Green Data Analytics of Supercomputing from Massive Sensor Networks: Does Workload Distribution Matter?

2023 | Information Systems Research | Citations: 0

Authors: Guo, Zhiling; Li, Jin; Ramesh, Ram

Abstract: Energy costs represent a significant share of the total cost of ownership in hig ... Expand

Abstract: Energy costs represent a significant share of the total cost of ownership in high-performance computing (HPC) systems. Using a unique data set collected by massive sensor networks in a petascale national supercomputing center, we first present an explanatory model to identify key factors that affect energy consumption in supercomputing. Our analytic results show that, not only does computing node utilization significantly affect energy consumption, workload distribution among the nodes also has significant effects and could effectively be leveraged to improve energy efficiency. Next, we establish the high model performance using in-sample and out-of-sample analyses. We then develop prescriptive models for energy-optimal runtime workload management and extend the models to consider energy consumption and job performance tradeoffs. Specifically, we present four dynamic resource management methodologies (packing, load balancing, threshold-based switching, and energy optimization), model their application at two levels (purely within-rack and jointly cross-rack resource allocation), and explore runtime resource redistribution policies for jobs under the emergent principle of computational steering and comparatively evaluate strategies that use computational steering with those that do not. Our experimental studies show that packing is preferred when the total workload of a rack is higher than a threshold and load balancing is preferred when it is lower. These results lead to a threshold strategy that yields near-optimal energy efficiency under all workload conditions. We further calibrate the energy-optimal resource allocations over the full range of workloads and present a bicriteria evaluation to consider energy consumption and job performance tradeoffs. We demonstrate significant energy savings of our proposed strategies under various workload conditions. We conclude with implementation guidelines and policy insights into energy-efficient computing resource management in large supercomputing data centers. History: Olivia Liu Sheng, Senior Editor; Xiaobai Li, Associate Editor. Funding: The authors thank the National Supercomputing Center of Singapore for enabling and supporting this research. J. Li received support from the National Natural Science Foundation of China [Grants 71901169, 72032006, and T2293774]. Supplemental Material: The online appendices are available at https://doi.org/10.1287/isre.2023.1208. Collapse

Topics: data center high-performance computing task productivity resource allocation analytical information system

Methods: optimization model experiment simulation descriptive statistic longitudinal research

Navigating Workload Compatibility Between a Recommender System and a NoSQL Database: An Interactive Tutorial

2023 | Communications of the Association for Information Systems | Citations: 0

Authors: Kayhan, Varol; Berndt, Donald

Abstract: Follow this and additional works at: https://aisel.aisnet.org/cais Recommended ... Expand

Abstract: Follow this and additional works at: https://aisel.aisnet.org/cais Recommended Citation Recommended Citation Abstract:In this tutorial, the issue of compatibility between a big data storage technology and an analytic workload is explored using a fictitious streaming company as an example. The tutorial offers an interactive approach to help students understand the importance of considering workload compatibility when adopting new technologies. We provide instructors with the following: two Jupyter Notebooks that analyze compatibility, a detailed instructor guide on how to execute these notebooks, lessons learned, and appendices containing solutions and explanations. This tutorial provides a valuable resource for instructors teaching courses in database systems, big data, and analytic concepts, helping students develop practical skills to navigate effectively the complexities of big data technologies. Collapse

Topics: database system document database recommender system collaborative filtering nosql database

Methods: database modeling design science survey

Developing a collaboration system for pancreatic cancer research: a clinical design science study

2023 | European Journal of Information Systems | Citations: 0

Authors: Rueckel, David; Krumay, Barbara; Dannerer, Ewald

Abstract: Great potential exists for the use of information systems (IS) in medical resear ... Expand

Abstract: Great potential exists for the use of information systems (IS) in medical research, as they already form the backbone of several long-term medical research projects. IS can potentially support practitioner-researchers (practitioners who conduct research in their areas of practice) by enabling collaboration, particularly when the number of cases is low and the treatment strategy is centralised. However, practitioner-researchers and academic researchers differ in their research approaches. In this clinical IS research effort, we designed an IS, together with a local hospital, for supporting research on pancreatic cancer by applying a design science research (DSR) approach. We also observed how DSR is accepted and used by practitioner-researchers in this context. Consequently, we not only developed an IS but also gained insights into the benefits of DSR in projects with heterogeneous participants, namely those from various fields with different experiences and requirements. The results show that DSR contributes to mutual understanding, transparency in the design process and long-term benefits for all parties involved, which was further confirmed by post-hoc interviews. Collapse

Topics: usability user interface database system MongoDB systems development

Methods: design artifact qualitative interview design process design science field study

OPTIONS FOR DESIGNING STABLECOINS AND CENTRAL BANK DIGITAL CURRENCIES

2022 | European Conference On Information Systems | Citations: 0

Authors: Jorgensen, Kim; Hays, Demelza

Abstract: Stablecoins are cryptographic digital currencies that follow a fixed or target ... Expand

