Towards the Ubiquitous Web

Paper Title: 
Towards the Ubiquitous Web
Andreas Hotho, Gerd Stumme
Today, we observe the amalgamation of the Social Web and the Mobile Web, which will ultimately lead to a Ubiquitous Web, whose applications will support us in many aspects of the daily life at any time and any place. The integration of the different kinds of available data, their integration and aggregation, and finally the extraction of useful knowledge and its representation has become an important challenge for different research communities, since it requires the confluence of previously separated lines of research. Consequently, the last years have seen increasing collaboration of researchers from the Semantic Web, Web 2.0, social network analysis and machine learning communities. In this position paper, we will discuss the Ubiquitous Web vision in two steps. First, we will address the challenge of bridging the gap between Web 2.0 and the Semantic Web, before widening the scope to mobile applications.
Full PDF Version: 
Submission type: 
Responsible editor: 
Krzysztof Janowicz

Review 1 by Marta Sabou:

The paper presents a vision of the Ubiquitous Web, enabled by mobile devices, enhanced sensor technology and web/web2.0. The authors also touch upon research challenges that the evolving Web technologies bring, e.g., developing minimal-invasive systems for large scale knowledge sharing, or mining the heterogeneous data sources that characterize the Ubiquitous Web (e.g., user generated content, sensor data, data from mobile devices).

While the ideas that are discussed are interesting, the paper could do with some more structuring and polishing.

Structure-wise, the paper might benefit from putting more focus on the actual research challenges of the Ubiquitous Web rather than on the evolution of Web technologies. In terms of the evolution of Web technologies, there is now a general consensus that various Web technologies (social, semantic, mobile, sensor) are increasingly used in combination leading to novel and exciting technological ecosystems, which are often referred to using different, partially overlapping terms (e.g., Web3.0., Ubiquitous Web, Semantic reality). Therefore, focusing the narrative on this evolution brings limited research novelty.

On the other hand, I liked the discussion on mining heterogeneous data sources, although this was quite brief and was only presented at the end of the paper. Such insights into future research challenges are really valuable and interesting. Therefore, I would suggest that the authors work out and extend these challenge parts even if this means reducing their forecast of the evolution of Web technologies.

In terms of polishing the paper, some parts of the text could be improved by a more careful phrasing. For example, the introduction largely reproduces the abstract, which is rather unusual for short papers as this one. Repetitions are also common within sentences and should be reduced as much as possible. For example, the second sentence of the abstract contains "integration" twice ("The integration of the different kinds of available data, their integration and aggregation ...").

Also, the paper lacks any references, although many of the statements that are made would require some support in terms of a reference, e.g., when stating that Wikipedia reaches the quality of traditional dictionaries (Section 2), or when referring to Ontology Learning techniques.

Review 2 by Boyan Brodaric:
The paper proposes the "Ubiquitous Web" as the integration of the Social Web (exemplified by wikis and blogs), Mobile Web (exemplified by web access from mobile devices), and the Semantic Web (exemplified by online structured knowledge). It argues that bridging the existing gap between the Social Web and the Semantic Web, and then amalgamating with the Mobile Web, will lead to rich new applications. This is an interesting idea, and in some ways so obvious as to be a "no-brainer". That said, the paper could be improved in several ways to better make apparent its contributions:

-the paper needs better grounding in existing literature. While this might not be a strict perquisite for a position paper, it would help distinguish the authors' contributions from the state-of-the-art, as well provide some key pointers to readers.

-in several places the authors introduce terms that require more explicit explanation or context to clarify meanings and better drive home their point e.g. "resource sharing systems", "deep semantics", "Social Semantic Web", "Internet of Things". In some cases these terms are implicitly explained, but it would be better to do so explicitly.

-the terms "data', "information", and "knowledge" need to be clarified, as they are used somewhat interchangeably throughout. For example, I find it a misrepresentation to suggest that blogs and wikis contain data—-data typically connotes (at least to me) representations of [scientific] observation and measurement, which rarely applies to the content of blogs and wikis. In section 2, a better substitute for "data" might be "content" or "information". In any case, the use of the various terms should be clarified and then used consistently.

-the paper would benefit from extending the discussion of "Ubiquitous Web" applications to show how they would be beneficial and an improvement on the social and mobile webs. For example, the discussion near the end poses the problem of mining over massive sources, but does not discuss inherent challenges and expected resultant benefits. Elaboration of the challenges and potential impacts would be very worthwhile.

-A minor point: Wikipedia is described as a "dictionary"—it would be more accurate to describe it as an "encyclopedia".

"are only first appearances" should be "are only the first appearances"
"This path allows or exciting" should be "This path allows for exciting"