Naver Corporation

About: Naver Corporation is a company organization based out in Seongnam-si, South Korea. It is known for research contribution in the topics: Terminal (electronics) & Computer science. The organization has 4038 authors who have published 4294 publications receiving 35045 citations. The organization is also known as: NAVER Corporation & NAVER.

Multimodal Dual Attention Memory for Video Story Question Answering

Abstract: We propose a video story question-answering (QA) architecture, Multimodal Dual Attention Memory (MDAM). The key idea is to use a dual attention mechanism with late fusion. MDAM uses self-attention to learn the latent concepts in scene frames and captions. Given a question, MDAM uses the second attention over these latent concepts. Multimodal fusion is performed after the dual attention processes (late fusion). Using this processing pipeline, MDAM learns to infer a high-level vision-language joint representation from an abstraction of the full video content. We evaluate MDAM on PororoQA and MovieQA datasets which have large-scale QA annotations on cartoon videos and movies, respectively. For both datasets, MDAM achieves new state-of-the-art results with significant margins compared to the runner-up models. We confirm the best performance of the dual attention mechanism combined with late fusion by ablation studies. We also perform qualitative analysis by visualizing the inference mechanisms of MDAM.

Variable Selection in General Frailty Models using Penalized H-likelihood

Abstract: Variable selection methods using a penalized likelihood have been widely studied in various statistical models. However, in semiparametric frailty models, these methods have been relatively less studied because the marginal likelihood function involves analytically intractable integrals, particularly when modeling multicomponent or correlated frailties. In this article, we propose a simple but unified procedure via a penalized h-likelihood (HL) for variable selection of fixed effects in a general class of semiparametric frailty models, in which random effects may be shared, nested, or correlated. We consider three penalty functions (least absolute shrinkage and selection operator [LASSO], smoothly clipped absolute deviation [SCAD], and HL) in our variable selection procedure. We show that the proposed method can be easily implemented via a slight modification to existing HL estimation approaches. Simulation studies also show that the procedure using the SCAD or HL penalty performs well. The usefulness of ...

Pressure assisted forward osmosis for shale gas wastewater treatment

Abstract: This study investigated the feasibility of using pressure assisted forward osmosis (PAFO) for shale gas wastewater treatment. PAFO combines osmotic gradient across a membrane with external pressure together, which was expected to obtain higher flux than forward osmosis (FO). Experiments were performed in a laboratory-scale PAFO system, which allows the application of external pressure up to 10 bar on the feed solution side. Deionized water and three kind of synthetic shale gas wastewater, including low range, medium range, and high range wastewaters, were used as feed solutions and NaCl was used as a draw solution. The water flux was improved up to 10–15% by applying external pressure to FO when low range and medium range wastewaters were treated. However, the effect of the external pressure was significantly reduced when the high-range wastewater was treated. After FO treatment, air gap membrane distillation was successfully applied to re-concentrate the draw solutes.

Method for providing answer for question on the internet

Abstract: The present invention relates to a method for providing answers to a question wherein the quality of the question and answer data posted by users is evaluated to improve the quality of the question and answer in an answer and question service in which the users can ask questions and give answers on the Internet. According to the present invention, there is provided a method for providing answers to a question, which is performed on the Internet, comprising the steps of: receiving a question from a first user, providing the question on a given web page, receiving evaluation data for the question from a second user or an operator of the web page, calculating evaluation result data by reflecting the evaluation data, and providing the evaluation result data on the web page in association with the question.

Method and system for matching advertising using seed

Abstract: A method of generating an advertising list using a seed, which can include a plurality of advertisers or a plurality of keywords with an identical property in a single seed and thereby can display various types of advertising information of the plurality of advertisers, which are included in the seed, according to a single condition, and an advertising matching system is provided. According to the present invention, it is possible to provided a method of generating an advertising list using a seed, which can determine a property of a content page using a seed, and thereby can display various types of advertising information of a plurality of advertisers with the determined property, as contextual advertising, for a user, and an advertising matching system using the method.