-
1.
公开(公告)号:US20210217109A1
公开(公告)日:2021-07-15
申请号:US17211992
申请日:2021-03-25
Inventor: Mingyang DAI , Yixuan SHI , Zixiang LIU , Shengwen YANG
Abstract: The present disclosure provides a method of constructing a fused relationship network. The method includes: obtaining interaction data from a plurality of data sources, wherein the interaction data contains a plurality of user relationship information; generating, based on the identification information for each user in each user relationship information, a node for the user in the fused relationship network, and generating, based on the interaction information between two users in each user relationship information, an edge between the nodes for the two users in the fused relationship network, wherein a same user identification is generated as one node. The present disclosure further provides an apparatus of constructing a fused relationship network, an electronic device, and a storage medium.
-
公开(公告)号:US20240362493A1
公开(公告)日:2024-10-31
申请号:US18770122
申请日:2024-07-11
Inventor: Yixuan SHI , Wei LI , Jiachen LIU , Xinyan XIAO
IPC: G06N3/092 , G06N3/045 , G06N3/0475 , G06N3/09 , G06T11/00
CPC classification number: G06N3/092 , G06N3/045 , G06N3/0475 , G06N3/09 , G06T11/00
Abstract: A method is provided that includes: obtaining a first Text-to-Image model and a pre-trained reward model, wherein the first Text-to-Image model is used to generate a corresponding image based on input text, and the pre-trained reward model is used to score a data pair composed of the input text and the corresponding generated image; and adjusting the parameters of the first Text-to-Image model based on the pre-trained reward model and a reinforcement learning policy to obtain a second Text-to-Image model.
-
公开(公告)号:US20210224269A1
公开(公告)日:2021-07-22
申请号:US17207872
申请日:2021-03-22
Inventor: Mingyang DAI , Yixuan SHI , Zixiang LIU , Shengwen YANG
IPC: G06F16/2458 , G06F21/62
Abstract: The present disclosure provides a method and an apparatus of recommending information based on a fused relationship network. The method includes: determining association relationships between an any node and other nodes in the fused relationship network based on at least one of a weight, interaction information and data source information of the interaction information of an edge; and recommending information to a user represented by the any node based on the association relationships. The present disclosure further provides an apparatus of recommending information based on a fused relationship network, and an electronic device and a storage medium.
-
-