The extraction of verb synonyms is a key technology to build a verb dictionary as a language resource. This paper presents a co-clustering-based verb synonym extraction approach that increases the number of extracted meanings of polysemous verbs from a large text corpus. For verb synonym extraction with a clustering approach dealing with polysemous verbs can be one problem issue because each polysemous verb should be categorized into different clusters depending on each meaning; thus there is a high possibility of failing to extract some of the meanings of polysemous verbs. Our proposed approach can extract the different meanings of polysemous verbs by recursively eliminating the extracted clusters from the initial data set. The experimental results of verb synonym extraction show that the proposed approach increases the correct verb clusters by about 50% with a 0.9% increase in precision and a 1.5% increase in recall over the previous approach.

While classification techniques can be applied for automatic unknown word recognition in a language without word boundary, it faces with the problem of unbalanced datasets where the number of positive unknown word candidates is dominantly smaller than that of negative candidates. To solve this problem, this paper presents a corpus-based approach that introduces a so-called group-based ranking evaluation technique into ensemble learning in order to generate a sequence of classification models that later collaborate to select the most probable unknown word from multiple candidates. Given a classification model, the group-based ranking evaluation (GRE) is applied to construct a training dataset for learning the succeeding model, by weighing each of its candidates according to their ranks and correctness when the candidates of an unknown word are considered as one group. A number of experiments have been conducted on a large Thai medical text to evaluate performance of the proposed group-based ranking evaluation approach, namely V-GRE, compared to the conventional naive Bayes classifier and our vanilla version without ensemble learning. As the result, the proposed method achieves an accuracy of 90.930.50% when the first rank is selected while it gains 97.260.26% when the top-ten candidates are considered, that is 8.45% and 6.79% improvement over the conventional record-based naive Bayes classifier and the vanilla version. Another result on applying only best features show 93.930.22% and up to 98.85 0.15% accuracy for top-1 and top-10, respectively. They are 3.97% and 9.78% improvement over naive Bayes and the vanilla version. Finally, an error analysis is given.

Initial growth of GaP on Si substrates using metalorganic vapor phase epitaxy was studied. Si substrates were exposed to PH3 preflow for 15 s or 120 s at 830 after they were preheated at 925. Atomic force microscopy (AFM) revealed that the Si surface after preflow for 120 s was much rougher than that after preflow for 15 s. After 1.5 nm GaP deposition on the Si substrates at 830, GaP islands nucleated more uniformly on the Si substrate after preflow for 15 s than on the substrate after preflow for 120 s. After 3 nm GaP deposition, layer structures were observed on a fraction of Si surface after preflow for 15 s. Island-like structures remained on the Si surface after preflow for 120 s. After 6 nm GaP deposition, the continuity of GaP layers improved on both substrates. However, AFM shows pits that penetrated a Si substrate with preflow for 120 s. Transmission electron microscopy of a GaP layer on the Si substrate after preflow for 120 s revealed that V-shaped pits penetrated the Si substrate. The preflow for a long time roughened the Si surface, which facilitated the pit formation during GaP growth in addition to degrading the surface morphology of GaP at the initial growth stage. Even after 50 nm GaP deposition, pits with a density on the order of 107 cm-2 remained in the sample. A 50-nm-thick flat GaP surface without pits was achieved for the sample with PH3 preflow for 15 s. The PH3 short preflow is necessary to produce a flat GaP surface on a Si substrate.

A precision in thickness recycling modus for a displays' color filter surface using a gear-form cathode in microelectrochemical removal is developed in the study. Through the precise removal processes of optical materials of nanostructure of Indium-Tin-Oxide crystallization, the optoelectronic semiconductor industry can effectively recycle defective products, and reducing production costs.

This letter proposes a new hybrid EM wave absorber with the crossed-wedge shape, which can be applied to 3 m semi anechoic chambers. In this study, we designed a new hybrid EM wave absorber with the crossed-wedge shape, which consisted of the inorganic and organic thin corrugated dielectric lossy sheet containing organic conductive fibers. Then the 3 m semi anechoic chamber is constructed in the size of 9.0 m6.0 m5.7 m (LWH) using these absorbers, and also the normalized site attenuation (NSA) is measured according to ANSI C63.4 in the frequency range of 30 MHz to 1 GHz. As a result, the measured NSA is obtained within 3 dB of the theoretical one.

