Applied Mechanics and Materials Vols. 571-572

Abstract: As the relational database is applied widely, how to protect the sensitive data and state secrets has become a problem which needs to be solved urgently. The conventional relational database encryption technology not only reduces the system’s performance, but also results that we can not quickly inquire the encryption data through the original index mechanism due to the loss of the original partial order relation among data. This paper proposes a kind of relational database encryption technology based on the main memory database. It has not only realized the transparent encryption, decryption and ciphertext storage of the relational database, but also still retained the original response and retrieval performance. This technical proposal has very strong adaptability as well as strong practice guidance meaning for protecting the internet business data and the war-industry informative state secrets.

Abstract: The application requirement of Geospatial data is increasing and complex as it is getting numerous as a result of furthering study on geosciences. Based on a deeply research on Oracle Spatial storage management mechanism, this paper proposed a method that applies the graph theory to domain of optimizing spatial query of massive geographical data, and established a geospatial data query model in order to settle a problem of lower spatial query efficiency in geospatial database. Combining with the practical applications, this paper did a conventional spatial query test and a spatial query based on geospatial data model respectively. The result is that the spatial query based on geospatial data query model has a better efficiency than that on conventional method. Besides, this model can greatly improve the spatial query performance and this improvement will be increasingly apparent as the data volume increases.

Abstract: Big data is becoming the focus of all circles, but disordered and ambiguous definition of big data will deeply affect the study of this field. On the basis of the literature review, this paper applies the method of concept replacement in analytic philosophy, and puts forward a strict, clarified and broadly-accepted definition of big data, based on the fair definition, the study of a series of subsequent problems becomes simple and easy. The paper further discusses the application perspective and modes of the definition of big data by cases study in management of enterprise technological innovation.

Abstract: Database as a service in cloud computing enables the user to create, store, modify, and retrieve data over the Internet. The user does not have to install and maintain the database himself. Instead, the database service provider takes responsibility for installing and maintaining the database. Storing data in a cloud database introduces security risks in data confidentiality, data integrity, and privacy. This research aims to design a method for storing data in a cloud database that provides data confidentiality and privacy and allows querying data. The comparison operations and aggregate functions can be performed on the cloud server. The research employs a combination of data masking and data encryption to achieve the objectives.

Abstract: Ensuring data privacy and improving query performance are two closely linked challenges for outsourced databases. Using mixed encryption methods to data attributes can reach an explicit trade-off between these two challenges. However, encryption cannot always conceal relations between attributes values. When the data tuples are accessed selectively, inferences based on comparing encrypted values could be launched and sensitive values may be disclosed. In this paper, we explore the attribute based inferences in mixed encrypted databases. We develop a method to construct private indexes on encrypted values to defend against inference while supporting efficient selective access to encrypted data. We have conducted some experiments to validate our proposed method.

Abstract: Recently system security monitoring has meet several challenges. Therefore a system security monitoring approach based on complex event processing and dynamic structure-based neural networks is proposed in this paper. Firstly, complex event processing is used to handle real-time event streams and extract complex events from system security sensors. Secondly, the complex events from CEP would be used for further study by dynamic structure-based neural network. Finally the process of system security monitoring is showed and experiments would be applied to validate the feasibility, efficiency and precision of the approach.

Abstract: With the rapid development of global information technology and the rapid popularization of the Internet and modern information system entered the era of big data, people's daily work and life are faced with too much data and information explosion, urgently need to be an effective technology to help people from these huge amounts of data excavated really need and valuable knowledge, so the study of interaction in the process of data processing, and data mining research work have the vital significance.

Abstract: During the last few decades, there has been a lot of interest on systems using large amounts of data. In practice, not every piece of information is available, and people have to deal with incomplete data. There have been a lot models proposed for this problem. In this paper, we propose a new approach for incomplete data recovery. This new approach is based on linear regression. We do some experiments on real-world data and show that this new approach is appropriate for incomplete data recovery and can produce very good results.

Abstract: Recently, systems with a large amount of data have received a lot of attention. But the reality is that information may not be complete, in other words, people may have missing data. In order to solve this problem, a lot of models have been proposed. In this paper, we propose a new approach, which is based on linear random model. We test our model using real-world data and show that this new approach is good for incomplete data recovery and can get us desirable results.

Abstract: The usual method of outlier analysis is mainly analyzing the outliers according to the Anomaly Index and Variable Contribution Measurement. But in the analysis of large samples of high-dimensional data, this method is difficult. Owing to this, this paper presents a method that weight value for outliers is introduced. The features of outliers are weighted by Analytic Hierarchy Process method. Through this method, the importance of each property of outlier for data mining’s target is rationed, namely the weight number of each property is calculated. And then the correlation values, which represent the degree of relevance between outliers and data mining target, are calculated by using the weight number multiplying by the property value. After correlation values computed, we array the correlation values of outlier from high to low then outlier analysis can become more efficient. At the end of this paper, an instance is presented to demonstrate the maneuverability and feasibility of the method.