Journal of Information Processing
Online ISSN : 1882-6652
Volume 25
Showing 101-119 articles out of 119 articles from the selected issue
  • Yuya Kaneda, Qiangfu Zhao, Yong Liu
    Type: Regular Papers
    Subject area: Knowledge Processing
    2017 Volume 25 Pages 820-830
    Published: 2017
    Released: August 15, 2017
    JOURNALS FREE ACCESS

    The decision boundary making (DBM) algorithm was proposed by us to induce compact and high performance machine learning models for implementation in portable/wearable computing devices. To upgrade performance of DBM-initialized models, we may use all observed data to retrain the model, but the computational cost is high. To reduce the cost, we may use the newly observed datum only, but this often degrades the performance of the model. To solve the problem, we propose on-line training algorithm with guide data (OLTA-GD) in this paper. OLTA-GD updates the model using only a few guide data along with the newest datum. The guide data are selected from all available data. Here, guide data selection is a key point. For this purpose, this paper investigates two methods. The first method is random selection, and the second one is cluster center based. In the second method, the cluster centers are obtained using k-means algorithm. Experimental results show that, OLTA-GD can upgrade the models more steadily than backpropagation (BP) algorithm, and the first selection method is better. For the guide data, around 5 data are usually enough to upgrade the performance steadily, and thus the computational cost is basically not increased compared with the BP.

    Download PDF (1932K)
  • Hidetaka Kamigaito, Akihiro Tamura, Hiroya Takamura, Manabu Okumura, E ...
    Type: Regular Papers
    Subject area: Natural Language Processing
    2017 Volume 25 Pages 831-840
    Published: 2017
    Released: August 15, 2017
    JOURNALS FREE ACCESS

    We propose a novel unsupervised word alignment method that uses a constraint based on Inversion Transduction Grammar (ITG) parse trees to jointly unify two directional models. Previous agreement methods are not helpful for locating alignments with long distances because they do not use any syntactic structures. In contrast, the proposed method symmetrizes alignments in consideration of their structural coherence by using the ITG constraint softly in the posterior regularization framework. The ITG constraint is also compatible with word alignments that are not covered by ITG parse trees. Hence, the proposed method is robust to ITG parse errors compared to other alignment methods that directly use an ITG model. Compared to the HMM, IBM Model 4, and the baseline agreement method, the experimental results show that, in word alignment evaluation, the IBM Model 4 with the proposed ITG constraint achieves the best performance on the Japanese-English KFTT and BTEC corpus, and in translation evaluation, the proposed method shows comparable or statistically significantly better performance on the Japanese-English KFTT, Japanese-English IWSLT 2007, and Czech/German-English WMT 2015 corpus.

    Download PDF (339K)
  • Reina Miyazaki, Kiminori Matsuzaki, Shigeyuki Sato
    Type: Regular Papers
    Subject area: Special Section on Programming
    2017 Volume 25 Pages 841-851
    Published: 2017
    Released: August 15, 2017
    JOURNALS FREE ACCESS

    MapReduce is a framework for large-scale data processing proposed by Google, and its open-source implementation, Hadoop MapReduce, is now widely used. Several language systems have been proposed to make developing MapReduce programs easier, for instance, Sawzall, FlumeJava, Pig, Hive, and Crunch. These language systems mainly target applications that can be naturally solved by using a MapReduce-like programming model. In this study, we propose a new MapReduce-program generator that accepts programs manipulating one-dimensional arrays. By using the proposed generator, users only need to write sequential programs to generate Hadoop MapReduce programs automatically. We applied some program optimization techniques to the generation of Hadoop MapReduce programs. In this paper, we also report our experiment results that compare programs generated by the proposed generator with hand-written MapReduce programs.

