This paper introduces a technique for automatically generating potential training data from sentences in which entity pairs are not apparently presented in a relation extraction. Most previous works on relation extraction by distant supervision ignored cases in which a relationship may be expressed via null-subjects or anaphora. However, natural language text basically has a network structure that is composed of several sentences. If they are closely related, this is not expressed explicitly in the text, which can make relation extraction difficult. This paper describes a new model that augments a paragraph with a “salient entity” that is determined without parsing. The entity can create additional tuple extraction environments as potential subjects in paragraphs. Including the salient entity as part of the sentential input may allow the proposed method to identify relationships that conventional methods cannot identify. This method also has promising potential applicability to languages for which advanced natural language processing tools are lacking.

Computing the partition function of the Ising model on a graph has been investigated from both sides of computer science and statistical physics, with producing fertile results of P cases, FPTAS/FPRAS cases, inapproximability and intractability. Recently, measurement-based quantum computing as well as quantum annealing open up another bridge between two fields by relating a tree tensor network representing a quantum graph state to a rank decomposition of the graph. This paper makes this bridge wider in both directions. An $O^*(2^{ rac{omega}{2} bw(G)})$-time algorithm is developed for the partition function on n-vertex graph G with branch decomposition of width bw(G), where O* ignores a polynomial factor in n and ω is the matrix multiplication parameter less than 2.37287. Related algorithms of $O^*(4^{rw( ilde{G})})$ time for the tree tensor network are given which are of interest in quantum computation, given rank decomposition of a subdivided graph $ ilde{G}$ with width $rw( ilde{G})$. These algorithms are parameter-exponential, i.e., O*(cp) for constant c and parameter p, and such an algorithm is not known for a more general case of computing the Tutte polynomial in terms of bw(G) (the current best time is O*(min{2n, bw(G)O(bw(G))})) with a negative result in terms of the clique-width, related to the rank-width, under ETH.

Multi-slot information extraction, also known as frame extraction, is a task that identify several related entities simultaneously. Most researches on this task are concerned with applying IE patterns (rules) to extract related entities from unstructured documents. An important obstacle for the success in this task is unknowing where text portions containing interested information are. This problem is more complicated when involving languages with sentence boundary ambiguity, e.g. the Thai language. Applying IE rules to all reasonable text portions can degrade the effect of this obstacle, but it raises another problem that is incorrect (unwanted) extractions. This paper aims to present a method for removing these incorrect extractions. In the method, extractions are represented as intuitionistic fuzzy sets, and a similarity measure for IFSs is used to calculate distance between IFS of an unclassified extraction and that of each already-classified extraction. The concept of k nearest neighbor is adopted to design whether the unclassified extraction is correct or not. From the experiment on various domains, the proposed technique improves extraction precision while satisfactorily preserving recall.

The deep neural named entity recognition model automatically learns and extracts the features of entities and solves the problem of the traditional model relying heavily on complex feature engineering and obscure professional knowledge. This issue has become a hot topic in recent years. Existing deep neural models only involve simple character learning and extraction methods, which limit their capability. To further explore the performance of deep neural models, we propose two character feature learning models based on convolution neural network and long short-term memory network. These two models consider the local semantic and position features of word characters. Experiments conducted on the CoNLL-2003 dataset show that the proposed models outperform traditional ones and demonstrate excellent performance.

A logo is a symbolic presentation that is designed not only to identify a product manufacturer but also to attract the attention of shoppers. Shoe logos are a challenging subject for automatic extraction and recognition using image analysis techniques because they have characteristics that distinguish them from those of other products; that is, there is much within-class variation in the appearance of shoe logos. In this paper, we propose an automatic extraction and recognition method for shoe logos with a wide variety of appearance using a limited number of training samples. The proposed method employs maximally stable extremal regions for the initial region extraction, an iterative algorithm for region grouping, and gradient features and a support vector machine for logo recognition. The results of performance evaluation experiments using a logo dataset that consists of a wide variety of appearances show that the proposed method achieves promising performance for both logo extraction and recognition.

