Professor Key-Sun Choi received a 2018 KAIST International Cooperation Award for his efforts in international cooperation. Prof. Choi is a Professor at the KAIST School of Computing. He served as the President of KSCS (Korean Society for Cognitive Science) in 2006, the President of AFNLP (Asian Federation of Natural Language Processing) between 2009 and 2010 and the Director of AAMT (Asia-Pacific Association for Machine Translation) from 2000 to the present. Major ongoing projects of Prof. Choi’s current research interests include “Exobrain WiseKB” which focuses on the construction of a wise knowledge base and question-answering platform, “Bio & Medical Informatics, PACS,” which focuses on the knowledge acquisition in the biomedical domain, “Eurostar” (Korean-Europe joint research project), which focuses on question-answering on mobile devices, and “VTT,” which focuses on edmotional intelligence.
Professor Dongsoo Han received a 2018 KAIST Innovation Award for his innovations in industry through technology transfer and commercialization and for sharing and spreading of research achievements. Prof. Han is a Professor at the KAIST School of Computing. His group, in collaboration with KTNET, built an indoor positioning and navigation system for COEX in 2010 and a KAIST Indoor Localization System (KAILOS) to realize an indoor GPS system in 2014. In 2011, Dongsoo Han founded the indoor positioning research center (IPRC) at KAIST. He is serving as director for the IPRC. In 2013, Dongsoo Han published the book “Infinite Challenge for Patents.” Dongsoo Han’s current research focus is on providing indoor positioning and navigation service for public buildings. Dongsoo Han is a member of IEEE and the IEEE Computer Society, and a board member of the ITS Korea Society.
Professor In-Young Ko received the 2018 KAIST Excellence in Teaching Award for his teaching efforts. Prof. Ko is an Associate Professor at the KAIST School of Computing. Prof. Ko received his Ph.D. in Computer Science from the University of Southern California (USC) in 2003. During 2005-2011, Prof. Ko was appointed as adjunct faculty for the Master of Software Engineering (MSE) program in the School of Computer Science at Carnegie Mellon University. Before joining KAIST, Prof. Ko worked as a postdoctoral research associate in the Distributed Scalable Systems Division at the USC Information Sciences Institute (ISI). His research interests span software engineering and Web engineering. He is currently conducting research on service engineering, service-oriented architecture, Web services, software reuse, service composition/mashups, Semantic Web, Web-based information management, and end-user software engineering.
Prof. Sue Bok Moon was appointed as a regular member of the National Academy Engineering of Korea on January 3, 2018. An article in Korean is available at the following: http://www.etnews.com/20180103000240
Three School of Computing faculty members won the 2018 Faculty Award in Commemoration of the 47th foundation anniversary of KAIST: International Cooperation Award to Prof. Key-Sun Choi Innovation Award to Prof. Dong Soo Han and Excellent Lecture Award to Prof. In-Young Ko. The International Cooperation Award is given to a faculty member who has contributed largely to the university’s international cooperation, the Innovation Award to a faculty member who has contributed to innovations to industry through technology transfer and commercialization and to sharing and spreading of research achievements, and the Excellent Lecture Award to a faculty member who had the highest course evaluation results for the previous year. The award ceremony was held on Jan 12th 2018 at the 47th School Anniversary Event.
Professor Eunho Yang was awarded a 2018 Songam Future Scholar Research Excellence Award. The Songam Future Scholar Research Excellence Award is granted according to the will of President Lee of Korea Former Co. and is given to Assistant Professors in Engineering Schools with less than five years in that position with an excellent research performance in order to encourage their research activity. Prof. Yang is an Assistant Professor at the KAIST School of Computing. Prior to joining KAIST in the summer 2016, he spent two years as a research staff member at IBM T.J. Watson Research Center. He obtained his Ph.D. in 2014 from the University of Texas at Austin, and earned M.S. and B.S degrees from Seoul National University in Korea in 2006 and 2004, respectively. From 2006 to 2009, he was a researcher at Korea Telecom as the alternative military service in Korea. His research interests are in statistical machine learning in general with a special focus on high-dimensional statistics.
Prof. Jinah Park received an appreciation plaque from the Dean of College of Engineering for her devotion to KAIST Breakthroughs, a biannual research newsletter of the College of Engineering.
