DCRL-KG: Distributed Multi-Modal Knowledge Graph Retrieval Platform Based on Collaborative Representation Learning

The knowledge graph with relational abundant information has been widely used as the basic data support for the retrieval platforms.Image and text descriptions added to the knowledge graph enrich the node information,which accounts for the advantage of the multi-modal knowledge graph.In the field of cross-modal retrieval platforms,multi-modal knowledge graphs can help to improve retrieval accuracy and efficiency because of the abundant relational infor-mation provided by knowledge graphs.The representation learning method is sig-nificant to the application of multi-modal knowledge graphs.This paper proposes a distributed collaborative vector retrieval platform(DCRL-KG)using the multi-modal knowledge graph VisualSem as the foundation to achieve efficient and high-precision multimodal data retrieval.Firstly,use distributed technology to classify and store the data in the knowledge graph to improve retrieval efficiency.Secondly,this paper uses BabelNet to expand the knowledge graph through multi-ple filtering processes and increase the diversification of information.Finally,this paper builds a variety of retrieval models to achieve the fusion of retrieval results through linear combination methods to achieve high-precision language retrieval and image retrieval.The paper uses sentence retrieval and image retrieval experi-ments to prove that the platform can optimize the storage structure of the multi-modal knowledge graph and have good performance in multi-modal space.

Retrieval and Regional Distribution Analysis of Ammonia,Sulfur Dioxide and Nitrogen Dioxide in the Urban Environment Using Ultraviolet DOAS Algorithm

Aiming at the in situ and mobile observation of urban environmental air pollution,a portable instrument using ultraviolet spectrum retrieval algorithm was developed based on the basis of Differential Optical Absorption Spectroscopy(DOAS)and multiple-pass cell technique.Typical trace gas pollutants,NH3,SO2,and NO2,were explored using their optical spectral characteristics in deep ultraviolet wavelength range from 210 to 215 nm.The gas concentration was retrieved by Lambert-Beer’s law and nonlinear least square method.With an optimized optical alignment,the detection limits of NH3,SO2,NO2 were estimated to be 2.2,2.3,and 36.2 ppb,respectively.The system was used in carrying out some cruise observations in Chengdu,China.During the entire period,the polluted gases showed varied distribution and typical daily average concentrations ofNH3,SO2,NO2 were 23.2,3.5,and 106.0 ppb,respectively.The contributions from different sources were analyzed combined with the HYSPLIT model.Results show that the portable DOAS system is a convenient and effective tool for regional distribution measurement and pollution source monitoring.

Similarity matching method of power distribution system operating data based on neural information retrieval

Operation control of power systems has become challenging with an increase in the scale and complexity of power distribution systems and extensive access to renewable energy.Therefore,improvement of the ability of data-driven operation management,intelligent analysis,and mining is urgently required.To investigate and explore similar regularities of the historical operating section of the power distribution system and assist the power grid in obtaining high-value historical operation,maintenance experience,and knowledge by rule and line,a neural information retrieval model with an attention mechanism is proposed based on graph data computing technology.Based on the processing flow of the operating data of the power distribution system,a technical framework of neural information retrieval is established.Combined with the natural graph characteristics of the power distribution system,a unified graph data structure and a data fusion method of data access,data complement,and multi-source data are constructed.Further,a graph node feature-embedding representation learning algorithm and a neural information retrieval algorithm model are constructed.The neural information retrieval algorithm model is trained and tested using the generated graph node feature representation vector set.The model is verified on the operating section of the power distribution system of a provincial grid area.The results show that the proposed method demonstrates high accuracy in the similarity matching of historical operation characteristics and effectively supports intelligent fault diagnosis and elimination in power distribution systems.

Optimization in the Migration Problem of Mobile Agents in Distributed Information Retrieval Systems

