Making Web-Scale Semantic Reasoning More Service- Oriented: The Large Knowledge Collider

Paper Title: 
Making Web-Scale Semantic Reasoning More Service- Oriented: The Large Knowledge Collider
Alexey Cheptsov, Zhisheng Huang
Reasoning is one of the essential application areas of the modern Semantic Web. Nowadays, the semantic reasoning algorithms are facing significant challenges when dealing with the emergence of the Internet-scale knowledge bases, comprising extremely large amounts of data. The traditional reasoning approaches have only been approved for small, closed, trustworthy, consistent, coherent and static data domains. As such, they are not well-suited to be applied in data-intensive applications aiming on the Internet scale. We introduce the Large Knowledge Collider as a platform solution that leverages the service-oriented approach to implement a new reasoning technique, capable of dealing with exploding volumes of the rapidly growing data universe, in order to be able to take advantages of the large-scale and on-demand elastic infrastructures such as high performance computing or cloud technology.
Full PDF Version: 
Submission type: 
Full Paper
Responsible editor: 
Reject and Resubmit

Submission in response to

Review by anonymous reviewer

Authors present an overview of the LarkC project. The article is a submission to the SWJ special issue about Big Data, and thus raises expectations accordingly.

The paper is partially quite clearly written. However, the style is sometimes really confusing, e.g. due to extremely long sentences. An example of this the sentence running from the end of the first page to the start of the second page. These and the overall "marketing style" writing make the paper less readable. Authors also recycle some parts of graphics from their other papers, but I am not sure what extra value these bring in the new paper.

In terms of content I was not convinced either. As a research paper this paper does not contribute much beyond describing the overall system architecture. From this we can learn that LarkC enables plug-ins, applications and workflows, but how this is new and different from other platforms? Where is the novelty? How is the (possible) novelty evaluated? What advantage does the LarkC platform give to the listed applications (in section IV) that could not be achieved otherwise? How exactly does it help?

For the above mentioned questions and comments I cannot recommend publishing this paper in the special issue, at least not in its current form.

Minor comments:

- there were some repeated parts of sentences in the introduction from the abstract. Please consider rewriting them.
- [4] does not sound like an optimal reference for the Semantic Web.
- page 2: "…data collections as well as application…" --> "…data collections as well as applications…"
- [2] is not an optimal reference for Jena. Besides, it lacks the publication details (e.g. the year)

Review by Natasha Noy

The content overlaps significantly with two other papers describing the LarKC project:

1.) Large Knowledge Collider - a Service-Oriented Platform
for Large-Scale Semantic Reasoning:


While this paper uses a different style to present the LarKC architecture, this alone does not seem to constitute a novel contribution. Readers who have read the other papers will not learn significant new information about the LarKC architecture from this presentation. According to the authors, the other difference is that they "focus to practical aspects of
constructing service-oriented Semantic Web services and introduce distinctive features of LarKC rather than discuss the implementation details of the platform", but I really didn't see much of that. There is very little discussion and it is not clear that the added value of this paper compared to the other publications warrants a publication in a journal.