检索结果-内蒙古大学图书馆

您好，读者！请登录

内蒙古大学图书馆

首页
概况
党建
资源
服务
科研支持
- 论文收录引用证明
- 科技查新
知识产权
档案馆
帮助

咨询与建议

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

您的常用邮箱：*

您的手机号码：*

问题描述：

当前已输入0个字，您还可以输入200个字

全部搜索
期刊论文
图书
学位论文
标准
纸本馆藏
外文资源发现
数据库导航
超星发现

高级检索

时间限定

出版年份：

文献类型

图书期刊文献学位论文多媒体

馆藏选择

电子馆藏纸本馆藏

核心期刊

全部期刊 SCI 收录期刊 SSCI 收录期刊 EI 收录期刊 CSCD 收录期刊 CSSCI 收录期刊

语言

中文英文

文献类型

期刊文献图书学位论文标准纸本馆藏

帮助

文字说明：

T=题名（书名、题名），A=作者（责任者），K=主题词，P=出版物名称，PU=出版社名称，O=机构（作者单位、学位授予单位、专利申请人），L=中图分类号，C=学科分类号，U=全部字段，Y=年（出版发行年、学位年度、标准发布年）

检索规则说明：

AND代表“并且”；OR代表“或者”；NOT代表“不包含”；(注意必须大写,运算符两边需空一格)

检索范例：

范例一：(K=图书馆学 OR K=情报学) AND A=范并思 AND Y=1982-2016
范例二：P=计算机应用与软件 AND (U=C++ OR U=Basic) NOT K=Visual AND Y=2011-2016

分类表

所选分类

>> <<

限定检索结果

文献类型

16,237 篇 会议
368 篇 期刊文献
22 册 图书

馆藏范围

16,627 篇 电子文献
0 种 纸本馆藏

日期分布

学科分类号

9,336 篇 工学
- 8,536 篇 计算机科学与技术...
- 4,019 篇 软件工程
- 1,982 篇 电气工程
- 1,383 篇 信息与通信工程
- 676 篇 电子科学与技术（可...
- 535 篇 控制科学与工程
- 228 篇 网络空间安全
- 188 篇 仪器科学与技术
- 141 篇 机械工程
- 115 篇 生物医学工程（可授...
- 106 篇 动力工程及工程热...
- 105 篇 测绘科学与技术
- 97 篇 光学工程
- 91 篇 生物工程
- 82 篇 建筑学
- 70 篇 土木工程
- 63 篇 环境科学与工程（可...
- 61 篇 安全科学与工程
1,973 篇 理学
- 1,505 篇 数学
- 245 篇 物理学
- 203 篇 统计学（可授理学、...
- 177 篇 系统科学
- 115 篇 生物学
- 100 篇 地球物理学
- 69 篇 化学
1,463 篇 管理学
- 1,205 篇 管理科学与工程(可...
- 468 篇 工商管理
- 321 篇 图书情报与档案管...
106 篇 医学
- 86 篇 临床医学
96 篇 经济学
- 93 篇 应用经济学
56 篇 法学
53 篇 农学
18 篇 教育学
12 篇 文学
9 篇 军事学
1 篇 艺术学

主题

2,212 篇 parallel process...
1,199 篇 computer archite...
1,130 篇 concurrent compu...
1,116 篇 distributed comp...
1,063 篇 computational mo...
1,037 篇 application soft...
1,017 篇 distributed proc...
990 篇 hardware
905 篇 computer science
708 篇 graphics process...
595 篇 runtime
527 篇 scalability
518 篇 parallel process...
507 篇 algorithm design...
494 篇 parallel program...
490 篇 parallel algorit...
470 篇 graphics process...
460 篇 kernel
446 篇 processor schedu...
440 篇 conferences

机构

38 篇 ibm thomas j. wa...
33 篇 college of compu...
31 篇 school of comput...
27 篇 oak ridge nation...
26 篇 university of ch...
26 篇 oak ridge natl l...
25 篇 georgia inst tec...
25 篇 ohio state univ ...
24 篇 department of co...
23 篇 tsinghua univers...
23 篇 pacific northwes...
21 篇 argonne national...
21 篇 oak ridge nation...
20 篇 georgia inst tec...
19 篇 college of compu...
19 篇 school of comput...
19 篇 department of co...
19 篇 argonne natl lab...
19 篇 pacific northwes...
19 篇 national laborat...

