http://sedici.unlp.edu.ar:80/handle/10915/140251 2026-06-13T00:47:41Z 2026-06-13T00:47:41Z Branca, Fernando Matoff, Pablo F. Gaona, Germán Pérez, Carlos A. http://sedici.unlp.edu.ar:80/handle/10915/140435 2022-08-11T20:03:01Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual) La actividad de captura de requerimientos es crítica para el desarrollo de software exitoso ya que cuando está bien ejecutada la probabilidad de obtener un software que cumpla con las necesidades de los usuarios aumenta, por el contrario, cualquier error cometido en esta etapa suele ser más costoso de sub-sanar posteriormente. Al mismo tiempo una parte importante de esta ejecución depende de las competencias que poseen las personas involucradas, del contexto en el que se desempeña y de los métodos o técnicas aplicados a tal fin. Por lo tanto, esta actividad puede verse afectada por factores no técnicos de índole individual o grupal. En este artículo se presenta una revisión sistemática de la literatura con el objetivo de conocer el estado del arte acerca de las propuestas existentes para la captura de requerimientos que contemplen factores humanos y socioculturales influyentes sobre esta actividad, así como también determinar qué adaptaciones se realizan en ese sentido.

2021-01-01T00:00:00Z La actividad de captura de requerimientos es crítica para el desarrollo de software exitoso ya que cuando está bien ejecutada la probabilidad de obtener un software que cumpla con las necesidades de los usuarios aumenta, por el contrario, cualquier error cometido en esta etapa suele ser más costoso de sub-sanar posteriormente. Al mismo tiempo una parte importante de esta ejecución depende de las competencias que poseen las personas involucradas, del contexto en el que se desempeña y de los métodos o técnicas aplicados a tal fin. Por lo tanto, esta actividad puede verse afectada por factores no técnicos de índole individual o grupal. En este artículo se presenta una revisión sistemática de la literatura con el objetivo de conocer el estado del arte acerca de las propuestas existentes para la captura de requerimientos que contemplen factores humanos y socioculturales influyentes sobre esta actividad, así como también determinar qué adaptaciones se realizan en ese sentido. Ponzio, Pablo Daniel Godio, Ariel Rosner, Nicolás Arroyo, Marcelo Aguirre, Nazareno Matías Frias, Marcelo F. http://sedici.unlp.edu.ar:80/handle/10915/140433 2022-08-11T20:03:03Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual) Software model checkers are able to exhaustively explore different bounded program executions arising from various sources of nondeterminism. These tools provide statements to produce non-determinis- tic values for certain variables, thus forcing the corresponding model checker to consider all possible values for these during verification. While these statements offer an effective way of verifying programs handling basic data types and simple structured types, they are inappropriate as a mechanism for nondeterministic generation of pointers, favoring the use of insertion routines to produce dynamic data structures when verifying, via model checking, programs handling such data types. We present a technique to improve model checking of programs handling heap-allocated data types, by taming the explosion of candidate structures that can be built when non-deterministically initializing heap object fields. The technique exploits precomputed relational bounds, that disregard values deemed invalid by the structure’s type invariant, thus reducing the state space to be explored by the model checker. Precomputing the relational bounds is a challenging costly task too, for which we also present an efficient algorithm, based on incremental SAT solving. We implement our approach on top of the CBMC bounded model checker, and show that, for a number of data structures implementations, we can handle significantly larger input structures and detect faults that CBMC is unable to detect.

