With the widespread applications of Massive Open Online Courses （MOOCs） platforms， an effective method is needed for personalized course recommendation. In view of the existing course recommendation methods， which usually use the course learning records to establish the overall static representation for users’ learning interests， while ignoring the dynamic changes of learning interests and users’ short-term learning interests， a Hierarchical and Phased Attention Network （HPAN） was proposed to carry out personalized course recommendation. In the first layer of the network， the attention network was used to obtain the user’s long- and short-term learning interests. In the second layer of the network， the user’s long- and short-term learning interests and short-term interaction sequence were combined to obtain the user’s interest vector through the attention network， then the preference value of the user’s interest vector with each course vector was calculated， and courses were recommended for the user according to the result. Experimental results on public dataset XuetangX show that， compared with the second best SHAN （Sequential Hierarchical Attention Network） model， HPAN model has the Recall@5 increased by 12.7%； compared with FPMC （Factorizing Personalized Markov Chains） model， HPAN model has the MRR@20 increased by 15.6%. HPAN model has better recommendation effect than the comparison models， and can be used for practical personalized course recommendation.

Generative steganography hides secret messages by generating sufficiently natural or true samples with secret，which is a hot research topic in information hiding， but there is little research in the field of video steganography. Combined with the idea of digital Cardan grille， a semi-generative video steganography scheme based on Deep Convolutional Generative Adversarial Net （DCGAN） was proposed. In this scheme， a dual-stream video generation network based on DCGAN was designed to generate three parts of videos： dynamic foreground， static background and spatio-temporal mask， and different videos were produced by the generation network driven by random noise. The sender in this scheme was able to set the steganography threshold and adaptively generate a digital Cardan grille in the mask， then the obtain digital cardan grille was used as the key for steganography and extraction； at same time， with the foreground as the carrier， the optimal embedding of information was realized. Experimental results show that the video-with-secret generated by the proposed scheme has good visual quality， with a Frechet Inception Distance score （FID） of 90， and the embedding capacity of the scheme is better than those of the existing generative steganography schemes， up to 0.11 bpp. It can be seen that the proposed scheme can transmit secret messages more efficiently.

In Internet Protocol Television （IPTV） applications， a television terminal is usually shared by several family members. The exiting recommendation algorithms are difficult to analyze the different interests and preferences of family members from the historical data of terminal. In order to meet the video-on-demand requirements of multiple members under the same terminal， a capsule network-based IPTV video-on-demand recommendation model， namely CapIPTV， was proposed. Firstly， a user interest generation layer was designed on the basis of the capsule network routing mechanism， which took the historical behavior data of the terminal as the input， and the interest expressions of different family members were obtained through the clustering characteristic of the capsule network. Then， the attention mechanism was adopted to dynamically assign different attention weights to different interest expressions. Finally， the interest vector of different family members and the expression vector of video-on-demand were extracted， and the inner product of them was calculated to obtain the Top-N preference recommendation. Experimental results based on both the public dataset MovieLens and real radio and television dataset IPTV show that， the proposed CapIPTV outperforms the other 5 similar recommendation models in terms of Hit Rate （HR）， Recall and Normalized Discounted Cumulative Gain （NDCG）.

To deal with Two-Sided Matching (TSM) problem based on Hesitant Fuzzy Value (HFV) of unknown weights, a multi-attribute matching decision making approach was proposed. To begin with, the weight information was determined by maximizing the sum of deviations of the given values in terms of HFVs with multi-attribute evaluated by both two-sided Agents. Then, the matching degree could be aggregated via an operation of adjusted hesitant fuzzy weighted averaging with obtained weights and multi-attribute information. In addition, a multi-objective optimization model was established based on the matching degree of two sides. By solving this model into single objective optimization model in min-max method, the matching scheme was generated. Finally, a numerical illustration and comparison was taken, the solutions of objectives by the proposed method were respectively 1.689 and 1.575, and a unique matching scheme was obtained. The experimental results show that the proposed method can avoid multiple solutions caused by subjective weights of goal functions.

