Deep learning-based GIM model and object ID matching method and system

• Main Authors: TONG CHONG, HUA XIA, ZHAN RUOPEI, MAI JINWEN, WU YUETAO, HONG YI, XIE ZHIMIN, WANG WEILIANG
• Format: Patent
• Language: Chinese; English
• Published: 24.05.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention discloses a deep learning-based GIM model and object ID matching method and system. The method comprises the following steps: preprocessing text data of a real object ID and a GIM equipment model, and encoding the text data into character vectors by using a long short-term memory network model; weighting the character vector by adopting an attention model to obtain a context vector, and forming an adaptive weight parameter of the character vector according to the context vector; and carrying out cosine similarity calculation on the corresponding character vectors of the GIM equipment model and the physical ID by utilizing a fuzzy matching algorithm based on cosine similarity in combination with the adaptive weight parameters, matching the corresponding items of the GIM equipment model and the physical ID according to the similarity, and optimizing the model parameters and the fuzzy matching algorithm. According to the scheme, the matching efficiency and compatibility of the GIM model and the phy