• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.34-36
• /
• 2022

In the existing text sentiment analysis models, the entire text is usually directly modeled as a whole, and the hierarchical relationship between text contents is less considered. However, in the practice of sentiment analysis, many texts are mixed with multiple emotions. If the semantic modeling of the whole is directly performed, it may increase the difficulty of the sentiment analysis model to judge the sentiment, making the model difficult to apply to the classification of mixed-sentiment sentences. Therefore, this paper proposes a sentiment analysis model BHGCN that considers the text hierarchy. In this model, the output of hidden states of each layer of BERT is used as a node, and a directed connection is made between the upper and lower layers to construct a graph network with a semantic hierarchy. The model not only pays attention to layer-by-layer semantics, but also pays attention to hierarchical relationships. Suitable for handling mixed sentiment classification tasks. The comparative experimental results show that the BHGCN model exhibits obvious competitive advantages.

• The KIPS Transactions:PartB
• /
• v.15B no.2
• /
• pp.137-146
• /
• 2008

Fault localization plays a significant role in enormous distributed system because it can identify root cause of observed faults automatically, supporting self-managing which remains an open topic in managing and controlling complex distributed systems to improve system reliability. Although many Artificial Intelligent techniques have been introduced in support of fault localization in recent research especially in increasing complex ubiquitous environment, the provided functions such as diagnosis and prediction are limited. In this paper, we propose fault localization for self-managing in performance evaluation in order to improve system reliability via learning and analyzing real-time streams of system performance events. We use probabilistic reasoning functions based on the basic Bayes' rule to provide effective mechanism for managing and evaluating system performance parameters automatically, and hence the system reliability is improved. Moreover, due to large number of considered factors in diverse and complex fault reasoning domains, we develop an efficient method which extracts relevant parameters having high relationships with observing problems and ranks them orderly. The selected node ordering lists will be used in network modeling, and hence improving learning efficiency. Using the approach enables us to diagnose the most probable causal factor with responsibility for the underlying performance problems and predict system situation to avoid potential abnormities via posting treatments or pretreatments respectively. The experimental application of system performance analysis by using the proposed approach and various estimations on efficiency and accuracy show that the availability of the proposed approach in performance evaluation domain is optimistic.

• Journal of KIISE:Information Networking
• /
• v.37 no.4
• /
• pp.272-283
• /
• 2010

Wireless Mesh Networks (WMNs) have gained a lot of attention recently. Based on the characteristic of WMNs as a highly connected wireless infrastructure, many efforts from research organizations are made in order to improve the performance of the flow throughput in WMNs. Therefore, it is very critical issue to establish efficient routing paths for multiple concurrent ongoing flows. In this paper, we propose a general modeling methodology to analyze the end-to-end throughput of multiple concurrent flows by analytical calculation taking into account the carrier sensing behaviors, interference and the IEEE 802.11 Distributed Coordination Function mechanism. After the comparison of the average service time for each successful transmission at each node, we analyze the bottlenecks of flows, and hence obtain the maximum end-to-end throughput of them. By using our proposed model, it is possible to predicate the throughput of several candidate routing paths for multiple concurrent ongoing data flows, so we can select the most efficient route that can achieve the highest throughput. We carry out simulations with various traffic patterns of multiple flows in WMNs to validate our modeling and our efficient route selection mechanism.

• Journal of KIISE:Information Networking
• /
• v.32 no.4
• /
• pp.515-524
• /
• 2005