In the modern day, MPEG-4 image compression technique have been commonly applied in various indoor wireless communication systems. The efficient system design mostly relies on the joint source channel coding algorithms, which aim to reduce the complexity of channel coding process, while maintaining the quality of the receiving images. In this paper, we design the MAP source-controlled channel decoders with both random and semirandom interleavers for Rician slow flat block-fading channels. The MAP-Viterbi decoder employs the residual redundancy from zerotree symbol sequences of MPEG-4 HFS packets. The interleaving processes are done after the overall channel coding process to combat the block-fading effects. The computer simulations summarize the system performance in terms of average WER and PSNR (dB). With the interleavers, the significant improvement in PSNR of about 15-17 dB is obtained for both ML and MAP decoding. Also in many cases, we obtain more improvement of about 0.2-0.4 dB for using MAP decoding with the interleavers.

A semi-distributed resource allocation scheme based on multihop equilibrium is proposed for OFDM-relay networks. This method aims to reduce the amount of feedback information from the relay nodes (RNs). Moreover, it utilizes radio resource by striking an efficient balance between the capacities of the BS-RN link and RN-MS link. Simulation results show that the proposed semi-distributed scheme achieves good performances in terms of throughputs and fraction of satisfied users.

In this study we introduce AdjScales, a method for scaling similar adjectives by their strength. It combines existing Web-based computational linguistic techniques in order to automatically differentiate between similar adjectives that describe the same property by strength. Though this kind of information is rarely present in most of the lexical resources and dictionaries, it may be useful for language learners that try to distinguish between similar words. Additionally, learners might gain from a simple visualization of these differences using unidimensional scales. The method is evaluated by comparison with annotation on a subset of adjectives from WordNet by four native English speakers. It is also compared against two non-native speakers of English. The collected annotation is an interesting resource in its own right. This work is a first step toward automatic differentiation of meaning between similar words for language learners. AdjScales can be useful for lexical resource enhancement.

Multi-source broadcasting is one of the information dissemination problems on communication networks such that some units disseminate distinct messages to all other units. In this paper, we study multi-source broadcasting on the de Bruijn and Kautz digraphs which are the models of interconnection networks. In [8] and [12], a cycle-rooted tree which has a large root-cycle is constructed by composition of isomorphic factors, and the multi-source broadcasting is executed on the cycle-rooted tree. On the other side, we execute multi-source broadcasting on each isomorphic factors at the same time. We present a method for multi-source broadcasting using isomorphic cycle-rooted trees which factorize these digraphs, and investigate its efficiency.

We have developed InGaAs-based VCSELs operating around 1.1 µm for high-speed optical interconnections. By applying GaAsP barrier layers, temperature characteristics were considerably improved compared to GaAs barrier layers. As a result, 25 Gbps 100 error-free operation was achieved. These devices also exhibited high reliability. No degradation was observed over 3,000 hours under operation temperature of 150 and current density of 19 kA/cm2. We also developed VCSELs with tunnel junctions for higher speed operation. High modulation bandwidth of 24 GHz and a relaxation oscillation frequency of 27 GHz were achieved. 40 Gbps error-free operation was also demonstrated.

In this paper, we solve the call admission control (CAC) and routing problem in an integrated network that handles several classes of calls of different values and with different resource requirements. The problem of maximizing the average reward (or cost) of admitted calls per unit time is naturally formulated as a semi-Markov Decision Process (SMDP) problem, but is too complex to allow for an exact solution. Thus in this paper, a policy gradient algorithm, together with a decomposition approach, is proposed to find the dynamic (state-dependent) optimal CAC and routing policy among a parameterized policy space. To implement that gradient algorithm, we approximate the gradient of the average reward. Then, we present a simulation-based algorithm to estimate the approximate gradient of the average reward (called GSMDP algorithm), using only a single sample path of the underlying Markov chain for the SMDP of CAC and routing problem. The algorithm enhances performance in terms of convergence speed, rejection probability, robustness to the changing arrival statistics and an overall received average revenue. The experimental simulations will compare our method's performance with other existing methods and show the robustness of our method.

Dimensionality reduction is one of the important preprocessing steps in practical pattern recognition. SEmi-supervised Local Fisher discriminant analysis (SELF)--which is a semi-supervised and local extension of Fisher discriminant analysis--was shown to work excellently in experiments. However, when data dimensionality is very high, a naive use of SELF is prohibitive due to high computational costs and large memory requirement. In this paper, we introduce computational tricks for making SELF applicable to large-scale problems.

Pairing-based cryptography provides us many novel cryptographic applications such as ID-based cryptosystems and efficient broadcast encryptions. The security problems in ubiquitous sensor networks have been discussed in many papers, and pairing-based cryptography is a crucial technique to solve them. Due to the limited resources in the current sensor node, it is challenged to optimize the implementation of pairings on sensor nodes. In this paper we present an efficient implementation of pairing over MICAz, which is widely used as a sensor node for ubiquitous sensor network. We improved the speed of ηT pairing by using a new efficient multiplication specialized for ATmega128L, called the block comb method and several optimization techniques to save the number of data load/store operations. The timing of ηT pairing over GF(2239) achieves about 1.93 sec, which is the fastest implementation of pairing over MICAz to the best of our knowledge. From our dramatic improvement, we now have much high possibility to make pairing-based cryptography for ubiquitous sensor networks practical.