In this paper, we employ genetic algorithms to solve the migration problem (MP). We propose a new encoding scheme to represent trees, which is composed of two parts: the pre-ordered traversal sequence of tree vertices and the children number sequence of corresponding tree vertices. The proposed encoding scheme has the advantages of simplicity for encoding and decoding, ease for GA operations, and better equilibrium between exploration and exploitation. It is also adaptive in that, with few restrictions on the length of code, it can be freely lengthened or shortened according to the characteristics of the problem space. Furthermore, the encoding scheme is highly applicable to the degree-constrained minimum spanning tree problem because it also contains the degree information of each node. The simulation results demonstrate the higher performance of our algorithm, with fast convergence to the optima or sub-optima on various problem sizes. Comparing with the binary string encoding of vertices, when the problem size is large, our algorithm runs remarkably faster with comparable search capability. Key words distributed information retrieval - mobile agents - migration problem - genetic algorithms CLC number TP 301. 6 Foundation item: Supported by the National Natural Science Foundation of China (90104005), the Natural Science Foundation of Hubei Province and the Hong Kong Polytechnic University under the grant G-YD63Biography: He Yan-xiang (1952-), male, Professor, research direction: distributed and parallel processing, multi-agent systems, data mining and e-business.

Research on the User Information Optimized Retrieval Technology in Distributed Database

The traditional method first classifies the user information and combines the query method to retrieve the interest information, but neglects to calculate the weight of the user interest information, which leads to the low retrieval accuracy. A retrieval method based on the fuzzy proximity classification technology is proposed. Approximation between the fuzzy sets is used to represent the consistency between the user interest information features, and the consistency calculation formula and the skewness confidence matrix between the user interest information features are given. The fuzzy classification of the user interest information can obtain the most consistent confidence data and eliminate the redundant approximation interference data. The probabilistic model of the information word frequency and the user interest information length calculates the weight of the user interest information, and adjusts the weight formula constantly.

Optimizing Big Data Retrieval and Job Scheduling Using Deep Learning Approaches

Big data analytics in business intelligence do not provide effective data retrieval methods and job scheduling that will cause execution inefficiency and low system throughput.This paper aims to enhance the capability of data retrieval and job scheduling to speed up the operation of big data analytics to overcome inefficiency and low throughput problems.First,integrating stacked sparse autoencoder and Elasticsearch indexing explored fast data searching and distributed indexing,which reduces the search scope of the database and dramatically speeds up data searching.Next,exploiting a deep neural network to predict the approximate execution time of a job gives prioritized job scheduling based on the shortest job first,which reduces the average waiting time of job execution.As a result,the proposed data retrieval approach outperforms the previous method using a deep autoencoder and Solr indexing,significantly improving the speed of data retrieval up to 53%and increasing system throughput by 53%.On the other hand,the proposed job scheduling algorithmdefeats both first-in-first-out andmemory-sensitive heterogeneous early finish time scheduling algorithms,effectively shortening the average waiting time up to 5%and average weighted turnaround time by 19%,respectively.

3D Engineering Model Retrieval Based on Enhanced Shape Distributions

To reuse and share the valuable knowledge embedded in repositories of engineering models for accelerating the design process, improving product quality, and reducing costs, it is crucial to devise search engines capable of matching 3D models efficiently and effectively. In this paper, an enhanced shape distributions-based technique of using geometrical and topological information to search 3D engineering models represented by polygonal meshes was presented. A simplification method of polygonal meshes was used to simplify engineering model as the pretreatment for generation of sample points. The method of sampling points was improved and a pair of functions that was more sensitive to shape was employed to construct a 2D shape distribution. Experiments were conducted to evaluate the proposed algorithm utilizing the Engineering Shape Benchmark (ESB) database. The experiential results suggest that the search effectiveness is significantly improved by enforcing the simplification and enhanced shape distributions to engineering model retrieval.

The retrieval of the nonlinear random waves from the non-Gaussian characteristics of the sea surface elevation distribution

In this paper, without recourse to the nonlinear dynamical equations of the waves, the nonlinear random waves are retrieved from the non-Gaussian characteristic of the sea surface elevation distribution. The question of coincidence of the nonlinear wave profile, spectrum and its distributions of maximum (or minimum) values of the sea surface elevation with results derived from some existing nonlinear theories is expounded under the narrow-band spectrum condition. Taking the shoaling sea wave as an example, the nonlinear random wave process and its spectrum in shallow water are retrieved from both the non-Gaussian characteristics of the sea surface elevation distribution in shallow water and the normal sea waves in deep water and compared with the values actually measured. Results show that they can coincide with the actually measured values quite well, thus, this can confirm that the method proposed in this paper is feasible.

