• KIISE Transactions on Computing Practices
• /
• v.23 no.4
• /
• pp.268-273
• /
• 2017

Recently, deep learning encoder-based approach has been actively applied in the field of sentiment classification. However, Long Short-Term Memory network deep learning encoder, the commonly used architecture, lacks the quality of vector representation when the length of the documents is prolonged. In this study, for effective classification of the sentiment documents, we suggest the use of attention method-based deep learning encoder that generates document vector representation by weighted sum of the outputs of Long Short-Term Memory network based on importance. In addition, we propose methods to modify the attention method-based deep learning encoder to suit the sentiment classification field, which consist of a part that is to applied to window attention method and an attention weight adjustment part. In the window attention method part, the weights are obtained in the window units to effectively recognize feeling features that consist of more than one word. In the attention weight adjustment part, the learned weights are smoothened. Experimental results revealed that the performance of the proposed method outperformed Long Short-Term Memory network encoder, showing 89.67% in accuracy criteria.

• KIPS Transactions on Software and Data Engineering
• /
• v.4 no.5
• /
• pp.225-230
• /
• 2015

In this paper, we propose an effective real-time visual loop closure detector, VILODE, which makes use of key frames and bag of visual words (BoW) based on SURF feature points. In order to determine whether the camera has re-visited one of the previously visited places, a loop closure detector has to compare an incoming new image with all previous images collected at every visited place. As the camera passes through new places or locations, the amount of images to be compared continues growing. For this reason, it is difficult for a visual loop closure detector to meet both real-time constraint and high detection accuracy. To address the problem, the proposed system adopts an effective key frame selection strategy which selects and compares only distinct meaningful ones from continuously incoming images during navigation, and so it can reduce greatly image comparisons for loop detection. Moreover, in order to improve detection accuracy and efficiency, the system represents each key frame image as a bag of visual words, and maintains indexes for them using DBoW database system. The experiments with TUM benchmark datasets demonstrates high performance of the proposed visual loop closure detector.

• KIPS Transactions on Computer and Communication Systems
• /
• v.2 no.6
• /
• pp.265-278
• /
• 2013

With the recent advancements in various wireless communication technologies, the importance of mobile multicasting is coming to the fore, in an effort to use network resources more efficiently. In the past, when various mobile IP-based multicast techniques were proposed, the focus was put on the costs needed for network delivery for providing multicast services, as well as on minimizing the multicast handover delay. For techniques using MIPv6 (Mobile IPv6), a host-based mobility management protocol, however, it is fundamentally difficult to resolve the problems of handover delay and tunnel convergence. To resolve these problems, a network-based mobility management protocol called PMIPv6 (Proxy Mobile IPv6) was standardized. Although performance is improved in PMIPv6 over MIPv6, it still suffers from the problems of handover delay and tunnel convergence. In this paper, to overcome these limitations, a technique called LFH (Low-cost and Fast Handoff) is proposed for fast and low-cost mobility management with multicasting support in PMIPv6 networks. To reduce the interactions between the complex multicast routing protocol and the multicast messages, a simplified proxy method called MLD (Multicast Listener Discovery) is implemented and modified. Furthermore, a TCR (Tunnel Combination and Reconstruction) algorithm was used in the multicast handover procedure within the LMA (Local Mobility Anchor) domain, as well as in the multicast handover procedure between domains, in order to overcome the problem of tunnel convergence. As a result, it was found that LFH has reduced multicast delay compared to other types of multicast techniques, and that it requires lower costs as well.

• Journal of KIISE:Computing Practices and Letters
• /
• v.10 no.1
• /
• pp.99-113
• /
• 2004

A new architecture called SAN(Storage Area Network) was developed in response to the requirements of high availability of data, scalable growth, and system performance. In order to use SAN more efficiently, most of SAN operating softwares support storage virtualization concepts that allow users to view physical storage devices attached to SAN as a large volume virtually h logical volume manager plays a key role in storage virtualization. It realizes the storage virtualization by mapping logical addresses to physical addresses. A logical volume manager also supports a snapshot that preserves a volume image at certain time and on-line reorganization to allow users to add/remove storage devices to/from SAN even while the system is running. To support the snapshot and the on-line reorganization, most logical volume managers have used table based mapping methods. However, it is very difficult to manage mapping table because the mapping table is large in proportion to a storage capacity. In this paper, we design and implement an efficient and flexible hybrid mapping method based on mathematical equations. The mapping method in this paper supports a snapshot and on-line reorganization. The proposed snapshot and on-line reorganization are performed on the reserved area which is separated from data area of a volume. Due to this strategy normal I/O operations are not affected by snapshot and reorganization. Finally, we show the superiority of our proposed mapping method through various experiments.

• Journal of KIISE:Software and Applications
• /
• v.31 no.9
• /
• pp.1233-1245
• /
• 2004

Though biometrics to authenticate a person is a good tool in terms of security and convenience, typical authentication algorithms using biometrics may not be executed on resource-constrained devices such as smart cards. Thus, to execute biometric processing on resource-constrained devices, it is desirable to develop lightweight authentication algorithm that requires only small amount of memory and computation. Also, among biological features, face is one of the most acceptable biometrics, because humans use it in their visual interactions and acquiring face images is non-intrusive. We present a new face authentication algorithm in this paper. Our achievement is two-fold. One is to present a face authentication algorithm with low memory requirement, which uses support vector machines (SVM) with the feature set extracted by genetic algorithms (GA). The other contribution is to suggest a method to reduce further, if needed, the amount of memory required in the authentication at the expense of verification rate by changing a controllable system parameter for a feature set size. Given a pre-defined amount of memory, this capability is quite effective to mount our algorithm on memory-constrained devices. The experimental results on various databases show that our face authentication algorithm with SVM whose input vectors consist of discriminating features extracted by GA has much better performance than the algorithm without feature selection process by GA has, in terms of accuracy and memory requirement. Experiment also shows that the number of the feature ttl be selected is controllable by a system parameter.

