During the software development cycle, there is often a need to run tests after correcting a bug. The tests to be executed are only a few times automated, demanding even more time from the tester. In addition to the testing time, there is also the difficulty in selecting the tests that should be executed at that moment, as depending on the size of the software, the number of tests reaches thousands. In this research, we propose a hybrid model for recommending manual tests, which, based on the bug description and test steps, selects the best tests to be carried out to check whether a given bug has been resolved. This work was developed using real-world data from software under development, which features manual user interface tests. During development, our application presented an accuracy rate of 60.82% through the Recall@20 metric using the training data, which was confirmed through the use of human users, where the system presented an accuracy rate of 59.70%.

Contextual word embeddings and the Transformers architecture have reached state-of-the-art results in many natural language processing (NLP) tasks and improved the adaptation of models for multiple domains. Despite the improvement in the reuse and construction of models, few resources are still developed for the Portuguese language, especially in the health domain. Furthermore, the clinical models available for the language are not representative enough for all medical specialties. This work explores deep contextual embedding models for the Portuguese language to support clinical NLP tasks. We transferred learned information from electronic health records of a Brazilian tertiary hospital specialized in cardiology diseases and pre-trained multiple clinical BERT-based models. We evaluated the performance of these models in named entity recognition experiments, fine-tuning them in two annotated corpora containing clinical narratives. Our pre-trained models outperformed previous multilingual and Portuguese BERT-based models for cardiology and multi-specialty environments, reaching the state-of-the-art for analyzed corpora, with 5.5% F1 score improvement in TempClinBr (all entities) and 1.7% in SemClinBr (Disorder entity) corpora. Hence, we demonstrate that data representativeness and a high volume of training data can improve the results for clinical tasks, aligned with results for other languages.

Text style transfer is a relevant task, contributing to theoretical and practical advancement in several areas, especially when working with non-parallel data. The concept behind non-parallel style transfer is to change a specific dimension of the sentence while retaining the overall context. Previous work used adversarial learning to perform such a task. Although it was not initially created to work with textual data, it proved very effective. Most of the previous work has focused on developing algorithms capable of transferring between binary styles, with limited generalization capabilities and limited applications. This work proposes a framework capable of working with multiple styles and improving content retention (BLEU) after a transfer. The proposed framework combines supervised learning of latent spaces and their separation within the architecture. The results suggest that the proposed framework improves content retention in multi-style scenarios while maintaining accuracy comparable to state-of-the-art.

Electronic health records (EHRs) contain patient-related information formed by structured and unstructured data, a valuable data source for Natural Language Processing (NLP) in the healthcare domain. The contextual word embeddings and Transformer-based models have proved their potential, reaching state-of-the-art for various NLP tasks. Although the performance for downstream NLP tasks with free-texts written in English has recently improved, less resource is available considering clinical texts and low-resource languages such as Portuguese. Our objective is to develop a Generative Pre-trained Transformer 2 (GPT-2) language model for Portuguese to support clinical and biomedical NLP tasks. We fine-tuned a generic Portuguese GPT-2 model to corpora of biomedical texts written in Portuguese, using transfer learning. We experimented on a public dataset, manually annotated for detecting patient fall, i.e., a classification task. Our in-domain GPT-2 model outperformed the generic Portuguese GPT-2 model by 3.43 in F1-score (weighted). Our preliminary results show that transfer learning with domain literature can benefit Portuguese biomedical NLP tasks, aligned with other languages' results.

Information Extraction techniques can retrieve useful information from unstructured data that improve data analytics’ effectiveness and play a key role in the consumer decision-making process. The growth of sponsored content videos on social media increases the demand on knowing the effectiveness of the sponsored investment in the engagement results obtained. This study aims to analyze an approach to infer sponsored content in videos from top digital influencers on YouTube using knowledge acquisition techniques in audio transcriptions. A dataset with 34,563 videos, among 103 different YouTubers channels, was used in a model comprising six stages: data acquisition, preprocessing of documents, identification of candidate videos, manual transcriptions processing, automatic transcriptions processing, and tuple filtering. Despite we perceived several difficulties during the recognition of speech from YouTube videos, such as the absence of clear boundaries between words, the presence of two or more people talking in the video, and colloquial expressions, as a result, the approach identifies sponsored videos from their audio transcripts in a feasible process by identifying keywords, obtaining knowledge from tuples, and recognizing named entities.