    Download PDF (576K)
  • Toshihiro Yamauchi
    Type: Special Issue of Computer Security Technologies against Sophisticated Cyber Attacks
    2017 Volume 25 Pages 852-853
    Published: 2017
    Released: September 15, 2017
    JOURNALS FREE ACCESS
    Download PDF (37K)
  • Takuro Yoshida, Kento Kawakami, Ryotaro Kobayashi, Masahiko Kato, Masa ...
    Type: Special Issue of Computer Security Technologies against Sophisticated Cyber Attacks
    Subject area: Countermeasures against DDoS attacks
    2017 Volume 25 Pages 854-865
    Published: 2017
    Released: September 15, 2017
    JOURNALS FREE ACCESS

    Water torture attacks are a recently emerging type of Distributed Denial-of-Service (DDoS) attack on Domain Name System (DNS) servers. They generate a multitude of malicious queries with randomized, unique subdomains. This paper proposes a detection method and a filtering system for water torture attacks. The former is an enhancement of our previous effort so as to achieve packet-by-packet, on-the-fly processing, and the latter is an application of our current method mainly for defending recursive servers. Our proposed method detects malicious queries by analyzing their subdomains with a naïve Bayes classifier. Considering large-scale applications, we focus on achieving high throughput as well as high accuracy. Experimental results indicate that our method can detect attacks with 98.16% accuracy and only a 1.55% false positive rate, and that our system can process up to 7.44Mpps of traffic.

    Download PDF (1192K)
  • Sebastian Eresheim, Robert Luh, Sebastian Schrittwieser
    Type: Special Issue of Computer Security Technologies against Sophisticated Cyber Attacks
    Subject area: Malwares
    2017 Volume 25 Pages 866-874
    Published: 2017
    Released: September 15, 2017
    JOURNALS FREE ACCESS

    Rootkits constitute a significant threat to modern computing and information systems. Since their first appearance in the early 1990's they have steadily evolved, adapting to ever-improving security measures. The main feature rootkits have in common is the ability to hide their malicious presence and activities from the operating system and its legitimate users. In this paper we systematically analyze process hiding techniques routinely used by rootkit malware. We summarize the characteristics of different approaches and discuss their advantages and limitations. Furthermore, we assess detection and prevention techniques introduced in operating systems in response to the threat of hidden malware. The results of our assessments show that defenders still struggle to keep up with rootkit authors. At the same time we see a shift towards powerful VM-based techniques that will continue to evolve over the coming years.

    Download PDF (437K)
  • Koji Shima, Hiroshi Doi
    Type: Special Issue of Computer Security Technologies against Sophisticated Cyber Attacks
    Subject area: Secret sharing
    2017 Volume 25 Pages 875-883
    Published: 2017
    Released: September 15, 2017
    JOURNALS FREE ACCESS

    Hierarchical secret sharing schemes are known for the way the secret is shared among a group of participants that is partitioned into levels. We examine these schemes in terms of how easily they delete a secret after it is distributed or namely for cases where the reliability of data deletion depends on deletion of the indispensable participants' share. In this paper, we consider Tassa's idea of using formal derivatives and Birkhoff interpolation so that his method will work well even over finite fields of characteristic 2, then we devise a method for derivatives. As a result, we propose a fast (k, n) hierarchical secret sharing scheme applicable to any level and report the software implementation evaluation. Moreover, taking practical use into consideration, we cover the optimization specialized for a ({1, 3}, n) hierarchical secret sharing scheme.

    Download PDF (209K)
  • Takeshi Miura, Takaaki Kaiga, Takeshi Shibata, Katsubumi Tajima, Hideo ...
    Type: Regular Papers
    Subject area: Body Information
    2017 Volume 25 Pages 884-887
    Published: 2017
    Released: September 15, 2017
    JOURNALS FREE ACCESS

    This paper proposes a method to obtain a low-dimensional feature vector appropriately representing the characteristics of a given motion-capture data stream. The feature vector is derived based on the concept of phase plane analysis. A set of phase plane trajectories are obtained from the temporal variation of the state variables representing the body-segment arrangement. The information on six motion-characteristic properties is extracted from the shapes of the trajectories, and used as the components of a six-dimensional feature vector. The experimental results showed the effectiveness and limitation of the proposed method.