作者

39 篇 jack dongarra
31 篇 dongarra jack
29 篇 zomaya albert y.
26 篇 bader david a.
23 篇 feng wu-chun
22 篇 boukerche azzedi...
19 篇 hoefler torsten
18 篇 gagan agrawal
18 篇 schulz martin
16 篇 dhabaleswar k. p...
16 篇 p. sadayappan
16 篇 wang yijie
15 篇 ito yasuaki
15 篇 yves robert
14 篇 h. casanova
14 篇 alexey lastovets...
14 篇 azad ariful
13 篇 dongsheng li
13 篇 wang guojun
13 篇 kishore kothapal...

语言

16,553 篇 英文
44 篇 其他
27 篇 中文
2 篇 土耳其文
1 篇 葡萄牙文

检索条件"任意字段=IEEE International Symposium on Parallel and Distributed Processing with Applications"

共 16627 条记录，以下是341-350 订阅

全选清除本页清除全部题录导出标记到"检索档案"

详细简洁

排序：

Computation Augmentation Techniques for Computing Continuum 24

Computation Augmentation Techniques for Computing Continuum

引用

24th ieee/ACM international symposium on Cluster, Cloud, and Internet Computing (CCGrid)

作者： Grosu, George-Mircea Nistor, Silvia-Elena Ciobanu, Radu-Ioan Kolodziej, Joanna Pop, Florin Natl Univ Sci & Technol Politehn Bucharest Bucharest Romania Natl Inst Res & Dev Informat ICI Bucharest Romania NASK Warsaw Warsaw Poland Cracow Univ Technol Krakow Poland Acad Romanian Scientists Bucharest Romania Cracow Univ Technol Dept Comp Sci Krakow Poland

ISBN: (纸本)9798350377521;9798350377514

We are witnessing a tremendous expansion in computational power, leading to monumental technological advances with various practical applications. The need for parallelism originated from the limitations of single-processing performance. Increasing the transistor density on integrated circuits was the way to go, thus creating faster and more complex monolithic processors. However, these circuits approached their limits of hardware expansion, so the single-processor model has become unreliable and insufficient overall. This is how multicore architectures were born and, with them, the notion of parallel computing. Our research background in the Drop Computing paradigm and Interest in studying parallel programming paradigms encouraged us to propose a new approach towards parallel processing in mobile, ad-hoc, opportunistic networks. This paper adapts one fundamental message-passing standard for parallel architectures within the Drop Computing context. We define a new library DroMPI. Besides the challenges of parallel programming, the solution has to address the challenges imposed by hardware constraints, limited resources, and the decentralized model specific to Drop Computing.

关键词： Drop Computing Computing Continuum Augmentation Multicore Architectures parallel Programming Performance Analysis

来源：评论

学校读者我要写书评

暂无评论

Peachy parallel Assignments (EduPar 2022) 36

Peachy Parallel Assignments (EduPar 2022)

引用

36th ieee international parallel and distributed processing symposium (ieee IPDPS)

作者： Buecker, H. Martin Casanova, Henri da Silva, Rafael Ferreira Lasserre, Alice Luyen, Derrick Namyst, Raymond Schoder, Johannes Wacrenier, Pierre-Andre Bunde, David P. Friedrich Schiller Univ Jena Inst Comp Sci Jena Germany Univ Hawaii Informat & Comp Sci Honolulu HI 96822 USA Oak Ridge Natl Lab Natl Ctr Computat Sci Oak Ridge TN USA Univ Bordeaux Comp Sci Dept Inria Bordeaux Sud Ouest Talence France Knox Coll Comp Sci Dept Galesburg IL USA

