Percorrer por autor "Azevedo, Renato Preigschadt de"
A mostrar 1 - 3 de 3
Resultados por página
Opções de ordenação
- Development of Q&A systems using AcQAPublication . Azevedo, Renato Preigschadt de; Pereira, Maria João; Henriques, Pedro RangelIn order to help the user to search for relevant information, Question and Answering (Q&A) Systems provide the possibility to formulate the question freely in a natural language, retrieving the most appropriate and concise answers. These systems interpret the user question to understand his information needs and return him the more adequate replies in a semantic sense; they do not perform a statistical word search like happens in the existing search engines. There are several approaches to developing and deploying Q&A Systems, making it hard to choose the best way to build the system. To turn easier this process, we are proposing a way to automatically create Q&A Systems (AcQA) based on DSLs, thus allowing the setup and the validation of the Q&A System independent of the implementation techniques. With our proposal (AcQA language), we want the developers to focus on the data and contents, instead of implementation details. We conducted an experiment to assess the feasibility of using AcQA. The study was carried out with people from the computer science field and shows that our language simplifies the development of a Q&A System.
- DSL based automatic generation of Q&A systemsPublication . Azevedo, Renato Preigschadt de; Pereira, Maria João; Henriques, Pedro RangelIn or der to help the user to search for relevant information,Question and Answering (Q&A) Systems provide the possibility to formulate the question freely in a natural language, retrieving the most appropriate and concise answers. These systems interpret the user’s question to understand his information needs and return him the more adequate replies in a semantic sense; they do not perform a statistical word search, thus differing from the existing search engines. There are several approaches to develop and deploy Q&A Systems, making hard to choose the best way to build the system. To turn easier this process, we are proposing a way to create automatically Q&A Systems based on DSLs (Domain-specific Languages), thus allowing the setup and the validation of the Q&A System to be independent of the implementation techniques. With our proposal, we want the developers to focus on the data and contents, instead of implementation details.
- Extending pythonQA with knowledge from stackOverflowPublication . Azevedo, Renato Preigschadt de; Henriques, Pedro Rangel; Pereira, Maria JoãoQuestion and Answering (QA) Systems provide a platform where users can ask questions in natural language to a system and get answers retrieved from a knowledge base. The work proposed in PythonQA create a Question and Answer System for the Python Programming Language. The knowledge is built from the Python Frequent Answered Questions (PyFAQ). In this paper, we extend the PythonQA system by enhancing the Knowledge Base with Question-Answer pairs from the StackExchange Python Question Answering Community Site. Some tests were performed to analyze the impact of a richer Knowledge Base on the PythonQA system, increasing the number of answer candidates.