• Title/Summary/Keyword: Two-Mode Network

Search Result 256, Processing Time 0.023 seconds

Resolving the 'Gray sheep' Problem Using Social Network Analysis (SNA) in Collaborative Filtering (CF) Recommender Systems (소셜 네트워크 분석 기법을 활용한 협업필터링의 특이취향 사용자(Gray Sheep) 문제 해결)

  • Kim, Minsung;Im, Il
    • Journal of Intelligence and Information Systems
    • /
    • v.20 no.2
    • /
    • pp.137-148
    • /
    • 2014
  • Recommender system has become one of the most important technologies in e-commerce in these days. The ultimate reason to shop online, for many consumers, is to reduce the efforts for information search and purchase. Recommender system is a key technology to serve these needs. Many of the past studies about recommender systems have been devoted to developing and improving recommendation algorithms and collaborative filtering (CF) is known to be the most successful one. Despite its success, however, CF has several shortcomings such as cold-start, sparsity, gray sheep problems. In order to be able to generate recommendations, ordinary CF algorithms require evaluations or preference information directly from users. For new users who do not have any evaluations or preference information, therefore, CF cannot come up with recommendations (Cold-star problem). As the numbers of products and customers increase, the scale of the data increases exponentially and most of the data cells are empty. This sparse dataset makes computation for recommendation extremely hard (Sparsity problem). Since CF is based on the assumption that there are groups of users sharing common preferences or tastes, CF becomes inaccurate if there are many users with rare and unique tastes (Gray sheep problem). This study proposes a new algorithm that utilizes Social Network Analysis (SNA) techniques to resolve the gray sheep problem. We utilize 'degree centrality' in SNA to identify users with unique preferences (gray sheep). Degree centrality in SNA refers to the number of direct links to and from a node. In a network of users who are connected through common preferences or tastes, those with unique tastes have fewer links to other users (nodes) and they are isolated from other users. Therefore, gray sheep can be identified by calculating degree centrality of each node. We divide the dataset into two, gray sheep and others, based on the degree centrality of the users. Then, different similarity measures and recommendation methods are applied to these two datasets. More detail algorithm is as follows: Step 1: Convert the initial data which is a two-mode network (user to item) into an one-mode network (user to user). Step 2: Calculate degree centrality of each node and separate those nodes having degree centrality values lower than the pre-set threshold. The threshold value is determined by simulations such that the accuracy of CF for the remaining dataset is maximized. Step 3: Ordinary CF algorithm is applied to the remaining dataset. Step 4: Since the separated dataset consist of users with unique tastes, an ordinary CF algorithm cannot generate recommendations for them. A 'popular item' method is used to generate recommendations for these users. The F measures of the two datasets are weighted by the numbers of nodes and summed to be used as the final performance metric. In order to test performance improvement by this new algorithm, an empirical study was conducted using a publically available dataset - the MovieLens data by GroupLens research team. We used 100,000 evaluations by 943 users on 1,682 movies. The proposed algorithm was compared with an ordinary CF algorithm utilizing 'Best-N-neighbors' and 'Cosine' similarity method. The empirical results show that F measure was improved about 11% on average when the proposed algorithm was used

    . Past studies to improve CF performance typically used additional information other than users' evaluations such as demographic data. Some studies applied SNA techniques as a new similarity metric. This study is novel in that it used SNA to separate dataset. This study shows that performance of CF can be improved, without any additional information, when SNA techniques are used as proposed. This study has several theoretical and practical implications. This study empirically shows that the characteristics of dataset can affect the performance of CF recommender systems. This helps researchers understand factors affecting performance of CF. This study also opens a door for future studies in the area of applying SNA to CF to analyze characteristics of dataset. In practice, this study provides guidelines to improve performance of CF recommender systems with a simple modification.

  • Effect of the type of resin cement on the fracture resistance of chairside CAD-CAM materials after aging