In this paper, we demonstrate how to subtract the intertrack interference (ITI) before the decoding process in multi-track multi-head bit-patterned media recording (BPMR) system, which can obtain a better bit error rate (BER) performance. We focus on the three-track/three-head BPMR channel and propose the ITI subtraction technique that performs together with a rate-5/6 two dimensional (2D) modulation code. Since the coded system can provide the estimated recorded bit sequence with a high reliability rate for the center track. However, the upper and lower data sequences still be interfered with their sidetracks, which results to have a low reliability rate. Therefore, we propose to feedback the data from the center and upper tracks for subtracting the ITI effect of the lower track. Meanwhile, the feedback data from the center and lower tracks will be also used to subtract the ITI effect of the upper track. The use of our proposed technique can effectively reduce the severity of ITI effect which caused from the two sidetracks. The computer simulation results in the presence of position and size fluctuations show that the proposed system yields better BER performance than a conventional coded system, especially when an areal density (AD) is ultra high.

As the manual creation of domain models and also of linked data is very costly, the extraction of knowledge from structured and unstructured data has been one of the central research areas in the Semantic Web field in the last two decades. Here, we look specifically at the extraction of formalized knowledge from natural language text, which is the most abundant source of human knowledge available. There are many tools on hand for information and knowledge extraction for English natural language, for written Thai language the situation is different. The goal of this work is to assess the state-of-the-art of research on formal knowledge extraction specifically from Thai language text, and then give suggestions and practical research ideas on how to improve the state-of-the-art. To address the goal, first we distinguish nine knowledge extraction for the Semantic Web tasks defined in literature on knowledge extraction from English text, for example taxonomy extraction, relation extraction, or named entity recognition. For each of the nine tasks, we analyze the publications and tools available for Thai text in the form of a comprehensive literature survey. Additionally to our assessment, we measure the self-assessment by the Thai research community with the help of a questionnaire-based survey on each of the tasks. Furthermore, the structure and size of the Thai community is analyzed using complex literature database queries. Combining all the collected information we finally identify research gaps in knowledge extraction from Thai language. An extensive list of practical research ideas is presented, focusing on concrete suggestions for every knowledge extraction task - which can be implemented and evaluated with reasonable effort. Besides the task-specific hints for improvements of the state-of-the-art, we also include general recommendations on how to raise the efficiency of the respective research community.

This paper introduces a filter level pruning method based on similar feature extraction for compressing and accelerating the convolutional neural networks by k-means++ algorithm. In contrast to other pruning methods, the proposed method would analyze the similarities in recognizing features among filters rather than evaluate the importance of filters to prune the redundant ones. This strategy would be more reasonable and effective. Furthermore, our method does not result in unstructured network. As a result, it needs not extra sparse representation and could be efficiently supported by any off-the-shelf deep learning libraries. Experimental results show that our filter pruning method could reduce the number of parameters and the amount of computational costs in Lenet-5 by a factor of 17.9× with only 0.3% accuracy loss.

Online chat systems, e.g.., Twitter and Slack, have been used in academic conferences or study meetings as a means of instant discussion and sharing related information alongside a real presentation. We propose a system for activating online discussion by providing a bot that suggests webpages related to current timeline of the discussion. Our system generates keyword vectors according to discussion timeline, searches best related webpages from several web sites, and timely provides these pages to the discussion timeline. This paper describes deployments of our system in two types of meetings: lightning talk format meetings and group meetings; and daily exchanges using online chat system. As a result, we could not find good enough reactions to the bot's postings from meeting participants at the lightning talk format meetings, but we could observe more reactions and progress of discussion caused by the bot's postings at the relaxed meetings and daily exchanges among group members.

The existing methods to estimate satellite attitude by using radar cross section (RCS) sequence suffer from problems such as low precision, computation complexity, etc. To overcome these problems, a novel model of satellite attitude estimation by the local maximum points of the RCS sequence is established and can reduce the computational time by downscaling the dimension of the feature vector. Moreover, a particle swarm optimization method is adopted to improve efficiency of computation. Numerical simulations show that the proposed method is robust and efficient.

In this paper, we study a variant of the MINIMUM DOMINATING SET problem. Given an unweighted undirected graph G=(V,E) of n=|V| vertices, the goal of the MINIMUM SINGLE DOMINATING CYCLE problem (MinSDC) is to find a single shortest cycle which dominates all vertices, i.e., a cycle C such that for the set V(C) of vertices in C and the set N(V(C)) of neighbor vertices of C, V(G)=V(C)∪N(V(C)) and |V(C)| is minimum over all dominating cycles in G [6], [17], [24]. In this paper we consider the (in)approximability of MinSDC if input graphs are restricted to some special classes of graphs. We first show that MinSDC is still NP-hard to approximate even when restricted to planar, bipartite, chordal, or r-regular (r≥3). Then, we show the (lnn+1)-approximability and the (1-ε)lnn-inapproximability of MinSDC on split graphs under P≠NP. Furthermore, we explicitly design a linear-time algorithm to solve MinSDC for graphs with bounded treewidth and estimate the hidden constant factor of its running time-bound.