Raindrop Size Distribution Parameters Retrieved from Xinfeng C-Band Polarimetric Radar Observations

The objective of this research was to acquire a raindrop size distribution(DSDs)retrieved from C-band polarimetric radar observations scheme for the first time in south China.An observation period of the precipitation process was selected,and the shape-slope(μ-Λ)relationship of this region was statistically analyzed using the raindrop sample observations from the two-dimensional video disdrometer(2DVD)at Xinfeng Station,Guangdong Province.Simulated data of the C-band polarimetric radar reflectivity ZHHand differential reflectivity ZDRwere obtained through scattering simulation.The simulation data were combined with DSD fitting to determine the ZDR-Λand log10(ZHH/N0)-Λrelationships.Using Xinfeng C-band polarimetric radar observations ZDRand ZHH,the raindrop Gamma size distribution parametersμ,Λ,and N0were retrieved.A scheme for using C-band polarimetric radar to retrieve the DSDs was developed.This research revealed that during precipitation process,the DSDs obtained using the C-band polarimetric radar retrieval scheme are similar to the 2DVD observations,the precipitation characteristics of rainfall intensity(R),mass-weighted mean diameter(Dm)and intercept parameter(Nw)with time obtained by radar retrieval are basically consistent with the observational results of the 2DVD.This scheme establishes the relationship between the observations of the C-band polarimetric radar and the physical quantities of the numerical model.This method not only can test the prediction of the model data assimilation system on the convective scale and determine error sources,but also can improve the microphysical precipitation processes analysis and radar quantitative precipitation estimation.The present research will facilitate radar data assimilation in the future.

Integrating Color and Spatial Feature for Content-Based Image Retrieval

In this paper, we present a novel and efficient scheme for extracting, indexing and retrieving color images. Our motivation was to reduce the space overhead of partition-based approaches taking advantage of the fact that only a relatively low number of distinct values of a particular visual feature is present in most images. To extract color feature and build indices into our image database we take into consideration factors such as human color perception and perceptual range, and the image is partitioned into a set of regions by using a simple classifying scheme. The compact color feature vector and the spatial color histogram, which are extracted from the seqmented image region, are used for representing the color and spatial information in the image. We have also developed the region-based distance measures to compare the similarity of two images. Extensive tests on a large image collection were conducted to demonstrate the effectiveness of the proposed approach.

An advanced wind vector retrieval algorithm for the rotating fan-beam scatterometer

The rotating fan-beam scatterometer （RFSCAT） is a new type of satellite scatterometer that is proposed approximately 10 a ago. However, similar to other rotating scatterometers, relatively larger wind retrieval errors occur in the nadir and outer regions compared with the middle regions of the swath. For the RFSCAT with the given parameters, a wind direction retrieval accuracy decreases by approximately 9 in the outer regions compared with the middle region. To address this problem, an advanced wind vector retrieval algorithm for the RFSCAT is presented. The new algorithm features an adaptive extension of the range of wind direction for each wind vector cell position across the whole swath according to the distribution histogram of a retrieved wind direction bias. One hundred orbits of Level 2A data are simulated to validate and evaluate the new algorithm. Retrieval experiments demonstrate that the new advanced algorithm can effectively improve the wind direction retrieval accuracy in the nadir and outer regions of the RFSCAT swath. Approximately 1.6 and 9 improvements in the wind direction retrieval are achieved for the wind vector cells located at the nadir and the edge point of the swath, respectively.

Shape Retrieval Using ECPDH with Dynamic Programming

The matching and retrieval of the 2D shapes are challenging issues in object recognition and computer vision. In this paper, we propose a new object contour descriptor termed ECPDH (Elliptic Contour Points Distribution Histogram), which is based on the distribution of the points on an object contour under the polar coordinates. ECPDH has the essential merits of invariance to scale and translation. Dynamic Programming (DP) algorithm is used to measure the distance between the ECPDHs. The effectiveness of the proposed method is demonstrated using some standard tests on MPEG-7 shape database. The results show the precision and recall of our method over other recent methods in the literature.