Prof. Insik Shin’s research team developed technology that can reduce the power consumption of smartphone web browsers down to 40％. It is expected to boost the battery time by reducing web browsing power consumption, which is a significant part of smartphone power consumption. Prof. Shin’s team said on etnews on Jan 8th that they have developed a low power mobile browsing technology by optimizing the frame rates of web pages. “Low power mobile browsing allows us to increase smartphone power performance with software only. It will help smartphone manufacturers and web browser developers who are trying to optimize power consumption,” said Prof. Shin in an interview. Etnews article in Korean is available at the following: http://www.etnews.com/20180105000383
Professor Sue Bok Moon was appointed as a regular member of the National Academy Engineering of Korea on January 3, 2018. The members of the National Academy Engineering of Korea are selected from universities, research institutions and industry based on their contribution to the country through advancement in science and engineering. The selection process, which spans four stages (search, nomination, review, election), considers not only the nominee’s academic contributions, but also the IPs, education output, and industry contributions.
Team ‘NLPCL ＊ CGV’ attended the ‘SW/Demo Contest’ and won the best demo award during the Korea Software Congress 2017 held in Busan on December 20th~22nd. Team ‘NLPCL ＊ CGV’, comprised of Ph.D students Jung-Ho Kim, Wonsuk Yang and Fitsum Gaim of NLPCL Lab (advisor: Jong Cheol Park) and Master’s student Youngjin Ahn of CGV Lab (advisor: Jinah Park), has demonstrated a ‘VR training and mobile feedback system to increase communication ability in work environment’ by merging VR and AI technology.
Ph.D student Jooyeon Kim (advisor: Alice Oh) develops algorithm to minimize fake news spread by leveraging the crowd The algorithm to minimize fake news spread researched by Ph.D student Jooyeon Kim of Prof. Alice Oh’s lab, together with Dr. Manuel Gomez-Rodriguez of Max Plank Institute, has been published in articles of Yonhapnews, Maeil Business News, and ETnews. This research is the first work on modeling techniques for crowd-powered reduction of fake news spread, whose concern have been growing recently. The work is to be presented in the ACM WSDM on February 2018. http://www.yonhapnews.co.kr/bulletin/2017/12/22/0200000000AKR20171222067600033.HTML http://news.mk.co.kr/newsRead.php?year=2017&no=849666 http://www.etnews.com/20171224000028
KAIST team has newly suggested the unit testing of C programs that can find errors automatically while minimizing false alarms. Prof. Moonzoo Kim and Dr. Yoon Ho Kim wrote a paper "Precise Concolic Unit Testing of C Programs with Alarm Filtering Using Symbolic Calling Contexts" in collaboration with Prof. Yoon Ja Choi at Kyung-Buk University. Their paper has been accepted to ACM/IEEE ICSE (Intl. Conf. on Software Engineering) which is one of the most prestigious conferences in the Software Engineering field. Congratulations！
The 9th Asian Conference on Machine Learning, where Prof. KeeEung Kim served as the General Chair, took place in Yonsei University from November 15th to 17th. In the conference experts and researchers shared results and ideas on machine learning, and discussed proposals centered around innovative machine learning ideas and paradigms. Since 2009, ACML has been a gathering of new ideas and technology about machine learning, which has taken important roles in various fields in the industry. ACML has started in Asian region and has been held in New Zealand, Hong Kong, Australia, Singapore, Taiwan, Japan and China. The conference has helped the scientists gain a broader understanding on machine learning and get insights for applying them in the industry.
On December 12th, KAIST undergraduate student Jaemin Shin and Dukgi Hong (Advisor Prof. Sue Moon) has earned the ‘Best Paper Award’ for their paper ‘Considerations on Deploying High-Performance Container-based NFV’ in the international conference ‘Cloud-Assisted networking Workshop at ACM CoNEXT 2017’ held at Incheon Paradice City Hotel. Congratulations！
Research team “Is this real story” lead by Prof. Meeyoung Cha was awarded the President of Artificial Intelligence Industry Association Award for their AI technology for identifying fake news by finishing first place in the university division of ‘AI R&D Challenge’ held in Gyeonggi Creative Economy Center. The ‘2017 AI R&D Challenge Competition’ aims to find problems that have high economic and social influence but are considered too challenging, and solve them using the AI technology. The first topic is this year’s identification of fake news. To identify fake news with coherence and inconsistency of news content two tasks were to be solved, and the competition had 6 million Won prize and 18 year R&D support totaling 1 billion Won. 71 teams of individuals, universities and companies participated in the qualifiers and final rounds. The top 3 teams were awarded the Minister of Science and ICT Award and the President of Institute for Information &communications Technology Promotions Award, with special awards of President of Artificial Intelligence Industry Association Award and Intel Award. Prof. Cha’s team (with PhD Student Keonwoo Park, Masters student Jeongmin Kim, Undergraduate student Taekyoon Kim and 3 students from Graduate School of Culture Technology) formed industrial-educational cooperation with Withcat and Jamong Lab, and developed a recurrent neural network that uses paragraph level embedding vectors to classify the coherence of information. The research team conducted the research with the support of KAIST 2017 AI Flagship project.
