SNAP (Stanford Network Analysis Platform) Large Network Dataset Collection,
Jure Leskovec and Anrej Krevl, http://snap.stanford.edu/data, June 2014.
email: jure at cs.stanford.edu
Stack Overflow temporal network
https://snap.stanford.edu/data/sx-stackoverflow.html
Dataset information
This is a temporal network of interactions on the stack exchange web site
Stack Overflow (http://stackoverflow.com/). There are three different types
of interactions represented by a directed edge (u, v, t):
user u answered user v's question at time t (in the graph
sx-stackoverflow-a2q) user u commented on user v's question at time t (in
the graph sx-stackoverflow-c2q) user u commented on user v's answer at time
t (in the graph sx-stackoverflow-c2a)
The graph sx-stackoverflow contains the union of these graphs. These graphs
were constructed from the Stack Exchange Data Dump
https://archive.org/details/stackexchange). Node ID numbers correspond to
the 'OwnerUserId' tag in that data dump.
Dataset statistics (sx-stackoverflow)
Nodes 2,601,977
Temporal Edges 63,497,050
Edges in static graph 36,233,450
Time span 2774 days
Dataset statistics (sx-stackoverflow-a2q)
Nodes 2,464,606
Temporal Edges 17,823,525
Edges in static graph 16,266,395
Time span 2774 days
Dataset statistics (sx-stackoverflow-c2q)
Nodes 1,655,353
Temporal Edges 20,268,151
Edges in static graph 11,226,829
Time span 2773 days
Dataset statistics (sx-stackoverflow-c2a)
Nodes 1,646,338
Temporal Edges 25,405,374
Edges in static graph 11,370,342
Time span 2773 days
Source (citation)
Ashwin Paranjape, Austin R. Benson, and Jure Leskovec. "Motifs in Temporal
Networks." In Proceedings of the Tenth ACM International Conference on Web
Search and Data Mining, 2017.
Files
File Description
sx-stackoverflow.txt.gz All interactions
sx-stackoverflow-a2q.txt.gz Answers to questions
sx-stackoverflow-c2q.txt.gz Comments to questions
sx-stackoverflow-c2a.txt.gz Comments to answers
Data format
SRC DST UNIXTS
where edges are separated by a new line and
SRC: id of the source node (a user)
TGT: id of the target node (a user)
UNIXTS: Unix timestamp (seconds since the epoch)
---------------------------------------------------------------------------
Notes on inclusion into the SuiteSparse Matrix Collection, July 2018:
---------------------------------------------------------------------------
The SNAP graph is 1-based, with nodes in all graphs numbered 1 to
n=6,024,270.
In the SuiteSparse Matrix Collection, the primary matrix, Problem.A, is
the overall static graph, with 36,233,450 edges, of size n-by-n with
n=2,601,977. These edges represent the 63,497,050 temporal edges. A(i,j)
is the number of times person u=nodeid(i) interacted with person
v=nodeid(j), with a temporal edge (u,v,t), with any kind of interaction.
Problem.aux.nodeid is a list of the node id's that appear in the SNAP data
set.
A2Q = Problem.aux.Q2A is the static sx-stackoverflow-a2q graph.
C2Q = Problem.aux.C2Q is the static sx-stackoverflow-c2q graph.
C2A = Problem.aux.C2A is the static sx-stackoverflow-c2a graph.
These sum together to give the the overall graph. That is,
A = A2Q + C2Q + C2A.
A2Q(u,v) is the number of times person u answered v's questions.
C2Q(u,v) is the number of times person u commented on v's question.
C2A(u,v) is the number of times person u commented on v's answer.
The temporal edges are held in:
Problem.aux.temporal_edges: [63,497,050 x 3]
Problem.aux.temporal_edges_a2q: [17,823,525 x 3]
Problem.aux.temporal_edges_c2q: [20,268,151 x 3]
Problem.aux.temporal_edges_c2a: [25,405,374 x 3]
Each row in these matrices is a single temporal edge, (u,v,t). Summing up
all entries in A gives 63,497,050 and likewise the sum of entries in the
other graphs gives the number of temporal edges they represent.