杭州地区大气臭氧浓度垂直分布

利用国产双池型电化学反应臭氧探空仪在杭州开展为期1年(2021年12月~2022年11月)观测,分析了该地区臭氧廓线垂直分布、臭氧柱总量以及对流层臭氧柱含量特征,并评估AIRS和MLS反演的臭氧廓线产品以及AIRS、OMIDOAS和TROPOMI反演的臭氧柱总量产品.结果表明,臭氧探空仪与地面原位臭氧分析仪观测结果具有高一致性.杭州地区臭氧垂直结构呈显著月变化特征,5月在上对流层-下平流层区域(12~16km高度范围)呈现高浓度臭氧的下传结构.4~9月在边界层附近1.5km高度以下观测到高浓度臭氧含量层.边界层内臭氧含量夏季最高,垂直变化不明显;其次是春季、秋季和冬季,均表现为随高度增加而增大.臭氧探空廓线与卫星反演臭氧垂直廓线吻合较好,在对比高度范围内相对偏差大多<10%.杭州地区臭氧柱总量和对流层柱含量平均值分别为302.1DU和40.5DU,呈明显季节变化特征.臭氧柱总量和对流层柱含量最高的季节分别为春季和夏季,5月对流层臭氧柱含量最高.臭氧探空柱总量与AIRS、OMI DOAS和TROPOMI反演臭氧柱总量一致性较好,相关系数分别为0.78、0.88和0.76.

基于ElasticSearch的医疗数据检索系统的设计与实现

随着医疗技术的发展,医疗业务场景越加复杂,由此产生的医疗数据也越来越多,其来源复杂,结构多变,信息冗余,数据不完整。这些特性使得系统在进行检索时,无法快速、有效、精确地查询数据。为了解决这个问题,设计并实现了基于ElasticSearch的医疗数据检索系统。该系统将医疗数据进行标准化,填补缺失值,选取合适的分词算法进行分词,将处理后的数据存入ElasticSearch中,同时使用SpringBoot构建系统应用,消耗多个医疗基础业务系统产生的数据,最终形成统一的医疗数据检索系统,给用户提供便捷、精确的查询服务。

基于雨滴谱参数反演的C波段双偏振雷达降水类型分类方法

降水类型分类对分析区域降水微物理特征、多源降水融合误差模型的构建以及雷达定量测量降水估计等都很重要。本文基于2015~2016年南京信息工程大学C波段双偏振雷达数据和南京地区滴谱仪观测资料,提出一种适用于南京地区的雷达降水类型分类方法,并对降水类型分类结果进行对比验证。首先,基于滴谱仪降雨率时序数据、地基雷达反射率因子平面位置显示数据和地基雷达反射率因子时间—高度显示数据,筛选出36次典型层状和对流降水过程。随后,统计3个滴谱仪站点典型层状(对流)降水的雨滴谱(DSD)参数,拟合得到适用于南京地区的降水类型分类线。将基于滴谱数据统计拟合的分类线应用于基于变分法反演的地基雷达DSD参数,进行地基雷达降水类型分类。根据典型层状(对流)过程降水类型分离指数的时间—高度分布,并对比星载双频测雨雷达(DPR)降水分类产品,对分类效果进行验证。最后,将分类结果应用于雷达分类定量降水估计,进一步说明降水分类的应用效果。结果表明,南京地区3个滴谱仪站点的拟合分类线非常一致,3个站点的典型层状(对流)过程均能够很好地分离在分类线两侧;与DPR降水分类产品进行对比分析,发现南京地区分类线的分类效果相对于其他典型降水分类方法,对层状和对流降水的识别率整体最高,分别为84.56%和72.64%;基于降水分类的雷达定量降水估计的测雨精度均优于未分类的测雨公式,且基于差分传播相移率的测雨公式[R(KDP)]在四种分类测雨公式中整体性能最优,基于水平反射率因子的测雨公式[R(ZH)]在层状云降水反演中性能最优,基于差分传播相移率的测雨公式[R(KDP)]在对流云降水反演中性能最优,基于水平反射率因子和差分反射率的测雨公式[R(ZH,ZDR)]对原有总体测雨公式降水精度的提升最为明显。

近20年间鄱阳湖湿地春秋季植被生物量时空动态

湿地生态系统是碳密度最大的生态系统之一,摸清湿地植被生物量及其动态变化对于实现“双碳”目标具有重要意义。基于Landsat遥感影像对鄱阳湖湿地2000—2020年间春秋季的植被地上生物量进行了反演和估算,分析了植被生物量随时间的变化趋势以及空间热点区域。结果表明:①近20年间鄱阳湖湿地春季和秋季植被地上生物量分别为0.85×10^(9)~4.20×10^(9) g和0.68×10^(9)~6.69×10^(9) g,春季植被地上生物量随时间变化不明显,秋季植被地上生物量随时间呈上升趋势。②鄱阳湖湿地春季植被地上生物量分布热点区域面积为754.15 km^(2),秋季为1085.49 km^(2),分别占鄱阳湖湿地总面积的21.58%和30.66%。③鄱阳湖湿地春季和秋季植被地上生物量与月均温度呈显著正相关关系。鄱阳湖湿地植被碳汇功能较强,且呈平稳甚至上升趋势,有利于碳中和目标的实现。

