Search

Abstract We consider incorporating topic information into message-response matching to boost responses with rich content in retrieval-based chatbots. To this end, we propose a topic aware attentive recurrent neural network in which representations of the message and the response are enhanced by the topic information. The model first leverages the message and the response represented by recurrent neural networks (RNNs) to weight topic words given by a pre-trained LDA model and forms topic vectors as linear combinations of the topic words. It then refines the representations of the message and the response with the topic vectors through an attention mechanism. The attention mechanism weights the hidden sequences of the message and the response not only by themselves but also by their topic vectors. Thus both the parts that are important to matching and the parts that are semantically related to the topics are highlighted in the representations.Empirical studies on public data and human annotated data show that our model can significantly outperform state-of-the-art methods and rank more responses with rich content in high positions. [ABSTRACT FROM AUTHOR]

• Push-out tests were performed on timber-concrete composite (TCC) shear connections with glued-in rod (GIR). • Most of the TCC shear connections with GIR mainly failed by the double-hinge yield mode of GIR. • Both the test configurations and loading rates showed influence on the shear performance of TCC shear connectors. • Calculation modes for shear capacity and shear stiffness were proposed according to typical failure modes of specimens. To examine the effect of test configurations and loading protocols on the performance of timber-concrete composite (TCC) shear connectors, a total of twenty-seven TCC shear connections with glued – in threaded steel rod (GIR) were designed to experience push-out tests. The experimental variables include rod diameter, test configurations (single - and double - shear), and loading protocols including the loading protocol (EN26891) and the displacement rate (0.5 mm/min, 2 mm/min, and 4 mm/min). The results indicate that both the test configurations and the displacement rates showed different influences on the performance of the shear connections. Most of the tested TCC shear connections with GIR mainly failed by the double – hinge yield mode of GIR. Double – shear specimens with 16 mm – diameter GIR exhibited a 49 % higher maximum load compared to the single – shear specimens. The slip modulus of the specimens tested at a displacement rate of 4 mm/min was about 40 % lower than that of the specimens tested at the displacement rates of 0.5 mm/min and 2 mm/min, respectively. Among the single-shear specimens, the 16 mm– diameter GIR showed the lowest ductility, by 65 % and 59 %, as compared to the 10 mm – and 12 mm – diameter specimens. Finally, theoretical analytical formulas were proposed to predict the load carrying capacity and the slip modulus of the TCC shear connections with GIR. The experimental slip modulus was also compared with the results obtained by existing calculation models. [ABSTRACT FROM AUTHOR]

Lattice-based cryptosystems have recently acquired much importance. In this work, we construct a fuzzy identity based signature (FIBS) scheme based on the Small Integer Solution (SIS) Problem. FIBS schemes allow a user with identity id to issue a signature which could be verified under identity id′ if and only if id and id′ are close to each other. To our best knowledge, no lattice based FIBS schemes were known before, and the existing security model of FIBS schemes is not correct indeed. We propose a modified security model and prove that our scheme is existentially unforgetable against adaptively chosen message and identity attacks in the random oracle model. To break the bottleneck of designing lattice-based FIBS scheme, the secret key of each identity bit is generated by employing the Bonsai Tree techniques in the fuzzy extract algorithm. We also use some techniques to prove its security. Then we show the performance comparisons of all existing FIBS schemes. Finally, we give its application in biometric authentication. [ABSTRACT FROM AUTHOR]

Abstract: In the traditional identity-based cryptography, when a user holds multiple identities as its public keys, it has to manage an equal number of private keys. The recent advances of identity-based cryptography allow a single private key to map multiple public keys (identities) that are selectable by the user. This approach simplifies the private key management. Unfortunately, the existing schemes have a heavy computation overhead, since the private key generator has to authenticate all identities in order to generate a resultant private key. In particular, it has been considered as a drawback that the data size for a user is proportional to the number of associated identities. Moreover, these schemes do not allow dynamic changes of user identities. When a user upgrades its identities, the private key generator (PKG) has to authenticate the identities and generate a new private key. To overcome these problems, in this paper we present an efficient dynamic identity-based key exchange protocol with selectable identities, and prove its security under the bilinear Diffie–Hellman assumption in the random oracle model. [Copyright &y& Elsevier]

