On Thu, Mar 25, 2010 at 3:21 AM, Chinmaya Poswalia < chinmaya.poswalia@gmail.com> wrote:

Hi Andrew, I am finding all the graph topic(my favorite topic to work on) very interesting that's why i asked a lot of questions regarding all in one mail.

- i want to work on semantic graph

- how about maintaining an adjacency list, denote V as nodes and E as edges, so the space complexity will be O(V+E).

- Also we can employ pointers of type void* to point to edges/vertices of different types so as to maintain the graph of different edge and vertex types.

- rather than employing a vector in each node and edge either we should allocate space during insertion or we maintain the size of vector array by ourself.

- Through generic programming, all the operations being implemented on graph up till now will be made applicable on these new members of the library. - various functions that can be induced on these graphs are

1. Distribution of node degree 2. finding clustering co-efficient of semantic graph 3. relevance of a link and node 4. scale on semantic graph 5. disparity of connected types/dispersion co-efficient.

- Semantic graph are generating huge level of importance in networking (network semantics on web), video screening, statistical analysis of data,etc so it will be a great thing to work on it.

Can you suggest some more links to read on for these

I am working on my proposal about the graph and your help is crucial for me. With regards Chinmaya