基于扩展主题模型的异常医疗处方检测方法

异常处方指的是医生为患者所开具的存在异常的处方。医疗处方中出现异常,如滥用药或者开错药等,会影响患者的治疗效率,甚至造成严重的后果。由于一些主观或者客观原因,医生总会开具一些异常处方。检测出这些异常处方能够提升患者就医效率,减少社会医疗成本,并且对药物滥用、多开药、错开药的有效管理等都有着重要意义。为此,提出了一种基于扩展主题模型的异常处方检测方法。该方法能够自动地从大量处方数据中检测出异常处方,并且对于每一个新的处方,该方法都能够判断其诊断和用药是否匹配,进而判断其是否正常。与其他异常检测算法相比,该方法具有更广泛的应用,不仅可以在医疗领域中使用,以检测异常处方,还可以在其他领域中使用,以检测其他特征之间的匹配关系异常。该方法已经得到了实现,并在真实的处方数据集中得到了验证。     

基于层次聚类的失眠处方用药分析

为了探寻层次聚类在失眠处方用药分析上的应用情况,进而分析失眠处方的用药规律,收集并整理了《方剂大辞典》中主治失眠的处方。对单味药物的四气、五味、归经及功效等数据,根据单连接、全连接和平均连接这三种不同的相似性度量方法进行层次聚类分析并比较。结果显示,基于全连接的层次聚类分组最为合理,将性味归经和功效有极大相似度的药物聚为一类,其聚类结果符合一定的中医理论。层次聚类结果客观地反映了失眠处方中药物间的关联关系,间接体现了失眠用药的药物组合规律,为临床用药提供新的研究方法和思路。     

An approach to accessing product data across system and software revisions

Long-term users of engineering product data are hampered by the ephemeral nature of CAD file formats and the applications that work with them. STEP, the Standard for the Exchange of Product Model Data (ISO 10303), promises to help with meeting this challenge, but is not without problems of its own. We present a complementary solution based on the use of lightweight file formats to preserve specific aspects of the product data, in conjunction with a registry of relevant representation information as defined by the Open Archival Information System Reference Model (ISO 14721). This registry is used to identify suitable destination file formats for different purposes, and provides a resource to aid in the recovery of information from these formats in the future.     

基于LDA主题模型的中医药方剂相似度计算

评价中医药方剂的相似度的常用方法是基于方剂的功效和主治功能进行相似度分析,但存在相似度无法定量化的问题,并且没有考虑中医方剂组成成分的影响。提出了利用LDA主题模型发掘“方剂-证型-组成成分”的隐含关系的方法,将“方剂-组成成分”转化成“方剂-证型”和“证型-组成成分”两个概率分布,并利用KL散度距离来计算相似度。实验结果表明基于LDA主题模型的方法能够更好地计算方剂间的相似度,并且能较好地反映中医辩证论治。     

一种基于始末距离的时间序列符号聚合近似表示方法

时间序列数据的特征表示方法是时间序列数据挖掘任务的关键技术,符号聚合近似表示(SAX)是特征表示方法中比较常用的一种。针对SAX算法在各序列段表示符号一致时无法区分时间序列间的相似性这一缺陷,提出了一种基于始末距离的时间序列符号聚合近似表示方法(SAX_SM)。由于时间序列有很强的形态趋势,因此文中提出的方法选用起点和终点来表示各个序列段的形态特征,并使用各序列段的形态特征和表示符号来近似表示时间序列数据,以将其从高维空间映射到低维空间;然后,针对起点和终点构建始末距离来计算两序列段间的形态距离;最后, 结合 始末距离和符号距离定义一种新的距离度量方式,以更客观地度量时间序列间的相似性。理论分析表明,该距离度量满足下界定理。在20组UCR时间序列数据集上的实验表明,所提SAX_SM方法在13个数据集中获得了最高的分类准确率(包含并列最大的),而SAX只在6个数据集中获得了最高的分类准确率(包含并列最大的),因此SAX_SM具有比SAX更优的分类效果。     

中文深层网络的模式匹配和接口集成

目前国内外在深层网络方面的研究几乎都围绕英文环境进行,还没有针对中文深层网络的研究．提出了对中文深层网络进行模式匹配和接口集成的方法．该方法首先创建一个用来存储同义词、超义词和子义词的字典,然后使用基于规则的分词算法将从接口中抽取的属性分成词．对于每一个属性,从定义的字典中找到其对应的所有同义词、超义词和子义词,生成一条相应的记录并存储到列表中,再从每条记录中选取出现次数最多的属性作为联合接口的属性．     