ISBN: (纸本)9781665497473

The presentation of Peachy parallel Assignments in several workshops on parallel and distributed computing education aims to promote the reuse of highquality assignments, both saving precious faculty time and improving the quality of course assignments. Presented assignments are selected competitively- they must have been successfully used in a real classroom, be easy for other instructors to adopt, and be "cool and inspirational" to encourage students to spend time on them and talk about them with others. Winning assignments are also archived on the Peachy parallel Assignments website. In this installment of Peachy parallel Assignments, we present three new assignments. The first assignment is to simulate an Abelian Sandpile, with grains of sand moving from tall piles to shorter ones. This is a discrete simulation that creates colorful and intricate images. The second assignment is a Big Data problem in which students use the MapReduce paradigm to recreate "Warming Stripes", a visualization of climate data that highlights climate change. The third assignment introduces climate-oriented optimization by asking students to schedule distributed workflows to minimize their carbon footprint.

关键词： Peachy parallel Assignments parallel computing education High-Performance Computing education parallel programming Curriculum Development Abelian Sandpile parallel Simulation MapReduce Big Data Warming Stripes distributed Workflow Scheduling Carbon Footprint

来源：评论

学校读者我要写书评

暂无评论

MIMDRAM: An End-to-End processing-Using-DRAM System for High-Throughput, Energy-Efficient and Programmer-Transparent Multiple-Instruction Multiple-Data Computing 30

MIMDRAM: An End-to-End Processing-Using-DRAM System for High...

引用

30th ieee international symposium on High-Performance Computer Architecture (HPCA)

作者： Oliveira, Geraldo F. Olgun, Ataberk Yaglikci, Abdullah Giray Bostanci, F. Nisa Gomez-Luna, Juan Ghose, Saugata Mutlu, Onur Swiss Fed Inst Technol Zurich Switzerland Univ Illinois Champaign IL USA

ISBN: (纸本)9798350393132;9798350393149

processing-using-DRAM (PUD) is a processing-in-memory (PIM) approach that uses a DRAM array's massive internal parallelism to execute very-wide (e.g., 16,384-262,144-bit-wide) data-parallel operations, in a single-instruction multiple-data (SIMD) fashion. However, DRAM rows' large and rigid granularity limit the effectiveness and applicability of PUD in three ways. First, since applications have varying degrees of SIMD parallelism (which is often smaller than the DRAM row granularity), PUD execution often leads to underutilization, throughput loss, and energy waste. Second, due to the high area cost of implementing interconnects that connect columns in a wide DRAM row, most PUD architectures are limited to the execution of parallel map operations, where a single operation is performed over equally-sized input and output arrays. Third, the need to feed the wide DRAM row with tens of thousands of data elements combined with the lack of adequate compiler support for PUD systems create a programmability barrier, since programmers need to manually extract SIMD parallelism from an application and map computation to the PUD hardware. Our goal is to design a flexible PUD system that overcomes the limitations caused by the large and rigid granularity of PUD. To this end, we propose MIMDRAM, a hardware/software co-designed PUD system that introduces new mechanisms to allocate and control only the necessary resources for a given PUD operation. The key idea of MIMDRAM is to leverage finegrained DRAM (i.e., the ability to independently access smaller segments of a large DRAM row) for PUD computation. MIMDRAM exploits this key idea to enable a multiple-instruction multiple-data (MIMD) execution model in each DRAM subarray (and SIMD execution within each DRAM row segment). We evaluate MIMDRAM using twelve real-world applications and 495 multi-programmed application mixes. Our evaluation shows that MIMDRAM provides 34x the performance, 14.3x the energy efficiency, 1.7x the throughp

关键词： DRAM energy-efficiency hardware/software co-design memory-centric computing processing-in-memory

来源：评论

学校读者我要写书评

暂无评论

The MIT Supercloud Workload Classification Challenge 36

The MIT Supercloud Workload Classification Challenge

引用

36th ieee international parallel and distributed processing symposium (ieee IPDPS)

作者： Tang, Benny J. Chen, Qiqi Weiss, Matthew L. Frey, Nathan C. McDonald, Joseph Bestor, David Yee, Charles Arcand, William Bergeron, William Byun, Chansup Edelman, Daniel Houle, Michael Hubbell, Matthew Jones, Michael Kepner, Jeremy Klein, Anna Michaleas, Adam Michaleas, Peter Milechin, Lauren Mullen, Julia Prout, Andrew Reuther, Albert Rosa, Antonio Bowne, Andrew McEvoy, Lindsey Li, Baolin Tiwari, Devesh Gadepally, Jiay Samsi, Siddharth MIT 77 Massachusetts Ave Cambridge MA 02139 USA MIT Lincoln Lab 244 Wood St Lexington MA 02173 USA Northeastern Univ Boston MA 02115 USA US Air Force Cambridge MA USA