2021-01-01T00:00:00Z Software model checkers are able to exhaustively explore different bounded program executions arising from various sources of nondeterminism. These tools provide statements to produce non-determinis- tic values for certain variables, thus forcing the corresponding model checker to consider all possible values for these during verification. While these statements offer an effective way of verifying programs handling basic data types and simple structured types, they are inappropriate as a mechanism for nondeterministic generation of pointers, favoring the use of insertion routines to produce dynamic data structures when verifying, via model checking, programs handling such data types. We present a technique to improve model checking of programs handling heap-allocated data types, by taming the explosion of candidate structures that can be built when non-deterministically initializing heap object fields. The technique exploits precomputed relational bounds, that disregard values deemed invalid by the structure’s type invariant, thus reducing the state space to be explored by the model checker. Precomputing the relational bounds is a challenging costly task too, for which we also present an efficient algorithm, based on incremental SAT solving. We implement our approach on top of the CBMC bounded model checker, and show that, for a number of data structures implementations, we can handle significantly larger input structures and detect faults that CBMC is unable to detect. Raura, Geovanny Pons, Claudia Fabiana Fonseca C., Efraín R. Dieste, Oscar http://sedici.unlp.edu.ar:80/handle/10915/140430 2022-08-11T20:03:06Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual) Contexto: Test-Driven Developement (TDD) es una técnica de desarrollo de software ágil que es ampliamente utilizada en la industria, aunque su efectividad ha generado incertidumbre si se compara con técnicas de desarrollo tradicional. Objetivo: Estudiar la efectividad de TDD considerando el grado de influencia de distintos factores humanos. Metodología: Experimento aleatorizado (crossover 2x2) realizado con sujetos profesionales en un ámbito académico. Resultados: La calidad y productividad al aplicar TDD es algo superior a lo obtenido con el desarrollo iterativo incremental (ITLD). La edad de los participantes, la función que desempeñaban en su trabajo y el conocimiento previo de la técnica de TDD ejercen influencia sobre las variables respuesta.

2021-01-01T00:00:00Z Contexto: Test-Driven Developement (TDD) es una técnica de desarrollo de software ágil que es ampliamente utilizada en la industria, aunque su efectividad ha generado incertidumbre si se compara con técnicas de desarrollo tradicional. Objetivo: Estudiar la efectividad de TDD considerando el grado de influencia de distintos factores humanos. Metodología: Experimento aleatorizado (crossover 2x2) realizado con sujetos profesionales en un ámbito académico. Resultados: La calidad y productividad al aplicar TDD es algo superior a lo obtenido con el desarrollo iterativo incremental (ITLD). La edad de los participantes, la función que desempeñaban en su trabajo y el conocimiento previo de la técnica de TDD ejercen influencia sobre las variables respuesta. Borrelli, Franco Martín Rouaux Servat, Candela Mariel Goin Plexevi, Francisco Challiol, Cecilia http://sedici.unlp.edu.ar:80/handle/10915/140427 2022-08-11T20:03:08Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual) El trabajo con equipos distribuidos se viene incrementando cada vez más en los últimos años. Sin embargo, la situación de pandemia generó que muchos equipos de trabajo que estaban acostumbrados a experiencias de codiseño presencial tuvieran que moverse rápidamente a un co-diseño virtual distribuido; generando así diferentes desafíos e impactando en la forma de abordaje. Este trabajo presenta una herramienta de autor que permite co-diseñar juegos móviles basados en posicionamiento en espacios indoor de manera distribuida. Además, se define un framework conceptual de co-diseño de este tipo de juegos, el cual permite combinar la herramienta de autor con recursos de Design Thinking. Usando este framework se describen dos experiencias de co-diseño distribuido; las cuales permiten explorar diferentes formas de utilizar un mismo artefacto de software en distintos abordajes de co-diseño. Asimismo, se presenta un espacio de discusión con algunas lecciones aprendidas y reflexiones sobre la temática.