Use of ebRIM-based CSW with sensor observation services for registry and discovery of remote-sensing observations

Recent advances in Sensor Web geospatial data capture, such as high-resolution in satellite imagery and Web-ready data processing and modeling technologies, have led to the generation of large numbers of datasets from real-time or near real-time observations and measurements. Finding which sensor or data complies with criteria such as specific times, locations, and scales has become a bottleneck for Sensor Web-based applications, especially remote-sensing observations. In this paper, an architecture for use of the integration Sensor Observation Service (SOS) with the Open Geospatial Consortium (OGC) Catalogue Service-Web profile (CSW) is put forward. The architecture consists of a distributed geospatial sensor observation service, a geospatial catalogue service based on the ebXML Registry Information Model (ebRIM), SOS search and registry middleware, and a geospatial sensor portal. The SOS search and registry middleware finds the potential SOS, generating data granule information and inserting the records into CSW. The contents and sequence of the services, the available observations, and the metadata of the observations registry are described. A prototype system is designed and implemented using the service middleware technology and a standard interface and protocol. The feasibility and the response time of registry and retrieval of observations are evaluated using a realistic Earth Observing-1 (EO-1) SOS scenario. Extracting information from SOS requires the same execution time as record generation for CSW. The average data retrieval response time in SOS+CSW mode is 17.6% of that of the SOS-alone mode. The proposed architecture has the more advantages of SOS search and observation data retrieval than the existing sensor Web enabled systems.     

基于三次样条插值的采样数据光滑曲线形成法

计算机辅助测试所得记录为一系列经滤波的离散采样数据,要把它恢复成不失真的连续光滑曲线打印输出,应对数据进行处理。对变化平缓的稳态过程可采用拟合法;而对波动较大,采样点较多的动态过程必须用插值法才能真实反映这种波动。介绍了三次样条插值函数的形成法。指出常规方法仅能指定过程始、终点被测物理量变化的速度或加速度,不能完整描述物理量的变化状况。提出在始端和末端规定两个不独立的附加虚拟采样节点可以增加方程组的两个自由度,使其能同时指定过程始终点被测物理量的速度和加速度,从而得到符合实际情况的完美结果。     

Dance movement: a focus on the technology

Dance notation systems, like music notes, enable documentation of symbolic representations of movement as signs on paper for individual analysis and interpretation. Today, dance notation systems operate within a digital environment in dance notation applications that facilitate the process of recording movement. The author argues that a key objective in the development of these applications should be to provide the user with an unambiguous method to record and represent movement. These applications offer varying functionality in their use of technology for the representation of movement and can be broadly defined in three different categories. Dance notation applications make up the first category - they help notate or record specific forms of movement using dance notation. Notation-based applications, the second category, include applications that use dance notation as a basis for their development. The last category, dance technology, consists of applications that use emerging technologies to record and visualize movement. While each application has a defined use, it's important to consider how effective the technologies they employ are in successfully achieving their objectives. In this article, the author focuses on dance applications in these three categories. The author considers the limitations of existing technologies in their ability to effectively describe and record movement within a specific context.     

中医药方剂相似度模型

针对当前方剂检索服务大多只提供简单查询功能,不能满足方剂知识整理与累积的需求,提出方剂4个方面的相似度模型。对于组成药、功效和主治的相似性,引入集合相似度模型。对于名称相似性,引入基于字符串的相似度模型,并进一步提出基于统计和词典的相似度模型。通过分析比较各模型,挖掘出方剂间的关联,完成方剂聚类。     

序贯LSB隐写术的提取攻击

序贯LSB隐写术在载体中通过连续LSB替换嵌入消息,其提取攻击问题本质上是消息嵌入起止点的估计问题。该文建立针对序贯LSB隐写术的提取攻击模型,将提取攻击问题转化为一类排序问题。对嵌入率未知情形,提出计算复杂度为O(n)的提取攻击算法。对嵌入率已知情形,给出计算复杂度为O(2logn)的快速提取攻击算法。实现了对序贯JSteg算法的提取攻击。