Article Preview
TopIntroduction
The vast amount of unstructured information available on the Internet has motivated the use and improvement of web semantic techniques so as to get a web with information more linked and meaningful. It also raised the need for providing tools and strategies aimed at publishing and accessing data on the web. As a result, the concept of linked data (LD) was adopted and a lot of information on the web is currently represented by using LD formats such as the Resource Description Framework (RDF). However, data published on the Internet is useful when there are no restrictions for its use, which has been stablished as the linked open data (LOD) concept.
The adoption of LOD sometimes is hindered by challenges such as those stated by Dadzie and Pietriga (2017): the large amount of LD, the heterogeneous nature of LD, and the technical knowledge required for generating, cleaning and linking new LD. Moreover, the lack of awareness with regard to the landscape of possibilities brought up by the LOD approach may result in less use of the LOD datasets publicly available for people and institutions interested in the analysis of information in particular topics.
Previous studies have explored in a systematic way some aspects in the topic of LD addressing in somehow the LOD concept but focusing on specific contexts of application or particular stages of its adoption rather than in a more general focus. For instance, Attard, Orlandi, Scerri, and Auer (2015) describe open government data initiatives by means of a systematic review. In that review, the authors followed guidelines for systematic literature reviews proposed by Dyba, Dingsoyr, and Hanssen (2007) and Kitchenham and Charters (2007). This review provides important insights into challenges and guidelines for publishing data in the governmental context. Barbosa, Bittencourt, Siqueira, Silva, and Calado (2017) present a systematic literature review in software tools used for publishing and consuming LD. This study explores 80 studies and also follows the literature review following criteria for systematic literature reviews proposed by Kitchenham and Charters (2007). In the same vein, Pereira, Siqueira, Nunes, and Dietze (2017) describe the actual research and future challenges of LD but focusing on the educational field. Apart from giving the current state of research on LD in education and future challenges, authors provide insights into the tools, vocabularies and datasets reported in the studies analyzed. In that review 114 studies were also analyzed by following the Kitchenham and Charters’s methodology (2007). In the study presented by Feitosa et al. (2017) about best practices for publishing LD, the authors suggest that further research is need in the field of LD with regard to the use of datasets, modelling methods, and the research process.
With the aim of contributing to the sphere of knowledge in LOD, this study focused on exploring (in a systematic way) areas of applications, methodologies and technologies used for LOD by giving a general overview of current state of research in studies dealing with this topic. 250 articles were selected in this systematic literature review and main findings include: i) most of the studies conducted so far focus on the use of semantic web technologies and tools applied to contexts such as biology, social sciences, libraries, research and education, ii) there is a lack of research with regard to a standardized methodology for managing LOD, and iii) a plenty of tools can be used for managing LOD but most of them lack of user-friendly interfaces for querying over datasets.