ISBN: (纸本)9781665497473

High-Performance Computing (HPC) centers and cloud providers support an increasingly diverse set of applications on heterogenous hardware. As Artificial Intelligence (AI) and Machine Learning (ML) workloads have become an increasingly larger share of the compute workloads, new approaches to optimized resource usage, allocation, and deployment of new AI frameworks are needed. By identifying compute workloads and their utilization characteristics, HPC systems may be able to better match available resources with the application demand. By leveraging datacenter instrumentation, it may be possible to develop AI-based approaches that can identify workloads and provide feedback to researchers and datacenter operators for improving operational efficiency. To enable this research, we released the MIT Supercloud Dataset, which provides detailed monitoring logs from the MIT Supercloud cluster. This dataset includes CPU and GPU usage by jobs, memory usage, and file system logs. In this paper, we present a workload classification challenge based on this dataset. We introduce a labelled dataset that can be used to develop new approaches to workload classification and present initial results based on existing approaches. The goal of this challenge is to foster algorithmic innovations in the analysis of compute workloads that can achieve higher accuracy than existing methods. Data and code will be made publicly available via the Datacenter Challenge website : https://***.

关键词： Technological innovation distributed processing File systems Instruments High performance computing Graphics processing units Machine learning

来源：评论

学校读者我要写书评

暂无评论

Efficient and Reliable Clustering by parallel Random Swap Algorithm 26

Efficient and Reliable Clustering by Parallel Random Swap Al...

引用

26th ieee/ACM international symposium on distributed Simulation and Real Time applications (DS-RT)

作者： Nigro, Libero Cicirelli, Franco Franti, Pasi Univ Calabria DIMES Dept Informat Modelling Elect & Syst Sci I-87036 Arcavacata Di Rende Italy Natl Res Council Italy CNR Inst High Performance Comp & Networking ICAR I-87036 Arcavacata Di Rende Italy Univ Eastern Finland Sch Comp Machine Learning Grp POB 111 Joensuu 80101 Finland

ISBN: (数字)9781665497992

ISBN: (纸本)9781665497992

Solving large-scale clustering problems requires an efficient algorithm which can be implemented also in parallel. Kmeans would be suitable but it can lead to an inaccurate clustering result. To overcome this problem, we present a parallel version of random swap clustering algorithm. It combines the scalability of k-means with high clustering accuracy. The new clustering method is experimented on top of Java parallel streams and lambda expressions, which offer interesting execution time benefits. The method is applied to standard benchmark datasets, with a varying population size and distribution of managed records, dimensionality of data points and the number of clusters. The experimental results confirm that high quality clustering can be obtained by parallel random swap together with a high time efficiency.

关键词： Clustering problem K-Means Random swap parallelism Streams Lambda Expressions Java

来源：评论

学校读者我要写书评

暂无评论

Synchronous parallel multisplitting method with convergence acceleration using a local Krylov-based minimization for solving linear systems 36

Synchronous parallel multisplitting method with convergence ...

引用

36th ieee international parallel and distributed processing symposium (ieee IPDPS)

作者： Tchakorom, Medane A. Couturier, Raphael Charr, Jean-Claude Univ Bourgogne Franche Comte UBFC CNRS FEMTO ST Inst Belfort France

ISBN: (纸本)9781665497473

Computer simulations of physical phenomena, such as heat transfer, often require the solution of linear equations. These linear equations occur in the form Ax = b, where A is a matrix, b is a vector, and x is the vector of unknowns. Iterative methods are the most adapted to solve large linear systems because they can be easily parallelized. This paper presents a variant of the multisplitting iterative method with convergence acceleration using the Krylov-based minimization method. This paper particularly focuses on improving the convergence speed of the method with an implementation based on the PETSc (Portable Extensible Toolkit for Scientific Computation) library. This was achieved by reducing the need for synchronization - data exchange - during the minimization process and adding a preconditioner before the multisplitting method. All experiments were performed either over one or two sites of the Grid5000 platform and up to 128 cores were used. The results for solving a 2D Laplacian problem of size 1024(2) components, show a speed up of up to 23X and 86X when respectively compared to the algorithm in [8] and to the general multisplitting implementation.

