• International Journal of Knowledge Content Development & Technology
• /
• v.11 no.3
• /
• pp.29-44
• /
• 2021

The study provides a quantitative and qualitative description of global research in the domain of quantum machine learning (QML) as a way to understand the status of global research in the subject at the global, national, institutional, and individual author level. The data for the study was sourced from the Scopus database for the period 1999-2020. The study analyzed global research output (1374 publications) and global citations (22434 citations) to measure research productivity and performance on metrics. In addition, the study carried out bibliometric mapping of the literature to visually represent network relationship between key countries, institutions, authors, and significant keyword in QML research. The study finds that the USA and China lead the world ranking in QML research, accounting for 32.46% and 22.56% share respectively in the global output. The top 25 global organizations and authors lead with 35.52% and 16.59% global share respectively. The study also tracks key research areas, key global players, most significant keywords, and most productive source journals. The study observes that QML research is gradually emerging as an interdisciplinary area of research in computer science, but the body of its literature that has appeared so far is very small and insignificant even though 22 years have passed since the appearance of its first publication. Certainly, QML as a research subject at present is at a nascent stage of its development.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.10a
• /
• pp.354-356
• /
• 2014

Recently, The sensor technology Has been used in many fields, it is detected data of various types of sensors, is very difficult to integrate due to differences in the standards and each other unit. Also, when we providing a service or program on a data cloud, it is important that integrates of such data in order to take advantage of the detected data in the similar field. In this paper, we propose a approaches to integrating data to be provided as a service in the cloud of data arising from a heterogeneous sensors. The approaches are generating a standard meta-data based on ontology, it is mapping with detected data by the sensor data. Accordingly, the detected data is possible to improve the efficiency of data transfer between the sensor and the application by sending an application in a standard format.

• Smart Media Journal
• /
• v.8 no.3
• /
• pp.31-40
• /
• 2019

To understand how a Convolutional Neural Network (CNN) model captures the features of a pattern to determine which class it belongs to, in this paper, we use Gradient-weighted Class Activation Mapping (Grad-CAM) to visualize and analyze how well a CNN model behave on the CNU weeds dataset. We apply this technique to Resnet model and figure out which features this model captures to determine a specific class, what makes the model get a correct/wrong classification, and how those wrong label images can cause a negative effect to a CNN model during the training process. In the experiment, Grad-CAM highlights the important regions of weeds, depending on the patterns learned by Resnet, such as the lobe and limb on 미국가막사리, or the entire leaf surface on 단풍잎돼지풀. Besides, Grad-CAM points out a CNN model can localize the object even though it is trained only for the classification problem.

• Proceedings of the National Institute of Ecology of the Republic of Korea
• /
• v.2 no.4
• /
• pp.298-304
• /
• 2021

This study aimed to determine the applicability of drones and air quality sensors in environmental monitoring of air pollutant emissions by developing and testing two new methods. The first method used orthoimagery for precise monitoring of pollutant-emitting facilities. The second method used atmospheric sensors for monitoring air pollutants in emissions. Results showed that ground sample distance could be established within 5 cm during the creation of orthoimagery for monitoring emissions, which allowed for detailed examination of facilities with naked eyes. For air quality monitoring, drones were flown on a fixed course and measured the air quality in point units, thus enabling mapping of air quality through spatial analysis. Sensors that could measure various substances were used during this process. Data on particulate matter were compared with data from the National Air Pollution Measurement Network to determine its future potential to leverage. However, technical development and applications for environmental monitoring of pollution-emitting facilities are still in their early stages. They could be limited by meteorological conditions and sensitivity of the sensor technology. This research is expected to provide guidelines for environmental monitoring of pollutant-emitting facilities using drones.

• Smart Media Journal
• /
• v.10 no.1
• /
• pp.79-87
• /
• 2021

In recent years research on automatic font generation with machine learning mainly focus on using transformation-based methods, in comparison, generative model-based methods of font generation have received less attention. Transformation-based methods learn a mapping of the transformations from an existing input to a target. This makes them ambiguous because in some cases a single input reference may correspond to multiple possible outputs. In this work, we focus on font generation using the generative model-based methods which learn the buildup of the characters from noise-to-image. We propose a novel way to train a conditional generative deep neural model so that we can achieve font style control on the generated font images. Our research demonstrates how to generate new font images conditioned on both character class labels and character style labels when using the generative model-based methods. We achieve this by introducing a modified generator network which is given inputs noise, character class, and style, which help us to calculate losses separately for the character class labels and character style labels. We show that adding the character style vector on top of the character class vector separately gives the model rich information about the font and enables us to explicitly specify not only the character class but also the character style that we want the model to generate.