During the development cycle of a project, it is common for software requirements and functionality to change and for code errors to occur. To deal with these unforeseen changes, the artifact known as change request, which is a formal proposal to alter a system, is used. Its assignment is an important step in the development process. Projects can receive a very high number of requests daily, which makes the automation of this process compelling. This work proposes a method for assigning unresolved requests, based on developer's profiles. The proposed method consists of three steps. The first step is to extract code quality metrics, commit data and previously resolved requests, in order to model developers through the mining of repositories. The second step concerns with the selection of the profile of potential developers through the application of natural language processing and information retrieval techniques. And finally, in the third step the appropriate developers are selected based on the quality of their code and the impact of their commits. Results from experimental evaluation show that the method is able to recommend more developers with a positive impact on the repository quality if compared to the iMacPro method.

Organizations have been increasingly relying on software development to manage their business, making it an essential activity. Companies have been using collaborative development environments to speed up their deliveries, resulting in projects tending to be produced by developers from different parts of the world with different cultures. In this context, controlling the software's overall quality by ensuring code quality is a crucial aspect of software development. Software quality can be analyzed based on software parameters like complexity, reusability, testability and maintainability. Such parameters are associated with different group of code metrics which are sets of software measures, providing better insights to developers about their code. In this research we aim to support software developers in better understanding code metrics information. We present a visual tool based on an original method which applies code metrics to an open source collaborative software development environment, grouping them into software product quality parameters. Developers are provided with dashboards and recommendations related to the part of code requiring quality improvement.

People involved in software development seek better ways of developing quality software. As software development is a collaborative activity dependent on technology and performed by group of people, the quality can be directly linked to the degree of collaboration and commitment of the development team. This research aims to identify and analyze software quality metrics that can measure software development team members' participation regarding their contribution in the project source code. Analyzing such information can contribute with software engineering project managers to conduct and better organize project development teams. To achieve this goal, we defined a set of quality metrics, applied them to open source code and analyzed them in other to evaluate members' contribution. Results show that it is possible to determine whether the team member contribution increase, decrease or had no influence on source code quality.

Cooperation is a fundamental trait of Systems of Systems (SoS). In this paper we discuss the design and implementation of proactive vocal interfaces with the goal of improving and stimulating cooperation between users and constituent systems. Our approach is based on intelligent personal assistants displaying proactive behavior and capable of interacting vocally. The paper discusses the main challenges for integrating proactive behavior into SoS interfaces, as well as for using them. The proposed approach is generic and can be applied to different domains.

Despite the great number of Systems of Systems (SoS) being developed, building them still remains hard and difficult. Currently, there is a lack of methods capable of supporting architects for building an actual SoS. In this paper we introduce an original method called GAMBAD for developing an SoS from a practical point of view. Our method guides the development of SoS on top of a multi-agent layer supported by ontologies. We tested GAMBAD by building an SoS in the domain of Health Care. Early results show that by using our method, architects can develop an SoS faster and more accurately.

Good Decision Support Systems require three main features: (i) a good handling of the domain data and information; (ii) an efficient user interface; and (iii) a good knowledge of past decisions. Usually such features are handled by different specialized systems difficult to integrate. In this research we keep specialized systems independent, focusing on interoperability. We propose a system of systems architecture (SoS) integrating a domain system in which users interact, a multi-agent system implementing an efficient user interface and taking into account results from the domain system, and a platform to capitalize and manage knowledge. Our approach extracts indicators from the interaction or behavior of users within the domain system, and provides them analyses, statistics and recommendations to help them reach good decisions. We built a prototype and applied it to two different domains: collaborative software development and healthcare. In this paper, we will focus on the multi-agent system which is a key component of the SoS architecture.

In a System of Systems (SoS), heterogeneous systems cooperate to achieve a higher purpose. Developing an SoS is complex and difficult, in particular interfacing systems coherently to make them interoperable. In this research, we propose a core architecture to simplify the development of SoS. Our approach is based on a model we call MBA: "Memory Broker Agent," consisting of agents for encapsulating systems, brokers for handling requests for functionalities and results, and memories for capitalizing and managing knowledge. In this paper we focus on the communication protocol of the proposed architecture, which provides semantic and multilingual interoperability, facilitating the exchange of information and the collaboration between system users. We have built a prototype implementing our approach and applied it to a case study in the health care domain.

