Upstage Wins Best Paper Award at 'WSDM 2023'

Upstage announced that it won the Best Paper Award at 'WSDM2023' held in Singapore.

Upstage announced on the 7th that it received an award at ‘WSDM2023’ held in Singapore.

WSDM is a global conference in the fields of web search and data mining, organized annually by ACM (Association for Computing Machinery, USA), and this year marked its 16th edition. It was held in Singapore from February 27 to March 3 this year. Researchers from industry and academia participated, submitting a total of 690 papers focused on topics such as recommendation systems, click-through rate prediction, and semantic search, of which only 123 papers (17.8%) were accepted.

Upstage, in collaboration with the research team of its Hong Kong branch, presented a paper titled “Efficient Utilization of Multi-step User Intent for Session-based Recommendation via Atten-Mixer Network.” This paper studies session-based recommendation, a method that predicts the next user action based on recent user behavior. It proposes modeling user intent in multiple steps and introduces a new network architecture called Atten-Mixer to improve recommendation performance.

For example, if a user searches for a wedding dress and then looks up furniture to purchase, the system understands the user’s intent related to marriage and newlywed home decoration, thereby recommending items necessary for the wedding and improving the accuracy of subsequent product recommendations.

Upstage’s research was selected for the 'Best Paper Honorable Mention Award.' This award is given only to the top 0.5% of all submitted papers, recognizing the research capabilities and achievements of the Upstage-Hong Kong research team.

Jae-beom Kim, AI Product Leader at Upstage who led this research, said, “We are very pleased that the excellence of Upstage’s recommendation technology has been recognized academically through this joint research with the Hong Kong research team, which goes beyond academic level to research on commercial models currently used in the field.” He added, “Not only with this achievement but also through various recommendation models tailored to each industry and customer service environment, we will create superior AI packs and contribute to customer success.”