Personal tools

Towards trust management for cloud-based ecosystems

Authors: Sini Ruohomaa, Lea Kutvonen

Category: research article

Keywords: trust management, cloud, reputation, inter-enterprise collaborations, open service ecosystems

Abstract: Inter-enterprise collaborations allow service providers to focus on their key competences while providing a composed service to end customers. The actors involved must determine whether the gains from participating outweigh the risks of depending on other autonomous collaboration participants, in order to make trust decisions on their willingness to collaborate. We define trust as the extent to which an actor is willing to participate in a given action with a given partner, considering the risks and incentives involved. In this paper, we present two high-level alternatives for trust management architectures for cloud-based service ecosystems: closed collaboration environments and open service ecosystems. Closed environments, such as traditional virtual organization breeding environments, are often built around a hub actor, centrally managed and apply pre-formed trust relationships in determining who is allowed into the breeding environment. Open service ecosystems, in contrast, allow service providers to enter the ecosystem by publishing a valid service offer, and trust relationships are formed and evolve within the ecosystem. We discuss the implications these choices have on further architecture refinement, and compare the strengths and weaknesses of the approaches, including their infrastructure needs, viability and ability to scale up in size.

Permanent link to this page:

File Initial submission
reviewer1571-2 says:
Mar 12, 2013 03:29 AM

This paper achieves something difficult but very valuable: the discussion of theoretical concepts in an industry context. The authors move from a description of the technology to what this actually means in the politics of the workplace and vice versa. I learn a lot from reading this paper and in a fashion that makes sense; what is appropriate for which types of context and how the medium can be adapted to suit a certain end goal.
• Second half very clear… The first sentence doesn’t contribute much to understanding the problem at hand… How about this?… Inter-enterprise collaborations allow actors to work together across organisations.
(I am providing a re-write suggestion because I recently got feedback about something that was wrong with little information about what could work).
• I think paragraph 5 (beginning with the sentence ‘The goal of our research is to explore trust management solutions’) would work well if it was brought up to be placed straight after the first sentence, paragraph 1.

• Despite only having software construction concepts available, cloud computing is left to support business-level needs with them as well, even in cases where business processes can be repeatably utilized as templates for new business cases in the style of BPaaS (Business Process as a Service) [6] and other alternatives.

This sentence is unclear…I don’t know who the ‘them’ is. I understand from this, that those responsible for the software construction end up designing for the business needs of the enterprise. Sometimes this work could be efficiently produced using a template, so the current arrangement wastes resources.

• The line … ‘Section 3 presents how objective reputation information is established to
drive trust decisions’. (Could be worth explaining how this fits in the flow of the argument (to join it to section 2 and 5)

• The definitions of open and closed systems is very clear and provides a basis for me to learn more. As a newcomer to this particular area, I am able to quickly see why different functionality are required by the different approaches (for instance, populator, negotiation support, reputation information, and the factor of scale on these different ventures.

• The actual section 3 could benefit from a sentence leading readers from section 2 into the main body of text in section 3.

• I wonder if it would work to swap around the position of sections 4 and 5… This would mean a transition from the general to the specific.

• I think that section 4 needs a minor re-write to fit more directly with the aims of the paper: to contrast the open and closed models. In section 4 I understand that there are some aspects that both models share, but then there are differences. Just a sentence or two orientating the reader would do the trick (in my opinion).

reviewer228-1 says:
Mar 22, 2013 04:42 PM

I have read this paper twice and I still couldn't get the main contribution. Granted, it discusses the theoretical background of inter-enterprose collaboration but still doesn't add to the work itself. It has some potentials but it's still too abstract. Also some part seem like a comparison between centralized and distributed trust management systems which has been extensively discussed in other publications.
- The definition of trust in the paper needs to be bound to a criterion (or a set of aspects) rather than "the extent to which an actor is willing to participate...."
- A very good part was the stress on the difference between the identities of service providers and the services itself. Most work in this area don't make that distinction though it is quite important
- in section 2.1 (3rd page) it wasn't clear how the open service ecosystem work with central reputation system --> "..and provides a central input for risk evaluations. All actors perform private trust and risk analysis on which collaborations to join..."
- Section 2.2. When talking about the negotiation phase; It wasn't clear how the assessment is done? what are the criteria of the assessment?
- Also it wasn't clear how to handle first time interactions!
All in all, I would recommend the authors to keep working on the paper and make it more concrete than the current version. Currently, it is an abstract concept that definitely needs further investigation and more details.

Pasi Tyrväinen
Pasi Tyrväinen says:
Oct 13, 2013 06:43 PM

Editor decision:
The reviewers state that a lot can be learned from reading this paper published as a discussion paper. They also recommend the authors to keep working on the paper with further investigation and elaborating it to make the content more concrete compared to this submission. This version will not be published in the journal issue and the review process is now finished. Thank you for submitting the discussion paper to

  • partners