Abstract: Stablecoins are cryptographic digital currencies that follow a fixed or target exchange rate regime. The objective of this paper is to introduce for the IS community the concept of stablecoins in their various forms and features and investigate their relationship to Central bank Digital Currencies. This seems timely as several central banks are considering launching a cryptographic currency. The central bank digital currency (CBDC) trend gained traction following the emergence of a number of privately-issued stablecoins. This paper distinguishes amongst thirteen design dimensions that are relevant for decision makers launching stablecoins and central bank digital currencies using distributed ledger technologies. We explain each dimension in the context of prominent CBDC and stablecoin projects. Finally, future areas of CBDC and stablecoin research are elaborated upon. This research is one of the first attempts to combine information systems and monetary policy research to explain the new phenomena of stablecoin and CBDC. Collapse

Topics: stablecoin price management distributed ledger technology privacy Bitcoin

Methods: design science database modeling literature study

What Makes Data Possible? A Sociotechnical View on Structured Data Innovations

2021 | HICSS | Citations: 12

Authors: Aaltonen, Aleksi; Penttinen, Esko

Abstract: Drawing from the theory of digital objects, this paper examines the distinction ... Expand

Abstract: Drawing from the theory of digital objects, this paper examines the distinction between structured and unstructured data as carriers of facts. We argue that data do not ‘have’ a structure but are made by a structure that confers data their capacity to represent contextual facts. We employ a case vignette involving XBRL (eXtensible Business Reporting Language) and its use in statutory financial reporting to illustrate and explore the sociotechnical nature of data and to describe what we call data innovations: new valuable ways to render phenomena as data. We find that data structure is best viewed as a matter that is relative to a purpose in a context. Theorizing data from a sociotechnical perspective could evolve to provide, in effect, the material science of digital economy. Collapse

Topics: electronic data interchange accounting domain-specific language innovation management information reporting system

Methods: natural language processing data modeling design science database modeling

Theories: theory of digital objects

PlanDigital: A Software Tool Supporting the Digital Transformation

2021 | DESRIST | Citations: 0

Authors: Hermann, Andreas; Gollhardt, Torsten; Cordes, Ann-Kristin; Kruse, Paul

Abstract: Due to the plethora of possible digital technologies, companies face difficultie ... Expand

Abstract: Due to the plethora of possible digital technologies, companies face difficulties in selecting appropriate technologies to innovate their business model. Therefore, we introduce PlanDigital: a software tool supporting the digital transformation. PlanDigital combines existing tools and concepts in an integrated manner to support the identification, evaluation, and selection of potential digitalization ideas. Thereby, the application helps practice and academia in managing digitalization ideas in the early phase of a company’s digital transformation. In the tool, the user models and organizes goals, the current business model, digitalization ideas, as well as digitalization roadmaps, which ultimately represent a company-specific digital transformation agenda. The paper accounts for the technological and visual design as well as the features of the tool. Collapse

Topics: web application database system open source Python business model

Methods: design artifact design principle group interview entity relationship modeling design science

Approaches to Enforce Privacy in Databases: Classical to Information Flow-Based Models

2021 | Information Systems Frontiers | Citations: 0

Authors: Shyamasundar, R.K.; Chaudhary, Pratiksha; Jaiswal, Arushi; Kuiri, Aniket

Abstract: Ever since databases became an ubiquitous part of enterprises or businesses, sec ... Expand

Abstract: Ever since databases became an ubiquitous part of enterprises or businesses, security and privacy became a requirement. Traditionally, privacy was realized through various methods of database access control and relied much on the use of statically defined views, which are essentially logical constructs imposed over database tables that can alter or restrict the data that can be viewed by an user. Privacy is about the responsible maintenance of private information. This responsibility is hard to define, which is why laws are necessary. With a vast accumulation of personal data in databases, there has been a heightened awareness and concern about the storage and use of private information leading to privacy-related guidelines, regulations and legislations, Compliance with these regulations has become one of the major concerns for organizations and companies. Traditionally, privacy in databases (DBs) have been addressed through access control techniques including multi-level security (MLS) based on mandatory access control (MAC), and restricted views to the users. As view definitions to comply with regulations became quite complex for accommodating all the restrictions in one view, explicit constructs for specifying privacy policies were introduced for complying with medical regulations like HIPAA (Health Insurance Portability and Accountability Act) from USA, in relational database systems. These enabled fine grained access control (FGAC) capable of enforcing disclosure control enunciated databases. Application of information flow control that is needed for multi-level security (MLS) databases to preserve privacy among multiple users but have their challenges like new abstractions for managing information flow in a relational database system, handling transactions and integrity constraints without introducing covert channels etc. As the DBs need to work alongside information flow controlled programming languages and operating systems for tracking flows, there is a need to enforce the security policy not only on the DBMS but also on the application platform. Due to the underlying requirement of decentralization, it calls for declassification/endorsement and santization requirements on the DB. In this paper, we shall first review some of the major privacy enhancing techniques used traditionally for DBs including MLS DBs, and then explore application of decentralized information flow control models for realizing information flow secure DBs in a robust manner. Towards the end, we shall also touch upon some of the roles of anonymization and psuedonymization including inference control and differential privacy in realizing privacy in practice. Collapse