    Download PDF (908K)
  • Kazushige Ouchi
    Type: Special Issue of Ubiquitous Computing Systems (VI)
    2017 Volume 25 Pages 888
    Published: 2017
    Released: October 15, 2017
    JOURNALS FREE ACCESS
    Download PDF (36K)
  • Yasuo Namioka, Akira Masuda, Takuya Maekawa
    Type: Special Issue of Ubiquitous Computing Systems (VI)
    Subject area: Mobile Computing
    2017 Volume 25 Pages 889-900
    Published: 2017
    Released: October 15, 2017
    JOURNALS FREE ACCESS

    A novel unconstrained person identification method is presented in this paper. This method is for tabletop systems which exist not only in daily life but also in working environments such as offices and factories. Recent state-of-the-art ubicomp, computer-vision, and CSCW studies have tried to recognize a user's activities and actions on a table using a ceiling-mounted device that overlooks the table, since we can install the ceiling-mounted device in an environment with limited space such as daily life environments and factory environments. Instead of conventional unconstrained person identification methods, such as face identification, we focus on a user's soft biometrics that can be captured from the ceiling such as the shoulder length, shape of the head, and posture of the back to achieve unconstrained person identification by using a ceiling-mounted depth camera. We achieve robust person identification by combining the soft biometrics within a framework of multiview learning. Multiview learning allows us to deal effectively with data consisting of features from multiple sources with different data distributions, i.e., multiple soft biometrics in our case. Our experimental evaluation revealed that our proposed method achieved high identification accuracy of about 94%.

    Download PDF (1637K)
  • Yasuo Namioka, Daisuke Nakai, Kazuya Ohara, Takuya Maekawa
    Type: Special Issue of Ubiquitous Computing Systems (VI)
    Subject area: User Interfaces and Interactive Systems
    2017 Volume 25 Pages 901-911
    Published: 2017
    Released: October 15, 2017
    JOURNALS FREE ACCESS

    In this paper, we attempt to estimate lead time (duration) of each period of an operation process by a factory worker using a wrist-worn accelerometer. In a factory line production system, a worker repetitively performs predefined operation processes, and the lead time greatly affects productivity of the line production system. Our proposed method automatically finds a frequent sensor data segment as a “motif” that occurs once in each operation period using only prior knowledge about predefined standard lead time of the operation process, and uses the occurrence intervals of the motif to estimate the lead time.

    Download PDF (1615K)
  • Hidetaka Kamigaito, Taro Watanabe, Hiroya Takamura, Manabu Okumura, Ei ...
    Type: Regular Papers
    Subject area: Natural Language Processing
    2017 Volume 25 Pages 912-923
    Published: 2017
    Released: October 15, 2017
    JOURNALS FREE ACCESS

    In hierarchical phrase-based machine translation, a rule table is automatically learned by heuristically extracting synchronous rules from a parallel corpus. As a result, spuriously many rules are extracted which may be composed of various incorrect rules. The larger rule table incurs more disk and memory resources, and sometimes results in lower translation quality. To resolve the problems, we propose a hierarchical back-off model for Hiero grammar, an instance of a synchronous context free grammar (SCFG), on the basis of the hierarchical Pitman-Yor process. The model can generate compact rules and phrase pairs without resorting to any heuristics, because longer rules and phrase pairs are automatically backing off to smaller phrases under SCFG. Inference is efficiently carried out using two-step synchronous parsing of Xiao et al. combined with slice sampling. In our experiments, the proposed model achieved a higher or at least comparable translation quality against a previous Bayesian model on various language pairs: German/French/Spanish/Japanese-English. When compared against heuristic models, our model achieved comparable translation quality on a full size German-English language pair in Europarl v7 corpus with a significantly smaller grammar size; less than 10% of that for heuristic models.

