| GIAN PAOLO TOPICO MAYUGA | 1361204: D, 中間発表 | 井上 美智子，中島 康彦，米田 友和，大和 勇太 |
|
title: Highly Reliable Memory Systems (Fault-Tolerant Design and Reliability Enhancement for ECC-Based Memory Architectures) abstract: Embedded memory is extensively being used in SoCs (System on Chips), and is rapidly growing in size and density. To keep up with the development pace of nanoscale devices, enhancement methods for yield and reliability must overcome the barriers set forth by advent of new technology. To address the issue of reliability, periodic in-field test and repair are implemented by using synergistic approach of employing redundancy and ECC to repair or correct both hard errors and soft errors. In this research, an online remap strategy for memory repair, which ensures ’fresh’ memory words are always used until the spare words run out, is proposed. The improvement of reliability for memory architectures and the area overhead introduced by the proposed scheme is evaluated. Extensions to this work are also introduced. language of the presentation: English | ||
| NACHAI LIMSETTHO | 1361206: D, 中間発表 | 松本 健一，中島 康彦，畑 秀明 |
|
title: Unsupervised Bug Report Categorization using Clustering & Labeling Algorithm abstract: Background: Bug reports are one of the most crucial information sources for software engineering offering answers to many important questions. Yet, getting these answers is not always easy; the information in bug reports is often implicit and some processes are required to extract the meaning of these reports. Most research in this area employ a supervised learning approach to classify bug reports so that required types of reports, such as security bugs and high impact bugs, could be identified. However, this approach often requires an immense amount of time and effort, the resources that already too scarce in many projects. Aims: We aim to develop an automated framework that can categorize bug reports, according to their grammatical structure without the need for labeled data. Method: Our framework categorizes bug reports according to their text similarity using topic modeling and a clustering algorithm. Each group of bug reports are labeled with our new clustering labeling algorithm specifically made for clusters in the topic space. Our framework is highly customizable with a modular approach and options to incorporate available background knowledge to improve its performance, while our cluster labeling approach make use of natural language process (NLP) chunking to create the representative labels. Results: Our experiment results demonstrate that the performance of our unsupervised framework is comparable to a supervised learning one. We also show that our labeling process is capable of labeling each cluster with phrases that are easy for humans to interpret and also representative for that cluster's characteristics. Conclusion: Our framework can be used to automatically categorize the incoming bug reports without any prior knowledge, as an automated labeling suggestion system or as a tool for obtaining knowledge about the structure of the bug report repository. language of the presentation: English | ||
| ZHANG ANNA | 1351213: M, 2回目発表 | 中島 康彦，笠原 正治，高前田 伸也，TRAN THI HONG |
|
title: Lowering the Complexity of k-means Clustering by BFS-dijkstra method for Graph Computing abstract: Currently there are many algorithms working on graph patitioning, a high-efficiency and accurate partition algorithm is required while the data set is large and asks for a well partition result. Our proposal which called K-dijstra, which used high efficiency BFS based on dijstra, has a rather fast speed compared with the traditional graph partitioning. And both the two algorithms have reasonable results that can partition the huge graph data into several groups. Due to the characteristic of traditional graph partitioning, some people prefer to use this traditional method because it use SVD for the second step. So I merge traditional graph partitioning method and K-dijkstra together, the mix-algorithms can keep the second step--SVD and has a much faster operating efficiency. language of the presentation: English | ||
| WITHAWAT TANGTRONGPAIROJ | 1361208: D, 中間発表 | 岡田 実，笠原 正治，東野 武史 |
|
title: Quality-based Channel Allocation Scheme with Predistortion in Multi-channel Radio-over-Fiber System abstract: Radio over Fiber (RoF) is a promising solution for providing wireless access services. Heterogeneous radio signals are transferring via optical fiber link with using analog transmission technique. When RoF and radio frequency (RF) devices have nonlinear characteristics, these create intermodulation products (IMPs) in the system and generate intermodulation distortion (IMD). In this paper, the IMD interference in the uplink RF signals from the coupling effect between the downlink and uplink antennas in remote antenna unit (RAU) has been newly addressed. We propose that dynamic channel allocation (DCA) algorithm with predistortion (PD) technique is applied to multi-channel RoF system to improve the throughput performance. The carrier to distortion plus noise power ratio (CDNR) is evaluated for all channel allocation combinations. The best channel combination is then assigned as a set of active channels to minimize the effect of IMD. The result shows the DCA with PD, which has the lowest IMD, can perform well throughput performance. language of the presentation: English | ||