The CrossMatch algorithm based on the idea of Dynamic Time Warping (DTW) can be used to solve the problems of local similarity between time series. However, due to the high complexity of time and space, large amounts of computing resources are required. Thus, it is almost impossible to be used for long sequences. To solve the above mentioned problems, a new algorithm for local similarity detection based on distributed platform was proposed. The proposed algorithm was a distributed solution for CrossMatch. The problem of insufficient computing resources including time and space requirement was solved. Firstly, the series should be splited and distributed on several nodes. Secondly, the local similarity of every node's own series was dealt with. Finally, the results would be merged and assembled in order to find the local similarity of series. The experimental results show that the accuracy between the proposed algorithm and the CrossMatch algorithm is similar, and the proposed algorithm uses less time. The improved distributed algorithm can not only solve the computation problem of long sequence of time series which can not be processed by a single machine, but also improve the running speed by increasing the number of parallel computing nodes.

The current research of bilateral matching problem is limited to single-period scenario. Aiming at the issue, an approach was proposed to study matching decision problem under multi-period and multi-attribute. First, through the orness, a measurement of Agent's preference, an optimal program was constructed to determine the cumulative weight of an Agent within each attribute. More specifically, the criteria of this program consisted of two parts: one part was to minimize the sum of deviation between an orness and corresponding cumulative weight of an Agent in different period; another part was to minimize the maximum disparity among cumulative weights of an Agent. Then, based on obtained cumulative weight, matching degree which represented by Agent's positive and negative ideal between the cumulative evaluation value and perceived expectation can be determined via the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). Furthermore, a double-objective optimization model based on perceived expectation was constructed and the minimax method was used to solve this model for obtaining matching results. Finally, a numerical example was given to compare the minimax method with the linear weighting method. The results show that difference of profit and loss of utility obtained by the former method was 0.33, less than 0.36 that obtained by the latter method. Moreover, it also demonstrates the proposed method can maximize the profit and loss of utility of inferior side.

Since high-speed network traffic can not be effectively coped with the network traffic capture system implemented by software, and the multiple network flow need to be collected simultaneously to improve the capturing efficiency, an high-speed network flow capture framework in combination of hardware and software was presented, and the implementation of network traffic capture system based on NetFPGA-10G, called HSNTCS, was discussed. A variety of network flow in hardware was filtered and classified by the exact string matching engine and the regular expression matching engine of this system. After being transmitted to the corresponding data buffer at the driver layer, the network flow was directly copied to the corresponding database in user space. The experiments show that the throughput of UDP (User Datagram Protocol）and TCP (Transmission Control Protocol）in the high-speed network traffic capture system implemented by the hardware under the condition of exact string matching achieved 1.2Gb/s, which is about 3 times of that implemented by the software; and the throughput of UDP and TCP in the system implemented by the hardware under the condition of regular expression matching achieved 640Mb/s, which is about 3 times of that implemented by the software. The results demonstrate that the capture performance by the method of hardware is better than the method of software.

A Web service Management System with Multiple Engines (WSMSME) was proposed to solve the problem of execution optimization for composite services in the system. The scheduler execution of composite services in system with multiple engines was analyzed, and a dynamic programming algorithm was put forward, which optimally minimized the heaviest load of engines by segmenting a pipelined execution plan into subsequences before they were dispatched and executed. Experiment with an initial prototype indicates that the algorithm can lead to significant performance improvement than the random algorithm.

At present, the embedded flash file systems specially designed for NAND Flash mainly have two disadvantages: too long booting time and insufficient consideration for wear leveling. According to these two aspects, a new NAND flash file system named SFFS was proposed, which shortened the booting time through changing the management of data nodes and storing the controlling information and data information separately，and made wear leveling better through using a block to store hot data and cold data alternatively.