Multi-lingual multi-partite product title matching

• Main Authors: TAY, Huan Lin, TAY, Wei Jie, LAUW, Hady Wirawan
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2023
• Subjects: Multi-lingual similarity; Multi-partite matching; Databases and Information Systems
• Online Access: https://ink.library.smu.edu.sg/sis_research/8308; https://ink.library.smu.edu.sg/context/sis_research/article/9311/viewcontent/webconf23.pdf
• Institution: Singapore Management University

In a globalized marketplace, one could access products or services from almost anywhere. However, resolving which product in one language corresponds to another product in a different language remains an under-explored problem. We explore this from two perspectives. First, given two products of different languages, how to assess their similarity that could signal a potential match. Second, given products from various languages, how to arrive at a multi-partite clustering that respects cardinality constraints efficiently. We describe algorithms for each perspective and integrate them into a promising solution validated on real-world datasets.