As the number of the mobile nodes (MNs) increases in the networks, the signaling traffic generated by mobility management for MNs will increase explosively, and such a phenomenon will probably affect overall network performance. In this paper, we propose a novel analytical approach using a continuous-time Markov chain model and hierarchical network model for the analysis on the signaling load of representative IPv6 mobility support Protocols such as Mobile IPv6 (MIPv6) and Hierarchical Mobile IPv6 (HMIPv6). According to these analytical modeling, this paper derives the various signaling costs, which are generated by an MN during its average domain residence time when MIPv6 and HMIPv6 are deployed under the same network architecture, respectively. In addition, based on these derived costs, we investigate the effects of various mobility/traffic-related parameters on the signaling costs generated by an MN under MIPv6 and HMIPv6. The analytical results show that as the average moving speed of an MN gets higher and the binding lifetime is set . to the larger value, and as its average packet arrival rate gets lower, the total signaling cost generated during its average domain residence time under HMIPv6 will get relatively lower than that under MIPv6, and that under the reverse conditions, the total signaling cost under MIPv6 will get relatively lower than that under HMIPv6.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.10
• /
• pp.724-739
• /
• 2009

This paper suggests the energy-efficient message delivery scheme and the software platform which exploits the multiple network interfaces of the mobile terminals and GPS in the current mobile devices. The mobile terminals determine the delivery method among 'direct', 'indirect', and 'WAN' based on the position information of itself and other terminals. 'Direct' method sends a message directly to the target terminal using local RAT. 'Indirect' method extends the service area by exploiting intermediate terminals as relay node. If the target terminal is too far to reach through 'direct' or 'indirect' method, the message is sent using wireless WAN technology. Our proposed scheme exploits the position information and, thus, power consumption is drastically reduced in determining handover time and direction. Network simulation results show that our proposed delivery scheme improves the message transfer efficiency and the handover detection latency. We implemented a message platform in a smart phone realizing the proposed delivery scheme. We compared our platform with other typical message platforms from energy efficiency aspect by observing the real power consumption and applying the mathematical modeling. The comparison results show that our platform requires significantly less power.

• Journal of the Korea Society of Computer and Information
• /
• v.16 no.6
• /
• pp.143-153
• /
• 2011

The goal of Ubiquitous Sensor Network(USN) environments is to provide users high quality services based on a variety of sensors. In this environment, sensor devices, sensor nodes and sensor networks are heterogeneous and have various characteristics. Therefore it is important for interoperability to define a metadata for describing USN resources. The OGC(Open Geospatial Consortium) proposes SensorML(Sensor Model Language) as a standard language for modeling sensors. However, SensorML provides a framework for describing a processing model among sensors rather than describing information of sensors. Therefore, to describe a USN metadata is not main purposes of SensorML. This paper defines a USN metadata which describes information about sensor device, sensor node, and sensor network. Also the paper proposes a method for efficiently storing and searching the USN metadata and implements a USN metadata management system based on our method. We show that our metadata management system is reasonable for managing the USN metadata through performance evaluation. Our USN metadata keeps the interoperability in USN environments because the metadata is designed on SensorML. The USN metadata management system can be used directly for a USN middleware or USN application.

• Journal of Intelligence and Information Systems
• /
• v.20 no.2
• /
• pp.109-122
• /
• 2014

