NAVER

NAVER Search Expands Its R&D Network in the U.S. to Strengthen Global Research Capability

- NAVER presented 40+ papers at leading global conferences last year—11-times the total five years ago, reflecting NAVER’s search and AI research capabilities

- NAVER provides new search experiences to users by applying research results to AI-based search and recommendation technology

- JinYoung Kim, newly appointed Head of Data Science and Director of U.S. R&D, to recruit global talents and upgrade research network

 

2021-05-07

 

NAVER today held its ‘2021 Naver Search Colloquium Global Re: Search’ online for Naver researchers to share research results on search, AI, and other cutting-edge technology trends with academics to promote technological advancement.

 

​​

 

Over 1,200 people—double the expected number—registered in advance for the conference. Experts, undergraduate and graduate students in search and AI fields from various countries around the world including Japan, France, the United States, the United Kingdom, and Germany attended the conference. As search and AI technologies are widely used across industries, professionals in medical care, autonomous driving, education, and fashion, were also in attendance.

 

NAVER presented 40+ papers at leading global conferences last year—11-times the total five years ago

At the conference, InHo Kang, Executive Officer at NAVER Search CIC, announced NAVER’s R&D achievements and upcoming plans to strengthen its global research network. 

 

Kang introduced NAVER Search’s academic achievements that have been recognized by leading global conferences. NAVER presented14 research papers at the 2021 International Conference on Acoustics, Speech, & Signal Processing(ICASSP; seven papers at 2020 Computer Vision and Pattern Recognition(CVPR), and six papers at the Associated for Computational Linguistics(ACL) and Empirical Methods in 2020. 

 

Kang elaborated that over 40 papers were presented at top academic conferences last year, demonstrating the significant growth of NAVER’s AI research capabilities. Compared to 2016, this is 11 times the number of presentations made at selected top-tier conferences. NAVER continues its notable accomplishments through long-standing technology exchanges between its R&D organizations, such as Search CIC, CLOVA CIC, and NAVER LABS Europe.

 

NAVER provides new experiences by applying AI-based search and recommendation technology

NAVER is not only conducting academic research but also ensuring these new technologies are widely used. NAVER provides its search engine service that incorporates new technologies based on insights gained from users’ service experience and continues to advance its search technology and business competitiveness. 

 

Key examples of NAVER’s AI technology being applied to its search services are ‘AiTEMS’, an AI technology-based product recommendation system, ‘AiRSPACE’, a location recommendation system, and ‘AiRS’, a content recommendation system—all focused on providing users personalized search results.

 

In addition, NAVER is offering users a differentiated search experience with new AI-based voice and visual search engine tools. Kang shared the company’s expectations for the growth potential of AI-based search tools, saying, “The use of AI search tools has increased by more than 24% in April compared to January this year through continuous improvement in usability.”

 

Kang also revealed plans to strengthen NAVER’s global research and development efforts, saying, “In applying deep learning models represented by BERT* and Big Generative Language Model GPT3* in the entire search service, we saw performance improvements in languages other than Korean through semantic understanding, and we believe our technology, research methods, and expertise will make inroads in the global market as well.”

*BERT: A natural language processing deep learning language model in which an algorithm reads a large amount of data and learns by itself to precisely implement a method of understanding words and contexts

*GPT3: AI language model with a wide range of functions unlike existing AI such as coding, writing, and translation

 

Establishing a Silicon Valley-based R&D center to facilitate Global Talent Recruitment

Kang emphasized that the Global Artificial Intelligence Research Belt—a technology research network created by NAVER in 2019—has continued to foster collaboration between researchers from around the world. To further strengthen this network, NAVER plans to expand its technology research efforts between South Korea, Japan, France, and Vietnam. 

 

At the colloquium, NAVER also introduced JinYoung Kim, its newly appointed Head of Data Science and Director of U.S. R&D. After obtaining a Ph.D. in information search in the U.S., Kim was responsible for evaluating search and recommendation systems and user modeling at Microsoft and Snap.

 

Kim plans to strengthen NAVER’s R&D network in North America by recruiting outstanding new developers mainly within Silicon Valley (regardless of nationality). All employees will have the flexibility to work remotely. For inquiries regarding recruitment, please contact the Global Jobs site (https://recruit.navercorp.com/global/).

 

“We want to apply the latest research breakthroughs from each field of data science to NAVER’s various search and recommendation services and improve their quality,” Kim said. “We look forward to working with competent people from Korea and abroad to join our team as NAVER leaps forward beyond Korean borders as a global company.”

 

A total of 24 AI technology presentations took place during the colloquium in six categories, including Vision, Language AI, Learning to Rank, User Modeling, e-Commerce, and Platform. NAVER also shared its experience of trial and error in the process of applying various AI technologies to its service that caters to a large user base.

 

NAVER Search Colloquium is an academic conference launched in 2016 to share research results on search and AI, as well as current technology trends with academics. The conference took place under the name ‘AI Colloquium’ for the last three years and was changed to ‘Search Colloquium’ beginning this year.

 

-END-