2021-01-01T00:00:00Z El trabajo con equipos distribuidos se viene incrementando cada vez más en los últimos años. Sin embargo, la situación de pandemia generó que muchos equipos de trabajo que estaban acostumbrados a experiencias de codiseño presencial tuvieran que moverse rápidamente a un co-diseño virtual distribuido; generando así diferentes desafíos e impactando en la forma de abordaje. Este trabajo presenta una herramienta de autor que permite co-diseñar juegos móviles basados en posicionamiento en espacios indoor de manera distribuida. Además, se define un framework conceptual de co-diseño de este tipo de juegos, el cual permite combinar la herramienta de autor con recursos de Design Thinking. Usando este framework se describen dos experiencias de co-diseño distribuido; las cuales permiten explorar diferentes formas de utilizar un mismo artefacto de software en distintos abordajes de co-diseño. Asimismo, se presenta un espacio de discusión con algunas lecciones aprendidas y reflexiones sobre la temática. Gutiérrez Brida, Simón Regis, Germán Zheng, Guolong Bagheri, Hamid Nguyen, ThanhVu Aguirre, Nazareno Matías Frias, Marcelo F. http://sedici.unlp.edu.ar:80/handle/10915/140426 2022-08-11T20:03:09Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual); https://ieeexplore.ieee.org/document/9402059 The rising popularity of declarative languages and the hard to debug nature thereof have motivated the need for applicable, automated repair techniques for such languages. However, despite significant advances in program repair for imperative languages, there is a dearth of repair techniques for declarative languages. We present BeAFix, an automated repair technique for faulty models written in Alloy, a first-order relational logic language. BeAFix has a number of distinguishing features. Firstly, it supports any kind of oracle, including assertions typically found in formal specifications, as well as “specification tests”. This is important since unit tests, widely available for programs, are not typically found in formal specifications. Secondly, given a defined set of mutation operations, a set of suspicious expressions and a maximum number of mutations to apply per expression, BeAFix's bounded-exhaustive approach will either find a fix, or guarantee that such a fix is not possible, within the provided bounds. With respect to fault localization, BeAFix is not tightly integrated to any specific technique. In fact, our technique is independent of the fault localization technique used, and the fault localization is run only once before the repair process begins. To support a bounded-exhaustive approach while keeping repair times reasonable, sound state pruning techniques (i.e., those that guarantee that no valid fixes are removed) are introduced. When a faulty model has more than one suspicious expression, we use both syntactic analysis and dynamically generated scenario-based assertions to check the feasibility of a particular repair candidate. A failing check will determine that this candidate would never lead to a fully repaired model, allowing BeAFix to prune significant parts of the search space. We evaluated our technique on two Alloy benchmarks, including one previously used in a state-of-the-art technique for Alloy repair. The results show that BeAFix is able to repair thousands of real-world faulty models, corroborating its ability to effectively, and efficiently generate correct repairs while also being less prone to overfitting than previous techniques.

2021-01-01T00:00:00Z The rising popularity of declarative languages and the hard to debug nature thereof have motivated the need for applicable, automated repair techniques for such languages. However, despite significant advances in program repair for imperative languages, there is a dearth of repair techniques for declarative languages. We present BeAFix, an automated repair technique for faulty models written in Alloy, a first-order relational logic language. BeAFix has a number of distinguishing features. Firstly, it supports any kind of oracle, including assertions typically found in formal specifications, as well as “specification tests”. This is important since unit tests, widely available for programs, are not typically found in formal specifications. Secondly, given a defined set of mutation operations, a set of suspicious expressions and a maximum number of mutations to apply per expression, BeAFix's bounded-exhaustive approach will either find a fix, or guarantee that such a fix is not possible, within the provided bounds. With respect to fault localization, BeAFix is not tightly integrated to any specific technique. In fact, our technique is independent of the fault localization technique used, and the fault localization is run only once before the repair process begins. To support a bounded-exhaustive approach while keeping repair times reasonable, sound state pruning techniques (i.e., those that guarantee that no valid fixes are removed) are introduced. When a faulty model has more than one suspicious expression, we use both syntactic analysis and dynamically generated scenario-based assertions to check the feasibility of a particular repair candidate. A failing check will determine that this candidate would never lead to a fully repaired model, allowing BeAFix to prune significant parts of the search space. We evaluated our technique on two Alloy benchmarks, including one previously used in a state-of-the-art technique for Alloy repair. The results show that BeAFix is able to repair thousands of real-world faulty models, corroborating its ability to effectively, and efficiently generate correct repairs while also being less prone to overfitting than previous techniques. Zoratto, Valeria Aranda, Gabriela N. Martínez Carod, Nadina Otermin, Facundo http://sedici.unlp.edu.ar:80/handle/10915/140424 2022-08-11T20:03:10Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual) La información contenida en foros de discusión disponibles en laWeb, suele ser considerada muy valiosa por usuarios que intentan resolver problemas similares. En función de esta necesidad, en los últimos años se ha generado un especial interés tanto en la búsqueda de técnicas de recuperación como en el análisis de la información que se extrae de hilos de discusión. La propuesta de investigación, a nivel general, es capturar y analizar hilos de discusión existentes en foros técnicos para, a partir de un problema particular, sugerir un conjunto de soluciones exitosas en menos intentos que utilizando buscadores multipropósito tradicionales, teniendo en cuenta diferentes técnicas para lograrlo. En este trabajo en particular, se utilizan métodos de Procesamiento del Lenguaje Natural (PNL) para evaluar la información textual de los hilos de discusión considerando sus categorías gramaticales y el rol de cada palabra en el contexto que es utilizada. Esta funcionalidad es lograda mediante la herramienta Stanford POS Tagger, y el agregado de sinónimos mediante el uso de WordNet. Respecto al caso inicial, se observa que agregar sinónimos de adjetivos, adverbios y verbos al hilo de discusión, permite clasificarlos de manera más precisa respecto a los documentos de referencia, que utilizando el texto sin el agregado de sinónimos.

