Perspectives on Negative Research Results in Pervasive Computing Ella Peltonen

2025-04-26 0 0 246.74KB 10 页 10玖币

侵权投诉

Perspectives on Negative Research Results in

Pervasive Computing

Ella Peltonen

University of Oulu, Finland

Nitinder Mohan

Technical University Munich, Germany

Peter Zdankin

University of Duisburg-Essen, Germany

Tanya Shreedhar

IIIT-Delhi, India

Tri Nguyen

University of Oulu, Finland

Suzan Bayhan

University of Twente, The Netherlands

Jon Crowcroft

University of Cambridge, United Kingdom

Jussi Kangasharju

University of Helsinki, Finland

Daniela Nicklas

University of Bamberg, Germany

Abstract—Not all research leads to fruitful results; trying new ways or methods may surpass the

state of the art, but sometimes the hypothesis is not proven or the improvement is insigniﬁcant.

In a systems discipline like pervasive computing, there are many sources of errors, from

hardware issues over communication channels to heterogeneous software environments.

However, failure to succeed is not a failure to progress. It is essential to create platforms for

sharing insights, experiences, and lessons learned when conducting research in pervasive

computing so that the same mistakes are not repeated. And sometimes, a problem is a symptom

of discovering new research challenges. Based on the collective input of the First International

Workshop on Negative Results in Pervasive Computing (PerFail 2022), co-located with the 20th

International Conference on Pervasive Computing and Communications (PerCom 2022), this

paper presents a comprehensive discussion on perspectives on publishing negative results and

lessons learned in pervasive computing.

arXiv:2210.05708v1 [cs.DC] 11 Oct 2022

RESEARCH

in ﬁelds such as pervasive com-

puting, networking, systems, etc., is usually driven

by results that push forward (or improve efﬁciency

of) the state-of-the-art or provide demonstrable

contributions. To achieve novel and practical so-

lutions, researchers often invest signiﬁcant efforts

into designing and evaluating several iterations

of design choices with hopes of achieving the

desired performance. However, the probability

of such efforts leading to a negative output

can be quite high due to a lack of complete

understanding/hindsight or unconvincing design

choices, some of which may be harder to rectify

at later stages. Such efforts are rarely rewarded

since negative results are usually hard to pub-

lish [

], [

], being considered not novel or lacking

signiﬁcant new knowledge. However, we argue

that negative results, if properly leveraged, can

be quite beneﬁcial to the research community at-

large in the perspective of lessons learned and

knowledge transfer between the research groups

tackling the same topics. While our arguments in

this article are from the perspective of pervasive

computing research, many of our insights are

valid in other ﬁelds of computing, such as human-

computer interaction, network/mobility measure-

ments, security and privacy, etc., thanks to the

inherently interdisciplinary nature of the pervasive

computing.

Most pervasive computing systems require

multiple, often non-standard components, like

hardware prototypes for sensing, developing com-

putation paradigms, novel energy-saving commu-

nication protocols, adaptive middleware, or multi-

modal user interfaces. Often, a new research idea

or hypothesis covers only one part of the system,

but to complete the experiment, you must (at least

partially) get everything to work. Each component

might be a source of error or the bottleneck that

prohibits the expected result. However, sometimes

the interplay of different components highlight

the symptom of an underlying, more signiﬁcant

challenge that could lead to new research ideas.

Hence, in this article, we assert that we must

treat negative results as a potential source for new

scientiﬁc insights.

While negative results might not directly con-

tribute to the advancement of the state-of-the-art,

the wisdom of hindsight could be an essential

contribution in itself, such that other researchers

could avoid falling into similar pitfalls. Such

“what not to do?” insights can also foster good

research practices, especially for handling gray

areas like ethical boundaries [

]. We consider

negative results to be outcomes of studies that

are run correctly (in the light of the current state-

of-the-art) and in good practice but fail in proving

the hypothesis (statistically signiﬁcantly). The

“badness” of the work might also emerge from

properly but unﬁttingly designed data collection

or (non-trivial) lapses of hindsight, especially in

studies involving real-world uncontrolled mea-

surements. Furthermore, the experiences that help

spot negative outcomes from intermediate results

beneﬁt the systems and networking community at

large.

Besides, it is crucial to share (especially for

junior researchers) the knowledge that not all

experiments will be successful, failures might

happen, and knowing how to overcome unexpected

outcomes is a skill to develop as a researcher.

The interest in this ﬁeld was evident from the

participation in the questionnaire for Percom 2022

participants, in which

≈

70 attendees participated.

The result of the questionnaire revealed that

≈

85%

of the conference participants already failed in

their research work. The perspectives noted in

this article are drawn from discussions with the

pervasive computing community during the First

International Workshop on Negative Results in

Pervasive Computing (PerFail 2022)

and from

authors’ past experiences. Speciﬁcally, this paper

discusses failure taxonomy in pervasive comput-

ing, considers mitigation and avoidance strategies,

and presents our call-for-arms for healthy failure

culture in pervasive research.

What is a Failure?

Traditional applied research aims to funda-

mentally understand how systems work or how to

make them work. Within the pervasive computing,

systems and networking research, solutions that

demonstrably work are naturally more attractive

to the community. Therefore, it becomes essential

to gauge the usefulness of research that failed to

work. Failure can be viewed as a necessary catalyst

to research – if there is no possibility to fail, the

1https://perfail-workshop.github.io/2022/

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

PerspectivesonNegativeResearchResultsinPervasiveComputingEllaPeltonenUniversityofOulu,FinlandNitinderMohanTechnicalUniversityMunich,GermanyPeterZdankinUniversityofDuisburg-Essen,GermanyTanyaShreedharIIIT-Delhi,IndiaTriNguyenUniversityofOulu,FinlandSuzanBayhanUniversityofTwente,TheNetherlandsJonCrowc...

展开>> 收起<<

Perspectives on Negative Research Results in Pervasive Computing Ella Peltonen.pdf

共10页,预览2页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

Perspectives on Negative Research Results in Pervasive Computing Ella Peltonen

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: