Upstage Ranks 12th in RecSys 2022 AI Recommendation Competition, Highest among Domestic Companies
2022/11/2Upstage, 12th place in the ACM RecSys Challenge 2022 competition, the most prestigious society in the recommended field
Nvidia, Intel, and other global big tech companies stand shoulder to shoulder... Reconfirmation of the highest level of recommended technology with a difference of 0.016 points from 1st place
“Proven upstage AI Seargest technology will help all businesses and industries that need hyper-personalized recommendations”
(Upstage = 2022/11/2) Upstage achieved a significant milestone in the ACM RecSys Challenge 2022 competition, hosted by the prestigious Association for Computing Machinery (ACM), with an overall ranking of 12th amongst global competitors.
Upstage announced on February 2nd its impressive performance in the ACM RecSys Challenge, a highly regarded competition in the field of recommendation systems hosted by the distinguished society ACM RecSys. Upstage ranked 12th overall among global competitors, while cementing its position as the best-performing Korean company in the contest. This is yet another triumph for the Upstage team who previously won the gold medal in Kaggle Shopee recommendation contest and the silver medal in the Kaggle H&M competition. With these accomplishments, Upstage has established itself as the leading team in the AI recommendation field.
The ACM RecSys Challenge, which has been an annual event since 2007, is widely recognized as a competitive platform for cutting-edge research and innovation in the field of recommendation systems across various domains, including music, social media, and commerce. This international event brings together leading global companies and universities to showcase their expertise. Past winners of the competition include renowned tech giants such as China's Alibaba (2016), Russia's Yandex (2015), and Nvidia (2020).
This year’s event was a fashion recommendation contest utilizing actual data from Dressipi, a fashion personalization AI solution. Contestants were tasked with developing an AI product recommendation model based on customer data, including user clicks and purchases. A total of 350 teams entered the intense competition, and the results were announced at last month's conference.
The top prize went to Xiaomen, China's largest commerce company. Upstage achieved a 12th place in the overall rankings, missing the first-place spot by only 0.016 points. This outstanding performance places Upstage on par with industry giants such as Nvidia, Intel, and Huawei, while also earning the distinction of being the highest-ranking Korean company in the competition.
Upstage achieved outstanding results by using the transformer model, a neural network method that tracks relationships within sequential data, learns context and meaning, and is capable of achieving higher performance. Compared to the traditional boosting method which selects candidate items first and then ranks them, the transformer model applies a mathematical technique that continues to evolve and detects even subtle differences in the meaning of data elements based on their relationships. This allowed Upstage to understand the relationships between click histories and generate personalized product recommendations, achieving notable differentiation.
AI-powered recommendation systems are now a core technology used extensively across various industries, with their effectiveness demonstrated on platforms like YouTube and Amazon. Recommendation technology excels in situations where customer decision-making is required. As the amount of information and data available grows, the importance of recommendation technology is also increasing. Advancements in AI-based technology are making super-personalized recommendations possible, further refining the capabilities of recommendation systems.
Upstage's recent accomplishment is a testament to its technological expertise in the field of "AI Seargest" (Search and Suggest). This cutting-edge technology employs personal data and preferences to search and recommend products or content using AI. In addition to its recent success, Upstage has previously achieved notable milestones, such as being the first Korean company to earn a double-digit gold medal in Kaggle competitions and having two Kaggle Grandmasters within the top 20, which is the highest achievable grade. These achievements have cemented Upstage's position as the top AI company in Korea.
Upstage CEO Kim Seong-hoon said, “Upstage recommendation technology was once again recognized for its global top technology by winning the highest domestic ranking in a challenge competition held by ACM RecSys, the most prestigious AI recommendation society. This will help all businesses and industries that need hyper-personalization.”
Upstage provides customized AI technologies such as OCR technology that can extract and use desired information from images, recommendation technology that considers customer information and product and service characteristics, and natural language processing search technology that enables meaning-based search. The no-code-low-code solution 'Upstage AI Pack' will be released at the end of this year. Using the 'Upstage AI Pack', you can easily utilize data processing, AI modeling, and indicator management, as well as support continuous updates and conveniently use the latest AI technology at all times.
* Source photo: Upstage, a leading AI company in Korea, ranked 12th in the 'ACM RecSys Challenge 2022' contest hosted by the world's most prestigious academic society in the recommended field, and recorded the highest ranking in Korea. The picture is the ACM RecSys 2022 conference website and upstage RecSys Pack conceptual diagram
Contact
Upstage | Geun-Kyo Kim PR Director | 010-8229-7240 | keunkyo@upstage.ai
-
Founded in October 2020, Upstage identifies areas that can be innovated through AI technology among corporate problems and provides consulting including basic AI model and system construction required for this. In addition, by supporting the nurturing of AI talents within customer companies, it will help companies with successful AI transformation by improving their own AI competitiveness. Furthermore, the company plans to actively design AI teams by nurturing differentiated professional talents who can be immediately put into AI business through hands-on training that melts AI business experience and solid AI basic training.
CEO Seong-Hoon Kim developed Kachine, an early search engine in Korea, and co-founded Nara Vision, Inc. Since then, while serving as a professor at Hong Kong University of Science and Technology, he won the ACM Sigsoft Distinguished Paper Award 4 times, the best paper award, for research on bug prediction and automatic source code generation that converges software engineering and machine learning. He is a world-class scholar in software engineering and AI, winning influential thesis awards. He is also widely known as a deep learning instructor for everyone, which has recorded more than 6 million views in total.
Upstage's co-founders include CTO Lee Hal-seok, who led Naver Visual AI / OCR and achieved global results, and CSO Park Eun-jung, who led the model team of Papago, the world's best translator.