Three-dimensional (3-D) scattering center models use a finite number of point scatterers to efficiently represent complex radar target signature. Using the CLEAN algorithm, 3-D scattering center model is extracted from the inverse synthetic aperture radar (ISAR) image, which is generated based on the shooting and bouncing ray (SBR) technique. The conventional CLEAN extracts the strongest peak iteratively based on the assumption that the scattering centers are isolated. In a realistic target, however, both interference from the closely spaced points and additive noise distort the extraction process. This paper proposes a matched filter-based CLEAN algorithm to improve accuracy efficiently. Using the matched filtering of which impulse response is the known point spread function (PSF), a point most correlated with the PSF is extracted. Thus, the proposed method optimally enhances the accuracy in the presence of massive distortions. Numerical simulations using canonical and realistic targets demonstrate that the extraction accuracy is improved without loss of time-efficiency compared with the existing CLEAN algorithms.

Knowledge graphs (KG) play a crucial role in many modern applications. However, constructing a KG from natural language text is challenging due to the complex structure of the text. Recently, many approaches have been proposed to transform natural language text to triples to obtain KGs. Such approaches have not yet provided efficient results for mapping extracted elements of triples, especially the predicate, to their equivalent elements in a KG. Predicate mapping is essential because it can reduce the heterogeneity of the data and increase the searchability over a KG. In this article, we propose T2KG, an automatic KG creation framework for natural language text, to more effectively map natural language text to predicates. In our framework, a hybrid combination of a rule-based approach and a similarity-based approach is presented for mapping a predicate to its corresponding predicate in a KG. Based on experimental results, the hybrid approach can identify more similar predicate pairs than a baseline method in the predicate mapping task. An experiment on KG creation is also conducted to investigate the performance of the T2KG. The experimental results show that the T2KG also outperforms the baseline in KG creation. Although KG creation is conducted in open domains, in which prior knowledge is not provided, the T2KG still achieves an F1 score of approximately 50% when generating triples in the KG creation task. In addition, an empirical study on knowledge population using various text sources is conducted, and the results indicate the T2KG could be used to obtain knowledge that is not currently available from DBpedia.

Feature visualization is of great significances in volume visualization, and feature extraction has been becoming extremely popular in feature visualization. While precise definition of features is usually absent which makes the extraction difficult. This paper employs probability density function (PDF) as statistical property, and proposes a statistical property guided approach to extract features for volume data. Basing on feature matching, it combines simple liner iterative cluster (SLIC) with Gaussian mixture model (GMM), and could do extraction without accurate feature definition. Further, GMM is paired with a normality test to reduce time cost and storage requirement. We demonstrate its applicability and superiority by successfully applying it on homogeneous and non-homogeneous features.

Knowledge graphs have been shown to be useful to many tasks in artificial intelligence. Triples of knowledge graphs are traditionally structured by human editors or extracted from semi-structured information; however, editing is expensive, and semi-structured information is not common. On the other hand, most such information is stored as text. Hence, it is necessary to develop a method that can extract knowledge from texts and then construct or populate a knowledge graph; this has been attempted in various ways. Currently, there are two approaches to constructing a knowledge graph. One is open information extraction (Open IE), and the other is knowledge graph embedding; however, neither is without problems. Stanford Open IE, the current best such system, requires labeled sentences as training data, and knowledge graph embedding systems require numerous triples. Recently, distributed representations of words have become a hot topic in the field of natural language processing, since this approach does not require labeled data for training. These require only plain text, but Mikolov showed that it can perform well with the word analogy task, answering questions such as, “a is to b as c is to __?.” This can be considered as a knowledge extraction task from a text for finding the missing entity of a triple. However, the accuracy is not sufficiently high when applied in a straightforward manner to relations in knowledge graphs, since the method uses only one triple as a positive example. In this paper, we analyze why distributed representations perform such tasks well; we also propose a new method for extracting knowledge from texts that requires much less annotated data. Experiments show that the proposed method achieves considerable improvement compared with the baseline; in particular, the improvement in HITS@10 was more than doubled for some relations.