Pobe:一种基于生成式模型的分布外文本检测方法

对于安全可靠的机器学习系统,具备检测训练集分布外(out-of-distribution,OOD)样本的能力十分必要.基于似然的生成式模型由于训练时不需要样本标签,是一类非常受欢迎的OOD检测方法.然而,近期研究表明通过似然来检测OOD样本往往会失效,并且失效原因与解决方案的探究仍较少,尤其是对于文本数据.从模型层面和数据层面分析文本上失效的原因:生成式模型的泛化性不足和文本先验概率的偏差.在此基础上,提出一种新的OOD文本检测方法Pobe.针对生成式模型泛化性不足的问题,引入KNN检索的方式,来提升模型的泛化性.针对文本先验概率偏差的问题,设计一种偏差校准策略,借助预训练语言模型改善概率偏差对OOD检测的影响,并通过贝叶斯定理证明策略的合理性.通过在广泛的数据集上进行实验,证明所提方法的有效性,其中,在8个数据集上的平均AUROC值超过99%,FPR95值低于1%.

保护搜索内容的字符串模式匹配算法研究

高效的在线字符串模式匹配算法对云数据库检索至关重要,然而搜索内容的泄露会威胁用户隐私。现有的字符串模式匹配算法没有考虑用户搜索内容的保护,可搜索加密方案虽然可以保护用户的搜索内容,但存在索引构建代价大、检索效率低等问题。因此,提出了两种保护用户搜索内容的模式匹配算法:基于分布式点函数的模式匹配(pattern matching based on distributed point function,PMDPF)算法和基于分布式点函数的跳跃式模式匹配(jumping pattern matching based on distributed point function,JPMDPF)算法。PMDPF算法利用指纹函数以及分布式点函数构造模式串真值表,并分发给两台独立的服务器,把搜索中字符对比操作转换为查表操作,从而保护搜索内容。为了提升搜索效率,提出了JPMDPF算法。通过字符跳转,JPMDPF算法以泄露更多信息为代价,其搜索效率比PMDPF算法平均提高了约m倍,其中m为搜索内容长度,同时显著降低了因指纹函数碰撞而导致的误判的概率。实验结果表明,PMDPF算法的搜索效率比基于指纹函数的经典算法提高约5%,并优于现有的可搜索加密方案,PMDPF算法的搜索耗时在搜索内容长度为4时是JPMDPF算法的4.2倍。

基于分布式技术的动车组车载大数据存储与检索技术研究

随着动车组的数据传输能力升级改造以及复兴号智能型动车组的上线运营,动车组监测项点数据快速扩充,如何高效、准确地对动车组的海量车载非实时数据进行存储、检索与查询,成为提升动车组车载数据管理和应用水平面临的一个挑战。文中提出一种基于分布式技术的海量非实时数据地面系统架构,将各动车运用所的数据服务站点作为子节点,全路数据服务中心作为主节点,对具备时空属性的海量数据快速检索算法进行了研究,构建了一种支持分布式数据存储、检索及数据分析的架构,开发了原型系统,验证了解决方案的可行性和有效性。

基于分布式缓存的科技文献信息动态检索系统设计

为便于科研人员研究科技发展前沿动态,设计基于分布式缓存的科技文献信息动态检索系统,为用户匹配最佳检索结果。用户层将用户检索需求传输至处理层数据采集模块,该模块采集数据信息后经网络层通信模块将信息传输至数据层科技文献信息数据库中存储;数据层分布式存储模块采用分布式缓存方法将以往的检索记录信息分别存储至热数据库及冷数据库中;利用网络层传输科技文献信息数据库内的检索需求信息至处理层检索处理模块,通过科技文献信息动态检索方法处理接收到的检索信息,将处理后信息与数据层分布式存储模块内的关联数据进行排序重组,获取检索结果,并通过用户层展示检索结果。结果表明,该系统可有效且准确地为用户检索其所需的科技文献信息,检索MAP值为0.95以上,系统具备较快的读写速度与较强的数据传输能力。