During his visit in Asiacrypt 2017 that was held in Hong Kong from December 3rd to 7th, Prof. Kwangjo Kim got the approval to hold the Asiacrypt 2020 in Korea after giving his proposal in the Coordination Committee meeting held in December 4th. Asiacrypt Coordination Committee decides the host country of Asiacrypt that takes place every December. Currently there are up to 3 representatives from each of the 13 countries, including Korea, Japan, China, Taiwan, Hong Kong, Australia, New Zealand, Malaysia, Singapore, Vietnam, UAE and Pakistan. The representatives from Pakistan could not attend this day’s meeting. Korea, Singapore and UAE gave their hosting proposal, and as the result of the voting procdure, Korea got 6 votes and UAE and Singapore got 3 votes each, which didn’t count as majority for Korea. In an additional meeting without the three running countries, the committee chair, Prof. Lai of Shanhai Jiao Tong University, chose Korea as the hosting country. The hosting of Asiacrypt2020 in Korea will be finalized once the IACR Board confirms during Eurocrypt held in Israel next April. Thus Asiacyrpt 2020 is to be held in Korea (Daejeon). The conference is expected to be attended by renowned cryptographers, and that research in cryptography in Korea will mature once more. Prof. Kwangjo Kim is currently IACR Fellow and will serve as the committee chair of Asiacrypt 2020.
Prof. Min H. Kim’s research project, compact hyperspectral imaging, is in press of EurekAlert Science News and Photonics Media. The traditional hyperspectral imaging technology has wide reach and is being applied in fields such as military combat, astronomy, agriculture, biomedical imaging, and geoscience. Scientists, for instance, rely on hyperspectral imaging to observe and analyze materials for mining and geology, or for various applications in the medical field. However, hyperspectral imaging systems are expensive—ranging from ＄25,000 to ＄100,000—and require complex specialized hardware to operate. Prof. Min H. Kim’s team of computer scientists from KAIST, South Korea, and Universidad de Zaragoza, Spain, has devised a way for low-cost accurate hyperspectral imaging, ridding of expensive equipment and complex coding. This novel, compact single-shot hyperspectral imaging method captures images using a conventional DSLR camera equipped with just an ordinary refractive prism placed in front of the lens. The new method was tested on a variety of natural scenes, and the results, according to the researchers, compared well with current state-of-the-art hyperspectral imaging systems, achieving quality images without compromising accuracy. “These hyperspectral imaging systems are generally built for specific purposes such as aerial remote sensing, or military applications, and as such they are not affordable nor practical for ordinary users,” said Min H. Kim, associate professor of computer science at KAIST and a lead author of the study. “Our system requires no advanced skills, and we are able to obtain hyperspectral images at virtually full resolution while making hyperspectral imaging practical.” Prof. Kim’s collaborators include Diego Gutierrez, associate professor at Universidad de Zaragoza; Seung-Hwan Baek, computer science PhD student at KAIST; and Incheol Kim, MS student at KAIST in Min H. Kim’s lab. The team presented their new method at SIGGRAPH Asia 2017 and also published it in the top computer graphics journal, ACM Transactions on Graphics (TOG). This annual conference and exhibition showcases the world’s leading professionals, academics and creative minds at the forefront of computer graphics and interactive techniques. EurekAlert Science News: https://www.eurekalert.org/pub_releases/2017-11/afcm-chi120417.php Photonics Media: https://www.photonics.com/Article.aspx?AID=62875 Technology Breaking News: https://www.technologybreakingnews.com/2017/12/compact-hyperspectral-imaging-at-low-cost/ R&D Magazine: https://www.rdmag.com/news/2017/12/compact-hyperspectral-imaging-low-cost
Professors, Insik Shin, Jaehyuk Huh and Min H. Kim, won MSRA Collaborative Research 2018 Grant Awards Microsoft Research Asia (MSRA) Collaborative Research Program is a grant program for faculty members in Asia, working together with MSRA researchers on mutually interesting topics. MSRA aspires to foster talents and establish Microsoft as a valuable research and technology partner for higher education in Asia. In addition, KAIST is the university that won the largest number of grant awards, along with Tsinghua, HKUST, and USTC. The selected research projects are: TouchSurface: Supporting Touch User Interface with Sound on Any Surface Insik Shin, KAIST Securing Key-Value Stores with Hardware Trusted Execution Environments Jaehyuk Huh, KAIST Capturing Intrinsic Material Appearance via Spectro-Polarimetric 3D Imaging Min H. Kim, KAIST Since the announcement of Call for Proposals in mid-July, there has been a large number of submissions. MSRA Collaborative Research Program Committee carefully reviewed them based on the state-of-the-art and common research interests. Now MSRA announces 62 awards. The awardees are from 35 universities and institutes in China mainland, Hong Kong, Taiwan, Japan, Korea and Singapore. We are pleased to announce that three professors of KAIST School of Computing, Insik Shin, Jaehyuk Huh and Min H. Kim, won MSRA Collaborative Research 2018 Grant Awards. Reference: http://www.microsoft.com/en-us/research/lab/microsoft-research-asia/articles/msra-collaborative-research-2018-award-announcement/
Two papers from the KAIST School of Computing got accepted for ACM POPL 2018, a top international conference on programming languages. The first paper is “On Automatically Proving the Correctness of math.h Implementations“ authored by Mr Wonyeol Lee in the school and his colleagues in Microsoft and Stanford (Sharma and Aiken). The second paper is “Denotational Validation of Higher-Order Bayesian Inference“ written by Prof Hongseok Yang in the school and his colleagues in Oxford, Cambridge, Tubingen and Edinburgh (Scibior, Kammar, Vakar, Staton, Cai, Ostermann, Moss, Heunen, Ghahramani).
