检索结果-内蒙古大学图书馆

Discrete Prompt Compression With Reinforcement Learning

IEEE ACCESS 2024年 12卷 72578-72587页

作者： Jung, Hoyoun Kim, Kyung-Joong Gwangju Inst Sci & Technol Sch Integrated Technol Gwangju 61005 South Korea

Compressed prompts aid instruction-tuned language models (LMs) in overcoming context window limitations and reducing computational costs. Existing methods, which are primarily based on training embeddings, face various challenges associated with interpretability, the fixed number of embedding tokens, reusability across different LMs, and inapplicability when interacting with black-box APIs. This study proposes prompt compression with reinforcement learning (PCRL), which is a discrete prompt compression method that addresses these issues. The proposed PCRL method utilizes a computationally efficient policy network that edits prompts directly. The training approach employed in the proposed PCRLs can be applied flexibly to various types of LMs, including both decoder-only and encoder-decoder architecture and it can be trained without gradient access to the LMs or labeled data. The proposed PCRL achieves an average reduction of 24.6% in terms of the token count across various instruction prompts while maintaining sufficient performance. In addition, we demonstrate that the learned policy can be transferred to larger LMs, and through a comprehensive analysis, we explore the token importance within the prompts. The source code is available at https://***/nenomigami/PromptCompressor.

关键词： Training Task analysis Optimization Closed box Computational efficiency Chatbots Reinforcement learning Large language models application programming interfaces prompt compression reinforcement learning computational efficiency black-box APIs

来源：评论

学校读者我要写书评

暂无评论

Decision as a Service for Transaction Banking Using Service-Oriented Modeling Architecture Methodology

引用

IEEE ACCESS 2023年 11卷 41455-41466页

作者： Gaol, Ford Lumban Nugraha, Grahita Djati Matsuo, Tokuro Bina Nusantara Univ Comp Sci Dept West Jakarta 11480 Indonesia Adv Inst Ind Technol AIIT Comp Sci Dept Tokyo 1400011 Japan

In order to gain a competitive edge in the banking industry, direct face-to-face interactions between customers and banks are being gradually replaced by virtual interactions. Instead, they communicate using electronic devices such as smartphones, tablets, and web applications. Transaction banking, which provides commercial services for banking products to small, medium, and large corporations, is a highly typical work unit in banking. Shipping and international payments, risk management for international trade, and other services are provided. We use the Service Oriented Modelling Architecture (SOMA) methodology to solve the problem of the transaction banking unit by relocating it from a system with general dependencies to a system that is independent of the entire service and operates on a small function scale. The study found that a decision as a service (DAAS) model with an Enterprise Service Bus (ESB), Business Process Management (BPM), and Business Rule Management (BRM) solution can provide guidelines for the design of decision rules pertaining to integrated or separate business modeling, thereby aiding business unit delivery.

关键词： Banking Business Service-oriented systems engineering Computer architecture System integration Collaboration Adaptation models application programming interfaces Process monitoring Service oriented modelling architecture (SOMA) enterprise service bus (ESB) business process management (BPM) business rule management (BRM) application programming interface (API) transaction banking

来源：评论

学校读者我要写书评

暂无评论

Enhanced Dynamic Analysis for Malware Detection With Gradient Attack

引用

IEEE SIGNAL PROCESSING LETTERS 2024年 31卷 2825-2829页

作者： Yan, Pei Tan, Shunquan Wang, Miaohui Huang, Jiwu Shenzhen Key Lab Media Secur Shenzhen 518000 Peoples R China Shenzhen Univ Guangdong Key Lab Intelligent Informat Proc Shenzhen 518060 Peoples R China Shenzhen City Polytech Sch Informat & Commun Technol Shenzhen 518116 Peoples R China Shenzhen MSU BIT Univ Guangdong Lab Machine Percept & Intelligent Comp Fac Comp Shenzhen 518116 Peoples R China

