Maristella Matera’s research relates to theoretical, methodological and experimental aspects of Web Engineering, with particular emphasis on model-based methods and tools for Web application development. Past research themes related to the conceptual models and CASE tools for the development of data-intensive Web applications. The research then evolved into investigating Web 2.0 technologies, with focus on the following themes:
Mashup models and development tools. During the last ten years Maristella has been investigating models, domain specific languages and tools for the development of Web and mobile mashups. The research has been focusing on the application of concepts typical of Web Services to autonomous components provided with a proper UI. The aim has been to define frameworks for the mashup-like integration of stand-alone modules or applications, where integration occurs especially at the presentation layer. The research has also led to the development of a platform that offers visual, live programming mechanisms for the definition of mashups and generates automatically application code that can be deployed on different devices, also with a distribution logic for the collaborative creation and execution of the resulting mashups.
The PEUDOM project site collects material on mashup composition paradigms and the enabling development platforms.
This research led to the publication of the book “Mashups: Concepts, Models and Architectures”, series on “Data Centric Systems and Applications”, Springer-Verlag, July 2014, written with F. Daniel.
Design methods and development tools for context-aware mobile mashups. Given the plethora of data and services today available online, it is often difficult to find on-the-fly the information or the applications that are appropriate to the current context of use. This is even more true in the mobile scenario, where device resources (memory, computational power, transmission budget) are still limited. Given this evidence, this research focuses on the definition of methods and tools for the design and development of Context-Aware Mobile mashUpS (CAMUS). CAMUS apps dynamically collect and integrate data from documental, social and Web resources (accessed by means of Web APIs) and adapt the integrated content to the users’ situational needs. They can offer multiple advantages thanks to their intrinsic capability of identifying pertinent data sources, selected on the basis of their adequateness with respect to the current users’ needs, and pervasively presenting them to the final user in form of integrated visualizations deployed as mobile apps. This application paradigm overcomes the limits posed by pre-packaged apps and offers to users flexible and personalized applications whose structure and content may even emerge at runtime based on the actual user needs and situation of use. The CAMUS framework does not only propose a new application paradigm; rather, it paves the way to novel design methodologies and related tools for fast prototyping of mobile mashups, where context becomes a first-class modeling dimension improving i) the identification of the most adequate resources that can satisfy the users’ information needs and ii) the consequent tailoring at runtime of the provided data and functions.
Elastic design principles for actionable, interactive workspaces. Due to the emergence of new technologies and the data deluge we are assisting today, there is an ever increasing need for effective approaches supporting people to find the “right” facts and make them “actionable” according to their tasks. The research on elastic design principles, conducted in cooperation with people from SAP (Walldorf, Germany) and University of Bari, focuses on methods for the exploration and seamless composition of heterogeneous data sources by end users, so that the retrieval of information as well as the transitions across different usage situations and the related possible analysis tasks are supported without requiring the end user to switch among different isolated applications. The goal is to define elastic environments where relevant information and functions that can be performed on it can be shaped by end users at runtime. The main idea is to exploit methods for the mashup of heterogeneous resources and elastic features that permit the easy transition of information between different task contexts according to the recently proposed notion of Transformative User Experience.
End User Development of Distributed Interactive Workspaces. This research theme relates to adoption of End-User Development (EUD) paradigms to provide visual, live programming paradigms letting users, not necessarily experts of technologies, create service-based, interactive Web applications. A novel contribution is to enable the collaborative creation and use of Distributed Interactive Workspaces (DIWs), i.e., component-based interactive applications, whose content is produced by end users via the aggregation and manipulation of data fetched from distributed online resources, that can be deployed as personal applications or can exploit a centralized, server-side execution logic to manage the sharing of workspaces among different users, the propagation of collaborative actions to active instances of a same workspace, and the distributed execution of a whole workspace or of selected components on different devices employed by different users. The resulting platform has been successfully applied to real usage scenarios in e-Learning and Cultural Heritage.
Quality models for Web 2.0. This research focuses on models, methods and technologies for the analysis the quality of Web 2.0 resource. The aim is to support users in the selection of quality resources. Covering these requirements raises a number of issues. First, relevant and authoritative Web sources must be selected. We have shown that grounding the analysis of sources on data quality dimensions improves this task. Second, the end users should be guided to compose the information access functionalities they need. We therefore propose a model for assessing the quality of mashup resources and of their composition that lead to a quality-aware development process. The method has been integrated within a platform for mashup development.