Lifelong Graph Learning for Graph Summarization

Summarizing web graphs is challenging due to the heterogeneity of the modeled information and its changes over time. We investigate the use of neural networks for lifelong graph summarization. After observing the web graph at a point in time, we train a network to summarize graph vertices. We apply this trained network to summarize the vertices of the changed graph at the next point in time. Subsequently, we continue training and evaluating the network to perform lifelong graph summarization. We use the GNNs Graph-MLP and GCN, as well as an MLP baseline, to summarize the temporal graphs. We compare 1-hop and 2-hop summaries. We investigate the impact of reusing parameters from a previous snapshot by measuring backward and forward transfer as well as forgetting rate. Our extensive experiments are on two series of ten weekly snapshots, from 2012 and 2022, of a web graph with over 100M edges. They show that all networks predominantly use 1-hop information to determine the summary, even when performing 2-hop summarization. Due to the heterogeneity of web graphs, in some snapshots, the 2-hop summary produces up to ten times as many vertex classes as the 1-hop summary. When using the network trained on the last snapshot from 2012 and applying it to the first snapshot of 2022, we observe a strong drop in accuracy. We attribute this drop over the ten-year time warp to the strongly increased heterogeneity of the web graph in 2022. The source code and additional resources are available at https://github.com/jofranky/Lifelong-Graph-Summarization-with-Neural-Networks.