Applied Mechanics and Materials Vols. 303-306

Abstract: The blast furnace faults will cause significant economic and human losses. Therefore, the study of blast furnace intelligent fault diagnosis technology is necessary and important. But some real fault data is difficult to obtain .The training set can not be provided enough samples.So a new virtual fault samples generation method is designed to get enough tainning samples. The designed method uses group discovery technique and the Box-Muller method to generate the candidate virtual fault sample set. And then with the help of manifold contraction and the semi-supervised learning algorithm to select the virtual samples with better performance. By adding the selected virtual samples to the original training set,a new svm fault diagnosis classifier is obtained. The results show that this method can simulate the fault samples effectively,and the new classifier’s accuracy is much more acceptable.

Abstract: This paper constructs function concept ontology for smart home to automate home service retrieval according to the functional properties. Firstly in order to construct function concept ontology, the function concept is divided into five categories by analyzing the lifecycle of information, and then each category is refined according to the ways the devices adopted. Then a scenario of audible alarm of gas detective is given to show the drawback of current retrieval mode. Finally an ontology-based architecture is proposed to present service registry, retrieval and invocation based on the function concept ontology.

Abstract: Cloud computing is a new computing and business paradigm with flexible and powerful computational architecture to offer universal services to users via Internet. The performance of the scheduling system influences the cost benefit of this computing paradigm. Thus, jobs should be scheduled efficiently to reduce the execution cost and time. In this paper, we present an intelligent scheduling system, which considers both the requirements of different service requests and the circumstances of the computing infrastructure which consists of various resource, then, the main components of the system are introduced in detail, at last, the conclusions are drawn and the further research directions of the scheduling systems are pointed out.

Abstract: The aquaculture IOT system consists of the water quality monitoring stations which are based on a wireless sensor network (WSN), meteorological station, water quality control station, on-site and remote monitoring center and the central cloud processing platform. If we apply this system effectively, we can modify the existing breeding mode at the cost of mass energy use and gain direct economic benefits. We got relative initial data from the field research in Yixing of Jiangsu province, and compared different aquaculture situation between farmers who used the system and who didn’t in 2011 through the parallel comparison method. We draw the conclusion that this system is helpful to save labor cost 1768.62 yuan/ha, increase river crab production 88.72 kg/ha and improve river crab sales revenue by 15.51%.

Abstract: This paper proposed a new facial feature points localization algorithm based on main characteristics of eyes.Use the result of pupil center position to initialize the model of hybrid improved active shape model (ASM) and active appearance model (AAM). The algorithm will use two-dimensional local gray information to update the feature point position when using ASM to locate the face contour feature points. As to the internal features point location, it establishes facial organs independent AAM model. At the same time, it optimizes measure functions of ASM and AAM to judge the convergence of search algorithm. The experimental results show that the new algorithm greatly improved the localization accuracy of facial feature points.

Abstract: Color histogram is an important technique for color database retrieving, but it often ignores color’s spatial distribution information. This paper proposes an improved color histogram algorithm based on the HSV space, whose subspaces are non-equally quantized. The algorithm first proceeds annular partition on the original image, and then uses the method presented by Aibing Rao etc. [1] to count each partition. At last, it calculates the weighted sums for the distances between distinct color histograms. Experimental results demonstrate that the algorithm reduces the feature dimensions and keeps a good accuracy as well as the spatial distribution information. Thus, a better retrieval result is obtained.

Abstract: An intelligent mobile petrol station recommender system based on context ontology and rule inference is deigned. The approach of context ontology modeling specific for mobile recommendation is discussed. And a two-level context ontology model including upper ontology and domain ontology used in petrol station recommender is developed. The generation of recommendation rules based on the context ontologies and the process of the rule inference for recommendation are also demonstrated.

Abstract: Emergency Resources Management System is an important part of emergency management system. There are many classification systems of emergency resources, which lead to inaccurate description of the structure of emergency resources and negative impacts of the emergency resources sharing and configuration scheduling. In this paper, we compared the strengths and weaknesses of existed emergency resources classification systems, constructed a emergency resources classification system suitable for the whole process of emergency rescue, built an upper emergency resource and domain ontology model rested on SUMO ontology, formally described the concepts in the field of emergency resources and the relationships between these concepts, and satisfied the emergency resources demand of quickly crisis dealing.

Abstract: Mobile personalized web search has been introduced for the purpose of distinguishing mobile user's personal different search interest. We first take the user's location information into account to do a geographic query expansion, then present an approach to personalizing web search for mobile users within language modeling framework. We estimate a user mixed model estimated according to both activated ontological topic model-based feedback and user interest model to re-rank the results from geographic query expansion. Experiments show that language model based re-ranking method is effective in presenting more relevant documents on the top retrieved results to mobile users. The main contribution of the improvements comes from the consideration of geographic information, ontological topic information and user interests together to find more relevant documents for satisfying their personal information need.

Abstract: This paper takes microblog as an example, recognising names in finance and economics field by the algorithm combining the use of rules and probability calculation. This method firstly get the candidate names through calculating the probabilities, then choose from the candidate names by matching them with corresponding rules which contain positive rules and negative rules .This method has good robustness and flexibility.