关键词： Iterative methods Krylov methods Multisplitting Linear Solvers PETSc MPI

来源：评论

学校读者我要写书评

暂无评论

Efficient distributed Secure Memory with Migratable Merkle Tree 29

Efficient Distributed Secure Memory with Migratable Merkle T...

引用

29th ieee international symposium on High-Performance Computer Architecture (HPCA)

作者： Feng, Erhu Du, Dong Xia, Yubin Chen, Haibo Shanghai Jiao Tong Univ Inst Parallel & Distributed Syst Shanghai Peoples R China Shanghai Artificial Intelligence Lab Shanghai Peoples R China

ISBN: (纸本)9781665476522

Hardware-assisted enclaves with memory encryption have been widely adopted in the prevailing architectures, e.g., Intel SGX/TDX, AMD SEV, ARM CCA, etc. However, existing enclave designs fall short in supporting efficient cooperation among cross-node enclaves (i.e., multi-machines) because the range of hardware memory protection is within a single node. A naive approach is to leverage cryptography at the application level and transfer data between nodes through secure channels (e.g., SSL). However, it incurs orders of magnitude costs due to expensive encryption/decryption, especially for distributed applications with large data transfer, e.g., MapReduce and graph computing. A secure and efficient mechanism of distributed secure memory is necessary but still missing. This paper proposes Migratable Merkle Tree (MMT), a design enabling efficient distributed secure memory to support distributed confidential computing. MMT sets up an integrity forest for distributed memory on multiple nodes. It allows an enclave to securely delegate an MMT closure, which contains both data and metadata of a subtree, to a remote enclave. By reusing the memory encryption mechanisms of existing enclaves, our design achieves secure data transfer without software re-encryption. We have implemented a prototype of MMT and a trusted firmware for management, and further applied MMT to real-world distributed applications. The evaluation results show that compared with existing systems using the AES-NI instruction, MMT can achieve up to 13x speedup on data transferring, and gain 12%similar to 58% improvement on the end-to-end performance of MapReduce and PageRank.

关键词： MapReduce

来源：评论

学校读者我要写书评

暂无评论

A Moving Target Defense Approach for the distributed Dynamic Network 21

A Moving Target Defense Approach for the Distributed Dynamic...

引用

21st ieee international symposium on parallel and distributed processing with applications, 13th ieee international Conference on Big Data and Cloud Computing, 16th ieee international Conference on Social Computing and Networking and 13th international Conference on Sustainable Computing and Communications, ISPA/BDCloud/SocialCom/SustainCom 2023

作者： Zhang, Lin Guo, Yunchuan Leng, Siyuan Li, Zifu Li, Fenghua Fang, Liang Institute of Information Engineering Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences School of Cyber Security Beijing China

ISBN: (纸本)9798350329223

The distributed dynamic network is vulnerable to scanning attacks due to the openness of wireless channels. Traditional defense systems tend to be passive and exhibit delayed responses. A moving target defense approach, namely distributed Network Address Shuffling (DNAS), is proposed to thwart attackers' network scanning through the shuffling of network addresses. To resolve address conflicts resulting from this shuffling, DNAS employs a dynamic diffusion method of allocated addresses before the shuffling process to reduce the probability of conflict generation, and utilizes a passive detection based conflict elimination algorithm after the shuffling process to eliminate any generated conflicts. To select low-risk addresses, DNAS leverages an artificial feature selection based Fully Connected Neural Network (FCNN) to recognize the attacker's scanning policy, and identifies low-risk addresses based on the scanning range of the policy. Empirical experiments and theoretical analysis indicate that DNAS significantly reduces the probability of address conflict generation at a minimal cost. It effectively eliminates all generated address conflicts within an average conflict resolution time of less than 500ms. Furthermore, DNAS exhibits an accuracy of 99.45% in recognizing scanning policies, surpassing pseudorandom address hopping in diminishing the success rate of sequential, local random, and mixed scanning. © 2023 ieee.