Abstract With the popularity of social network, social media data embedding has attracted extensive research interest and boomed many applications, such as image classification and cross-modal retrieval. In this paper, we examine the scenario of social images containing multimodal content (e.g., visual content and textual tags) and connecting with each other (e.g., two images submitted to the same group). In such a case, both the multimodal content and link information provide useful clues for representation learning. Therefore, simply learning the embedding from network structure or data content results in sub-optimal social image representation. In this paper, we propose a Deep Multimodal Attention Networks (DMAN) to combine multimodal content and link information for social image embedding. Specifically, to effectively incorporate the multimodal content, a visual-textual attention model is proposed to encode the fine-granularity correlation between multimodal content, i.e., the alignment between image regions and textual words. To incorporate the network structure for embedding learning, a novel Siamese-Triplet neural network is proposed to model the first-order proximity and the second-order proximity among images. Then the two modules are integrated into a joint deep model for social image embedding. Once the representation has been learned, a wide variety of data mining problems can be solved by using the task-specific algorithms designed for handling vector representations. Extensive experiments are conducted to demonstrate the effectiveness of our approach on multi-label classification and cross-modal search. [ABSTRACT FROM AUTHOR]

Power grid is a complex system which closely links the power generation and power consumer through transmission and distribution networks. With the development of smart grid, smart grid is more open to external communication systems, it also has exposed some problems in the network attacks. A new false data injection attack (called the unobservable attack) that can bypass the traditional BDD and inject random errors into state estimation. We propose an improved extreme learning machine (ELM) for attack detection. The artificial bee colony (ABC) incorporates the thought of differential evolution algorithm (DE) to optimize ELM for improving detection precision. In this paper, Autoencoder is used to reduce the dimensionality of the measurement data, which makes the low-dimensional data information basically and fully represent high-dimensional data. We verify the performance of the proposed method on IEEE bus systems, and prove that the proposed method can effectively detect such unobservable attack. [ABSTRACT FROM AUTHOR]

Named entity disambiguation (NED) refers to the task of mapping entity mentions in running texts to the correct entries in a specific knowledge base (e.g., Wikipedia). Although there has been a lot of work on NED for long and formal texts like Wikipedia and news, the task is not well studied for questions in community question answering (CQA). The challenges of the task include little context for mentions in questions, lack of ground truth for learning, and language gaps between CQA and knowledge bases. To overcome these problems, we propose a topic modelling approach to NED for questions. Our model performs learning in an unsupervised manner, but can take advantage of weak supervision signals estimated from the metadata of CQA and knowledge bases. The signals can enrich the context of mentions in questions, and bridge the language gaps between CQA and knowledge bases. Besides these advantages, our model simulates people’s behavior in CQA and thus is intuitively interpretable. We conduct experiments on both Chinese and English CQA data. The experimental results show that our method can significantly outperform state-of-the-art methods when we apply them to questions in CQA. [ABSTRACT FROM AUTHOR]

Abstract: Social image tagging is becoming increasingly popular with the development of social website, where images are annotated with arbitrary keywords called tags. Most of present image tagging approaches are mainly based on the visual similarity or mapping between visual feature and tags. However, in the social media environment, images are always associated with multi-type of object information (i.e., visual content, tags, and user contact information) which makes this task more challenging. In this paper, we propose to fuse multi-type of information to tag social image. Specifically, we model social image tagging as a “ranking and reinforcement” problem, and a novel graph-based reinforcement algorithm for interrelated multi-type objects is proposed. When a user issue a tagging request for a query image, a candidate tag set is derived and a set of friends of the query user is selected. Then a graph which contains three types of objects (i.e., visual features of the query image, candidate tags, and friend users) is constructed, and each type of objects are initially ranked based on their weight and intra-relation. Finally, candidate tags are re-ranked by our graph-based reinforcement algorithm which takes into consideration both inter-relation with visual features and friend users, and the top ranked tags are saved. Experiments on real-life dataset demonstrate that our algorithm significantly outperforms state-of-the-art algorithms. [Copyright &y& Elsevier]