• International Journal of Contents
• /
• v.17 no.4
• /
• pp.1-15
• /
• 2021

Named entity recognition (NER) is a crucial task for NLP, which aims to extract information from texts. To build NER systems, deep learning (DL) models are learned with dictionary features by mapping each word in the dataset to dictionary features and generating a unique index. However, this technique might generate noisy labels, which pose significant challenges for the NER task. In this paper, we proposed DL-dictionary features, and evaluated them on two datasets, including the OntoNotes 5.0 dataset and our new infectious disease outbreak dataset named GFID. We used (1) a Bidirectional Long Short-Term Memory (BiLSTM) character and (2) pre-trained embedding to concatenate with (3) our proposed features, named the Convolutional Neural Network (CNN), BiLSTM, and self-attention dictionaries, respectively. The combined features (1-3) were fed through BiLSTM - Conditional Random Field (CRF) to predict named entity classes as outputs. We compared these outputs with other predictions of the BiLSTM character, pre-trained embedding, and dictionary features from previous research, which used the exact matching and partial matching dictionary technique. The findings showed that the model employing our dictionary features outperformed other models that used existing dictionary features. We also computed the F1 score with the GFID dataset to apply this technique to extract medical or healthcare information.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.8
• /
• pp.2473-2489
• /
• 2022

Urban development has brought about the increasing saturation of urban traffic demand, and traffic congestion has become the primary problem in transportation. Roads are in a state of waiting in line or even congestion, which seriously affects people's enthusiasm and efficiency of travel. This paper mainly studies the discrete domain path planning method based on the flow data. Taking the traffic flow data based on the highway network structure as the research object, this paper uses the deep learning theory technology to complete the path weight determination process, optimizes the path planning algorithm, realizes the vehicle path planning application for the expressway, and carries on the deployment operation in the highway company. The path topology is constructed to transform the actual road information into abstract space that the machine can understand. An appropriate data structure is used for storage, and a path topology based on the modeling background of expressway is constructed to realize the mutual mapping between the two. Experiments show that the proposed method can further reduce the interpolation error, and the interpolation error in the case of random missing is smaller than that in the other two missing modes. In order to improve the real-time performance of vehicle path planning, the association features are selected, the path weights are calculated comprehensively, and the traditional path planning algorithm structure is optimized. It is of great significance for the sustainable development of cities.

• Annual Conference of KIPS
• /
• 2011.04a
• /
• pp.21-24
• /
• 2011

Ad-hoc 네트워크란 지정된 이동성 지원 기반 시설의 도움 없이 임시 망을 구성하는 이동 호스트와 무선 인터페이스의 집합이다. Ad-hoc 네트워크는 Access Point 없이 이동성이 지원 된다는 점에서 최근 많은 각광을 받고 있으며, 현재 많은 연구들이 진행되고 있는 기술이다. 본 논문에서는 이러한 Ad-hoc 네트워크의 특징을 이용하여 소수의 관리자가 다수의 이동 가능한 관리 대상들을 효율적으로 제어하기 위한 사용자 맵핑 시스템을 설계 및 구현 한다. 본 논문에서 하드웨어적으로 직접 구현한 사용자 맵핑 시스템은 사용자 위치 정보 노드(u_LIN, User Location Information Node)라 명명된 장치들이 네트워크를 구축하고, 그 네트워크에서 하나 이상의 u_LIN이 사라지면 경고 시스템을 활성화 하여 사라진 u_LIN을 찾도록 도와주는 시스템이다. 본 논문에서 구현한 사용자 맵핑 시스템의 성능 평가 결과 u_LIN 간 1:1 통신을 직선거리에서 수행하였을 때 약 250m까지 정상적으로 통신하였으며, 1:N 통신 역시 100m 이내의 거리에서 안정적으로 정보를 주고받았다. 본 시스템은 미아 방지 시스템, 등산객 조난 방지 시스템, 유치원 아동관리 시스템, 관광 가이드의 관광객 관리 시스템 등 여러 시스템에 유연하게 적용이 가능하여 앞으로도 활용도가 높을 것으로 예상된다.

• Educational Technology International
• /
• v.15 no.2
• /
• pp.117-141
• /
• 2014

This qualitative study investigated different interaction patterns in an online discussion. The data was collected from asynchronous discussion occurred in a graduate course. The data analysis methods include inductive analysis and mapping strategy. The results of the study suggest three layers of interaction: response sequences, interaction amongst participants, and concept map of messages. The visualization of response sequences enabled the researcher to discover complex and dynamic interaction patterns amongst participants. The many-to-many communication feature of online discussion does not always enable direct one-on-one interaction between two participants. Rather, one message contributed to multiple threads in the stream of conversation. In terms of interaction amongst participants, the interaction amongst participants, as indicated in the data, the messages also bind each participant and consequently a group(s) of participants together. It appears that the contribution of one message may not only enable a response to one participant, but also connect many participants to each other. The concept map of messages proposes that response sequences and interaction amongst participants can also be viewed between concepts within messages in the discussion. On the surface, the messages posted by individuals are linked by the system in a linear fashion as they are posted. However, the interaction extends to collaborative conversation amongst participants. Ultimately, a conceptual network of interrelated ideas including multiple perspectives is built in asynchronous discussion.

• ETRI Journal
• /
• v.46 no.2
• /
• pp.277-289
• /
• 2024

In this paper, we describe a neural network-based application that recommends multiple items using dialog context input and simultaneously outputs a response sentence. Further, we describe a multi-item recommendation by specifying it as a set of clothing recommendations. For this, a multimodal fusion approach that can process both cloth-related text and images is required. We also examine achieving the requirements of downstream models using a pretrained language model. Moreover, we propose a gate-based multimodal fusion and multiprompt learning based on a pretrained language model. Specifically, we propose an automatic evaluation technique to solve the one-to-many mapping problem of multi-item recommendations. A fashion-domain multimodal dataset based on Koreans is constructed and tested. Various experimental environment settings are verified using an automatic evaluation method. The results show that our proposed method can be used to obtain confidence scores for multi-item recommendation results, which is different from traditional accuracy evaluation.