From May 29 to June 1, Professor Junehwa Song and Professor Wang-Chien Lee of Pennsylvania State University co-chaired the IEEE International Conference on Mobile Data Management (MDM), hosted by KAIST. Celebrating its 18th year, MDM is an academic conference that focuses on data management in mobile, ubiquitous and pervasive environment. MDM deals with topics addressed by both academia and industry such as IoT, autonomous vehicles, crowdsourcing and mobile sensing. A total of 50 papers were presented by 105 researchers from 13 countries. In addition to domestic and international researchers, companies such as Google, IBM, and Delta Electronics participated in discussions about the potentials of mobile data in the future. The conference was held with the support of IEEE, KAIST School of Computing, IEEE Technical Committee on Data Engineering(TCDE), University of Pittsburgh, Daejeon International Marketing Enterprise, Software Research Center of Chungnam National University and Daejeon Metropolitan City.
Seung-Hwan Baek and Chang Hyun Park won MSRA PhD Fellowship 2017 [link]. Congratulations to Seung-Hwan Baek (advisor: Prof. Min H. Kim) and Chang Hyun Park (advisor: Prof. Jaehyuk Huh), on winning the 2017 Microsoft Research Asia Fellowship Award！ Their excellent guidance and continued support for the Microsoft Research Asia Fellowship program are highly appreciated. This year, 107 distinguished Ph.D. candidates from 38 leading research universities or institutions were nominated for fellowships. Applicants included candidates from the China mainland, Hong Kong, Taiwan, Japan, South Korea, and Singapore. Each candidate’s credentials, publications, and research projects were thoroughly evaluated by a review committee and only 10 extremely outstanding students have been awarded fellowships. A cash award is also provided to support the fellow’s research and academic endeavors. Moreover, Microsoft Research Asia would like to provide Moreove the PhD student a 3-month visiting research opportunity, which helps the winner to broaden his or her horizons by doing research at a top international research institute under the guidance of an experienced researcher. Seung-Hwan and Chang Hyun were invited to attend the 2017 MSRA Fellowship award ceremony held in Harbin, China, on Oct 19th. The ceremony was one of the key sessions of the 19th Computing in the 21st Century Conference, during which they could talk from distinguished scientists from around the world, and have a lunch meeting with Turing Award winner. Since its inception in 1999, the program has attracted applications from approximately 1,100 PhD candidates from more than 40 universities in the Asia-Pacific region, with 396 outstanding students having been awarded Microsoft Research Asia Fellowships. Many of them have become influential researchers in their area. We are glad to have Seung-Hwan and Chang Hyun as Fellowship winners this year, and we look forward to their future achievements！ Chang Hyun Park (left) and Seung-Hwan Baek (right)
On September 19th, all three projects proposed by KAIST Computer School of Computing were accepted in the Next Generation Information Computing Technology Development Project, which was announced by the Ministry of Science and ICT and National Research Foundation of Korea. The project’s mission is to "support development of basic SW technology that differentiates itself from existing R&D in IT, in order to secure long-term national competitiveness" The detailed projects assigned to our department are as follows. - SuggestBot: Context-based smart interaction Research Goal: Development of software technology including big data of interaction, engine for situation inference and action suggestion, wearable UI and computing subsystem to realize in-situ, intelligent, and in-time smart interaction in wearable computing environment. Project manager: Geehyuk Lee Participants from School of Computing: Juho Kim, Sung Hyon Myaeng, Alice Oh, Junehwa Song Participants from other departments: Uichin Lee -Development of reliable intelligent CPS complex system and on-the-fly verification technology Research Goal: Development of software engineering technology for reliable and safe software development in Cyber-Physical System (CPS) environments such as Vehicle to Everything (V2X) and Smart Factory Project manager: In-Young Ko Participants from School of Computing: DooHwan Bae, Sungwon Kang, Junehwa Song, Jongmoon Baik Participants from other institutions: Jang Ui Hong(Chungbuk Univ.), Junbeom Yoo(Konkuk Univ.), Seung-Woo Kang(KoreaTech) -Multilanguage Verification and Debugging of Full Stack SW through Intelligent Automation Project manager: Moonzoo Kim Participants from School of Computing: Sukyoung Ryu, Shin Yoo Participants from other institutions: Yunja Choi(Kyungpook Nat. Univ.), Kyungmin Bae(POSTECH), Shin Hong(Handong Univ.) The project runs from September 1, 2017 to December 31, 2020 (3 years and 4 months), with funding total of 9.53 billion won.
