Improving the Decision-Making Support in Context-Aware Applications: The Case of an Adaptive Virtual Education Learning Management System
Jose Bocanegra, Luis Garcia-Paucar, Jaime Pavlich-Mariscal and Nelly Bencomo
Goal:
Context entities/Information Items: monitoring student's ability information such as critical thinking, problem solving, assertive communication may lead to recommendations for students to do essays, presentations or simulations.
Recommendation Model:
They defined metrics regarding ranges of satisfaction levels regarding the student's skills. Based on those ranges, they have also developed a transition model to determine the probability of enhancing the student's ability in case he/she performs each type of activity (essays, presentation, survey).
They have made some experiments to adjust the probabilities and weights of their recommendation model.
This work has provided them with directions for future work:
- the use of decision maker in real environments;
- incorporate the decision maker in a framework to develop context-aware applications (Midas);
- Use the decision maker in other domains (e.g. supply chains).
Ufahamu: Um Framework para Aplicações Móveis Android Conscientes de Contexto
Rafael Souza and Cleidson De Souza
Context aware systems adapt their functionalities to the environment without explicit intervention of the user from data captured from the context. E.g. Moto Assist (context-aware app to control smartphones) has a limited set of contexts (Sleeping, Driving, Home and Meeting) and actions (Talk to me, Silence, etc.).
Ufahamu allows people to build context-aware mobile applications (the framework's users are thus software developers)
Ufahamu is a java framework developed for Android; it was built on top of the Google's Awareness API; and it was designed as a context-rule-action model based on the trigger-action programming paradigm (uses if this then that - https://ifttt.com).
Validation:
1) performed by the user itself, highlights some properties of the framework;
2) comparison with other systems, applying Metrix, a tool that compares the complexity of APIs, i.e. how complex to use an API is (for developers); the result was positive in terms of complexity; on the other hand, the tools that were more complex also provided more functionality. In the future, it would be a fairer evaluation to compare complexity vs. functionality.
3) Feasibility study with students.
The results of both validations have been positive.
Future work: controlled experiment for validation; improve the framework as a result of this validation.
Sýntixi - A Generative Approach to Dynamic Fusion of Software Components
Juan-Alberto Hernández-Martínez, Ulises Juárez-Martínez and Nicolás Cardozo
- Current software applications must be able to respond immediately and automatically to constant changes.
- This changes must be made without stopping the execution of the application. E.g. bank systems, airplane controlling system, etc.
- The approach uses a dynamic component fusion analogy (inspired in a simplified chemical model for molecule creation)
1. each atom corresponds to a software component;
2. Each atom provides its elements partially or completely;
3. each molecule corresponds to a new software component adaptation or generation.
- The approach also uses the idea of generator to create and manage new requirements for the application, which should adapt in runtime.
- Syntixi is a generator whose main goal is the dynamic fusion of software components from high-level specifications.
In Syntixi:
- Requirements - requirements are specified in XML.
- Fusionable Component - from the XML specification, checks the buildability of a component given such specification, completes the specification in case the buildability check indicates incompleteness, assembles the component and finally results in a concrete system component.
- Generator - works following a cycle: monitoring - analysis - planning - execution - knowledge update
- There are different fusion alternatives, i.e. different aways to fuse components to generate a new component.
Future work:
- strengthen the mechanism for creating new behavior based on the selected elements of each component;
- enhance the requirements specification approach.
Quito SmartSafe City: Un Sistema en Tiempo Real de Rescate Usando Comunicación IoT
Ana Zambrano, Eduardo Ortiz, Xavier Calderón and Marcelo Zambrano.
Kidnaping is increasing in global scope.
Goal of this project: decrease the rescuing time in case of a kidnapping.
Related works:
- Natalia Estemirova Project in Russia;
- Child Abduction Alert System
- Amber Alert in USA
The project proposes a socio-technical architecture in which the parent of a kidnapped child can contact the authorities with a multimedia info (photo, video etc.) characterizing the kidnapped child and the authorities share this multimedia info with the surrounding community. Someone identifying the kid may share this information in the system so that the authorities can take action.
This architecture includes web-applications, cloud servers and android apps.
The author presented a few applications that have been developed in the scope of this project, having such architecture into account. They have used Amazon Web Services for the cloud part.
Future work:
- make a more generic architecture that serves to other cities - not only Quito
- make the applications themselves more secure
They have a straight collaboration with the Interior Ministry to have expert knowledge about kidnapping. Thus, there is also explicit interest from the authorities in their project.
They plan to involve the community with some marketing campaigns and surveys in order to understand if they would like to use such app and advertise its use, aiming at obtaining adoption from the public.
