Article Preview
Top1. Introduction
Web science has emerged as the study of using the large amounts of information relating to people and their activities that is located in various regions. The evolution of web science has offered numerous methods by which to analyse activities on the web, such as social network analysis and trend analysis. A recent development in web science is to link different datasets through linked data in order to publish structured groups of data using semantic queries (Aslam et al., 2016). Recently, analysing web data has become quite attractive for researchers, practitioners and academics due to its huge impact on business, academic and social communities (Price et al., 2018). Web data analysis involves not only counting the statistics of web traffic; it is more about business perspectives, users’ behaviour, cultural reporting and more, to explore users’ involvement in current trends and happenings on the web. To better understand web data, different methodologies have been proposed by various analysts to analyse the web, such as opinion analysis (Walha et al., 2016), sentiment analysis (Cambria, 2016), trend observations (Rotabi & Kleinberg, 2016) and so on. Researchers have come up with many ways of better understanding today’s web, and one of the most recently introduced concepts is that of the Web Observatory (WO) system.
A WO is defined as a platform to share data, applications, analyses, and resources related to the world wide web, allowing various communities to connect by using special terms and agreements. The availability of shared data sources with live monitoring and visualization tools is the state of the art in understanding user-generated content on the web better. The motivation behind this paper is to provide insights into WOs and discuss challenges associated with these WOs. The paper contributes:
- •
A study of existing WOs, with their associated host name;
- •
An architecture of WOs described by different hosts;
- •
Components of WOs, from a literature review;
- •
Insights into the limitations of WOs, based on the findings of the literature review.
The rest of the paper is organised as follows: section 2 provides various definitions of WOs available in literature. Section 3 provides the terminologies associated with WOs. Section 4 provides general motivation behind creating WOs. Section 5 discusses existing WOs in detail. Section 6 highlights challenges associated with WOs and section 7 provides discussions and concludes the paper.
Top2. What Is A Web Observatory System?
WO has emerged as a new term to conceptualize how the data on the web can be collected from various sources in order to collaborate, analyse and generate knowledge. Several studies have revealed the idea of WO according to their understanding. WO has a major impact on web data sources such as social networks, linked data and government data. The most widespread definition in the literature is that the WO system is useful to create a vision for prospective development, as emphasized by web science,1 which is the concept of providing distributed data linked by various interfaces in such a way as to analyse the past and predict potential developments in the future. Moreover, it provides a way to access linked datasets on the web through dedicated portals (Tiropanis et al., 2013; Madaan et al., 2016), and it is further explained by Gloria et al. (2013) that a WO aims to group different research communities in order to control authority over multiple disciplines, methodologies and theoretical frameworks. Several definitions of WO have been proposed in previous literature, further composed in Table 1. The common aspects among different definitions include: sharing, analysis, and scalability.