Malware detection is an effective way to prevent the intrusion of malware into computer systems, and the API-based dynamic analysis method can effectively detect obfuscated and packaged malware. However, existing methods still suffer from limited detection accuracy and weak generalization. To address this issue, this paper presents a gradient attack-based malware dynamic analysis method. Through exerting adversarial noise into the embedding layer, the malware detection model can learn more robust representations of API sequences during training, achieving broader coverage of sample representations. The strategy of normalizing attack noise and recovering attacked representation is designed, which controls the strength of the gradient attack within a reasonable range and prevents a negative impact on the model's detection performance. The proposed method can be applied to existing API-based malware detection models to enhance their detection performance, indicating the strong generality of the proposed method. Experimental results on two benchmark datasets (i.e., Aliyun and Catak) demonstrate the effectiveness of the proposed gradient attack method, which further improves the detection performance of the mainstream API-based models, with an average accuracy increase of 2.80% and 3.66% on these two datasets, respectively.

关键词： Malware Training Noise Feature extraction application programming interfaces Perturbation methods Backpropagation Vocabulary Vectors Accuracy Adversarial method dynamic analysis malware detection network security

来源：评论

学校读者我要写书评

暂无评论

Accelerating Digital Transformation Visual Insights from the API Ecosystem

引用

IT PROFESSIONAL 2016年第6期18卷 20-25页

作者： Basole, Rahul C. Georgia Inst Technol Sch Interact Comp Atlanta GA 30332 USA Georgia Inst Technol Tennenbaum Inst Atlanta GA 30332 USA Georgia Inst Technol GVU Ctr Atlanta GA 30332 USA

Digital transformation is accelerating due to the rapid emergence of open APIs. Using a curated dataset of nearly 15,000 APIs, the author visualizes the structure of the API ecosystem, revealing distinct clusters with... 详细信息

关键词： Ecosystems Companies Technological Innovation Mashups Mobile Communication Cloud Computing Digital Systems application programming interfaces Information Management API Digital Innovation Digital Transformation Visualization

来源：评论

学校读者我要写书评

暂无评论

Enhancing Security Testing for Identity Management Implementations: Introducing Micro-Id-Gym Language and Micro-Id-Gym Testing Tool

引用

IEEE SECURITY & PRIVACY 2024年第6期22卷 50-61页

作者： Bisegna, Andrea Bitussi, Matteo Carbone, Roberto Ranise, Silvio Fdn Bruno Kessler Trento Italy Univ Trento Trento Italy

We introduce MIG-L, a declarative language for the specification of security tests, and MIG-T, a testing tool, for identity management solutions based on SAML and OAuth/OpenID Connect. We verify compliance with best c... 详细信息

关键词： Security Threat modeling Testing HTTP application programming interfaces Standards Browsers

来源：评论

学校读者我要写书评

暂无评论

Real-Time Communications Security on the Web

引用

IEEE INTERNET COMPUTING 2014年第6期18卷 8-10页

作者： Desmet, Lieven Johns, Martin Univ Leuven iMinds DistriNet Res Grp Leuven Belgium SAP AG Secur & Trust Grp Walldorf Germany SAP AG Secur & Trust Grp Web Applicat Secur Team Walldorf Germany

With the introduction of Web Real-Time Communications (WebRTC), both Web technology and RTC have entered a new era. WebRTC is a joint effort of the W3C and IETF, supported by a large industry consortium. The technology offers real-time, peer-to-peer communication capabilities directly in the user's browser, accessible to websites via JavaScript APIs. Because of the numerous potential applications and the accessibility of the technology, WebRTC is expected to quickly become a game-changing technology in the communication landscape.