On September 12th, KAIST PhD candidate Sangeun Oh was awarded the 2017 Google PhD Fellowship in Mobile Computing area, in recognition of his latest work "Designing Multi-device Mobile Platform for Cross-device Functionality Sharing". Google has been conducting Google PhD Fellowship Program to spotlight PhD candidates who are doing outstanding work in the field of computer science. The awardees are granted ten thousand dollars worth of scholarship and become connected with Google's specialists each field as mentors to discuss the research and get feedback. This year, a total of 47 PhD students worldwide were awarded, including 6 students in the East Asia, including Korea, China and Japan. Throughout his PhD years, Sangeun Oh developed platforms for mobile devices to share functionalities such as log in, payment and sensors, enabling user experiences previously impractical to implement. His research was presented last June 21th in ACM Mobysis, the most prestigious academic conference in Mobile computing.
KAIST Convergence AMP program (KCAMP, supervisor Prof. Young Whan Kim) held the 15th entrance ceremony on September 4th. The reason CEOs continue to study is to be responsive to new trends and to have wider human networks. If so, a CEO program could be evaluated according to the contents of the curriculum and the composition of the study group. In this regard, the KCAMP is a prestigious program with a curriculum that is tailored to the tastes of corporate executives. The KCAMP program aims to be a distinguished CEO course by integrating KAIST’s capabilities as a global research-oriented university with the fields of management and humanities. It is known to have Korea’s best faculties, differentiated curriculum, diverse students, wide network with alumni, detailed learning support service, active support from KAIST and geographical advantages of Gangnam. The convergence of industries and the transition into a hyper-connected society made by novel technologies such as artificial intelligence, IoT, Big Data and 5G inform us that the fourth industrial revolution is emerging. The KCAMP course aims to educate CEOs who have to prepare for the era of the 4th industrial revolution and to foresee the big trends of industry and society. The KCAMP course aims to train convergence type leaders in preparation for the fourth industrial revolution era. The students were well suited for the convergence program. Among the 40 students were 15 representatives/officers of mid and small sized companies, 8 officers of large size companies, 6 from government and public institutions and 5 from finance, with some people from the media, lawyers and other professionals. The entrance ceremony was attended by more than 50 senior students who have completed the program. They welcomed and socialized with the new students, and organized monthly gathering of all students and alumni to form networks. Professor Kim, the manager of the program, emphasized that KCAMP is striving to become a platform for industry-university cooperation through his speech in the ceremony. He proposed specific directions for development such as linking KAIST students and companies for recruitment, conducting joint research and projects between KAIST and companies and training human resources. KCAMP is a 19-week long program. Upon completion of the program in next January, students of the 15th cycle will be able to participate in various projects held in KAIST and utilize advanced research facilities as KAIST alumni.
Prof. Sung-Ju Lee's SCAN and Prof. Alice Oh's Eliph system were selected to be presented in KAIST Breakthroughs. Prof. Sung-Ju Lee's SCAN system (http://breakthroughs.kaist.ac.kr/?post_no=913) Prof. Alice Oh's Eliph system (http://breakthroughs.kaist.ac.kr/?post_no=908) Congratulations on both teams！
KAIST SoC Prof. Hongseok Yang presented a keynote at QONFEST’17 as a joint invited speaker for CONCUR, QUEST, and FORMATS. Prof. Hongseok Yang at the KAIST School of Computing presented a keynote talk on probabilistic programming at QONFEST’17, as a joint invited speaker for CONCUR, QUEST, and FORMATS. QONFEST’17 is the umbrella event of the four major international conferences CONCUR, QEST, FORMATS, and EPEW, whose topics jointly cover theory, formal modeling, verification, performance evaluation and engineering of concurrent, timed and other systems. QONFEST '17 was held in Berlin, Germany from September 4, 2017, until September 9, 2017.
