I have dataset of author and their coauthor. I made a graph of this dataset. It contains a huge no. of cycles. I have tried JGRAPHT java API to find these cycles by using Johnsons Algorithm but failed on real data of about 577 publications. It is working on quite small set of authors and coauthors data.
If anybody know some good API in java or in any language or suggest me some other good algorithm for the said purpose.
or
Can I use community detection algorithm for this purpose or strongly connected component??
Please guide..
Thanks in advance..
I think topological sort Applies only on directed acyclic graphs (DAGs).
First of all thanks for your interest. Yes these are symmetric relations and list grows exponentially. I need longest non-overlapping list of cycles or longest 6 list are required not short lists.
Now i further explain about it with a simple example.
Basically these are the coauthors relationships.
As if in a paper there are 4 authors a,b,c,d. then the relationships would be
a-b, a-c, a,d
b-c,b,d
c-d
and reverse of all of the above. I mean these all relationships are two way. It forms complete graph.a-b is a cycle and b-a is also cycle of two nodes. It contains a huge number of cycles of small to large number of nodes cycles.
Say in a paper there are 10 authors then it's graph would be complete graph of 10 nodes and there would be 90 relationships between them. However no of cycles may be more than 1000 or more not sure about it.
If you still need some explanation I would be happy to clarify.
- Badges
- Science topic
- Similar topics
- Engineering