Inspired on decision trees and evolutionary algorithms, this paper proposes a learning algorithm of constructive neural networks that relies on three principles: to layout the neurons in a tree-like structure; to train each neuron individually; and, to optimize all the weights using an evolutionary approach. This way, it is expected to advance in two main questions concerning multilayer perceptrons (MLPs): how to determine the network architecture and how to build models that are more comprehensible. Based on the normalized information gain of each attribute, the algorithm builds the network architecture. In the process, it automatically creates a set of training examples for each individual neuron and executes single-cell learning. Once the network is created and trained, particle swarm optimization is utilized to evolve the connections of the network. Five metrics were utilized to validate the method when compared to decision trees and MLPs: accuracy, sensitivity, specificity, precision and comprehensibility. The experiments were executed in thirteen different databases and the results suggest that the proposed algorithm can generate neural networks with good classification performance and more comprehensible.

Collaborative software development is a complex activity. An important factor that needs to receive attention in collaborative software development is software quality. High quality software reduces the development and the maintenance; improves delivery schedules; and reduces repairs and rework. In order to measure, evaluate, control and improve the software quality, software metrics can be used. In this research we present an advanced collaborative environment for software development currently being built, called ACE4SD, which intends to support the improvement of the code quality during collaborative software development. ACE4SD is a system of systems composed of a software development environment, a multi-agent system and a platform to capitalize and manage knowledge, all of them being integrated in the same environment. ACE4SD can provide personalized support to team members to improve the code quality and encourage its reuse, it can answer questions or doubts arisen during the development, record document problems and solutions, and improve the awareness and collaboration between the participants.

This paper introduces the use of a modified feedforward neural network to cope with the problem of predicting protein functions. Since this kind of classification task is inherently hierarchical, this work proposes the use of two different architectures for the modified feedforward neural network, both mimicking the hierarchical nature of the classes (protein functions) to be predicted. The first approach consists of four feed-forward neural networks in cascade, each one taking as input the classification obtained by the previous network, which means, the input to a network is the classes that could be assigned to the protein at the immediately higher (parent) level in the class hierarchy. The second approach is an extension of the first one, which also adds as input to each sub-network the attributes of the protein being classified. In both situations, it was used two kinds of feed-forward architectures: an Adaline network, which is composed of a single layer of adjustable weights, and a MLP ("Multi-Layer Perceptron"), composed by two layers of adjustable weights. Both approaches were compared with a baseline consisting of a single MLP that maps the input attributes to the classes of the lowest level in the hierarchy. The MLP was built with the input layer, plus one hidden layer and one output layer. The three approaches were compared on eight datasets, the first four involving the prediction of GPCR (G-Protein Coupled Receptor) functions and the second four datasets involving the prediction of enzymes functions. The results show that a big-bang hierarchical neural network, based on the MLP paradigm, using a top-down evaluation for new instances has better behavior in hierarchical problems, when compared to its flat version.

This article presents a method to assess whether folksonomies are useful shared conceptualizations for conceptual modeling. A method based on the tripartite model (actors, concepts and instances) induces from social applications a folksonomy related to a domain of knowledge. The hypothesis of this paper is that if folksonomies are shared conceptualizations, then using them in conceptual modeling should reduce the number of divergences between actors when they elicit terms to be part of a model (a concept map in this article). A controlled experiment of conceptual modeling was performed with experimental groups who received tags inducted from data of categorizations from Delicious and control groups who received terms extracted from Web pages categorized on Delicious. The results show that the experimental groups obtained less divergences (on average) in elicitation of terms when compared to control groups.

Software development is a collaborative activity, dependent on technology and performed by groups of people. The software technology involved is an important factor, since it provides the necessary tools for the development of the work. This paper presents a collaborative virtual workspace that follows the code development, comparing it with the models developed in earlier stages. It provides useful information for professionals involved in the task of developing code and even managing the project. It aims at allowing monitoring of the project, helping users to be aware of the context in which they are working. We present its main features and architecture. Results of experimentations show the value of the workspace as a tool to support awareness and software verification.

Dialogues are created by the interaction between people, who speak different kinds of topics using natural language.Task-oriented dialogue aims the solution of a given task in a given domain.Folksonomies are knowledge structures composed of users, tags and resources.Folksonomies emerge from the tagging process in collaborative tagging systems.Dialogues and folksonomies have in common their social dimension.One of the main characteristics of the folksonomies is its social dimension (users), which is also presented in dialogues, through the interaction between human beings.In this research, we describe a method that performs the learning of folksonomies, represented by a quadripartite model, from task-oriented dialogues.Using the learned folksonomies, we propose an approach for trend detection (those topics being discussed more than others).The main difference from others approaches is that we use the content of each resource in this process.This can be useful for instance, to retrieve the topics addressed by the interlocutors of the dialogues, in different time intervals.Experiments with a real-world task-oriented dialogue corpus were done.