IEEE/ACM ASE 2017 (Automated Software Engineering), a top international conference on software engineering, accepted a paper from School of Computing’s Graduate School of Information Security. Professor Sang Kil Cha, alongside Soomin Kim and KAIST Cyber Security Research Center, worked on the paper, “Testing Intermediate Representations for Binary Analysis”. The research was a part of Prof. Cha’s project with the Ministry of Science and ICT, a R&D project carried out since 2016 in cooperation with KAIST Cyber Security Research Center. The project is on comparing the expressiveness of existing intermediate representations used for binary analysis. The paper will be presented at IEEE/ACM ASE 2017 conference, held Oct. 30 to Nov. 3 at Illinois, US. Reference: http://ase2017.org
Our very own School of Computing alumnus Jeongseop Ahn, Ph.D (class of 2015, advisor: Prof. Jaehyuk Huh), was appointed an Assistant Professor at Ajou University’s Department of Software. Dr. Ahn worked at a postdoc position in University of Michigan after graduating, and afterwards found a research job at Oracle Labs in Silicon Valley. For his Ph.D, Dr. Ahn worked on fundamental cloud computing technologies such as virtualization and convergence in computer architecture technologies. At Oracle Labs, he worked on systems for big data analysis platforms, both SW and HW. The resultant research found themselves published on top architecture conferences, including ISCA and MICRO, where Dr. Ahn published multiple papers. Dr. Jeongseop Ahn’s achievements are available on his website, at: https://jeongseob.github.io/
ㅁ KAIST Professors Dongman Lee, Wonjae Lee, Juyong Park, Meeyoung Cha, and their respective research teams unveiled their “Placeness Data Mining and Inference”, a part of their research on the development of core technologies for placeness based data mining for use in real time intelligent information suggestion services in the context of smart spaces. ㅇ The API is available at placeness.kaist.ac.kr:8080/, with the wiki on relevant information available at placeness.kaist.ac.kr/wiki ㅇ The development of core technologies for placeness based data mining for use in real time intelligent information suggestion services in the context of smart space was a government funded R&D effort, running from 2015 July to 2017 Aug, supported by the Ministry of Science and ICT (former Ministry of Science, ICT, and Future Planning). ㅇ This research analyzes the vast amounts of geotagged multimedia and text data generated by online social networking services to garner the purpose and social context of specific commercial locations’ visitors, and through this data infer the social actions, emotion, and the relationship the user has with the location. ㅇ The recently unveiled API relies on data collected and analyzed from the current major R&D objective locations (e.g. COEX, IPark Mall) and their constituent locations. Social multimedia data collected from the locations are analyzed to provide locational context from various perspectives (passers-by, visitor, time, emotion, etc). A REST API allows access via HTTP requests for the general public, providing the placeness information of the major spaces. ㅁ The above API provides 4 different inferences arising from the process of placeness inference: 1) placeness of the location of which this place is a constituent of, 2) placeness of the location within the context of the aforementioned parent location, 3) emotion based inference on the location and its parent location’s ambience, and 4) the user-location relationship. ㅇ The 4 inferences include the following information: - Fundamental technology for real time placeness matching via the analysis of users’ social context within the locality - Fundamental technology for the inference of granular placeness inference via the application of standardized data structures - Fundamental technology for the inference of placeness for reflecting the user’s diachronic / immediate context, thereby presenting a multidimensional relationship connectivity graph between localities and users - Fundamental technology for emotion inference for real time location suggestion and user specific locality emotion score output based on the inference ㅇ The information provided enables developers to improve the accuracy and satisfaction of location suggestion services, as well as mine social data for social contexts and the resultant public visitation patterns and changes in preferences, with which specific locations can be suggested, and advertising and coupons created for the purpose of promoting the consumption of specific contents within the locality. ㅁ Prof. Dongman Lee (Project Lead) claims the API developed during the research effort improves the quality of existing geography based location search and suggestion services, automatically providing changing location suggestions based on urban visitors’ changing location visit trends. The research is expected to form the core of technologies to overcome the limitations in the previous non-standardized text data analysis, harnessing both image and text simultaneously to infer the social information of the location, leading to a leap in existing location based suggestion services and AI based personal assistant services. ㅁ Placeness data mining and inference technology is applicable in various location based information search and suggestion service providers and location based social commerce services, allowing information search companies and mobile coupon providers to offer intelligent information suggestion functionalities and their improvement.