    • Laura Vitoria Rizzatto;Daniel Meneghetti;Marielle Di Domenico;Julia Cadorin Facenda;Katia Raquel Weber;Pedro Henrique Corazza;Marcia Borba
      • The Journal of Advanced Prosthodontics
      • /
      • v.15 no.3
      • /
      • pp.136-144
      • /
      • 2023
    • PURPOSE. The study objective was to evaluate the influence of the type of resin cement on the flexural strength and load to fracture of two chairside CADCAM materials after aging. MATERIALS AND METHODS. A polymer-infiltrated ceramic network (PICN) and a nanoceramic resin (RNC) were used to produce the specimens. Two types of dual-cure resin cements, a self-adhesive and a universal, were investigated. Bilayer specimens were produced (n = 10) and aged for 6 months in a humid environment before the biaxial flexural strength test (σf). Bonded specimens were subjected to a mechanical aging protocol (50 N, 2 Hz, 37℃ water, 500,000 cycles) before the compressive load test (Lf). σf and Lf data were analyzed using two-way ANOVA and Tukey tests (α = .05). Chi-square test was used to analyze the relationship between failure mode and experimental group (α = .05). RESULTS. The type of resin cement and the interaction between factors had no effect on the σf and Lf of the specimens, while the type of restorative material was significant. RNC had higher σf and Lf than PICN. There was a significant association among the type of cracks identified for specimens tested in Lf and the restorative material. CONCLUSION. The type of resin cement had no effect on the flexural strength and load to fracture of the two investigated CAD-CAM chairside materials after aging.

    On Promoting the Coastal Transport of Container (컨테이너 연안해송 활성화에 관한 연구 -부산항을 중심으로-)

    • Roh H.S.;Lee C.Y.
      • Journal of Korean Port Research
      • /
      • v.7 no.1
      • /
      • pp.43-58
      • /
      • 1993
    • There has been fast progress in economy in Korea derived by a consecutive five-year plan program for economic development started in the early 1960's. In the field of transportation, rapid changes in the technological environment of transportation and communication have brought a revolution of the transport system, of which inter-modal transportation through containerisation is typical. Because of the rapidly growing traffic volumes of cargo, especially container traffic, and lack of investment into transport infrastructure in the past, both road and railway are beyond their capacity. As a result, the public-road network has suffered a serious congestion problem. For instance, in relation to the corridor between Seoul and Pusan, today, it takes about 14 hours for the journey of container trailer through Kyongbu Expressway, for which it used to take only 7 hours in 1986. For the railway, though the congestion problem is not very serious compared with the road sector, a shortage of capacity on certain main lines has emerged as a problem as railway traffic has increased. Furthermore, the further expansion of the system in near future is difficult due to burden of higher construction the cost. Unlike these two modes, coastal shipping, which has been paid relatively less attention for commodity transport in Korea, shows no constraint in this respect. In addition, it is the most cost efficient mode of transport. This work therefore aims to make a proposal for the alternative inland transportation mode, which is to promote the coastal transport of container. Three obstructing factors for the promotion of the coastal transport are investigated and some solutions for those are suggested as follows : First, it appears to be essential to provide exclusive ports for the coastal shipping, that comply with simplification, specialization and rationalization. The optimum size of berths on the exclusive ports in Pusan port is estimated as 16-20. We found that it needs periodical study and publicity on the advantages from the adoption of the coastal mode. Inducing competition in the coastal shipping market is also necessary. For the supply of the fleet in the coastal shipping, chartering of the surplus ships in the oversea shipping is found to be more desirable than new shipbuilding. Second, to solve the fragmentation of the companies which wish to participate in the coastal transport, government has to implement the subsidy policy. The encouragement of participation of the shipping lines engaging in Korea-Japan run and Korea-East South Asia run, into coastal shipping also needs to be considered cautiously. Third, simplification of the document for entry in ports is needed for rational coastal shipping management. We can use B/L (Bill of Lading) for coastal shipping as a prerequisite to get the indemnity by P & I Club. The reduction of the government controls on entering and leaving the ports also needs.

    • PDF

    Alarm Processing for Efficient Fault Management on High Speed Information Networks (초고속정보통신망에서 효율적인 장애관리를 위한 알람처리)

    • 김동일;오창석
      • The Journal of the Korea Contents Association
      • /
      • v.2 no.1
      • /
      • pp.86-97
      • /
      • 2002
    • In this thesis, we propose a high-speed information network management mode based on layering and partitioning concepts of ITU-T G.805 We describe an information mood of the configuration of VP and VC networks and an information mode to manage faults in high-speed information networks. In order to efficient manage faults, us distinguish between an alarm with a fault that has negative effects on services and an alarm with an alert that barely affects services. Thus, we propose two types of computational models to treat all of these alarms, a fault alarm processing system and an alert alarm processing system. We also describe procedures of alarm processing and we describe the scope and processing course of NMS and SubNMS when rerouting a fault VP and VC fault connection. Lastly, we describe the implementation and analysis of the performance of the proposed fault alarm processing system. We compare the alarm processing tire of each priority according to the alarm occurrence ratio.