In this work, two new proposed NMF models are developed for facial expression recognition. They are called maximum volume constrained nonnegative matrix factorization (MV_NMF) and maximum volume constrained graph nonnegative matrix factorization (MV_GNMF). They achieve sparseness from a larger simplicial cone constraint and the extracted features preserve the topological structure of the original images.

In recent years, deep learning has been widely applied in relation extraction task. The method uses only word embeddings as network input, and can model relations between target named entity pairs. It equally deals with each relation mention, so it cannot effectively extract relations from the corpus with an enormous number of non-relations, which is the main reason why the performance of relation extraction is significantly lower than that of relation classification. This paper designs a deep reinforcement learning framework for relation extraction, which considers relation extraction task as a two-step decision-making game. The method models relation mentions with CNN and Tree-LSTM, which can calculate initial state and transition state for the game respectively. In addition, we can tackle the problem of unbalanced corpus by designing penalty function which can increase the penalties for first-step decision-making errors. Finally, we use Q-Learning algorithm with value function approximation to learn control policy π for the game. This paper sets up a series of experiments in ACE2005 corpus, which show that the deep reinforcement learning framework can achieve state-of-the-art performance in relation extraction task.

Image quality assessment (IQA) plays an important role in quality monitoring, evaluation and optimization for image processing systems. However, current quality-aware feature extraction methods for IQA can hardly balance accuracy and complexity. This paper introduces multi-order local description into image quality assessment for feature extraction. The first-order structure derivative and high-order discriminative information are integrated into local pattern representation to serve as the quality-aware features. Then joint distributions of the local pattern representation are modeled by spatially enhanced histogram. Finally, the image quality degradation is estimated by quantifying the divergence between such distributions of the reference image and those of the distorted image. Experimental results demonstrate that the proposed method outperforms other state-of-the-art approaches in consideration of not only accuracy that is consistent with human subjective evaluation, but also robustness and stability across different distortion types and various public databases. It provides a promising choice for image quality assessment development.

Appearance-based generic object recognition is a challenging problem because all possible appearances of objects cannot be registered, especially as new objects are produced every day. Function of objects, however, has a comparatively small number of prototypes. Therefore, function-based classification of new objects could be a valuable tool for generic object recognition. Object functions are closely related to hand-object interactions during handling of a functional object; i.e., how the hand approaches the object, which parts of the object and contact the hand, and the shape of the hand during interaction. Hand-object interactions are helpful for modeling object functions. However, it is difficult to assign discrete labels to interactions because an object shape and grasping hand-postures intrinsically have continuous variations. To describe these interactions, we propose the interaction descriptor space which is acquired from unlabeled appearances of human hand-object interactions. By using interaction descriptors, we can numerically describe the relation between an object's appearance and its possible interaction with the hand. The model infers the quantitative state of the interaction from the object image alone. It also identifies the parts of objects designed for hand interactions such as grips and handles. We demonstrate that the proposed method can unsupervisedly generate interaction descriptors that make clusters corresponding to interaction types. And also we demonstrate that the model can infer possible hand-object interactions.

This paper first reviews the state-of-the-art noise reduction methods and points out their vulnerability in noise reduction performance and speech quality, especially under the low signal-noise ratios (SNR) environments. Then this paper presents an improved perceptual multiband spectral subtraction (MBSS) noise reduction algorithm (NRA) and a novel robust voice activity detection (VAD) based on the amended sub-band SNR. The proposed SNR-based VAD can considerably increase the accuracy of discrimination between noise and speech frame. The simulation results show that the proposed NRA has better segmental SNR (segSNR) and perceptual evaluation of speech quality (PESQ) performance than other noise reduction algorithms especially under low SNR environments. In addition, a fully operational digital hearing aid chip is designed and fabricated in the 0.13 µm CMOS process based on the proposed NRA. The final chip implementation shows that the whole chip dissipates 1.3 mA at the 1.2 V operation. The acoustic test result shows that the maximum output sound pressure level (OSPL) is 114.6 dB SPL, the equivalent input noise is 5.9 dB SPL, and the total harmonic distortion is 2.5%. So the proposed digital hearing aid chip is a promising candidate for high performance hearing-aid systems.