From July 24 to 28, KAIST’s School of Computing organized and hosted the 14th International Conference on Computability and Complexity in Analysis (CCA2017) and Workshop on Real Verification: http://complexity.kaist.edu/CCA2017 48 leading senior experts, rising young scientists, and eager students have attended and contributed to this unique event. They came from all over the world: Amsterdam, Birmingham, Brussels, Buenos Aires, Bulgaria, Cambridge, Connecticut, Cornell, Darmstadt, Hagen, Kyoto, Maastricht, Munich, Nagoya, Okinawa, Pohang, Pretoria, Saarbrücken, Seoul, Singapore, Tokyo, Trier, and Versailles. Over the course of five days we had 33 presentations of recent research, discussions of current challenges, and explorations of future developments in the Algorithmic Foundations of Numerics. The organizers gratefully acknowledge generous financial support from KAIST's School of Computing, from the International Relations Team (IRT), and from the National Research Foundation of Korea (NRF). Co-located with CCA 2017, a Workshop on Real Verification was organized by Prof. Gyesik Lee (Hankyong National University) and Prof. Martin Ziegler (KAIST) with invited speakers from Yonsei University, INRIA, Aston University, Trier University, KAIST, SNU, and AdaCore. For many participants this was the first time to visit Korea: They have particularly enjoyed the impressive experience of this technologically advanced country as well as of its warm hospitality, culinary richness, and cultural heritage. In fact, as part of a joint NRF/EU H2020 project, five conference participants continue staying at KAIST throughout August for further collaborative research
(Pictured in the middle: School of Computing student Hanbyul Lee) Aug 16 Segye Daily, “KAIST produces Korea’s first 3 ‘Master Degrees of Startup’” Article in Korean: http://www.segye.com/newsView/20170814000382 Linked is an article on SoC student Hangyul Lee (advisor: Prof. DooHwan Bae), who received his Master’s Degree of Startup.
Pictured: KAIST personnel involved in the SCAN technology research effort. From the left, KAIST SoC Professors Sung-Ju Lee, Dongman Lee, and Juho Kim, Ph.D. student Cheonjong Park, and Samsung Electronics SW Center researcher Junsung Lim Our very own School of Computing Professors Sung-Ju Lee and Dongman Lee, with their respective research teams, developed technology to automatically detect the current user’s situation and adjust smartphone notification settings accordingly. The technology, which they call SCAN (Social Context-Aware smartphone Notification system, or Freedom from Notifications), was featured in an Aug. 7 iPnomics article. http://www.ipnomics.co.kr/?p=63868
KAIST School of Computing Associate Professor Jinah Park presented the keynote at the 21st Medical Image Understanding and Analysis Conference (MIUA 2017) last July 11 through 13, held at John McIntyre Centre, Pollock Halls, Edinburgh, UK. The presentation was on “model-based approach to 3D shape recovery and analysis.” MIUA 2017 is a medical image analysis forum for experts in the field held every year in the UK, boasting attendees from various European countries as well as the US, Australia, and Asia. This year, of the 150 organizations attending the conference, 46 were not from the UK, and out of the 105 papers and 22 clinical abstracts submitted, 82 were from overseas. KAIST SoC Prof. Jinah Park was invited as a keynote presenter alongside Prof. Ingela Nyström (Uppsala University) and Prof. Daniel Rueckert (Imperial College London). The conference also invited Sir Michael Brady, a Professor at the University of Oxford, as the Honorary Guest Speaker. Prof. Jinah Park’s lecture was on her 3 dimensional modeling technique for extracting clinical understanding from clinical imaging data, a subject she has worked on for two decades.
KAIST School of Computing Ph.D. graduate Junghwan Lee, and M.S. student Heechan Shin (advisor: Prof. Sungeui Yoon) won the Best Paper in Robotic Planning award at the ICAR 2017, held last July 10 through 12 at Hong Kong. The award recognized their research on data-driven kinodynamic RRT. This research was on a method for constructing a database of known possible states and inputs of a kidonynamic robot’s movements for future reference, such that robots can simply search the database. By removing the need to calculate the inputs in real time, the research greatly reduces the computational load of robotic movement. Lee has a history of outstanding academic achievement, having published multiple papers at various internationally renowned journals and conferences during the course of his Ph.D. years, including the IEEE Transaction on Robotics (TRO), International Conference on Robotics and Automation (ICRA), and the International Conference on Intelligent Robotics and Systems (IROS). Advisor: Prof. Sungeui Yoon, 1 st Author: Junghwan Lee, 2 nd Author: Heechan Shin IICAR 2017 Website: http://www.ee.cuhk.edu.hk/~qhmeng/icar2017/index.htm We offer our most sincere congratulations on their success.