This paper presents the FOLKUS-SD a module of CSCW-SD, intend to build Folksonomies from source codes. The CSCW-SD is an architecture to integrate tools in a development environment through a Multi-Agent System. The Folksonomies are built using dynamic data collected during codification. The Folksonomies' entities are represented by the module as: the users are the developers, the tags are names of classes, methods and attributes, and the resources are source codes. The Folksonomies can be used for software documentation. The Folksonomies produced by FOLKUS-SD can be used in the software documentation for monitoring the status of some project and to get quantitative data about the project and its participants.

The lack of tools for small software development teams leads us to propose an architecture to better support it. The multi-agent system (MAS) based architecture is called CSCW-SD. CSCW-SD has a module (MODUS-SD) that models users for better system customization and usability. In this paper, we present an enhancement to this module to help dynamically modeling the users as they interact with the system. In order to do that, we introduce an algorithm called GSOM that clusters user's behavior (the extracted features) and automatically detects cluster boundaries in the resulting trained SOM network.

We have been researching in CSCW for small teams for the last years. As we presented in previous works, small software development teams have special needs and requirements that must be taken into account when designing tools for supporting cooperation of their participants. We have already proposed the architecture of a system (called CSCW-SD) to support small collocated teams developing software. We developed a module, called MODUS-SD, capable of modeling users. In this paper, we present an ongoing project aiming at, automatically, recognize the activities performed by developers during their daily work. Recognizing those activities might facilitate the execution of the same activities in the near future.

The Abstract Computational Model of Awareness for Community Identification (AMACI) is based on the contents of resources created or used by users, which allows noticing others that perform or have performed activities in similar contexts, thereby identifying potential communities and teams. The model was evaluated through experiments using data from a sample of teachers and students at two universities. In order to evaluate the performance of the proposed model in community identification, two metrics were proposed: local and global performance. As a main result of experimentation, different uses for the model were found by combining the possible values for the local and global metrics. Such uses are the identification of new communities, identification of existing communities, expansion of existing communities, and the identification of teams. One of the main benefits of the model is providing an architectural pattern for developing this kind of application. Future works include the elimination of some manual steps in order to analyze larger communities than those observed ones; to look for other clues in e-texts and other types of resources that can help in identifying members of a community, make a temporal analysis of communities what can provide important information on the intellectual capital of an organization.

Collaboration is an important issue when developing software, because it involves working together towards a common goal. This work presents OPERAM, a collaborative semantic workspace that allows comparing the modeling performed at earlier stages of software development with JAVA code. OPERAM provides useful information for professionals involved in the task of developing code and even managing the project. OPERAM aims to allow monitoring of the project, helping users to be aware of the context they are working, allowing carry out verification of software development, confronting the modeling performed with source-code produced during the programming phase. We present its main features and architecture. In this paper the main features and architecture of OPERAM are presented, as well as a validation was conducted to prove the effectiveness of OPERAM. The results show the value of the workspace as a tool to support collaboration and software verification.

During the software development cycle, artifacts (source-code, documentation, user manuals, etc.) are written, most of them, cooperatively. Each participant in a software development team plays a specific role, but may write an artifact cooperatively with participants playing different roles. In large and distributed teams the roles are well-defined and mainly respected. In small teams however, a participant may perform different roles simultaneously. We have already proposed the architecture of a system (called CSCW-SD) to support small collocated teams developing software. We are now adding to CSCW-SD a module, called MODUS-SD, capable of modeling users. In this paper we present this module, how it was implemented and present a case study, with preliminary results, in the context of modeling Java developers in a collaborative software development team.

This paper presents Noctua, a tool to assist in Knowledge Acquisition and Collaborative Knowledge Construction processes. Noctua contains an innovation: a virtual catalyst designed to facilitate the task of eliciting and validating knowledge. The virtual catalyst queries collaborators, proposing new knowledge, seeking confirmation to the knowledge already elicited, and showing conflicting opinions. Noctua takes into account collaborators' profiles in order to automatically ask them questions related to each one's field of knowledge or interest.

