Applied Mechanics and Materials Vols. 380-384

Abstract: Information explosion brings lots of challenges to text classification. The dimension disaster led to a sharp increase of computational complexity and lower classification accuracy. Therefore, it is critical to use feature selection techniques before actual classification. Automatic classification of English text has been researched for many years, but little on Chinese text. In this paper, several classic feature selection methods, namely TF, IG and CHI, are compared on classifying Chinese text. Meanwhile, we take imbalanced data into consideration in the paper. Experimental results show that CHI performed better than IG and TF when the dataset is imbalanced, but no obvious difference on balanced data.

Abstract: In this paper, we present a new method to build public-key Cryptosystem. The method is based on the state explosion problem occurred in the computing of average number of tokens in the places of Stochastic Petri Net (SPN). The reachable markings in the coverability tree of SPN are used as the encryption keys. Accordingly, multiple encryption keys can be generated, thus we can perform multiple encryption to get as strong security as we expect. The decryption is realized through solving a group of ordinary differential equations from Continuous Petri Net (CPN), which has the same underlying Petri net as that of SPN. The decipherment difficulty for attackers is in exponential order. The contribution of this paper is that we can use continuous mathematics to design cryptosystems instead of discrete mathematics.

Abstract: Because of uncertainty data, traditional algorithm of mining frequent items in certain dataset is difficult to apply to uncertain dataset. Considering characteristics of uncertain data, an improved vertical mining algorithm to find frequent items in uncertain dataset was proposed with the algorithm thought of classic vertical algorithm-Eclat in certain dataset. The improved algorithm merged TID field and corresponding probability field into probability vector. During the expansion of itemset and probability vector, itemset tree was established, and the support of candidate itemsets was calculated by means of vector operations. The improved algorithm is proved to be feasible and efficient according to experimental comparison and analysis.

Abstract: Twitter is one of the largest social networks in the world. People could share contents on it. When we interact with each other, the information spreads. And its users retweet behavior that makes information spread so fast. So there comes an important question: Whats about users retweet behavior Could we simulate information spreading in twitter by retweeting behavior We crawl twitter and mine information spreading based on users retweet behavior in it. Through our dateset, we verify the power-law distribution of the retweet-width and retweet-depth. At the same time, we study the correlation between retweet-width and retweet-depth. Finally, we propose an information spreading model to simulate the information spreading process in twitter and get a good result.

Abstract: On the basis of hardware development platform ARM9,the paper describes the porting of Embedded Linux on ARM9 by constructing cross-compile environment, and ramdisk has been produced as a file system . The experiments and running verify that the system has the advantages of small volume, stable operation, high reliability, and the system is a useful technology platform for developing the further embedded products.

Abstract: Spatiotemporal data are widely visible in everyday life. This paper proposes an algorithm to represent them in a granular wayinformation granules. Information granules can be regarded as a collection of conceptual landmarks using which people can view the data and describe them in a semantic way. The key objective of this paper is to introduce a new granular way of data analysis through their granulation. Several experiments are done with synthetic data and the results show a clear way how our algorithm performs.

Abstract: Spiking neural P systems are a new class of bio-inspired computing devices incorporating the ideas of spiking neural networks into P systems. Homogeneous spiking neural P systems are a variant of spiking neural P systems, where each neuron has the same set of rules. The universality of homogeneous spiking neural P systems without delays is investigated and proved. This paper provides an answer to an open problem formulated by Zeng Xiangxiang, etal.

Abstract: Classical cryptography theory holds that the true random sequence is better than any pseudorandom sequence on the security of stream cipher. So people prefer the pseudorandom sequence with long-period to the pseudorandom sequence with short-period. In this paper, it is proved through power spectrum analysis that the pseudorandom sequence with long-period cant effectively conceal message characteristics in frequency-domain. To combine the pseudorandom sequences with long cycle and short cycle may be the better encryption key.

Abstract: The 3D meshes deformation is the deformation on 3D mesh models under users want, and keeping the mesh details at the same time. In this paper, we present a technique for multi-resolution meshes deformation based on the offset of sketching. In order to decrease the computational cost, the proposed method reduces the number of vertices involved in the deformation. First, we construct a simplified mesh using a modified CPM algorithm; and then correct Laplacian coordinates rotation by the offset of sketching and generate the deformed models by solving linear system in least squares sense with Gauss-Seidel algorithm; finally, series of details are added to the deformed simplified mesh. In contrast with the traditional deformation, our method can achieve natural deformed models more efficiently.

Abstract: An authentication protocol is a procedure by which a sender tries to convey n bits of information, which we call an input message, to a receiver. An adversary controls the network over which the sender and the receiver talk and may change any message before it reaches its destination. This paper deals with the simple scenario where two parties, Alice and Bob, communicate and want to assure that the message received by Bob is the one sent by Alice. Unconditional cryptographic security cannot be generated simply from scratch, but must be based on some given primitive to start with. In this paper, we introduce a message authentication protocol with unconditional cryptographic security. More precisely, we give protocols using a weakly secret key allowing for the exchange of authenticated messages.