Standard neural machine translation (NMT) is on the assumption that the document-level context is independent. Most existing document-level NMT approaches are satisfied with a smattering sense of global document-level information, while this work focuses on exploiting detailed document-level context in terms of a memory network. The capacity of the memory network that detecting the most relevant part of the current sentence from memory renders a natural solution to model the rich document-level context. In this work, the proposed document-aware memory network is implemented to enhance the Transformer NMT baseline. Experiments on several tasks show that the proposed method significantly improves the NMT performance over strong Transformer baselines and other related studies.

Tor's hidden services provide both sender privacy and recipient privacy to users. A hot topic in security of Tor is how to deanonymize its hidden services. Existing works proved that the recipient privacy could be revealed, namely a hidden server's real IP address could be located. However, the hidden service's circuit is bi-directionally anonymous, and the sender privacy can also be revealed. In this letter, we propose a novel approach that can transparently discover the client of the hidden service. Based on extensive analysis on the hidden service protocol, we find a combination of cells which can be used to generate a special traffic feature with the cell-padding mechanism of Tor. A user can implement some onion routers in Tor networks and monitor traffic passing through them. Once the traffic feature is discovered, the user confirms one of the controlled routers is chosen as the entry router, and the adjacent node is the client. Compared with the existing works, our approach does not disturb the normal communication of the hidden service. Simulations have demonstrated the effectiveness of our method.

Tor is the most popular and well-researched low-latency anonymous communication network provides sender privacy to Internet users. It also provides recipient privacy by making TCP services available through “hidden service”, which allowing users not only to access information anonymously but also to publish information anonymously. However, based on our analysis of the hidden service protocol, we found a special combination of cells, which is the basic transmission unit over Tor, transmitted during the circuit creation procedure that could be used to degrade the anonymity. In this paper, we investigate a novel protocol-feature based attack against Tor's hidden service. The main idea resides in fact that an attacker could monitor traffic and manipulate cells at the client side entry router, and an adversary at the hidden server side could cooperate to reveal the communication relationship. Compared with other existing attacks, our attack reveals the client of a hidden service and does not rely on traffic analysis or watermarking techniques. We manipulate Tor cells at the entry router to generate the protocol-feature. Once our controlled entry onion routers detect such a feature, we can confirm the IP address of the client. We implemented this attack against hidden service and conducted extensive theoretical analysis and experiments over Tor network. The experiment results validate that our attack can achieve high rate of detection rate with low false positive rate.

In this paper, we study simultaneous wireless information and power transfer (SWIPT) in two-way relay channels where two users exchange information with each other via a multi-antenna relay node. The signals forwarded by the relay node are also used to supply the power to two users. We formulate a max-min optimization problem aiming to maximize the minimum harvested energy between two users to achieve fairness. We jointly optimize the relay beamforming matrix and allocating powers at the two users subject to the quality of service (QoS) constraints. To be specific, we consider the amplify-and-forward (AF) relay strategy and the time splitting SWIPT strategy. To this end, we propose two different time splitting protocols to enable relay to supply power to two users. To solve the non-convex joint optimization problem, we propose to split the original optimization problem into two subproblems and solving them iteratively to obtain the final solution. It is shown that the first subproblem dealing with the beamforming matrix can be optimally solved by using the technique of relaxed semidefinite programming (SDR), and the second subproblem, which deals with the power allocation, can be solved via linear programming. The performance comparison of two schemes as well as the one-way relaying scheme are provided and the effectiveness of the proposed schemes is verified.

Time performance optimization and resource conflict resolution are two important challenges in multiple project management contexts. Compared with traditional project management, multi-project management usually suffers limited and insufficient resources, and a tight and urgent deadline to finish all concurrent projects. In this case, time performance optimization of the global project management is badly needed. To our best knowledge, existing work seldom pays attention to the formal modeling and analyzing of multi-project management in an effort to eliminate resource conflicts and optimizing the project execution time. This work proposes such a method based on PRT-Net, which is a Petri net-based formulism tailored for a kind of project constrained by resource and time. The detailed modeling approaches based on PRT-Net are first presented. Then, resource conflict detection method with corresponding algorithm is proposed. Next, the priority criteria including a key-activity priority strategy and a waiting-short priority strategy are presented to resolve resource conflicts. Finally, we show how to construct a conflict-free PRT-Net by designing resource conflict resolution controllers. By experiments, we prove that our proposed priority strategy can ensure the execution time of global multiple projects much shorter than those without using any strategies.

Attention mechanism, which selectively focuses on source-side information to learn a context vector for generating target words, has been shown to be an effective method for neural machine translation (NMT). In fact, generating target words depends on not only the source-side information but also the target-side information. Although the vanilla NMT can acquire target-side information implicitly by recurrent neural networks (RNN), RNN cannot adequately capture the global relationship between target-side words. To solve this problem, this paper proposes a novel target-attention approach to capture this information, thus enhancing target word predictions in NMT. Specifically, we propose three variants of target-attention model to directly obtain the global relationship among target words: 1) a forward target-attention model that uses a target attention mechanism to incorporate previous historical target words into the prediction of the current target word; 2) a reverse target-attention model that adopts a reverse RNN model to obtain the entire reverse target words information, and then to combine with source context information to generate target sequence; 3) a bidirectional target-attention model that combines the forward target-attention model and reverse target-attention model together, which can make full use of target words to further improve the performance of NMT. Our methods can be integrated into both RNN based NMT and self-attention based NMT, and help NMT get global target-side information to improve translation performance. Experiments on the NIST Chinese-to-English and the WMT English-to-German translation tasks show that the proposed models achieve significant improvements over state-of-the-art baselines.

With the surge of social media platform, users' profile information become treasure to enhance social network services. However, attributes information of most users are not complete, thus it is important to infer latent attributes of users. Contemporary attribute inference methods have a basic assumption that there are enough labeled data to train a model. However, in social media, it is very expensive and difficult to label a large amount of data. In this paper, we study the latent attribute inference problem with very small labeled data and propose the SRW-COND solution. In order to solve the difficulty of small labeled data, SRW-COND firstly extends labeled data with a simple but effective greedy algorithm. Then SRW-COND employs a supervised random walk process to effectively utilize the known attributes information and link structure of users. Experiments on two real datasets illustrate the effectiveness of SRW-COND.

AlphaSeq is a new paradigm to design sequencess with desired properties based on deep reinforcement learning (DRL). In this work, we propose a new metric function and a new reward function, to design an improved version of AlphaSeq. We show analytically and also through numerical simulations that the proposed algorithm can discover sequence sets with preferable properties faster than that of the previous algorithm.