O trabalho dos usuários em sistemas de informação é uma atividade social que envolve grupos de pessoas que cooperam entre si para desempenhar as mais variadas funções. A natureza da cooperação, por si só é complexa e depende dos indivíduos envolvidos, do ambiente físico e da organização onde o trabalho se desenvolve. Os aspectos ligados ao trabalho cooperativo dos usuários não são considerados no enfoque tradicional da engenharia de software, uma vez que o usuário é visto de modo independente do meio ou grupo em que está inserido, com o modelo individual generalizado para o estudo do comportamento coletivo envolvendo todos os usuários. O objetivo deste trabalho é propor um processo de requisitos de software para tratar as questões envolvendo o trabalho cooperativo em sistemas de informação que apresentem coordenação distribuída nas ações dos usuários e a comunicação entre eles ocorre, preponderantemente, de modo indireto por meio dos dados inseridos no uso do software. Para tanto, a pesquisa faz uso de conceitos da ergonomia, da cognição e da engenharia de software. Utiliza-se a pesquisa-ação como metodologia de pesquisa em três ciclos, aplicada durante o desenvolvimento de um sistema de workflow corporativo em uma empresa de pesquisa tecnológica. No primeiro ciclo, o processo trata da definição dos requisitos do domínio do problema e das contribuições individuais dos usuários. No segundo ciclo, as contribuições do grupo (suas ações e inter-relações) são consideradas com as contribuições individuais pela simulação da solução proposta. No terceiro ciclo, o processo trata do refinamento dos requisitos do trabalho cooperativo, com o software em uso real no ambiente de trabalho. Os resultados obtidos no final do ciclo 2 e início do ciclo 3 durante a aplicação do processo em campo, mostraram a necessidade de melhoria do processo. Esta evolução é necessária, visto que a inclusão do sistema informatizado altera o ambiente de trabalho dos usuários, passando da interação face a face para a interação mediada pelo software. Os resultados obtidos evidenciaram que o maior grau de consciência dos usuários sobre como os inter-relacionamentos de suas atividades são realizados contribuem para um decréscimo em seus erros individuais, diminuindo o retrabalho de recodificação do software e acima de tudo o uso inadequado do sistema, evitando a propagação das consequências desses erros nos resultados finais do trabalho em grupo. / Users\' work in information systems is a social activity that involves people groups cooperating to perform many different functions. The nature of cooperation itself is complex and depends on the people involved, on the workplace environment and on the organization in which the work develops. Aspects related to the users\' cooperative work are not considered in the traditional approach of software engineering, since the user is viewed independently of his/her workplace environment or group, with the individual model generalized to the study of collective behavior of all users. This work proposes a process for software requirements to address issues involving cooperative work in information systems that provide distributed coordination in the users\' actions and the communication among them occurs indirectly through the data entered while using the software. To achieve this goal, this research uses ergonomics, cognition and software engineering concepts. Research-action is used as a research methodology applied in three cycles during the development of a corporate workflow system in a technological research company. In the first cycle, the proposed process exposes the definition of the problem domain requirements and the users\' individual contributions. In the second cycle, the contributions of the group (their actions and inter-relationships) are considered together with the individual contributions through the simulation of the proposed solution. In the third cycle, the process deals with the refinement of the cooperative work requirements with the software in actual use in the workplace. The results at the end of cycle 2 and the beginning of cycle 3 during the process application in the field show the need for process improvement. This is necessary because the inclusion of a computer system changes the users workplace, from the face to face interaction to the interaction mediated by the software. The results show that the highest degree of users\' awareness as the interrelationship of their activities are carried out contributes to a decrease in their individual errors, reducing software recoding rework and above all the inappropriate use of the system, avoiding the spread of the consequences of these errors in the final results of the group work.