关键词： Network security

来源：评论

学校读者我要写书评

暂无评论

AxoNN: An asynchronous, message-driven parallel framework for extreme-scale deep learning 36

AxoNN: An asynchronous, message-driven parallel framework fo...

引用

36th ieee international parallel and distributed processing symposium (ieee IPDPS)

作者： Singh, Siddharth Bhatele, Abhinav Univ Maryland Dept Comp Sci College Pk MD 20742 USA

ISBN: (纸本)9781665481069

In the last few years, the memory requirements to train state-of-the-art neural networks have far exceeded the DRAM capacities of modern hardware accelerators. This has necessitated the development of efficient algorithms to train these neural networks in parallel on large-scale GPU-based clusters. Since computation is relatively inexpensive on modern GPUs, designing and implementing extremely efficient communication in these parallel training algorithms is critical for extracting the maximum performance. This paper presents AxoNN, a parallel deep learning framework that exploits asynchrony and message-driven execution to schedule neural network operations on each GPU, thereby reducing GPU idle time and maximizing hardware efficiency. By using the CPU memory as a scratch space for offloading data periodically during training, AxoNN is able to reduce GPU memory consumption by four times. This allows us to increase the number of parameters per GPU by four times, thus reducing the amount of communication and increasing performance by over 13%. When tested against large transformer models with 12-100 billion parameters on 48-384 NVIDIA Tesla V100 GPUs, AxoNN achieves a per-GPU throughput of 49.4-54.78% of theoretical peak and reduces the training time by 2237 days (15-25% speedup) as compared to the state-of-the-art.

关键词： parallel deep learning asynchrony message driven scheduling memory optimizations

来源：评论

学校读者我要写书评

暂无评论

High-Speed Serial and Semi-parallel IMPLY-based Approximate Adders through Memristors for In-Memory Computing

High-Speed Serial and Semi-Parallel IMPLY-based Approximate ...

引用

ieee international symposium on Circuits and Systems (ISCAS)

作者： Kaushik, Nandit Krishna, Hemanth Bodapati, Srinivasu Indian Inst Technol Mandi Sch Comp & Elect Engn SCEE Mandi 175005 Himachal Prades India

ISBN: (纸本)9798350330991;9798350331004

This paper introduces three memristive approximate adder designs using Material Implication (IMPLY) logic. Two designs follow a serial approach, while the third adopts a semiparallel technique. The proposed design achieves the approximate output in just 2 steps, utilizing only 3 memristors, in contrast to the 8 steps and 5 memristors required by the best competing IMPLY-based approximate adder. Furthermore, when integrated into an 8-bit Ripple Carry Adder (RCA), the proposed designs enhance latency by 4% to 28% compared to the existing IMPLYbased approximate design. Compared to exact serial and semiparallel adders in literature, latency is improved by 27% to 56.8%. Similarly, an improvement in the range of 4% to 42% is achieved for energy consumption. The performance of the proposed approximate adder designs is then evaluated in image processing applications, specifically Image Masking and RGB to Grayscale conversion. The results demonstrate better Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) values compared to the existing approximate IMPLY design.

关键词： Memristor IMPLY Approximate Adder IMC

来源：评论

学校读者我要写书评

暂无评论

没有更多数据了...

全选清除本页清除全部题录导出标记到“检索档案”

共500页 << < 31 32 33 34 35 36 37 38 39 40 > >>

检索报告对象比较合并检索0

隐藏清空

合并搜索

回到顶部

执行限定条件

内容：

评分：

请选择保存的检索档案：

请选择收藏分类：

订阅名称：

通借通还

温馨提示：

图书名称：

借书校区：

取书校区：

手机号码：

邮箱地址：

一卡通帐号：

电话和邮箱必须正确填写，我们会与您联系确认。

联系人：

所在院系：

联系邮箱：

联系电话：

内蒙古自治区呼和浩特市赛罕区大学西街235号邮编: 010021

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：