    Download PDF (1008K)
  • Che Huang, Chawanat Nakasan, Kohei Ichikawa, Yasuhiro Watashiba, Hajim ...
    Type: Regular Papers
    Subject area: Special Section on Advanced Computing Systems
    2017 Volume 25 Pages 924-933
    Published: 2017
    Released: October 15, 2017
    JOURNALS FREE ACCESS

    A large amount of data needs to be transferred from one site to another as fast as possible in the computational science fields. To achieve high-speed data transfer, many applications utilize multiple TCP streams. However, since multiple TCP streams of applications are usually routed according to the default IP routing protocol, only a single shortest path among the multiple paths can be utilized for the data transfer. This research proposes a multipath controller that increases the performance of data transfer by leveraging multiple paths simultaneously for parallel TCP streams. For this purpose, we utilize the Software-Defined Networking (SDN) technology and its implementation, OpenFlow. Furthermore, we proposed a method to determine optimal numbers of parallel TCP streams to be assigned for each path according to its own network condition. This paper presents the design and implementation of the proposed system. As a case study, we applied our proposed system on GridFTP and evaluated the performance improvement. The results demonstrate that our proposed system accelerates the data transfer of GridFTP in both a virtual and a real global-scale environment.

    Download PDF (946K)
  • Atsushi Keyaki, Jun Miyazaki, Kenji Hatano
    Type: Regular Paper
    Subject area: Special Section on Databases
    2017 Volume 25 Pages 934-944
    Published: 2017
    Released: November 15, 2017
    JOURNALS FREE ACCESS

    We applied an element-based retrieval approach in this study to improve search accuracy for information retrieval on small screen devices. Finer-grained retrieval units compared with document granules are expected in mobile information retrieval because of limited screen size. An information retrieval system employing information units (iUnits), i.e., text fragments that are relevant to a query and atomic in interpreting information, is one of the solutions to finer-grained retrieval units. An iUnit-based mobile information retrieval task, called MobileClick, assumes two-layered information access where the most important iUnits are arranged in the first layer and detailed iUnits for individual intents are arranged in the second layer. This two-layered information access enabled useful information to be presented with fewer page transitions. We first adopt element-based retrieval to identify elements containing relevant descriptions in a document to tackle MobileClick. The next step is scoring the iUnits where we assigned higher scores to iUnits that are similar to elements with higher scores. Our experimental evaluations demonstrated that our method is more accurate than a baseline by 9.7%.

    Download PDF (701K)
  • Hirotoshi Honma, Yoko Nakajima, Shigeru Masuyama
    Type: Regular Papers
    Subject area: Algorithm Theory
    2017 Volume 25 Pages 945-948
    Published: 2017
    Released: December 15, 2017
    JOURNALS FREE ACCESS

    Let G = (V, E) be a simple connected graph. A vertex uV is called a hinge vertex if there exist two vertices x and y in G whose distance increases when u is removed. Finding all hinge vertices of a given graph is called the hinge vertex problem. This problem can be applied to improve the stability and robustness of communication network systems. In a number of graph problems, it is known that more efficient sequential or parallel algorithms can be developed by restricting classes of graphs. Circular trapezoid graphs are proper super-classes of trapezoid graphs. In this paper, we propose an O(n2) time algorithm for the hinge vertex problem of circular trapezoid graphs.

    Download PDF (357K)
  • Shahidatul Sadiah, Toru Nakanishi, Nasima Begum, Nobuo Funabiki
    Type: Regular Papers
    Subject area: Security Infrastructure
    2017 Volume 25 Pages 949-961
    Published: 2017
    Released: December 15, 2017
    JOURNALS FREE ACCESS

    An anonymous credential system allows a user to convince a service provider anonymously that he/she owns certified attributes. Previously, a system to prove AND and OR relations simultaneously by CNF formulas was proposed. To achieve a constant-size proof of the formula, this system adopts an accumulator that compresses multiple attributes into a single value. However, this system has a problem: the proof generation requires a large amount of computational time in the case of lots of OR literals in the formula. Therefore, we convert the attribute relation from the CNF formula to a monotone formula to decrease the number of OR literals. The monotone formula is a logic formula that contains any combination of AND and OR relations. In this paper, we propose an extended accumulator to prove the monotone formula, and apply it to the anonymous credential system. Our approach to prove the monotone formula is that the tag assignment in the accumulator is extended to be adapted to the tree expressing the monotone formula. Using this type of formula, the number of public parameters multiplied in the accumulator is decreased, which greatly impacts the reduction of authentication time.

