Under the surface of the world wide web, layers of interfaces are used to connect clients with applications and data of various kinds. This is where Francis Palma’s research takes place, with the aim of giving client developers a better software development experience.
The web isn’t just about showing texts and images. A lot of the pages you visit on the internet use programs or applications to present their content to you and to interact with you.
In order to make these pages work in the best possible way, designers and developers need to ensure interface designs of high quality for web-based applications.
“Thus, research efforts should be made on assessing the quality, both syntactical and semantic, of web interfaces. This quality should be judged from both structural and linguistic/semantic perspectives”, says Francis Palma.
To develop a unified approach for modelling and evaluating the design of so-called service-based systems (SBSs) is what Francis Palma’s research targets. He joined Linnaeus University in late 2018, making him one of the newest senior lecturers at the Department of computer science and media technology. Francis is Canadian with Bangladeshi origin, with a PhD in software engineering from the University of Montreal.
“Using research of this type, several design quality issues have been identified and validated. Well-known web interfaces such as Dropbox, Facebook, and Twitter all suffer from either syntactic or semantic issues in their design”, says Francis.
Explore new areas
At Linnaeus University, Francis will focus on both research and teaching at the undergraduate and graduate levels. In particular, his research will explore new areas – the application of reinforcement learning in the domain of Internet of Things (IoT), for example.
“I also plan to focus on studying and exploring the best practices in Web of Things (WoT), that is, software patterns and programming styles that allow real-world objects to be part of the web”.
Francis thinks that the most exciting thing about working at Linnaeus University is its flexible research and teaching environment leveraging, which will allow him to build his own research sphere.
“Most importantly, Linnaeus University is offering me brand new cultural and societal perspectives that will make this journey more exciting and self-discovering”, he concludes.
More on Francis’s research field
The description of the problem of interest
Web application programming interfaces (APIs) and service-based systems (SBSs) evolve to accommodate new user requirements and execution contexts. As such, the changes resulting from the evolution of SBSs may degrade their design and quality of service (QoS) and may often cause the appearance of common poor solutions in their architecture – antipatterns – in opposition to design patterns, which are good solutions to recurring problems.
Moreover, antipatterns resulting from these changes may hinder the future maintenance and evolution of SBSs. The automatic detection and correction of antipatterns are thus crucial to assess the design and QoS of SBSs to facilitate their maintenance and evolution.
Structural design problems
The SBSs are developed on top of diverse service-oriented architecture (SOA) technologies or architectural styles. The current literature lacks a unified approach for modelling and evaluating the design of SBSs in term of their structural design quality.
To address this gap, my research targets a meta-model unifying the three main service technologies: REST, SCA, and SOAP. Using the meta-model, we plan to utilize a unified approach supported by a framework, for modelling and evaluating the design quality and QoS of SBSs. More details on my research are available here.
Linguistic design problems
Identifier lexicon has a direct impact on software understandability and reusability and, thus, on the quality of the final software product. Understandability and reusability are two critical characteristics of software quality. The REST is a de facto standard adopted by software organizations to build their web applications. Understandable and reusable uniform resource identifiers (URIs) are essential to attracting client developers of RESTful APIs because good URIs support the client developers to understand and reuse the APIs.
The use of proper lexicon in RESTful APIs also has a direct impact on the quality of web applications that integrate these APIs. Linguistic antipatterns represent poor practices in the naming, documentation, and choice of identifiers in the APIs as opposed to linguistic patterns that represent the corresponding best practices. We plan to employ both syntactic and semantic analyses for the detection of linguistic patterns and antipatterns in RESTful APIs. More details on the linguistic aspects of REST APIs are available here.
Further research trends and applicability
As the computing paradigm has already moved to the cloud environment, cloud APIs of high structural and linguistic quality is of natural demand from the cloud consumers and developers. The proposed methods and tools can be adapted to cloud APIs, with further research potentials.