One of the most convenient ways to query XML data is a keyword search because it does not require any knowledge of XML structure or learning a new user interface. However, the keyword search is ambiguous. The users may use different terms to search for the same information. Furthermore, it is difficult for a system to decide which node is likely to be chosen as a return node and how much information should be included in the result. To address these challenges, we propose an XML semantic search based on keywords called XSemantic. On the one hand, we give three definitions to complete in terms of semantics. Firstly, the semantic term expansion, our system is robust from the ambiguous keywords by using the domain ontology. Secondly, to return semantic meaningful answers, we automatically infer the return information from the user queries and take advantage of the shortest path to return meaningful connections between keywords. Thirdly, we present the semantic ranking that reflects the degree of similarity as well as the semantic relationship so that the search results with the higher relevance are presented to the users first. On the other hand, in the LCA and the proximity search approaches, we investigated the problem of information included in the search results. Therefore, we introduce the notion of the Lowest Common Element Ancestor (LCEA) and define our simple rule without any requirement on the schema information such as the DTD or XML Schema. The first experiment indicated that XSemantic not only properly infers the return information but also generates compact meaningful results. Additionally, the benefits of our proposed semantics are demonstrated by the second experiment.

Previous approaches to question retrieval in community-based question answering rely on statistical translation techniques to match users' questions (queries) against collections of previously asked questions. This paper presents a simple but effective method for computing word relatedness to improve question retrieval based on word co-occurrence information directly extracted from question and answer archives. Experimental results show that the proposed approach significantly outperforms translation-based approaches.

Resource search is a fundamental problem in large-scale and highly dynamic Peer-to-Peer (P2P) systems. Unstructured search approaches are widely used because of their flexibility and robustness. However, such approaches incur high communication cost. The index-dissemination-based search is a kind of efficient unstructured search approach. We investigate such approaches with respect to minimize the system communication cost. Based on a dynamic system model that peers continuously leave and join, we solve two problems. One problem is how to efficiently disseminate and maintain a given number of indices. Another is to determine the optimal number of indices for each resource object of a given popularity. Finally, we propose an optimized index dissemination scheme which is fully decentralized and self-adaptive. A remarkable advantage is that the scheme yields no additional communication cost to achieve the self-adaptive feature.

We propose a gap-less optical interconnection between BGA package and board for practical on-board, chip-to-chip optical interconnection. The optical interconnect consists of polymer optical waveguides, an integral mirror on the PWB (printed wiring board), an optical via hole through package, and a connection structure and method requiring no alignment process. Optical waveguide, mirror, waveguide extensions and alignment studs were fabricated on the PWB as horizontal optical interconnect. Coaxial structured optical vias with core and cladding were formed through the package and with precise holes for alignment. Two packages were attached onto the PWB using standard BGA technology utilizing passive optical alignment. The optical characteristics and 10 Gbit/s open-eye diagram were measured. A completely gap-less three dimensional optical interconnect between package-PWB-package was demonstrated.

Assume that 2m red points and 2n blue points are given on the lattice Z2 in the plane R2. We show that if they are in general position, that is, if at most one point lies on each vertical line and horizontal line, then there exists a rectangular cut that bisects both red points and blue points. Moreover, if they are not in general position, namely if some vertical and horizontal lines may contain more than one point, then there exists a semi-rectangular cut that bisects both red points and blue points. We also show that these results are best possible in some sense. Moreover, our proof gives O(N log N), N=2m+2n, time algorithm for finding the desired cut.

We identify a broadband equivalent circuit of an on-chip self-complementary antenna integrated with a µm-sized semiconductor mesa structure whose circuit elements can be interpreted by using closed-form analysis. Prior to the equivalent circuit analysis, an electromagnetic simulation is done to investigate frequency independency of the input impedance for the integrated self-complementary antenna in terahertz range.

An expectation for more intelligent Web is recently being reflected through the new research field called Semantic Web. In this paper, related with Semantic Web security, we introduce an RDF triple based access control model having explicit authorization propagation by inheritance and implicit authorization propagation by inference. Especially, we explain an authorization conflict problem between the explicit and the implicit authorization propagation, which is an important concept in access control for Semantic Web. We also propose a novel conflict detection algorithm using graph labeling techniques in order to efficiently find authorization conflicts. Some experimental results show that the proposed detection algorithm has much better performance than the existing detection algorithm when data size and number of specified authorizations become larger.