    • PDF

    Implementation of Next Generation DSL Networks (차세대 DSL망의 구현 방안)

    • Park Seung-Chul
      • Journal of KIISE:Information Networking
      • /
      • v.32 no.2
      • /
      • pp.236-243
      • /
      • 2005
    • Most of the existing legacy DSL networks, about 80 million lines world-widely, adopted ‘ATM(Asynchronous Transfer Mode) over DSL’protocol architecture for the effective interworking with the ATM-based regional backbone networks and for the effective support of native ATM-based multimedia application services. Recently, however, the regional backbone networks are moving towards Metro Ethernet networks, instead of the ATM networks, and most of the current multimedia applications are If-based, not native ATM-based. These environmental changes push the architecture of DSL networks to be accordingly changed, and the‘Ethernet over DSL’protocol architecture, instead of existing‘ATM over DSL’, is tried to be applied to the implementation of next generation DSL networks such as VDSL(Very high-rate DSL) . In this paper, we propose two different implementation models for next generation DSL networks in Metro Ethernet backbone environments, respectively EA(Ethernet-to-ATM) implementation model and RE(Ethernet-to-Ethernet) implementation model. And, a comparative analysis focused on the performance and the backward compatability with the legacy DSL networks will be presented.

    Iterative Detection and ICI Cancellation for MISO-mode DVB-T2 System with Dual Carrier Frequency Offsets

    • Jeon, Eun-Sung;Seo, Jeong-Wook;Yang, Jang-Hoon;Paik, Jong-Ho;Kim, Dong-Ku
      • KSII Transactions on Internet and Information Systems (TIIS)
      • /
      • v.6 no.2
      • /
      • pp.702-721
      • /
      • 2012
    • In the DVB-T2 system with a multiple-input single-output (MISO) transmission mode, Alamouti coded orthogonal frequency division multiplexing (OFDM) signals are transmitted simultaneously from two spatially separated transmitters in a single frequency network (SFN). In such systems, each transmit-receive link may have a distinct carrier frequency offset (CFO) due to the Doppler shift and/or frequency mismatch between the local oscillators. Thus, the received signal experiences dual CFOs. This not only causes dual phase errors in desired data but also introduces inter-carrier interference (ICI), which cannot be removed completely by simply performing a CFO compensation. To overcome this problem, this paper proposes an iterative detection with dual phase errors compensation technique. In addition, we propose a successive-iterative ICI cancellation technique. This technique successively eliminates ICI in the initial iteration by exploiting pre-detected data pairs. Then, in subsequent iterations, it performs a fine interference cancellation using a priori information, iteratively fed back from the channel decoder. In contrast to previous works, the proposed techniques do not require estimates of dual CFOs. Their performances are evaluated via a full DVB-T2 simulator. Simulation results show that the DVB-T2 receiver equipped with the proposed dual phase errors compensation and the successive-iterative ICI cancellation techniques achieves almost the same performance as ideal dual CFOs-free systems, even for large dual CFOs.

    Ultra low-power active wireless sensor for structural health monitoring

    • Zhou, Dao;Ha, Dong Sam;Inman, Daniel J.
      • Smart Structures and Systems
      • /
      • v.6 no.5_6
      • /
      • pp.675-687
      • /
      • 2010
    • Structural Health Monitoring (SHM) is the science and technology of monitoring and assessing the condition of aerospace, civil and mechanical infrastructures using a sensing system integrated into the structure. Impedance-based SHM measures impedance of a structure using a PZT (Lead Zirconate Titanate) patch. This paper presents a low-power wireless autonomous and active SHM node called Autonomous SHM Sensor 2 (ASN-2), which is based on the impedance method. In this study, we incorporated three methods to save power. First, entire data processing is performed on-board, which minimizes radio transmission time. Considering that the radio of a wireless sensor node consumes the highest power among all modules, reduction of the transmission time saves substantial power. Second, a rectangular pulse train is used to excite a PZT patch instead of a sinusoidal wave. This eliminates a digital-to-analog converter and reduces the memory space. Third, ASN-2 senses the phase of the response signal instead of the magnitude. Sensing the phase of the signal eliminates an analog-to-digital converter and Fast Fourier Transform operation, which not only saves power, but also enables us to use a low-end low-power processor. Our SHM sensor node ASN-2 is implemented using a TI MSP430 microcontroller evaluation board. A cluster of ASN-2 nodes forms a wireless network. Each node wakes up at a predetermined interval, such as once in four hours, performs an SHM operation, reports the result to the central node wirelessly, and returns to sleep. The power consumption of our ASN-2 is 0.15 mW during the inactive mode and 18 mW during the active mode. Each SHM operation takes about 13 seconds to consume 236 mJ. When our ASN-2 operates once in every four hours, it is estimated to run for about 2.5 years with two AAA-size batteries ignoring the internal battery leakage.

