Paper Type
Short
Paper Number
1563
Description
Previous studies have proposed various methods for recommending academic patents to companies. However, these methods predominantly focus on analyzing the technology information of companies, neglecting to integrate alliance technology information crucial for future development. Consequently, previous methods fail to adequately predict the potential preferences of alliance companies. To overcome this challenge, we introduce a novel deep patent recommendation method capable of modeling the sequential patents and alliance technology influence context of the companies. Specifically, we employ bidirectional long short-term memory (Bi-LSTM) to model the sequential historical patents of the companies. Additionally, we utilize a graph convolutional network (GCN) to capture the alliance technology influence context of the companies by modeling the interaction of technology topics among alliance companies over the past year. Experimental results based on a real dataset demonstrate that our proposed method outperforms benchmark methods.
Recommended Citation
Liu, Zhaobin; Zhu, Peihu; Zeng, Jicheng; Shi, Zijing; and Ma, Jian, "A Deep Patent Recommendation Method Based on Sequential Patents and Alliance Technical Influence Modeling" (2024). PACIS 2024 Proceedings. 4.
https://aisel.aisnet.org/pacis2024/track04_dessci/track04_dessci/4
A Deep Patent Recommendation Method Based on Sequential Patents and Alliance Technical Influence Modeling
Previous studies have proposed various methods for recommending academic patents to companies. However, these methods predominantly focus on analyzing the technology information of companies, neglecting to integrate alliance technology information crucial for future development. Consequently, previous methods fail to adequately predict the potential preferences of alliance companies. To overcome this challenge, we introduce a novel deep patent recommendation method capable of modeling the sequential patents and alliance technology influence context of the companies. Specifically, we employ bidirectional long short-term memory (Bi-LSTM) to model the sequential historical patents of the companies. Additionally, we utilize a graph convolutional network (GCN) to capture the alliance technology influence context of the companies by modeling the interaction of technology topics among alliance companies over the past year. Experimental results based on a real dataset demonstrate that our proposed method outperforms benchmark methods.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
Design