Network embedding, as a promising way of node representation learning, is capable of supporting various downstream network mining tasks, and has attracted growing research interests recently. Existing approaches mostly focus on learning the low-dimensional node representations by preserving the local or global topology information of a static network. It is difficult for such methods to learn desirable features for the nodes in an incomplete graph whose topology information is sparse or the new nodes in a dynamic graph. It is also challenging for them to deeply incorporate node attributes as the complementary information to improve the network embedding performance. To this end, in this paper we propose a Multi-View Adversarial learning based Network Embedding model named MVANE to deeply fuse the network topology information and node attributes to better perform network embedding on incomplete graphs. The insight is that the network topology and the node attributes are treated as two correlated views. The learned embedding vector of a node should be able to reveal its unique characteristics in both views. Specifically, the adversarial autoencoder is introduced as the basic model of MVANE. Autoencoder can learn a projection function to directly map the input feature vectors into the latent space, which ensures the MVANE learn embeddings for the new nodes through features projection without the need of retraining the model. Meanwhile, the adversarial learning strategy is also applied to better capture the cross-view correlations. The idea is that the learned embeddings in one view can not only reconstruct the inputs in this view, but also generate the features in another view. Under a unified learning framework, the latent representations in different views are fused and jointly reinforced by the proposed self/cross-view learning model. Empirically, we evaluate MVANE over multiple network datasets, and the results demonstrate the superiority of our proposal. • We investigate the novel problem of network embedding on incomplete graphs. • We propose an adversarial learning based network embedding model MVANE. • The experimental results demonstrate the superiority of the proposed model. [ABSTRACT FROM AUTHOR]

Social network has become an inseparable part of our daily lives and thus the automatic sentiment analysis on social media content is of great significance to identify people's viewpoints, attitudes, and emotions on the social websites. Most existing works have concentrated on the sentiment analysis of single modality such as image or text, which cannot handle the social media content with multiple modalities including both image and text. Although some works tried to conduct multi-modal sentiment analysis, the complicated correlations between the two modalities have not been fully explored. In this paper, we propose a novel Bi-Directional Multi-Level Attention (BDMLA) model to exploit the complementary and comprehensive information between the image modality and text modality for joint visual-textual sentiment classification. Specifically, to highlight the emotional regions and words in the image–text pair, visual attention network and semantic attention network are proposed respectively. The visual attention network makes region features of the image interact with multiple semantic levels of text (word, phrase, and sentence) to obtain the attended visual features. The semantic attention network makes semantic features of the text interact with multiple visual levels of image (global and local) to obtain the attended semantic features. Then, the attended visual and semantic features from the two attention networks are unified into a holistic framework to conduct visual-textual sentiment classification. Proof-of-concept experiments conducted on three real-world datasets verify the effectiveness of our model. • Bi-directional attention to highlight the emotional regions and words. • Multiple levels to excavate the emotional correlations between image and text. • The experimental results demonstrate the superiority of the proposed model. [ABSTRACT FROM AUTHOR]

Abstract Embedding the network into a low-dimensional space has attracted extensive research interest as well as boomed a lot of applications, such as node classification and link prediction. Most existing methods learn the network embedding simply from the network structure. However, the social media data, such as social images, usually contain both multimodal contents (e.g., visual content and text description) and social links among the images. To address this problem, we propose a novel model Attention-based Multi-view Variational Auto-Encoder (AMVAE) to fuse both the links and the multimodal contents for more effectively and efficiently network embedding. Specifically, Bi-LSTM (bidirectional long short-term memory) with attention model is proposed to capture the fine-granularity correlation between different data modalities, such as some words are reflected by specific visual regions. A joint representation of the multimodal contents is accordingly learned. Then, the network structure information and the learned representation for the multimodal contents are considered as two views. To fuse the two views, a multi-view correlation learning based Variational Auto-Encoder (VAE) is proposed to learn the representation of each node. By jointly optimizing the two components into a holistic learning framework, the embedding of network structure and multimodal contents are integrated and mutually reinforced. Experiments on three real-world datasets demonstrate the superiority of the proposed model in two applications, i.e., multi-label classification and link prediction. [ABSTRACT FROM AUTHOR]