KAIST School of Computing Professor Meeyoung Cha presented the keynote for the IEEE/ACM International Conference on ASONAM (Advances in Social Networks Analysis and Mining) with a talk on “The Propagation of Rumors and Fake News”. The conference, held this Aug 1 through 3 at Sydney, Australia, awarded Prof. Cha a plaque of appreciation for her services. The keynote presented an AI based method for detecting fake news online, and the validation of the theory using sociology. This research pointed out the problem that social media, being devoid of any fact checking stage, might propagate false information as if it was true, especially in a society where phones and social media enable a convenient and habitual consumption of information. The research team examined this social phenomenon scientifically, thus proposing a big data and AI algorithm based unsupervised method for online information credibility assessment. The research results were published on various renowned conferences and journals, including ICDM, IJCAI, and PLoS One.
※ Pictured, from left to right: Professors Hongseok Yang, Meeyoung Cha, and Sooel Son We at the School of Computing welcome the appointment of Professors Hongseok Yang, Meeyoung Cha, and Sooel Son. □ Prof. Hongseok Yang was appointed a professor at the SoC as of July 24 His areas of interest include programming languages and machine learning A more detailed introduction is available on his page https://cs.kaist.ac.kr/people/view?idx=552&kind=faculty&menu=160 □ Prof. Meeyoung Cha was appointed a professor at the SoC as of Aug. 1 Her areas of interest include network science, computational social science, and statistical inference A more detailed introduction is available on her page https://cs.kaist.ac.kr/people/view?idx=418&kind=faculty&menu=160 □ Prof. Sooel Son was appointed a professor at the SoC as of Aug. 1 His areas of interest include web security & privacy, and web authentication A more detailed introduction is available on his page https://cs.kaist.ac.kr/people/view?idx=553&kind=faculty&menu=160 We extend our most sincere congratulations on their appointment.
Linked are articles on School of Computing Professor Insik Shin’s lab and their development oㄹ mobile platform technology for sharing app functionalities on a smart device. The research paper was also published on ACM MobiSys. Articles in Korean. Segye Daily, 2017 July 28: http://www.segye.com/newsView/20170726003462 Sedaily, 2017 July 26: http://www.sedaily.com/NewsView/1OIMI5D8UD Herald Business, 2017 July 26: http://biz.heraldcorp.com/view.php?ud=201707260002
A KAIST School of Computing research team led by Professor Shin Yoo won the 2017 Human Competitiveness Award (Humies) Silver Award at Genetic and Evolutionary Computation Conference (GECCO), a conference held by ACM SIGEVO. The event recognizes human parity achievements in research by evolutionary computation. Prof. Yoo’s research contribution involved machine learning based defect location technology (technology for automatically finding defective code found during software testing), which is on parity with existing methods researched by man, and also proved via machine learning that a more accurate method does not exist. Prof. Yoo provided the core elements of the theoretical proof, as well as led an international research team composed of researchers from China’s Wuhan University, Australia’s Swinburne University of Technology, and Briton’s University College London, and then submitted the paper to ACM Transactions on Software Engineering and Methodology in the Humies track. Fourteen teams competed in this year’s Humies track, out of which 9 passed the final selection from a panel of judges. The gold award went to a University of Sidney research team for using genetic algorithms for explaining the causal model of recent quantum entanglement research results. The bronze went to two teams, for a technology to optimize a structure’s truss arrangement, and a technology to optimize A/B testing on websites to maximize customer loyalty. ACM GECCO is a large scale conference with a proud 20 year history in the field of genetic and evolutionary computation research. Its 13 tracks cover various topics, including genetic and evolutionary computation, artificial life, artificial immune systems among others, and welcoms topics on both theory and application. This Humies Award was the 14th such award, created by ex-Stanford Professor John Koza, who greatly contributed to the research of evolutionary computation. Entries must prove via research papers that their evolutionary computation technique shows human parity in their performance in a specific field. Prof. Yoo is a world renown expert in the field of search based software engineering, which uses evolutionary computation techniques to solve software engineering problems.
Our very own School of Computing undergraduate student Seung Geun Baek (Advisor: Prof. Moonzoo Kim) won silver in the 7th TOPCIT Periodic Evaluation held jointly by the Ministry of Science and ICT, and the IITP. We extend our most sincere congratulations. Article in Korean: http://www.newstap.co.kr/news/articleView.html?idxno=56908
Linked is a 2017 July 12 article about Professor Sungho Jo on the Electronic Times. Article in Korean: http://www.etnews.com/20170712000212