In view of the lack of effective resource demand forecasting and optimal allocation in Platform-as-a-Service (PaaS) platform, a resource demand forecasting model and an allocation method were proposed. Firstly, according to the periodicity of the application demand for resources in PaaS platform, the resource sequence was segmented. And on the basis of short-term prediction, combined with the multi-periodicity characteristics of the application, a comprehensive prediction model was established by using the multiple regression algorithm. Then, based on MapReduce architecture, a PaaS platform resource allocation system based on Master-Slave mode was designed and implemented. Finally, the resources were allocated based on current task request and resource demand prediction results. The experimental results show that, compared with autoregressive model and exponential smoothing algorithm, the proposed resource demand forecasting model and allocation method has the mean absolute percentage error drop of 8.71 percentage points and 2.07 percentage points respectively, root mean square error drop of 2.01 percentage points and 0.46 percentage points respectively. It can be seen that the prediction result of the prediction model has little error and its fitting degree with real value is high, while high accuracy costs little time. Besides, the average waiting time of PaaS platform with the proposed prediction model for resource requests decreases significantly.

There are many illegal Web pages on the Internet, which may have pornographic, violent, gambling or reactionary content. Without being filtered effectively, they will exercise a malign influence on the searching services. An improved K-Nearest Neighbors (KNN) classification algorithm to promote the recognition accuracy was proposed and implemented on a virtualized platform following the MapReduce model provided by the open source software Hadoop, which made it distributed and parallel. Through experiments and comparison with the existing work, it is proved that the proposed recognition method improves the accuracy and efficiency greatly. The algorithm is implemented on a virtualized platform following the MapReduce model provided by the open source software Hadoop, which makes it distributed and parallel. Through experiments and comparison with existing work, it is proved that the recognition method we propose improves the accuracy and efficiency greatly.

The influence of social network is closely related with its structural characteristics. Based on the data from Sina microblog, the distributions of the number of followers and followings were analyzed and found that the number of followers and followings both were power-law distributed. The distance characteristic between different pairs of nodes was discussed, and it was found and proved that there was "small-world" phenomenon in the microblog network. At last, the links between nodes in the network were investigated and found that the formation of the link satisfied triple closure principle. The investigation results on the above three topics are important for us to explore the relationship between the influence of micro-blog and the structural characteristics of its underlying social network, as well as to the design of mechanisms to control the influence.