This thesis is focused on ways of defining the requirements for performance testing software, both in practice (based on survey) as well as in theory. The aim of this thesis is to analyze ways of defining performance requirements, which are actually used on IT projects. In order to achieve this goal it is necessary to define concepts of performance testing, implement and evaluate the survey of the using ways of defining performance requirements in practice and then analyze ways of defining performance requirements in terms of their applicability to different types of IT projects. The contribution of this thesis is the comprehensive introduction performance testing software issues and mainly insight into ways of defining performance requirements used in practice and problems associated with them, through the survey, which was implemented and evaluated by myself. The conclusions resulting from this survey summarize which ways of defining performance requirements are applied to specific types of IT projects, which of these ways worked and which problems in certain cases that occur in practice. The thesis is divided into theoretical and practical part. The theoretical part explains the basic concepts associated with performance testing software. In this part, there is also described the methodology of defining performance requirements according to Tom Gilb. The practical part is focused on the realization and evaluation of survey of the using ways of defining performance requirements in practice and on the analysis ways of defining performance requirements with respect to certain types of projects.

No description available.

Cette thèse porte sur l’utilisation d’une langue contrôlée pour les spécifications des besoins du logiciel en thaï. L’étudedécrit les ambiguïtés syntaxiques et sémantiques ainsi que les problèmes rencontrés dans les spécifications des besoins dulogiciel en thaï. Ce travail explique également la nature de la langue thaïe. Le modèle de la langue contrôlée pour lesspécifications des besoins du logiciel en thaï, proposé dans cette étude, comprend trois composantes: l’analyse lexicale,l’analyse syntaxique et l’analyse sémantique. Pour l’analyse syntaxique, une syntaxe contrôlée est conçue en utilisant laforme du Backus-Naur (BNF). Quant à l’analyse lexicale, nous créons une ressource lexicale sous forme de langage XMLpour stocker tous les mots classés selon leur domaine. Les mots reçus de la ressource XML sont corrects d’un point de vueconceptuel mais ne sont pas pertinents d’un point de vue sémantique. Pour résoudre ce problème, nous faisons alors usage dematrices booléennes pour aligner les phrases sémantiquement. Ainsi les phrases produites par le modèle serontsyntaxiquement et sémantiquement correctes.Après avoir créé le modèle, nous avons construit un logiciel pour tester son efficacité. Il est ainsi évalué par quatreméthodes d’évaluation : 1. le test de fonctionnement syntaxique pour vérifier la syntaxe de la phrase; 2. le test defonctionnement sémantique pour tester la sémantique de la phrase; 3. le test d’acceptation en terme de satisfaction desutilisateurs avec le logiciel; et 4. le test d’acceptation en terme d’acception des données de sortie.Des résultats positifs montrent que : 1. les phrases produites par le modèle proposé sont syntaxiquement correctes; 2. lesphrases produites par le modèle proposé sont sémantiquement correctes; 3. les utilisateurs sont satisfaits et acceptent lelogiciel; et 4. les utilisateurs acceptent et comprennent les phrases produites par ce modèle. / This thesis focuses on using controlled language for Thai software requirements specifications. The studydescribes the ambiguities and problems encountered in Thai software requirements specifications; both syntacticambiguity and semantic ambiguity. The study also describes the nature of the Thai language. The model of controlledlanguage for Thai software requirements specifications is composed of three main components: lexical analysis,syntactic analysis, and semantic analysis. For syntactic analysis, a controlled syntax is created using Backus-NaurForm (BNF). In the lexical analysis stage, an XML format lexical resource is built to store words according to theirdomain. The words received from the XML resource are conceptually correct but may be semantically irrelevant. Tosolve this issue, the model applies Boolean Matrices to align sentences semantically. As a result, the sentencesproduced from the model are guaranteed to be syntactically and semantically correct.After having created this model, a program for testing the efficiency of the model is developed. The model isevaluated using four testing methods as follows: 1. functional testing for the correctness of the sentence’s syntax, 2.functional testing for the semantic correctness of the sentences produced by the model, 3. acceptance testing in termsof user satisfaction with the program, and 4. acceptance testing in terms of the validity of the outputs.The positive results signify that: 1. the sentences produced by the proposed model are syntactically correct, 2. thesentences produced by the proposed model are semantically correct, 3. the users are satisfied and accept the softwarecreated, and 4. the users approve and understand the sentences produced from this model.