I am a master major in computer science and graduated from Sun Yat-sen University, and currently working at Baidu with a focus on graph neural networks and natural language processing.
I have worked on many cool projects at Baidu, and have had the opportunity to use my expertise in graph neural networks to develop cutting-edge solutions. Here are a few highlights:
- Created PGL, a graph neural network framework based on PaddlePaddle. With PGL, you can easily build and train graph neural networks. Check out the sample code below:
import pgl
g = pgl.Graph(num_nodes=num_nodes, edges=edges, node_feat={"h": feature})
g.tensor()
def send_func(src_feat, dst_feat, edge_feat):
return src_feat
def recv_func(msg):
return msg.reduce_sum(msg["h"])
msg = g.send(send_func, src_feat=g.node_feat)
ret = g.recv(recv_func, msg)-
With PGL, we developed TransformerConv, a novel graph neural network architecture that has achieved top performance on multiple OGB benchmarks, as well as winning first place at KDD Cup 2021 and NeurIPS 2022 OGB-LSC.
-
Our teams created two powerful tools, Graph4Rec and PGLBox (Github links), for running recommender systems with graph neural networks.
-
We applied graph neural networks to a variety of applications, including geographic location search, ETA prediction, and cross-modal document understanding.
If you are interested in collaborating with me or interning at Baidu, feel free to send me an email at huangzhengjie@baidu.com.