Jose Bocanegra, Luis Garcia-Paucar, Jaime Pavlich-Mariscal and Nelly Bencomo
Goal:
Context entities/Information Items: monitoring student's ability information such as critical thinking, problem solving, assertive communication may lead to recommendations for students to do essays, presentations or simulations.
Recommendation Model:
They defined metrics regarding ranges of satisfaction levels regarding the student's skills. Based on those ranges, they have also developed a transition model to determine the probability of enhancing the student's ability in case he/she performs each type of activity (essays, presentation, survey).
They have made some experiments to adjust the probabilities and weights of their recommendation model.
This work has provided them with directions for future work:
- the use of decision maker in real environments;
- incorporate the decision maker in a framework to develop context-aware applications (Midas);
- Use the decision maker in other domains (e.g. supply chains).
Ufahamu: Um Framework para Aplicações Móveis Android Conscientes de Contexto
Rafael Souza and Cleidson De Souza
Context aware systems adapt their functionalities to the environment without explicit intervention of the user from data captured from the context. E.g. Moto Assist (context-aware app to control smartphones) has a limited set of contexts (Sleeping, Driving, Home and Meeting) and actions (Talk to me, Silence, etc.).
Ufahamu allows people to build context-aware mobile applications (the framework's users are thus software developers)
Ufahamu is a java framework developed for Android; it was built on top of the Google's Awareness API; and it was designed as a context-rule-action model based on the trigger-action programming paradigm (uses if this then that - https://ifttt.com).
Validation:
1) performed by the user itself, highlights some properties of the framework;
2) comparison with other systems, applying Metrix, a tool that compares the complexity of APIs, i.e. how complex to use an API is (for developers); the result was positive in terms of complexity; on the other hand, the tools that were more complex also provided more functionality. In the future, it would be a fairer evaluation to compare complexity vs. functionality.
3) Feasibility study with students.
The results of both validations have been positive.
Future work: controlled experiment for validation; improve the framework as a result of this validation.
Sýntixi - A Generative Approach to Dynamic Fusion of Software Components
Juan-Alberto Hernández-Martínez, Ulises Juárez-Martínez and Nicolás Cardozo
- Current software applications must be able to respond immediately and automatically to constant changes.
- This changes must be made without stopping the execution of the application. E.g. bank systems, airplane controlling system, etc.
- The approach uses a dynamic component fusion analogy (inspired in a simplified chemical model for molecule creation)
1. each atom corresponds to a software component;
2. Each atom provides its elements partially or completely;
3. each molecule corresponds to a new software component adaptation or generation.
- The approach also uses the idea of generator to create and manage new requirements for the application, which should adapt in runtime.
- Syntixi is a generator whose main goal is the dynamic fusion of software components from high-level specifications.
In Syntixi:
- Requirements - requirements are specified in XML.
- Fusionable Component - from the XML specification, checks the buildability of a component given such specification, completes the specification in case the buildability check indicates incompleteness, assembles the component and finally results in a concrete system component.
- Generator - works following a cycle: monitoring - analysis - planning - execution - knowledge update
- There are different fusion alternatives, i.e. different aways to fuse components to generate a new component.
Future work:
- strengthen the mechanism for creating new behavior based on the selected elements of each component;
- enhance the requirements specification approach.
Quito SmartSafe City: Un Sistema en Tiempo Real de Rescate Usando Comunicación IoT
Ana Zambrano, Eduardo Ortiz, Xavier Calderón and Marcelo Zambrano.
Kidnaping is increasing in global scope.
Goal of this project: decrease the rescuing time in case of a kidnapping.
Related works:
- Natalia Estemirova Project in Russia;
- Child Abduction Alert System
- Amber Alert in USA
The project proposes a socio-technical architecture in which the parent of a kidnapped child can contact the authorities with a multimedia info (photo, video etc.) characterizing the kidnapped child and the authorities share this multimedia info with the surrounding community. Someone identifying the kid may share this information in the system so that the authorities can take action.
This architecture includes web-applications, cloud servers and android apps.
The author presented a few applications that have been developed in the scope of this project, having such architecture into account. They have used Amazon Web Services for the cloud part.
Future work:
- make a more generic architecture that serves to other cities - not only Quito
- make the applications themselves more secure
They have a straight collaboration with the Interior Ministry to have expert knowledge about kidnapping. Thus, there is also explicit interest from the authorities in their project.
They plan to involve the community with some marketing campaigns and surveys in order to understand if they would like to use such app and advertise its use, aiming at obtaining adoption from the public.
Nenhum comentário:
Postar um comentário