JWE Abstracts 

Vol.17 No.3&4 June 1, 2018

Web Engineering Technologies in the Era of BigData

Editorial (pp181-182)
       
Francisco José Domínguez-Mayo, Julián Alberto García-García,
       
and
Laura García-Borgoñón

Challenges for the Adoption of Model-Driven Web Engineering Approaches in Industry (pp183-205)
       
Esteban Robles Luna, Juan Miguel Sánchez-Begines, Jose Matías Rivero,
       
Leticia Morales-Trujillo, Jose G. Enríquez, and Gustavo Rossi
Model-Driven Web Engineering approaches have become an attractive research and technology solution for Web application development. However, for more than 20 years of development, the industry has not adopted them due to the mismatch between technical versus research requirements. In the context of this joint work between academia and industry, the authors conduct a survey among hundreds of engineers from different companies around the world and, by statistical analysis, they present the current problems of these approaches in scale. Then, a set of guidelines is provided to improve Model-Driven Web Engineering approaches in order to make them viable industry solutions.

MARIA: A Process to Model Entity Reconciliation Problems (pp206-223
       
J.G. Enríquez, M. Olivero, A. Jimenez-Ramirez, M.J. Escalona, and M. Mejías
Within the development of software systems, the development of web applications may be one of the most widespread at present due to the great number of advantages they provide such as: multiplatform, speed of access or the not requiring extremely powerful hardware among others. The fact that so many web applications are being developed, makes enormous the volume of information that it is generated daily. In the management of all this information, the entity reconciliation (ER) problem occurs, which is to identify objects referring to the same real-world entity. This paper proposes to give a solution to this problem through a web perspective based on the Model-Driven Engineering paradigm. To this end, the Navigational Development Techniques (NDT) methodology, that provides a formal and complete set of processes that bring support to the software lifecycle management, has been taken as a reference and it has been extended adding new activities, artefacts and documents to cover the ER. All these elements are defined by a process named Model-Driven Entity ReconcilIAtion (MaRIA), that can be integrated in any software development methodology and allows one to define the ER problem from the early stages of the development. In addition, this proposal has been validated in a real-world case study helping companies to reduce costs when a software product that must give a solution to an ER problem has to be developed.

An Approach for Guesstimating the Deployment Cost in Cloud Infrastructures at Design Phase in Web Engineering (pp224-240)
       
J.C. Preciado, R. Rodríguez-Echeverría, J.M. Conejero, F. Sánchez-Figueroa
       
and A.E. Prieto
Nowadays, the total cost of cloud computing infrastructures for Web applications is calculated in deployment and production phases. Recently, the scientific community offers several methodologies to calculate the most suitable infrastructure at these stages to minimize its monetary costs while covering Service Level Agreement (SLA) constraints. On the other hand, Model Driven Web Engineering is taking advantages of code generation from Design level. With both concepts in the scene, in this work we show the first stage toward an approach to estimate the production costs in cloud computing infrastructures at Design phase, choosing the right infrastructure for the job. The process we have performed started defining the variables of the analysis, measuring the time needed in each different combination obtained, validating the confidence of results obtained and finally applying them to an illustrative example to exemplify the proposal in practical terms.

Other Research Articles
 

 OntoNavShop: An Ontology-Based Approach for Web-Shop Navigation (pp241-269)
       
Philip Ruijgrok, Flavius Frasincar, Damir Vandic, and Frederik Hogenboom

Existing literature shows that navigation and visualization features play a significant role in successful Web shop design. Traditional Web shops, however, often lack a uniform, intuitive interface to navigate through products, while also providing an insightful overview of the product assortment. In this article, we employ ontologies for the presentation of product assortments in Web shops in order to ease the users' process of finding their desired products. OntoNavShop visualizes the product assortment ontology directly in a Web browser using a circular view algorithm that outputs SVG graphics. Consumers can navigate uniformly through the ontology and zoom into its categories. The visualisation is evaluated on efficiency, user satisfaction, and specific problems against a classical tree-based Web shop. Our evaluations under a representative group of users show that users maintain a better overview of the structure of the product assortment, while being able to find products more quickly (i.e., less time) and more efficiently (i.e., less clicks) than in our benchmark Web shop. The participants prefer the OntoNavShop over the classical approach, and the identified problems are rather minor.

 

Hidden Webpages Detection Using Distributed Learning Automata (pp270-283)
       
Manish Kumar and Rajesh Bhatia
Webpages directly connected to each other on the Web can be reached easily by following hyperlinks. Those webpages that are not linked by hyperlinks comprises hidden Web and it is challenging to find them. Furthermore, most of webpages in hidden Web are generated dynamically. This paper proposes first time an algorithm to find webpages in hidden Web using distributed learning automata. Learning automata use its self-learning characteristic of taking action based on the action probabilities using <keyword-value> pairs. These actions may lead the current webpage to hidden webpages that are generated dynamically. At each stage of the proposed algorithm, we determine the edge that should be chosen to reach webpage of interest. The proposed algorithm is validated on four different websites from dmoz.org. Precision-recall curve and coverage plot in the results section shows the effectiveness of the proposed algorithm.

 

Implementation and Evaluation of a Resource-based Learning Recommender based on Learning Style and Web Page Features (pp284-304)
       
Mohammad Tahmasebi, Faranak F. Ghazvini, and Mahdi Esmaeili
It is generally believed that recommender systems are a suitable key to overcome the information overload problem. In recent years, a special research area in this domain has emerged that concerns recommender systems for Technology Enhanced Learning, in particular, self-regulated learning with resources on the web, known as Resource-Based Learning. Grey-sheep users are a major challenge in RecsysTEL. This group of users have completely different opinions from other users. They do not profit from collaborative algorithms, so they must be supported in discovering learning resources relevant to their characteristics and needs. The main contribution of this work is to develop a feature-based educational recommender system which interacts with the user based on his or her learning style. The learning style dimensions would be determined based on Felder-Silverman theory. In addition, the system crawls and extracts the necessary meta-data of sample OCW’s web pages. Based on the proposed web page ranking formula, the user’s learning style dimension and web page feature’s vector would be accommodated to generate learning object suggestions. The general satisfaction, perception and motivation towards the proposed method measured among 77 science and engineering students by a questionnaire. Moreover, the system has been evaluated to provide feedbacks on its suitability. The research findings imply that the proposed method outperforms the general search algorithm. This system can be used as a template in formal and informal learning and educational environments as a RecsysTEL.

Back to JWE Online Front Page