Abstract Sentiment analysis of social media data is crucial to understand people's position, attitude, and opinion toward a certain event, which has many applications such as election prediction and product evaluation. Though great effort has been devoted to the single modality (image or text), less effort is paid to the joint analysis of multimodal data in social media. Most of the existing methods for multimodal sentiment analysis simply combine different data modalities, which results in dissatisfying performance on sentiment classification. In this paper, we propose a novel image–text sentiment analysis model, i.e., Deep Multimodal Attentive Fusion (DMAF), to exploit the discriminative features and the internal correlation between visual and semantic contents with a mixed fusion framework for sentiment analysis. Specifically, to automatically focus on discriminative regions and important words which are most related to the sentiment, two separate unimodal attention models are proposed to learn effective emotion classifiers for visual and textual modality respectively. Then, an intermediate fusion-based multimodal attention model is proposed to exploit the internal correlation between visual and textual features for joint sentiment classification. Finally, a late fusion scheme is applied to combine the three attention models for sentiment prediction. Extensive experiments are conducted to demonstrate the effectiveness of our approach on both weakly labeled and manually labeled datasets. [ABSTRACT FROM AUTHOR]

Recently, a large number of geo-tagged landmark images have been uploaded through various social media services. Usually, these geo-tagged images are annotated by users with GPS and tags related to the landmarks where they are taken. Landmark tagging aims to automatically annotate an image with the tags to describe the landmark where the image is taken. It has been observed that the images and tags show strong correlation with the geographical locations. The widely used assumption by many existing tagging methods is that images are independently and identically distributed is not effective to capture the geographical correlation. In this paper, we study the novel problem of utilizing the geographical correlation among images and landmarks for better tagging landmark images. In particular, we propose an unsupervised feature learning approach to learn the geographically discriminative features across geographical locations, by integrating latent space learning and geographically structural analysis (LSGSA) into a joint model. A latent space learning model is proposed to effectively fuse the heterogeneous features of visual content and tags. Meanwhile, the geographical structure analysis and group sparsity are applied to learn the geographically discriminative features. Then, a geo-guided sparse reconstruction method is proposed to tag images by utilizing the discriminative information of features, in which the landmark-specific tags are boosted by a weighting method. Experiments on the real-world datasets demonstrate the superiority of our approach. [ABSTRACT FROM AUTHOR]

This paper studies response selection for human-computer conversation systems. Existing retrieval-based human-computer conversation systems are intended to reply to user utterances based on existing utterance-response pairs. However, collecting sufficient utterance-response pairs is intractable in practical situations, especially for many specific domains. We introduce DocChat a novel information retrieval approach for human-computer conversation systems that can use unstructured documents rather than semi-structured utterance-response pairs, to react to user utterances. The key of DocChat is a learning to rank model with features designed at various levels of granularity which is proposed to quantify the relevance between utterances and responses directly. We conduct comprehensive experiments on both sentence selection and real human-computer conversation scenarios. Empirical studies of sentence selection datasets shows reasonable improvements and the strong adaptability of our model. We compare DocChat with Xiaoice, a famous open domain chitchat engine in China. Side-by-side evaluation shows that DocChat is a good complement for human-computer conversation systems using utterance-response pairs as the primary source of responses. Furthermore, we release a large scale open-domain dataset for sentence selection which contains 304,413 query-sentence pairs. [ABSTRACT FROM AUTHOR]

• A novel token selection mechanism based on the dynamic self-attention network is proposed to automatically extract important video features. • A vision synchronization network is proposed to align appearance and motion features at the time slice level. • Extensive experiments and analysis confirm the superiority of the proposed model DSAVS. Video Question Answering (VideoQA) has gained increasing attention as an important task in understanding the rich spatio-temporal contents, i.e., the appearance and motion in the video. However, existing approaches mainly use the question to learn attentions over all the sampled appearance and motion features separately, which neglect two properties of VideoQA: (1) the answer to the question is often reflected on a few frames and video clips, and most video contents are superfluous; (2) appearance and motion features are usually concomitant and complementary to each other in time series. In this paper, we propose a novel VideoQA model, i.e., Dynamic Self-Attention with Vision Synchronization Networks (DSAVS), to address these problems. Specifically, a gated token selection mechanism is proposed to dynamically select the important tokens from appearance and motion sequences. These chosen tokens are fed into a self-attention mechanism to model the internal dependencies for more effective representation learning. To capture the correlation between the appearance and motion features, a vision synchronization block is proposed to synchronize the two types of vision features at the time slice level. Then, the visual objects can be correlated with their corresponding activities and the performance is further improved. Extensive experiments conducted on three public VideoQA data sets confirm the effectivity and superiority of our model compared with state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Smart substation is a crucial Cyber-Physical system and is prone to cyber-attack. In this paper, we propose a novel anomaly detection mechanism tailored for detecting the IEC 61850-based network traffic. Three types of traffic features are taken into account for comprehensively characterizing the network traffic during a time window. To eliminate the subjectivity of manually selecting the traffic features, we exploit Discrete Wavelet Transform (DWT) algorithm to secondarily extract the deep features. An improved Locally Linear Embedding (LLE) algorithm is proposed to reduce the dimension of deep feature vectors with more effective dimensionality reduction ability. By doing so, the LSTM (Long Short Term Memory)-based Autoencoder network that can learn to reconstruct the normal traffic time-series behavior, and thereafter uses the reconstruction error to detect the anomalies. We assess the performance of our proposed mechanism with the comprehensive experiments on the real smart substation. The results indicate that the proposed mechanism can be performed in a fast way with satisfactory detection performance. [ABSTRACT FROM AUTHOR]