    Download PDF (384K)
  • Kazuhiro Abe, Keisuke Nakano
    Type: Regular Papers
    Subject area: Special Section on Programming
    2017 Volume 25 Pages 962-974
    Published: 2017
    Released: December 15, 2017
    JOURNALS FREE ACCESS

    Macro tree transducers (MTTs) and macro forest transducers (MFTs) have been used as good models of tree-structured data transformations such as XML transformations. Typechecking of transformations in these models is performed to verify if any tree of an input type is always transformed into a tree of an output type, which is useful for validating XML transformations against given XML schemata. In typechecking problems for MTTs and MFTs, each “type” is usually given by a tree automaton. A naive implementation of a typechecking algorithm is very inefficient because its time complexity is beyond exponential to the number of states of tree automata, and a large number of equivalence checking operations over finite maps are required. For typechecking of MTTs, Frisch and Hosoya proposed an efficient and practical algorithm by using alternating tree automata as an internal representation of types and reducing the problem to satisfiability checking over first-order logic formulae. In this paper, we extend their typechecking method to apply it to MFTs that are more expressive than MTTs. Our implementation of the proposed method shows that it performs typechecking for relatively simple cases in a reasonable time.

    Download PDF (324K)
  • Nariyoshi Chida, Kimio Kuramitsu
    Type: Regular Papers
    Subject area: Special Section on Programming
    2017 Volume 25 Pages 975-982
    Published: 2017
    Released: December 15, 2017
    JOURNALS FREE ACCESS

    Parsing expression grammars (PEGs) were formalized by Ford in 2004, and have several pragmatic operators (such as ordered choice and unlimited lookahead) for better expressing modern programming language syntax. In addition, PEGs can be parsed in a linear time by using recursive-descent parsing and memoization. In this way, PEGs have a lot of positive aspects. On the other hand, it is known that ordered choices defy intuition. They may cause bugs. This is due to a priority of an ordered choice. To avoid this, unordered choices are required. In this paper, we define a parsing expression grammar with unordered choices (PEGwUC), an extension of a PEG with unordered choices. By the extension, it is expected that a PEGwUC includes both a PEG and a context-free grammar (CFG), and this allows us to write a grammar more intuitively. Furthermore, we show an algorithm for parsing a PEGwUC. The algorithm runs in a linear time when a PEGwUC does not include unordered choice and in a cubic time in worst-case running time.

    Download PDF (882K)
  • Teruo Tanimoto, Takatsugu Ono, Koji Inoue
    Type: Regular Papers
    Subject area: Special Section on Advanced Computing Systems
    2017 Volume 25 Pages 983-992
    Published: 2017
    Released: December 15, 2017
    JOURNALS FREE ACCESS

    The dependence graph model of out-of-order (OoO) instruction execution is a powerful representation used for the critical path analysis. However, most, if not all, of the previous models are out-of-date and lack enough detail to model modern OoO processors, or are too specific and complicated which limit their generality and applicability. In this paper, we propose an enhanced dependence graph model which remains simple but greatly improves the accuracy over prior models. The evaluation results using the gem5 simulator with configurations similar to Intel's Haswell and Silvermont architecture show that the proposed enhanced model achieves CPI errors of 2.1% and 4.4% which are 90.3% and 77.1% improvements from the state-of-the-art model.

    Download PDF (1412K)
feedback
Top