarities can be thought of as weighted edges. Then a force directed layout is used to layout the graph. Take a look at:
Todd
On 2/10/08, 전광성 <deltakam@naver.com> wrote:
How are you?
I'm korean students who is interested in developing recommender system.(not commecially, just for my study and learning)
I have a question about your visualization of netflix movie data set.
I understand that you used movie similarity extracted from the article you mentioned in your web site.
However, how did you use it? I mean, we need 2 coordinate, x and y axis... 1 axis (x) can be drawn from the similarity, but, how about y axis? did you use different form of similarity indicator for another axis?
I got a plan on exhibiting my recommender system at IT festival in South Korea held by SAMSUNG SDS, and I DO need to visualize my system, but struggling with how to show the principle of the algorithm....
plz help me