Topics: privacy PostgreSQL database system differential privacy authorization

Methods: design science case study design artifact database modeling data modeling

AUTOMATED ASSISTANCE FOR DATA MODELERS: A HEURISTICS-BASED NATURAL LANGUAGE PROCESSING APPROACH

2021 | European Conference On Information Systems | Citations: 0

Authors: Ternes, Benjamin; Rosenthal, Kristina; Strecker, Stefan

Abstract: Identifiers of model elements convey semantics of conceptual models essential t ... Expand

Abstract: Identifiers of model elements convey semantics of conceptual models essential to interpretation by human viewers, e.g., the identifier 'Customer' labeling an entity type in an Entity-Relationship diagram. Devising meaningful and purposeful identifiers has proven a particular challenge for data modelers, and has repeatedly been observed among the most common difficulties learners of conceptual data modeling face. The Automated Assistant reported in this paper provides data modelers with modelingtime suggestions on identifying and signifying entity types, relationship types, and attributes with meaningful and expedient identifiers. Different from earlier approaches, the Automated Assistant implementation does not rely on fixed reference solutions for modeling tasks, and works with (m)any natural language descriptions of a universe of discourse. We report on the design artifact, the Automated Assistant, highlight two selected design iterations, demonstrate practical applications, and evaluate its performance in typical application cases. Collapse

Topics: electronic mail enterprise architecture

Methods: design artifact design science data modeling natural language processing entity relationship modeling

TextBenDS: a Generic Textual Data Benchmark for Distributed Systems

2021 | Information Systems Frontiers | Citations: 0

Authors: Truică, Ciprian-Octavian; Apostol, Elena-Simona; Darmont, Jérôme; Assent, Ira

Abstract: Extracting top-k keywords and documents using weighting schemes are popular tech ... Expand

Abstract: Extracting top-k keywords and documents using weighting schemes are popular techniques employed in text mining and machine learning for different analysis and retrieval tasks. The weights are usually computed in the data preprocessing step, as they are costly to update and keep track of all the modifications performed on the dataset. Furthermore, calculation errors are introduced when analyzing only subsets of the dataset, i.e., wrong weighting are computed as weighting schemes use the number of documents for scoring keywords and documents. Therefore, in a Big Data context, it is crucial to lower the runtime of computing weighting schemes, without hindering the analysis process and the accuracy of the machine learning algorithms. To address this requirement for the task of computing top-k keywords and documents (which largely relies on weighting schemes), it is customary to design benchmarks that compare weighting schemes within various configurations of distributedframeworks and database management systems. Thus, we propose TextBenDS - a generic document-oriented benchmark for storing textual data and constructing weighting schemes. Our benchmark offers a generic data model designed with a multidimensional approach for storing text documents. We also propose using aggregation queries with various complexities and selectivities for constructing term weighting schemes, that are utilized in extracting top-k keywords and documents. We evaluate the computing performance of the queries on several distributed environments set within the Apache Hadoop ecosystem. Our experimental results provide interesting insights. As an example, MongoDB shows the best overall performance, while Spark’s execution time remains almost constant regardless of weighting schemes. Collapse

Topics: information retrieval MongoDB database system distributed system MapReduce

Methods: term frequency–inverse document frequency Okapi BM25 experiment design science qualitative content analysis

Beyond the Block: A Novel Blockchain-Based Technical Model for Long-Term Care Insurance

2021 | Journal of Management Information Systems | Citations: 0

Authors: Zhang, Wenping; Wei, Chih-Ping; Jiang, Qiqi; Peng, Chih-Hung; Zhao, J. Leon

Abstract: The insurance business is characterized by complicated transactional interrelati ... Expand

Abstract: The insurance business is characterized by complicated transactional interrelationships among various stakeholders involved in insurance-related activities. Given this unique nature, the century-old challenge in the insurance industry is to effectively reduce transaction costs among the stakeholders while maintaining business privacy and trust. Although blockchain is a promising technology to mitigate this challenge, two technical issues, namely (1) inefficiency in data auditing and (2) difficulty in verifying encrypted data, are of strategic importance when applying blockchain to the insurance industry. To address these technical challenges, we propose an innovative blockchain-based technical model, InsurModel, in the context of newly initiated long-term care insurance in China. Specifically, we utilize cryptographical methods including "zero-knowledge-proof" to 1) represent business interdependence and 2) verify confidential business information without disclosure of specifics. We demonstrate the scalability and applicability of InsurModel and explore its strategic implications in constraining adverse behaviors of the stakeholders. Collapse

Topics: blockchain Bitcoin database system privacy IT manager

Methods: experiment design science applicability check computational algorithm database modeling

Theories: transaction cost economics