People are nowadays creating a tremendous amount of data on Social Network Service (SNS). In particular, the incorporation of SNS into mobile devices has resulted in massive amounts of data generation, thereby greatly influencing society. This is an unmatched phenomenon in history, and now we live in the Age of Big Data. SNS Data is defined as a condition of Big Data where the amount of data (volume), data input and output speeds (velocity), and the variety of data types (variety) are satisfied. If someone intends to discover the trend of an issue in SNS Big Data, this information can be used as a new important source for the creation of new values because this information covers the whole of society. In this study, a Twitter Issue Tracking System (TITS) is designed and established to meet the needs of analyzing SNS Big Data. TITS extracts issues from Twitter texts and visualizes them on the web. The proposed system provides the following four functions: (1) Provide the topic keyword set that corresponds to daily ranking; (2) Visualize the daily time series graph of a topic for the duration of a month; (3) Provide the importance of a topic through a treemap based on the score system and frequency; (4) Visualize the daily time-series graph of keywords by searching the keyword; The present study analyzes the Big Data generated by SNS in real time. SNS Big Data analysis requires various natural language processing techniques, including the removal of stop words, and noun extraction for processing various unrefined forms of unstructured data. In addition, such analysis requires the latest big data technology to process rapidly a large amount of real-time data, such as the Hadoop distributed system or NoSQL, which is an alternative to relational database. We built TITS based on Hadoop to optimize the processing of big data because Hadoop is designed to scale up from single node computing to thousands of machines. Furthermore, we use MongoDB, which is classified as a NoSQL database. In addition, MongoDB is an open source platform, document-oriented database that provides high performance, high availability, and automatic scaling. Unlike existing relational database, there are no schema or tables with MongoDB, and its most important goal is that of data accessibility and data processing performance. In the Age of Big Data, the visualization of Big Data is more attractive to the Big Data community because it helps analysts to examine such data easily and clearly. Therefore, TITS uses the d3.js library as a visualization tool. This library is designed for the purpose of creating Data Driven Documents that bind document object model (DOM) and any data; the interaction between data is easy and useful for managing real-time data stream with smooth animation. In addition, TITS uses a bootstrap made of pre-configured plug-in style sheets and JavaScript libraries to build a web system. The TITS Graphical User Interface (GUI) is designed using these libraries, and it is capable of detecting issues on Twitter in an easy and intuitive manner. The proposed work demonstrates the superiority of our issue detection techniques by matching detected issues with corresponding online news articles. The contributions of the present study are threefold. First, we suggest an alternative approach to real-time big data analysis, which has become an extremely important issue. Second, we apply a topic modeling technique that is used in various research areas, including Library and Information Science (LIS). Based on this, we can confirm the utility of storytelling and time series analysis. Third, we develop a web-based system, and make the system available for the real-time discovery of topics. The present study conducted experiments with nearly 150 million tweets in Korea during March 2013.

• Journal of Korean Society of Transportation
• /
• v.17 no.2
• /
• pp.55-68
• /
• 1999

Many ITS services require map databases in digital form to meet desired needs. Due to the dynamic nature of ITS and the sheer diversity of applications, the design and development of spatial databases to meet those needs pose a major challenge to both the public and private sectors. This challenge is further complicated by the necessity to transfer locationally referenced information between different kinds of databases and spatial data handling systems so that ITS products will work seamlessly across the region and nation. The Purpose of this paper is to develop the framework-models commonly to reference locations in the various applications and systems-the ITS Map Datum and LRS(Location Referencing System). The ITS Map Datum consists of the around control points which are the prime intersections (nodes) of the nationwide road network In this study, the major points have been determined along wish link-node modeling procedure. LRS, defined as a system for determining the position (location) of an entity relative to other entities or to some external frame of reference, has also been set up using CSOM type method. The method has been implemented using ArcView GIS software over the Kangnam and Seocho districts in the city of Seoul, showing that the implemented LRS scheme can be used successfully elsewhere. With the proper advent of the K.ITS architecture and services, the procedure can be used to improve the data sharing and to inter operate among systems, enhancing the efficiency both in terms of money and time.

• Journal of Advanced Navigation Technology
• /
• v.15 no.6
• /
• pp.1015-1024
• /
• 2011

Mobile P2P protocols in ad-hoc networks have gained large attention recently. Although there has been much research on P2P algorithms for wired networks, existing P2P protocols are not suitable for mobile ad-hoc networks because they do not consider mobility of peers. This study proposes a new cluster-based P2P protocol for ad hoc networks which utilizes peer mobility. In typical cluster-based P2P algorithms, each cluster has a super peer and other peers of the cluster register their file list to the super peer. High mobility peers would cause a lot of file list registration traffic because they hand-off between clusters frequently. In the proposed scheme, while peers with low mobility behave in the same way as the peers of the typical cluster-based P2P schemes, peers with high mobility behave differently. They inform their entrance to the cluster region to the super peer but they do not register their file list to the super peer. When a peer wishes to find a file, it first searches the registered file list of the super peer and if fails, query message is broadcasted. We perform mathematical modeling, analysis and optimization of the proposed scheme regarding P2P traffic and associated routing traffic. Numerical results show that the proposed scheme performs much better than or similar to the typical cluster-based P2P scheme and flooding based Gnutella.