Abstract. UML-B is a 'UML-like' graphical front end for Event-B that provides support for object-oriented modelling concepts. In particular, UML-B supports class diagrams and state machines, concepts that are not explicitly supported in plain Event-B. In Event-B, refinement is used to relate system models at different abstraction levels. The same abstraction-refinement concepts can also be applied in UML-B. This paper introduces the notions of refined classes and refined state machines to enable refinement of classes and state machines in UML-B. Together with these notions, a technique for moving an event between classes to facilitate abstraction is also introduced. Our work makes explicit the structures of class and state machine refinement in UML-B. The UML-B drawing tool and Event-B translator are extended to support the new refinement concepts. A case study of an auto teller machine (ATM) is presented to demonstrate application and effectiveness of refined classes and refined state machines.

UML-B is a 'UML-like' graphical front end for Event-B that provides support for object-oriented and state-machine modelling concepts, which are not available in Event-B. In particular, UML-B includes class diagram and state-machine diagram editors with automatic generation of corresponding Event-B. In Event-B, refinement is used to relate system models at different abstraction levels. The same refinement concepts are also applicable in UML-B but require special consideration due to the higher-level modelling concepts. In previous work we described a case study to introduce support for refinement in UML-B. We now provide a more complete presentation of the technique of refinement in UML-B including a formalisation of the refinement rules and a definition of the extensions to the abstract syntax of UML-B notation. The provision of gluing invariants to discharge the proof obligations associated with a refinement is a significant step in providing verifiable models. We discuss and compare two approaches for constructing gluing invariants in the context of UML-B refinement.

Web portals are emerging as a key component for government agencies in Malaysia to served public citizen in providing information and services in a more convenient, prompt, and secured platform. Usability of the web portal is one of the most important attributes for web portal quality because it generates persuasive factor to attract and satisfy users. This study evaluates the usability of the Ministry of Education Malaysia (MOE) web portal by conducting the usability testing in order to measure the usability level of the web portal and provides the relevant usability enhancement based on the evaluation's l'esult. It also focused on evaluation of seven (7) usability attributes adopted from ISO/TEe 25010 and Website Analysis and Measurement Inventory (WAMMI) which are effectiveness, efficiency, learn ability, controllability, attractiveness, helpfulness and satisfaction. The evaluation is limited to frequent tasks that ordinarily performed by the users on the MOE web portal which gathered from pilot study. Two stages of usability testing was conducted in this study which are pre-usability testing to evaluate the usability of the current MOE web portal and post-usability testing on the enhanced version of the MOE web portal. The pre-usability testing result showed that the usability of the current MOE web portal was at a moderate level for all the usability attributes and the post-usability testing result showed that the usability score improved significantly for all the usability attributes. This study proved that the usability testing method is one of the useful methods in enhancing web portal usability.

As there is an enormous amount of online research material available, finding pertinent information for specific purposes has become a tedious chore. So there is a requirement of the research paper recommendation system to facilitate research scholars in finding their interested and relevant research papers. There are many paper recommendation systems available, most of them are depending on paper assemblage, references, user profile, mind maps. This information is generally not easily available. The majority of the prevailing recommender system is based on collaborative filtering that rely on other user's proclivity. On the other hand, content-based methods use information regarding an item itself to make a recommendation. In this paper, we present a research paper recommendation method that is based on single paper. Our method uses content-based recommendation approach that employs information extraction and text categorization. . It performs the profile learning by using naive Bayesian text classifier and generates recommendation on the basis of an individual's preference.