The Evaluation of DyHATR Performance for Dynamic Heterogeneous Graphs

The Evaluation of DyHATR Performance for Dynamic Heterogeneous Graphs

dynamic; graph mining; heterogeneous; link prediction; performance evaluation

Dynamic heterogeneous graphs can represent real-world networks. Predicting linksinthesegraphsis more complicated than in static graphs. Until now, research interest of link prediction has focusedon static heterogeneous graphsor dynamically homogeneous graphs. A link prediction technique combining temporal RNN and hierarchical attentionhas recently emerged, called DyHATR. Thismethod is claimed to be able to work on dynamic heterogeneous graphs by testing themon four publicly available data sets (Twitter, Math-Overflow, Ecomm, and Alibaba). However, after further analysis, it turned out that the four data sets did not meet the criteria of dynamic heterogeneous graphs. In the presentwork, we evaluated the performance of DyHATR ondynamic heterogeneous graphs. We conducted experiments with DyHATR based on the Yelp data set representedas a dynamic heterogeneous graph consistingof homogeneous subgraphs. The results show that DyHATR can be applied to identify link prediction on dynamic heterogeneous graphs by simultaneously capturing heterogeneous information and evolutionarypatterns,and then considering them to carry out linkpredicition. Compared to the baseline method, the accuracy achieved by DyHATR is competitive, although the results can still be improved.

Nasy`an Taufiq Al Ghifari,Gusti Ayu Putri Saptawati*, Masayu Leylia Khodra & Benhard Sitohang

https://journals.itb.ac.id/index.php/jictra/article/view/20502/6314

School of Electrical Engineering and Informatics, Institut Teknologi Bandung, Jalan Ganesa 10 Bandung 40132, Indonesia

2023

Fajar bagus W

PDF

English

Text