2021-01-01T00:00:00Z La información contenida en foros de discusión disponibles en laWeb, suele ser considerada muy valiosa por usuarios que intentan resolver problemas similares. En función de esta necesidad, en los últimos años se ha generado un especial interés tanto en la búsqueda de técnicas de recuperación como en el análisis de la información que se extrae de hilos de discusión. La propuesta de investigación, a nivel general, es capturar y analizar hilos de discusión existentes en foros técnicos para, a partir de un problema particular, sugerir un conjunto de soluciones exitosas en menos intentos que utilizando buscadores multipropósito tradicionales, teniendo en cuenta diferentes técnicas para lograrlo. En este trabajo en particular, se utilizan métodos de Procesamiento del Lenguaje Natural (PNL) para evaluar la información textual de los hilos de discusión considerando sus categorías gramaticales y el rol de cada palabra en el contexto que es utilizada. Esta funcionalidad es lograda mediante la herramienta Stanford POS Tagger, y el agregado de sinónimos mediante el uso de WordNet. Respecto al caso inicial, se observa que agregar sinónimos de adjetivos, adverbios y verbos al hilo de discusión, permite clasificarlos de manera más precisa respecto a los documentos de referencia, que utilizando el texto sin el agregado de sinónimos. Molina, Facundo Ponzio, Pablo Daniel Aguirre, Nazareno Matías Frias, Marcelo F. http://sedici.unlp.edu.ar:80/handle/10915/140422 2022-08-11T20:03:15Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual); https://ieeexplore.ieee.org/document/9402148 Software reliability is a primary concern in the construction of software, and thus a fundamental component in the definition of software quality. Analyzing software reliability requires a specification of the intended behavior of the software under analysis, and at the source code level, such specifications typically take the form of assertions. Unfortunately, software many times lacks such specifications, or only provides them for scenario-specific behaviors, as assertions accompanying tests. This issue seriously diminishes the analyzability of software with respect to its reliability, emphasizing the well known Oracle Problem. We tackle this problem by proposing a technique that, given a Java method, automatically produces a postcondition assertion of the method's current behavior. This mechanism is based on generating valid and invalid pre/post state pairs (i.e., state pairs that represent, and do not represent, the method's current behavior, respectively), which guide EvoSpex, a genetic algorithm, to produce a JML-like assertion characterizing the valid pre/post pairs, while leaving out the invalid ones. The generation of valid pre/post pairs is based on executing the method on a bounded exhaustive test set, generated by exercising the method inputs' APIs. The invalid pairs, on the other hand, are obtained by mutating valid pairs modifying the post-states so that each resulting pair does not belong to the set of valid pairs. EvoSpex searchs for an adequate postcondition assertion expressed in a JML-like language that involves quantification, object navigation and reachability expressions, making our approach particularly well-suited for reference-based class implementations with (implicit) strong representation invariants, such as heap-allocated structural objects, and complex custom types. We evaluated our technique on a benchmark of open source Java projects, featuring complex implementations of reference-based classes. In these case studies, EvoSpex was able to generate post-conditions that were stronger and more accurate, than those generated by related specification-inference approaches, as evaluated by an automated oracle assessment tool. Moreover, EvoSpex was also able to infer an important part of manually written rich postconditions (strong contracts used for verification) present in verified classes, and reproduce contracts for methods whose class implementations were automatically synthesized from specifications.