Divergences in conceptual modeling choices are inherent to the collaborative ontology development. Such divergences have been typically solved through some process of negotiation among the participants of the development process. When negotiating, the participants argue to defend their ideas based on their past experiences. We propose to support and to guide the argumentation using philosophical notions brought by the OntoClean methodology to show the participants the consequences of their conceptual modeling choices. So the participants can choose the modeling option that better represents the intended model of the domain. This paper presents an approach for collaborative ontology development in a distributed way, specifically to support activities of the conceptualization phase and the whole process of discussion to reach consensus, aiming at developing ontologies that better fit the intended models for a domain.

Software developers often face the task of documenting source code. For many of them, documenting code development is a boring task. However, source code documentation is an important task, especially when dealing with groups of developers. An updated documentation allows group members to have greater visibility on what has been and is being developed, allowing the reuse of source code. This research aims at designing, developing and validating a semi-automatic documentation method for source code from the existing design documentation on a particular project being developed by a small team, as well as updating this documentation from information gathered from the source code under development. It is understood as design documentation, those documents or parts of documents that are linked directly to the code under construction.

In this paper, we discuss the construction of dialogs for Personal Assistant Agents that are in charge of the interface between users and a Multi-Agent System. Such a system aims at providing support for small teams developing software collaboratively. These small teams have specific needs such as the integration of free or open-source tools or the support to elaborate project documentation. Considering such specific needs, we have elaborated a Multi-Agent architecture that has been implemented using a platform called OMAS. We present the structure that OMAS offers to handle dialogs with users and discuss some implementation details. We also describe some of the dialogs that represent interactions between members of small software development teams and their Personal Assistant Agents. We consider that the use of Personal Assistant Agents can help small teams to handle documentation issues in an integrated and undemanding way.

The necessity of lowering the execution of system tests' cost is a consensual point in the software development community. The present study presents an optimization of the regression tests' activity, by adapting a test cases prioritization technique called Failure Pursuit Sampling-previously used and validated for the prioritization of tests in general-improving its efficiency for the exclusive execution of regression test. For this purpose, the clustering and sampling phases of the original technique were modified, so that it becomes capable of receive information from tests made on the previous version of a program, and can use this information to drive de efficiency of the new developed technique, for tests made on a present version. The adapted technique was implemented and executed using the Schedule program, of the Siemens suit. By using Average of the Percentage of Faults Detected charts, the modified Failure Pursuit Sampling technique presented a high level of efficiency improvement.

Most CSCW and groupware systems focus the activities of distributed teams involved in large projects by means of tools for communication and awareness. The activities of small collocated teams are often neglected. Analyzing preliminary requirements of small teams, it is possible to observe the need of tools to help the elaboration of project documentation. This paper presents a multi-agent system architecture to support software development in small collocated teams. The architecture proposed in this paper tackles the problem of elaborating documentation, by facilitating the elaboration of the Small Project Management Plan [S.K. Land and J.W. Walz, 2006]. Tools used in the software development are encapsulated by agents that extract and organize useful information for the elaboration of such a document.

The arrhythmias or abnormal rhythms of the heart are common cardiac riots and may cause serious risks to the life of people, being one of the main causes on deaths. These deaths could be avoided if a previous monitoring of these arrhythmias were carried out, using the Electrocardiogram (ECG) exam. The continuous monitoring and the automatic detection of arrhythmias of the heart may help specialists to perform a faster diagnostic. The main contribution of this work is to show that self-organized artificial neural networks (ANNs), as the ART2, can be applied in arrhythmias automatic detection, working with Wavelet transforms for feature extraction. The self-organized ANN allows, at any time, the inclusion of other groups of arrhythmias, without the need of a new complete training phase. The paper presents the results of practical experimentations.

WebAnima is an interface agent specially designed to assist team members of a CSCW application during their daily work based on computers. In WebAnima, the intelligent behavior is guaranteed thanks to a conversational interface and ontologies that support semantic interpretation. We believe that embodied conversational assistants will improve the quality of assistance and increase collaboration between project members. In this paper, we present the embodied conversational assistant and its insertion into a multi- agent system designed for research and development projects. We describe the design of the agent, highlighting the role of ontologies for semantic interpretation and the dynamic behavior of the embodied animated agent.

This paper describes the design of an ontology-based speech interface for personal assistants applied in the context of cooperative projects. We believe that this type of interface could improve the quality of assistance that personal assistants may offer. We present such an interface and apply it in a multi-agent system in the context of research and development projects. We describe the design of the interface, highlighting the role of ontologies for semantic interpretation. As a result of this conversational speech interface, we expect an increase in the quality of assistance and a reduction of the time needed to answer user's requests.