Luis Berdún

Assisting novice software designers by an expert designer agent

Object-oriented patterns are enjoying much popularity as mechanisms to address flexibility and reusability in object-oriented designs. Consequently, many troubles have appeared to incorporate these practices to novice designers. The selection of appropriate patterns for a given design context is left to the developers criterion. This activity can be problematic for the developer, and thus, he/she is amenable for tool assistance. Along this line, this paper proposes the use of interface agents, describing an agent called PatternAdvisor that is able to help a novice developer with the application of design patterns in his/her projects. This agent works on expert knowledge captured in a Bayesian network, which models knowledge from both design pattern catalogs and expert developers feedback.

An Intelligent tutor for teaching software design patterns (Indexed SCI)

How to teach students to design in the classroom? When is experience crucial to do design? In particular, how to teach design patterns to students who are beginning to know the importance of a good design? Experience is essential to understand and apply patterns in an effective way. Generally, novice users are not good at working in real experiences while they are good at learning new techniques and methods for designing. In this work, we show the results of teaching patterns using an artificial intelligent assistant that helps novice developers during the design process. Our assistant is an interface agent that observes novice users working, and when it detects that a design pattern can be applied, it makes a suggestion justifying its opinion. Thus, students understand when and where a pattern could be applied.© 2011 Wiley Periodicals, Inc. Comput Appl Eng Educ 22:583–592, 2014; View this article online at wileyonlinelibrary.com/journal/cae; DOI 10.1002/cae.20582

An intelligent tutor for teaching software design patterns

How to teach students to design in the classroom? When is experience crucial to do design? In particular, how to teach design patterns to students who are beginning to know the importance of a good design? Experience is essential to understand and apply patterns in an effective way. Generally, novice users are not good at working in real experiences while they are good at learning new techniques and methods for designing. In this work, we show the results of teaching patterns using an artificial intelligent assistant that helps novice developers during the design process. Our assistant is an interface agent that observes novice users working, and when it detects that a design pattern can be applied, it makes a suggestion justifying its opinion. Thus, students understand when and where a pattern could be applied.© 2011 Wiley Periodicals, Inc. Comput Appl Eng Educ 22:583–592, 2014; View this article online at wileyonlinelibrary.com/journal/cae; DOI 10.1002/cae.20582

An agent specific planning algorithm

Planning algorithms are often applied by intelligent agents for achieving their goals. For the plan creation, this kind of algorithm uses only an initial state definition, a set of actions, and a goal; while agents also have preferences and desires that should to be taken into account. Thus, agents need to spend time analyzing each plan returned by these algorithms to find one that satisfies their preferences. In this context, we have studied an alternative in which a classical planner could be modified to accept a new conceptual parameter for a plan creation: an agent mental state composed by preferences and constraints. In this work, we present a planning algorithm that extends a partial order algorithm to deal with the agents preferences. In this way, our algorithm builds an adequate plan in terms of agent mental state. In this article, we introduce this algorithm and expose experimental results showing the advantages of this adaptation.

Unsupervised Learning for Detecting Refactoring Opportunities in Service-Oriented Applications

Service-Oriented Computing (SOC) has been widely used for building distributed and enterprise-wide software applications. One major problem in this kind of applications is their growth; as size and complexity of applications increase, the probability of duplicity of code increases, among other refactoring issues. This paper proposes an unsupervised learning approach to assist software developers in detecting refactoring opportunities in service-oriented applications. The approach gathers non-refactored Web Service Description Language (WSDL) documents and applies clustering and visualization techniques to deliver a list of refactoring suggestions to start working on the refactoring process. We evaluated our approach using two real-life case-studies by using internal validity criteria for the clustering quality.

Social networks and genetic algorithms to choose committees with independent members

We propose an approach for the committee selection problem.We define a novel social network group independence performance function.We build a social network from an R&D public agency.We compare results with current committees of an R&D public agency. Choosing committees with independent members in social networks can be regarded as a group selection problem where independence, as the main selection criterion, can be measured by the social distance between group members. Although there are many solutions for the group selection problem in social networks, such as target set selection or community detection, none of them have proposed an approach to select committee members based on independence as group performance measure. In this work, we propose a novel approach for independent node group selection in social networks. This approach defines an independence group function and a genetic algorithm in order to optimize it. We present a case study where we build a real social network with on-line available data extracted from a Research and Development (R&D) public agency, and then we compare selected groups with existing committees of the same agency. Results show that the proposed approach can generate committees that improve group independence compared with existing committees.

Discrete sequences analysis for detecting software design patterns

A design pattern names, abstracts and identifies the key aspects of a common design structure that make it useful for creating a reusable object-oriented design. Designers with little or no experience in this area are forced to read long catalogs of patterns to acquire this knowledge, missing the learning that is only obtained from practice. In this paper we propose to analyze the sequence of actions needed to be executed in a CASE tool in order to model different design patterns. The purpose of this analysis is to create a model that can be used by an interface agent to detect the design pattern an inexperienced user is trying to create in the tool and to assist him/her in this procedure.

Soploon: A virtual assistant to help teachers to detect object-oriented errors in students’ source codes

When checking students’ source codes, teachers tend to overlook some errors. This work introduces Soploon, a tool that automatically detects novice programmer errors. By using this tool, teachers can reduce the number of overlooked errors. Thus, students receive a more complete and exhaustive feedback about their errors and misconceptions.

Classification of collaborative behavior from free text interactions

Abstract In a computer-supported collaborative work (CSCW) environment, a group of people can work together to fulfill a given assignment. In this context, collaborative problems might naturally arise. The detection of these problems is extremely important for teachers or team leaders to help the team members to improve their collaborative skills and the resolution of the collaboration problems. The observation and analysis of the interaction process of several groups is a time-consuming and difficult task for any teacher or team leader. In this article, we propose a multi-phase classification approach to automatically classify free text observed form the chat of a CSCW environment into different collaborative categories of behavior. We obtained promising results that can be used to help teachers or team leaders with the interaction process analysis in order to focus only in the resolution’s actions of the problems that might arise.

Detección de incidentes de tránsito en Twitter

Nowadays, traffic has become a real chaos in big cities, affecting the mobility of millions of people. On the other hand, social networks handle large amounts of publications dealing divers topics. In particular, many of these publications are shared with the aim of warning about traffic incidents. In this paper, an approach that combines Machine Learning and Natural Language Processing techniques to detect traffic incidents posted on Twitter is proposed. The viability and effectiveness of this approach was evaluated in a study case showing promising results.