2021-01-01T00:00:00Z Software reliability is a primary concern in the construction of software, and thus a fundamental component in the definition of software quality. Analyzing software reliability requires a specification of the intended behavior of the software under analysis, and at the source code level, such specifications typically take the form of assertions. Unfortunately, software many times lacks such specifications, or only provides them for scenario-specific behaviors, as assertions accompanying tests. This issue seriously diminishes the analyzability of software with respect to its reliability, emphasizing the well known Oracle Problem. We tackle this problem by proposing a technique that, given a Java method, automatically produces a postcondition assertion of the method's current behavior. This mechanism is based on generating valid and invalid pre/post state pairs (i.e., state pairs that represent, and do not represent, the method's current behavior, respectively), which guide EvoSpex, a genetic algorithm, to produce a JML-like assertion characterizing the valid pre/post pairs, while leaving out the invalid ones. The generation of valid pre/post pairs is based on executing the method on a bounded exhaustive test set, generated by exercising the method inputs' APIs. The invalid pairs, on the other hand, are obtained by mutating valid pairs modifying the post-states so that each resulting pair does not belong to the set of valid pairs. EvoSpex searchs for an adequate postcondition assertion expressed in a JML-like language that involves quantification, object navigation and reachability expressions, making our approach particularly well-suited for reference-based class implementations with (implicit) strong representation invariants, such as heap-allocated structural objects, and complex custom types. We evaluated our technique on a benchmark of open source Java projects, featuring complex implementations of reference-based classes. In these case studies, EvoSpex was able to generate post-conditions that were stronger and more accurate, than those generated by related specification-inference approaches, as evaluated by an automated oracle assessment tool. Moreover, EvoSpex was also able to infer an important part of manually written rich postconditions (strong contracts used for verification) present in verified classes, and reproduce contracts for methods whose class implementations were automatically synthesized from specifications. Martínez, Roxana Pons, Claudia Fabiana Rodríguez, Rocío Vera, Pablo Martín http://sedici.unlp.edu.ar:80/handle/10915/140421 2022-08-11T20:03:16Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual) Este trabajo presenta la metodología utilizada para la validación de métricas de calidad de datasets abiertos gubernamentales. Estas son sugeridas en base a los distintos aspectos definidos sobre estudios y relevamientos realizados. Para ello se explica cada métrica propuesta a través de un prototipo de desarrollo propio llamado HEVDA, el cual realiza un estudio de un dataset y verifica el estado de éste para cada una de las métricas sugeridas (registros repetidos, cantidad de casos con registros incompletos, entre otras), obteniendo así, un “estado de salud de los datos abiertos disponibilizados”. Además, se plantean los criterios estándares y dimensiones de calidad que son utilizados para la justificación de dichas métricas basadas en Normas ISO, buenas prácticas y estudios de criterios estándares universales en aspectos de calidad.

2021-01-01T00:00:00Z Este trabajo presenta la metodología utilizada para la validación de métricas de calidad de datasets abiertos gubernamentales. Estas son sugeridas en base a los distintos aspectos definidos sobre estudios y relevamientos realizados. Para ello se explica cada métrica propuesta a través de un prototipo de desarrollo propio llamado HEVDA, el cual realiza un estudio de un dataset y verifica el estado de éste para cada una de las métricas sugeridas (registros repetidos, cantidad de casos con registros incompletos, entre otras), obteniendo así, un “estado de salud de los datos abiertos disponibilizados”. Además, se plantean los criterios estándares y dimensiones de calidad que son utilizados para la justificación de dichas métricas basadas en Normas ISO, buenas prácticas y estudios de criterios estándares universales en aspectos de calidad. Blas, María Julia Gonnet, Silvio M. Tebes, Guido Becker, Pablo Olsina Santos, Luis Antonio http://sedici.unlp.edu.ar:80/handle/10915/140420 2022-08-11T20:03:18Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual) Este trabajo presenta los términos, propiedades, relaciones y axiomas de ParticularEventCO como un modelo semántico basado en la noción de eventos causados por los comportamientos de las entidades. La ontología propuesta se encuentra situada en el nivel “Core” de una arquitectura ontológica de cuatro capas llamada FCD-OntoArch (Foundational, Core, and Domain Ontological Architecture for Sciences), donde en el nivel fundacional se encuentra la ontología ThingFO. Luego, se detalla la forma en la cual los principales elementos del nivel fundacional han sido redefinidos en ParticularEventCO. Además, para ilustrar la utilidad de la propuesta, presentamos la instanciación de un ejemplo.