关键词： Special issues and sections WebRTC Peer-to-peer computing Browsers Web services Real-time systems application programming interfaces Network security Web technologies WebRTC real-time communications peer-to-peer communications

来源：评论

学校读者我要写书评

暂无评论

When to Use Standards-Based APIs (Part 1)

引用

IEEE CLOUD COMPUTING 2015年第5期2卷 76-80页

作者： Sill, Alan Texas Tech Univ Ctr High Performance Comp Lubbock TX 79409 USA

This column provides examples of data access patterns ranging from completely local user-driven methods to cloud-based tools to illustrate concepts related to application programmer interface (API) design, and the use... 详细信息

关键词： application Program interfaces Cloud Computing Standards Based API Data Access Pattern Local User Driven Method Cloud Based Tools application Programmer Interface API Design API Concept Cloud Computing programming application programming interfaces Context Modeling Technological Innovation Design Methodology Cloud Standards programming Data Access API Design

来源：评论

学校读者我要写书评

暂无评论

A Case for SIP in JavaScript

引用

IEEE COMMUNICATIONS MAGAZINE 2013年第4期51卷 28-33页

作者： Singh, Kundan Krishnaswamy, Venkatesh Avaya Laboratories USA

This article presents the challenges and compares the alternatives to interoperate between SIP-based systems and the emerging standards for web real-time communication (WebRTC). We argue for an endpoint- and web-focused architecture, and present both sides of the SIP in JavaScript approach. Until WebRTC has ubiquitous cross-browser availability, we suggest a fall-back strategy for web developer: detect and use HTML5 if available;otherwise, fall back to a browser plugin.

关键词： Browsers Logic gates Media Interoperability Codecs Web sites Protocols Web services Telecommunication services Java application programming interfaces

来源：评论

学校读者我要写书评

暂无评论

A Performance Evaluation of Machine Learning-Based Streaming Spam Tweets Detection

引用

IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS 2015年第3期2卷 65-76页

作者： Chen, Chao Zhang, Jun Xie, Yi Xiang, Yang Zhou, Wanlei Hassan, Mohammad Mehedi AlElaiwi, Abdulhameed Alrubaian, Majed Deakin Univ Sch Informat Technol Melbourne Vic 3125 Australia Sun Yat Sen Univ Sch Informat Sci & Technol Guangzhou 510000 Guangdong Peoples R China King Saud Univ Coll Comp & Informat Sci Riyadh 12372 Saudi Arabia

The popularity of Twitter attracts more and more spammers. Spammers send unwanted tweets to Twitter users to promote websites or services, which are harmful to normal users. In order to stop spammers, researchers have proposed a number of mechanisms. The focus of recent works is on the application of machine learning techniques into Twitter spam detection. However, tweets are retrieved in a streaming way, and Twitter provides the Streaming API for developers and researchers to access public tweets in real time. There lacks a performance evaluation of existing machine learning-based streaming spam detection methods. In this paper, we bridged the gap by carrying out a performance evaluation, which was from three different aspects of data, feature, and model. A big ground-truth of over 600 million public tweets was created by using a commercial URL-based security tool. For real-time spam detection, we further extracted 12 lightweight features for tweet representation. Spam detection was then transformed to a binary classification problem in the feature space and can be solved by conventional machine learning algorithms. We evaluated the impact of different factors to the spam detection performance, which included spam to nonspam ratio, feature discretization, training data size, data sampling, time-related data, and machine learning algorithms. The results show the streaming spam tweet detection is still a big challenge and a robust detection technique should take into account the three aspects of data, feature, and model.

关键词： Twitter Feature extraction Machine learning algorithms Uniform resource locators Performance evaluation Real-time systems application programming interfaces

来源：评论

学校读者我要写书评

暂无评论

Analytics Goes on Location with New Approaches

引用

COMPUTER 2013年第4期46卷 14-17页

作者： Garber, Lee IEEE Comp Soc Washington DC 20036 USA

Businesses are increasingly using location analytics-which integrates organizational and geographic data-to help identify trends, develop more accurate predictions, and make better decisions.

关键词： application programming interfaces Internet Software Development Computer interfaces Cloud Computing API Service Providers Software API application programming Interface Web applications Mobile applications Representational State Transfer REST Cloud Computing Analytics Machine To Machine M 2 M O Auth Open ID Connect API Frameworks Back End As A Service Baa S

来源：评论

学校读者我要写书评

暂无评论

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：

请选择收藏分类：

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

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：