• Journal of KIISE
• /
• v.42 no.3
• /
• pp.307-319
• /
• 2015

Current ontology studies use the Hadoop distributed storage framework to perform map-reduce algorithm-based reasoning for scalable ontologies. In this paper, however, we propose a novel approach for scalable Web Ontology Language (OWL) Horst Lite ontology reasoning, based on distributed cluster memories. Rule-based reasoning, which is frequently used for scalable ontologies, iteratively executes triple-format ontology rules, until the inferred data no longer exists. Therefore, when the scalable ontology reasoning is performed on computer hard drives, the ontology reasoner suffers from performance limitations. In order to overcome this drawback, we propose an approach that loads the ontologies into distributed cluster memories, using Spark (a memory-based distributed computing framework), which executes the ontology reasoning. In order to implement an appropriate OWL Horst Lite ontology reasoning system on Spark, our method divides the scalable ontologies into blocks, loads each block into the cluster nodes, and subsequently handles the data in the distributed memories. We used the Lehigh University Benchmark, which is used to evaluate ontology inference and search speed, to experimentally evaluate the methods suggested in this paper, which we applied to LUBM8000 (1.1 billion triples, 155 gigabytes). When compared with WebPIE, a representative mapreduce algorithm-based scalable ontology reasoner, the proposed approach showed a throughput improvement of 320% (62k/s) over WebPIE (19k/s).

• Korean Journal of Remote Sensing
• /
• v.36 no.5_1
• /
• pp.653-666
• /
• 2020

Understanding the vegetation structure is important to manage forest resources for sustainable forest development. With the recent development of technology, it is possible to apply new technologies such as drones and deep learning to forests and use it to estimate the vegetation structure. In this study, the vegetation structure of Gongju, Samchuk, and Seoguipo area was identified by fusion of drone-optical images and LiDAR data using Artificial Neural Networks(ANNs) with the accuracy of 92.62% (Kappa value: 0.59), 91.57% (Kappa value: 0.53), and 86.00% (Kappa value: 0.63), respectively. The vegetation structure analysis technology using deep learning is expected to increase the performance of the model as the amount of information in the optical and LiDAR increases. In the future, if the model is developed with a high-complexity that can reflect various characteristics of vegetation and sufficient sampling, it would be a material that can be used as a reference data to Korea's policies and regulations by constructing a country-level vegetation structure map.

• Journal of Sensor Science and Technology
• /
• v.27 no.5
• /
• pp.317-327
• /
• 2018

Terahertz (THz) time-domain spectroscopy(TDS), imaging techniques, and related systems have become mature technologies, widely used in many universities and research laboratories. However, the development of creative technologies still requires improved THz application systems. A few key points are discussed, including the innovative advances of mode-locking energy-emitting semiconductor lasers and better photoconductive semiconductor quantum structures. To realize a compact, low cost, and high performance THz system, it is essential that THz spectroscopy and imaging technologies are better characterized by semiconductor and nano-devices, both static and time-resolved. We introduce the THz spectroscopy and imaging systems, the OSCAT(Optical Sampling by laser CAvity Tuning) system and the ASOPS(ASynchronous Optical Sampling) system, are constructed by our research team. We report on the THz images obtained from their use.

• Archives of design research
• /
• v.18 no.1 s.59
• /
• pp.223-232
• /
• 2005

By supply of high performance computer becomes all industries automation and convenient of our life. Can speak that is pursuit of swiftness and variety that is prior most development of information society and production system in competitive industrial design product development process. Specially, students who consumer personality studies design according to variableness must present idea in a short time. Therefore, find a new tool that can show quick modeling result for present of new idea or fast design specifications examination and production at college or corporation now. According to such environment change did request by cost time and cost-cutting until reach in production from product design because competition is gone vigorously. University arrived to think RP induction by result presentation and so on for third dimension model by various design process. This research purpose is going to compare difference methods(Manual processing, CNC machine) with students' design result manufacture process on RP appliance that is using in college. Consequently, think that this study is helpful in the college of which is making use of RP appliance or planning installation. By result of this research, model of big curved line is profitable in hand process, by CNC Is profitable when model is going to manufacture finely, RP work could know that size of product is profitable small thing as is complicated.

• KIISE Transactions on Computing Practices
• /
• v.23 no.5
• /
• pp.275-283
• /
• 2017

A large amount of RDF data has been generated along with the increase of semantic web services. Various distributed storage and query processing schemes have been studied to efficiently use the massive amounts of RDF data. In this paper, we propose a distributed SPARQL query processing scheme that considers the data locality and query execution path of large RDF data. The proposed scheme considers the data locality and query execution path in order to reduce join and communication costs. In a distributed environment, when processing a SPARQL query, it is divided into several sub-queries according to the conditions of the WHERE clause by considering the data locality. The proposed scheme reduces data communication costs by grouping and processing the sub-queries through the index based on associated nodes. In addition, in order to reduce unnecessary joins and latency when processing the query, it creates an efficient query execution path considering data parsing cost, the amount of each node's data communication, and latency. It is shown through various performance evaluations that the proposed scheme outperforms the existing scheme.