2021-01-01T00:00:00Z Este trabajo presenta los términos, propiedades, relaciones y axiomas de ParticularEventCO como un modelo semántico basado en la noción de eventos causados por los comportamientos de las entidades. La ontología propuesta se encuentra situada en el nivel “Core” de una arquitectura ontológica de cuatro capas llamada FCD-OntoArch (Foundational, Core, and Domain Ontological Architecture for Sciences), donde en el nivel fundacional se encuentra la ontología ThingFO. Luego, se detalla la forma en la cual los principales elementos del nivel fundacional han sido redefinidos en ParticularEventCO. Además, para ilustrar la utilidad de la propuesta, presentamos la instanciación de un ejemplo. Zheng, Guolong Nguyen, ThanhVu Gutiérrez Brida, Simón Regis, Germán Frias, Marcelo F. Aguirre, Nazareno Matías Bagheri, Hamid http://sedici.unlp.edu.ar:80/handle/10915/140263 2022-08-09T20:03:27Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual) Alloy is a specification language that has been used in a wide range of applications, such as program verification, test case generation, IoT and Android security, etc. Unlike imperative languages, such as C or Java, Alloy is declarative, which describes the logic of a computation without describing its control flow and does not generate traces during the execution. Thus, traditional fault localization techniques developed for imperative programs based on analyzing the control flows of passing and failing tests do not directly apply to Alloy. To aid developers in debugging Alloy models, we develop FLACK, a tool to automatically localize Alloy buggy expressions. Given an Alloy model with violated assertions, FLACK automatically outputs a ranking list of expressions based on their spaciousness to the assertions violations. For each assertion, FLACK first queries the Alloy analyzer for counterexamples, i.e. instances of the model that violate the asserted property. FLACK then uses a Partial Max-SAT (PMAXSAT) solver to find instances that satisfy the asserted property and are most similar to the counterexamples. FLACK then identifies the relations and atoms that are different between the counterexamples and the satisfying instances. The differences illustrate how the counterexamples violate the assertion. The PMAXSAT solver guarantees that these differences are “minimal”, containing only essential information related to the assertion violation. By finding expressions most related to these differences, FLACK identifies the potential expressions causing the assertion violation. FLACK is different than the state of the art on Alloy fault localization in that it does not rely on unit tests which are not commonly found accompanying Alloy models. Instead, FLACK relies on assertions and constraint solvers to obtain counterexamples and satisfying instances, which are the main underlying technology in Alloy and commonly used by the Alloy developers.

2021-01-01T00:00:00Z Alloy is a specification language that has been used in a wide range of applications, such as program verification, test case generation, IoT and Android security, etc. Unlike imperative languages, such as C or Java, Alloy is declarative, which describes the logic of a computation without describing its control flow and does not generate traces during the execution. Thus, traditional fault localization techniques developed for imperative programs based on analyzing the control flows of passing and failing tests do not directly apply to Alloy. To aid developers in debugging Alloy models, we develop FLACK, a tool to automatically localize Alloy buggy expressions. Given an Alloy model with violated assertions, FLACK automatically outputs a ranking list of expressions based on their spaciousness to the assertions violations. For each assertion, FLACK first queries the Alloy analyzer for counterexamples, i.e. instances of the model that violate the asserted property. FLACK then uses a Partial Max-SAT (PMAXSAT) solver to find instances that satisfy the asserted property and are most similar to the counterexamples. FLACK then identifies the relations and atoms that are different between the counterexamples and the satisfying instances. The differences illustrate how the counterexamples violate the assertion. The PMAXSAT solver guarantees that these differences are “minimal”, containing only essential information related to the assertion violation. By finding expressions most related to these differences, FLACK identifies the potential expressions causing the assertion violation. FLACK is different than the state of the art on Alloy fault localization in that it does not rely on unit tests which are not commonly found accompanying Alloy models. Instead, FLACK relies on assertions and constraint solvers to obtain counterexamples and satisfying instances, which are the main underlying technology in Alloy and commonly used by the Alloy developers. Capece, Cintia Tatiana Der Boghosian, Reyna Romero, Ciro Edgardo Hazan, Pablo Colque, Jon http://sedici.unlp.edu.ar:80/handle/10915/140261 2022-08-09T20:03:29Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual) Dentro del mercado de software, orientar el desarrollo en la resolución de problemáticas es importante para ser competitivo. Enforcar los esfuerzos en el análisis de problemáticas es complejo cuando no se cuenta con una arquitectura pre definida asi como generar una estructura de un proyecto desde cero, puede ser una tarea que ocupe tiempo no necesariamente productivo. Desde esta premisa, el departamento de Investigación, Desarrollo e Innovación de C&S analiza el desarrollo de una herramienta propia que permita crear una estructura base, orientada a proyectos de software web.