Abstract: Designing elliptic curve password-based authenticated key agreement (ECPAKA) protocols for wireless mobile communications is a challenging task due to the limitation of bandwidth and storage of the mobile devices. Some well-published ECPAKA protocols have been proved to be insecure. We notice that until now none of the existing ECPAKA protocols for wireless mobile communication is provided any formal security analysis. In this paper, we propose a novel protocol and conduct a formal security analysis on our protocol. Compared with other ECPAKA protocol, our protocol meets all basic security properties and is the first ECPAKA protocol with formal security proof for wireless communication. We also explore the suitability of the novel protocol for 3GPP2 specifications and improve the A-Key (Authentication Key) distribution for current mobile cellular systems. [Copyright &y& Elsevier]

• A novel model based on the self-attention mechanism is proposed to learn more effective multi-modal representations. • The DSACA model is proposed to capture the internal dependencies and cross-modal correlation between the image and question sentence. • Extensive experiments and analysis confirm the superiority of the proposed DSACA. Visual Question Answering (VQA) as an important task in understanding vision and language has been proposed and aroused wide interests. In previous VQA methods, Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN) are generally used to extract visual and textual features respectively, and then the correlation between these two features is explored to infer the answer. However, CNN mainly focuses on extracting local spatial information and RNN pays more attention on exploiting sequential architecture and long-range dependencies. It is difficult for them to integrate the local features with their global dependencies to learn more effective representations of the image and question. To address this problem, we propose a novel model, i.e., Dual Self-Attention with Co-Attention networks (DSACA), for VQA. It aims to model the internal dependencies of both the spatial and sequential structure respectively by using the newly proposed self-attention mechanism. Specifically, DSACA mainly contains three submodules. The visual self-attention module selectively aggregates the visual features at each region by a weighted sum of the features at all positions. The textual self-attention module automatically emphasizes the interdependent word features by integrating associated features among the sentence words. Besides, the visual-textual co-attention module explores the close correlation between visual and textual features learned from self-attention modules. The three modules are integrated into an end-to-end framework to infer the answer. Extensive experiments performed on three generally used VQA datasets confirm the favorable performance of DSACA compared with state-of-the-art methods. [ABSTRACT FROM AUTHOR]

• Define three types of features extracted from URLs, domains, etc. • Exploit a method to balance the majority and minority class samples. • Adopt an improved DAE-based method to reduce the dimension of the dataset. • Boost the detection performance by using the improved ELM-based classifier. • Do experiments to verify the feasibility and effectiveness of the proposed approach. In this paper, a novel approach based on non-inverse matrix online sequence extreme learning machine (NIOSELM) for phishing detection is presented, which takes into account three types of features to comprehensively characterize a website. For the NIOSELM algorithm, we use Sherman Morriso Woodbury equation to avoid the matrix inversion operation, and introduce the idea of online sequence extreme learning machine (OSELM) to update the training model. In order to reduce the dependence of the detection model on the majority class, we use Adaptive Synthetic Sampling (ADASYN) algorithm to generate the synthetic minority class samples to balance the distribution between the samples of the majority and minority classes. Furthermore, an improved denoising auto-encoder (SDAE) is designed to reduce the dimension of the experimental dataset. The experimental results show the efficiency and feasibility of the proposed detection mechanism. Moreover, the overall detection performance of NIOSELM is better than that of other existing methods, especially in training speed and the detection accuracy. [ABSTRACT FROM AUTHOR]