    Essential Competencies for Digital Workforce of Provincial Office in Thailand Using Delphi Technique

    • Rujira Rikharom;Wirapong, Chansanam
      • Journal of Information Science Theory and Practice
      • /
      • v.11 no.4
      • /
      • pp.51-81
      • /
      • 2023
    • This study aimed to study its required performance requirements and proposes a competency framework necessary for the digital workforce of the Provincial Offices in Thailand. The specific primary informants were determined as 17 people. The collecting process was performed using the Delphi technique and the electronic Delphi technique in two phases, totaling four rounds. In the first time, a structured interview was used to conduct online interviews for 15 people. Content validation was performed to determine issues of the competency framework essential for the digital workforce with 7-level scaled questionnaires, and then online reviews were collected between 10-15 people (2nd to 4th times). A consensus was found and confirmed four times with descriptive statistics, namely frequency, mean, standard deviation, mode, median, and the absolute value of the difference between mode and median, interquartile range, and application of the conceptual framework. The research findings revealed that the essential competency requirements for the digital workforce were covered in digital literacy (six aspects), digital skills (four aspects), and digital characteristics (four aspects). Consensus was confirmed for 84 issues. Therefore, it was concluded that 61 points for building an essential competency framework for the digital workforce made them effective in using digital technology as a labor-saving instrument, as well as for expanding the breadth of development of digital expertise to include members of the organization's digital practitioner network. This development will benefit government agencies and the private sector, both national and international, in the future.

    A Comparative Performance Study of Speech Coders for Three-Way Conferencing in Digital Mobile Communication Networks (이동통신망에서 삼자회의를 위한 음성 부호화기의 성능에 관한 연구)

    • Lee, Mi-Suk;Lee, Yun-Geun;Kim, Gi-Cheol;Lee, Hwang-Su;Jo, Wi-Deok
      • The Journal of the Acoustical Society of Korea
      • /
      • v.14 no.1E
      • /
      • pp.30-38
      • /
      • 1995
    • In this paper, we evaluated the performance of vocoders for three-way conferencing using signal summation technique in digital mobile communication network. The signal summation technique yields natural mode of three-way conferencing, in shich the mixed voice signal from two speakers are transmitted to a third person, though there has been no useful speech coding technique for the mixed voice signal yet. We established Qualcomm code term prediction (RPE-LTP) vocoders to provide three-way conferencing using signal summation techinique. In addition, as the conventional speech quality measures are not applicable to the vocoders for mixed voice signals, we proposed two kinds of subjective quality measures. These are the sentence discrimination (SD) test and the modified degraded mean opinion score (MDMOS) test. The experimental results show that the output speech quality of the VSELP vocoder is superior to other two.

    • PDF

    Search for Very Fast Variability in AGN Radio Light Curves

    • Lee, Taeseok;Trippe, Sascha;Sohn, Bong Won;Lee, Sang-Sung;Byun, Do-Young;Oh, Junghwan
      • The Bulletin of The Korean Astronomical Society
      • /
      • v.38 no.2
      • /
      • pp.43.2-43.2
      • /
      • 2013
    • AGN are known for irregular variability on all time scales, down to intra-day variability with relative variations of a few percent within minutes to hours. In the shortest time scale of AGN activity, unexplored territory still exists: first, the existence of a shortest time scale of AGN activity and secondly the shape of high frequency end of AGN power spectrum. Also the spectral variations at the shortest timescale. Here, we present the preliminary results of AGN fast photometry performed with Korean VLBI Network(KVN). Observations were done in a 'anti-correlated' mode using two antennas, with always either one antenna pointing at the target. This results in an effective time resolution of 3 minutes. We used all four KVN frequencies, 22, 43, 86, and 129 GHz, in order to trace spectral variability. We have been able to derive high-quality light curves for 3C111 and 3C279 at 22 and 86 GHz observed on 31st of May and 30th of Nov. in 2012. We are currently performing detailed statistical analysis in order to assess the levels of variability or the corresponding upper limits.

    • PDF

    (34141) Korea Institute of Science and Technology Information, 245, Daehak-ro, Yuseong-gu, Daejeon
    Copyright (C) KISTI. All Rights Reserved.