2021-01-01T00:00:00Z Dentro del mercado de software, orientar el desarrollo en la resolución de problemáticas es importante para ser competitivo. Enforcar los esfuerzos en el análisis de problemáticas es complejo cuando no se cuenta con una arquitectura pre definida asi como generar una estructura de un proyecto desde cero, puede ser una tarea que ocupe tiempo no necesariamente productivo. Desde esta premisa, el departamento de Investigación, Desarrollo e Innovación de C&S analiza el desarrollo de una herramienta propia que permita crear una estructura base, orientada a proyectos de software web. Ruiz de Mendarozqueta, Álvaro Bustos, Fabio O. Colla, Pedro Ernesto http://sedici.unlp.edu.ar:80/handle/10915/140258 2026-03-05T17:36:56Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual) This paper reports on research work on Argentinean software development organizations. The analysis pro-vides insights on the profile of the companies regarding the usage of agile methods and software engineering practices trends, their motivations, and drivers. The conclusions can be used to understand what drivers facilitate the understanding of bonds between both in order to increase their competitiveness in domestic and off-shore markets.

2021-01-01T00:00:00Z This paper reports on research work on Argentinean software development organizations. The analysis pro-vides insights on the profile of the companies regarding the usage of agile methods and software engineering practices trends, their motivations, and drivers. The conclusions can be used to understand what drivers facilitate the understanding of bonds between both in order to increase their competitiveness in domestic and off-shore markets. Campo, Marcelo Amandi, Analía Acosta, Nelson Guerrero, Matías http://sedici.unlp.edu.ar:80/handle/10915/140253 2022-08-09T20:03:32Z 2021-01-01T00:00:00Z

Objeto de conferencia XXII Simposio Argentino de Ingeniería de Software (ASSE 2021) - JAIIO 50 (Modalidad virtual) In this article, the main motivations and conceptual basis of using chatbot-based technology as the platform to support education and training on agile development are presented, using ScrumTalk as an example. ScrumTalk is a prototype of an intelligent conversational core for the AgileTalk platform under development for an Intelligent Synthetic Agile Software Development Environment. AgileTalk represents an ambitious evolution of Virtual Scrum (a Virtual Reality-based platform for Scrum training we developed some years ago) towards a synthetic 3D conversational-based full-fledged process training and development environment. This evolution means a step further "Virtual" environments towards "Synthetic" ones where actors can be whatever human or synthetic alter egos that interact through natural language to carry out the activities involved in a defined process according to their role. ScrumTalk represents the proof of concept for the AgileTalk platform's underlying paradigm.

2021-01-01T00:00:00Z In this article, the main motivations and conceptual basis of using chatbot-based technology as the platform to support education and training on agile development are presented, using ScrumTalk as an example. ScrumTalk is a prototype of an intelligent conversational core for the AgileTalk platform under development for an Intelligent Synthetic Agile Software Development Environment. AgileTalk represents an ambitious evolution of Virtual Scrum (a Virtual Reality-based platform for Scrum training we developed some years ago) towards a synthetic 3D conversational-based full-fledged process training and development environment. This evolution means a step further "Virtual" environments towards "Synthetic" ones where actors can be whatever human or synthetic alter egos that interact through natural language to carry out the activities involved in a defined process according to their role. ScrumTalk represents the proof of concept for the AgileTalk platform's underlying paradigm.