ID	原文	译文
44666	并对解析结果中的异常现象进行了分析，发现了 DNS 劫持、权威域名服务器配制错误等问题。	Analyzing all abnormal phenomena in parsing results, some problems such as DNS hijacking and wrong configuration were discovered in DNS authority name servers.
44667	在研究 Web Service 事务的补偿、替换、重试、时间约束等应用语义之上，重点提出了一个适用于复杂流程型 Web Service 事务模型，	Analyze the characteristics and constraints of Web Service transaction, and then propose a Web Service transaction model combined with application semantics including compensation, retry, and alternation, which is suitable for business process.
44668	引入抽象事务来协助管理这些语义信息，	This model defines an abstract transaction to help manage these semantics mentioned above.
44669	并通过事务的多级特性来适应不同的服务粒度，	It also uses the multi-level characteristic of Web Service transaction to adapt to different service granularity.
44670	采用有色 Petri 网对该模型进行了形式化描述。	A colored Petri netis used to describe the model.
44671	采用贝叶斯网络，从业务流程的角度，对 Web Service 事务的可靠性进行理论分析，给出了分析模型算法和边界理论值。	The reliability of Web Service transaction was analyzed from business process perspective with Bayes net.
44672	日益增长的网络流量使得有效识别恶意访问成为亟待解决的网络安全问题之一，	The growing network traffic makes effective identification of malicious access one of the network security issues that need to be addressed.
44673	现有的检测方法多是基于域名黑名单展开研究的，忽略了非黑名单中也可能存在着隐藏的恶意访问。	Most of the existing detection methods are based on the domain name blacklist, ignoring the hidden malicious access in the non-blacklist.
44674	为了解决上述问题，利用了时间序列的分析方法建立了一种基于 URL 的恶意访问检测模型。	In order to solve the above problems, a URL-based malicious access dejection model by using time series analysis method was proposed.
44675	首先，以用户访问某域名的 URL 日志为研究对象，从域名访问相似度、信息熵、功率谱密度等多维度挖掘恶意访问的表现特征，	Firstly, the performance was studied and quantified characteristics of malicious access from multiple dimensions by the user accessing the URL log of a domain name, such as domain name access similarity, information entropy and power spectral density.