| 石井 将大 | 1361001: D, 中間発表 | 藤川 和利，山口 英，楫 勇一，猪俣 敦夫 |
|
title: Pairings over genus 2 hyperelliptic curves at high security levels
abstract: Bilinear pairings on (hyper)elliptic curves have been applied into many cryptographic applications such as functinal encryption including identity-based encryption and attribute-based encryption. The security of pairings is based on the difficulty of discrete logarithm problems (DLP) in the subgroup of Jacobian of the curve and the finite field. Recently, the theoretical and practical advancements of the efficient DLP algorithm in finite fields have been made. Especially, there were major breakthroughs in computing DL in finite fields of small characteristics, as a result the symmpetric pairings which is defined by using such finite fields became unsuitable for cryptography. In this paper, we focus on the hyperelliptic curves of genus 2 and the optimal pairing algorithms at high (192-bit) security level on such curves. We first propose the family of pairing-friendly curves of genus 2 by using the existing construction methods, and we provide the optimal pairing algorithm for each curve. Furthermore, we propose the super-optimal pairing on the genus 2 curve to provide a secure and fast pairing on genus 2 curve. language of the presentation: Japanese | ||
| DO QUOC TRUONG | 1351206: M, 2回目発表 | 中村 哲，松本 裕治，戸田 智基，SAKRIANI SAKTI，GRAHAM NEUBIG |
|
title: Preserving Word-level Emphasis in Speech-to-speech Translation
abstract: In speech, emphasis is an important type of paralinguistic information that helps convey the focus of an utterance, new information, and emotion. If emphasis can be incorporated into a speech-to-speech (S2S) translation system, it will be possible to convey this information across the language barrier. However, previous related work focuses only on the translation of particular prosodic features, such as F0 or works with emphasis but focuses on extremely small vocabularies, such as the 10 digits. In this paper, we describe a new S2S method that is able to translate the emphasis across languages and consider multiple features of emphasis such as power, F0, and duration over larger vocabularies. We do so by introducing two new components: word-level emphasis estimation using linear regression hidden semi-Markov models, and emphasis translation that translates the word-level emphasis to the target language with conditional random fields. The text-to-speech synthesis system is also modified to be able to synthesize emphasized speech. The result shows that our system can translate the emphasis correctly with 91.6% F-measure for objective test, and 87.8% for subjective test. language of the presentation: English | ||
| PHILIP ARTHUR | 1351210: M, 2回目発表 | 中村 哲，松本 裕治，戸田 智基，GRAHAM NEUBIG，SAKRIANI SAKTI |
|
title: Semantic Parsing via Paraphrasing and Verification
abstract: We propose a new method for semantic parsing of ambiguous and ungrammatical input, such as search queries. We do so by building on an existing semantic parsing framework that uses synchronous context free grammars (SCFG) to jointly model the input sentence and output meaning representation. We generalize this SCFG framework to allow not one, but multiple outputs. Using this formalism, we construct a grammar that takes an ambiguous input string and jointly maps it into both a meaning representation and a natural language paraphrase that is less ambiguous than the original input. This paraphrase can be used to disambiguate the meaning representation via verification using a language model that calculates the probability of each paraphrase. language of the presentation: English. | ||
| LIANG JUN | 1351208: M, 2回目発表 | 松本 裕治，中村 哲，新保 仁，KEVIN DUH |
|
title: Extracting Bilingual Multi-word Terms from Comparable Corpora
abstract: Our proposed approach focus on bilingual multi-word terms extraction using hybrid method combination the topic model based appraoch with multi-word term extraction, and then extracting multi-word term pairs using a word-alignment method. The key points of the combination are as follows: Firstly, we use multi-word term extraction method to extract the candidate multi-word terms from comparable corpora. Secondly, we use multi-lingual topic model to extract semantic clusters of the large-scale comparable corpora. Our approach does not require any prior knowledge; it improves the precision of multi-word term pairs and also proves the topic model can be used to deal with multi-word term extraction problem. language of the presentation: English | ||
| 秋間 大輔 | 1351003: M, 2回目発表 | 松本 裕治，中村 哲，新保 仁，進藤 裕之 |
|
title: The Ordering of Cross-Document relations with Time-Series information abstract: In speech, we propose a new processing method for summarization about one topic that is obtained in multi-documents.In documents describe some events is occurred for a specific topic, like news articles, is characterized:Creation date of the article does not match occurrence dates of sentence events, and article contains same description in common with the other documents.In performing a summary of topics from the set of documents, we need for estimating the sequence of events; Aggregate the description of same events, and reveal the time-series of events are extracted from articles. We propose to realize the method by estimating Cross-Document relations with time-series. language of the presentation: Japanese | ||