Browsing by Issue Date, starting with "2025-10-17"
Now showing 1 - 3 of 3
Results Per Page
Sort Options
- Application of NLP techniques for the optimization of SQL driven data analysis in ERP softwarePublication . VIOLANTE, DIOGO DE SÁ; Martinho, Diogo Emanuel PereiraData management in Industry 4.0 has become a growing complexity process, leading to industries increasingly relying on large-scale datasets, which results in traditional analysis methods becoming inefficient and even inaccessible for end users. Enterprise Resource Planning systems deal with heterogeneous data from multiple modules and processes, which creates a need for more accessible and sophisticated tools. Recently, the growth of Artificial Intelligence solutions has played a central role in addressing these challenges. Fields like Natural Language Processing, Computer Vision and Machine Learning have helped the development of systems that create more value from complex datasets, making information more manageable across industrial environments. The objective of this thesis is the exploration, implementation and validation of NLP solutions with generative capabilities that can integrate into these systems, by proposing a solution that aims at providing a more efficient and optimized way of analyzing SQL data, through a pipeline that transforms user natural queries into SQL queries used for data retrieval. A conversational chatbot, capable of translating natural language queries into SQL statements, was developed, with the central feature of this project being a RAG component used to search files with database tables schema to provide context to a LLM, for it to generate SQL statements that can be used to retrieve information, without compromising the user experience or the database itself. The user’s intent is detected and the RAG component is adapted according to it. A mechanism to search the Web for information was also developed, to help provide context, when there is not enough to create a valid answer. The generated queries are analyzed, to prevent potential dangers for the integrity of the database and, if they are considered as valid, they are persisted by another component, to be used in future context to formulate other queries. The chosen LLM model, as the backbone for this pipeline, allows not only for the generation of the queries but also for providing text answers for several matters, including user manuals or simple informal conversations, depending on the need. Also, it’s multi-language support helps in enhancing the overall user experience and accessibility. A test set with real-world examples was created, to help validate the system, by using evaluation metrics like Exact Match Accuracy, Execution Accuracy and Valid Efficiency Score. A manual validity test was also conducted, to determine if the queries that did not achieve a good Exact Match Accuracy score, could still be considered as valid, given the ambiguity of the SQL language. The results demonstrate that the system is capable of handling queries with simple to medium complexity, but needs further optimization for higher ones. This helps to conclude that NLP-driven text-to-SQL solutions can enhance data accessibility for both technical and non-technical users, while compliance with privacy and security requirements.
- Otimização do algoritmo de correspondência de prestadores de serviços para uma plataforma de serviços domésticosPublication . GASPAR, DIOGO FRANCISCO SOUSA; Malheiro, Nuno Filipe TeixeiraThis thesis addresses the optimization of the service provider–client matching algorithm in Oscar, a growing start-up offering on-demand and scheduled home services. The existing algorithm relied on rules with a low flexibility, resulting in high cancellation rates, poor provider engagement, and limited visibility into decision-making. To overcome these challenges, the work combines two complementary directions: algorithmic improvement through a fairnessaware boosting framework, and the integration of observability to enhance monitoring and transparency. A systematic mapping study was conducted to analyze state-of-the-art matching algorithms, supply–demand compatibility strategies, and observability methodologies. Based on the findings, a Conceptual Boosting Framework was designed, incorporating modular heuristics such as task performance and recent activity boosts, while ensuring configurability and fairness monitoring. Observability was integrated via structured logging, telemetry, and tracing, enabling detailed insights into algorithmic decisions and operational metrics. The solution was implemented and deployed in the production system of Oscar. Evaluation included unit and integration tests, as well as one-week of A/B testing comparing the new algorithm against the baseline. The final results demonstrated significant improvements: cancellations decreased by 17.9%, pool time by 18.5%, and time-to-accept by 26.1%, while acceptance rate increased by 15.3%. These outcomes validate the effectiveness of the approach in improving efficiency and user satisfaction. The thesis contributes by introducing a practical framework that balances efficiency, fairness, and observability, empirically validating it on a live service platform, and documenting a design that is extensible to other gig-economy domains. Limitations include low prioritization for testing, reliance on heuristics rather than machine learning, and lack of quantitative fairness auditing. Future work should address these aspects through long-term experiments, machine learning-based adaptive boosting, and formal fairness evaluation.
- Desenvolvimento de uma cânula flexível de retenção automática e estudo da força de retenção pelo método de elementos finitosPublication . VILARIÇA, ÍRIS CATARINA CÉSAR; Marques, Maria ArcelinaA laparoscopia é uma técnica cirúrgica minimamente invasiva, amplamente usada em diversas intervenções. A eficácia destes procedimentos depende, em grande parte, da estabilidade dos instrumentos de acesso, os trocartes. Um dos principais desafios associados a estes dispositivos médicos é a força de retenção, responsável por manter a cânula estável após a sua inserção na parede abdominal. Com o objetivo de otimizar este parâmetro, foi desenvolvida uma cânula flexível de retenção automática no SolidWorks, combinando materiais biocompatíveis e técnicas de fabrico inovadoras. Para avaliar o desempenho do novo design, realizaram-se simulações pelo método de elementos finitos, modelando o ambiente clínico com a cânula inserida em tecidos simulados. Foram considerados dois modelos teciduais (1 - maior espessura de tecido adiposo e baixa tonicidade muscular; 2 - menor espessura de tecido adiposo e maior tonicidade muscular) e três modelos de cânula (lisa, de rosca e a desenvolvida neste projeto). As simulações incluíram condições de contacto realistas, atrito e forças aplicadas, avaliando-se os resultados tanto das cânulas como dos tecidos, a nível de stress, deformação, deslocamento e, por fim, da força de retenção. Os resultados demonstraram que a cânula flexível desenvolvida apresentou a maior força de retenção, com 3,10 e 1,03 N para os modelos teciduais 1 e 2, respetivamente, superando a cânula de rosca (4,2x10-1 e 6,53x10-2 N) e a lisa (2,26x10-3 e 4,42x10-4 N). Os resultados indicaram também que o novo design proporciona maior estabilidade e melhor adaptação aos tecidos, evidenciando o seu potencial para uma melhor segurança e eficácia. Este trabalho propõe e valida uma nova abordagem no design de trocartes através do desenvolvimento de uma cânula flexível. Contribui para a compreensão do seu comportamento mecânico, fornecendo bases sólidas para o desenvolvimento e contínua evolução de procedimentos la laparoscópicos.