Capturing and anticipating user intents in complex analytics workflows
In today's data-driven world, the ability to extract meaningful information from data is becoming essential for businesses, organizations and researchers alike. For that purpose, a wide range of tools and systems exist involving data-related tasks, from data integration, preprocessing and model...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis de maestría |
| Fecha de publicación: | 2023 |
| País: | España |
| Institución: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/398992 |
| Acceso en línea: | https://hdl.handle.net/2117/398992 |
| Access Level: | acceso abierto |
| Palabra clave: | Data mining mineria de dades graf de coneixement anàlisi d'intencions data mining knowledge graphs intent analytics knowledge graph embeddings Mineria de dades Àrees temàtiques de la UPC::Informàtica::Sistemes d'informació |
| Sumario: | In today's data-driven world, the ability to extract meaningful information from data is becoming essential for businesses, organizations and researchers alike. For that purpose, a wide range of tools and systems exist involving data-related tasks, from data integration, preprocessing and modeling, to the interpretation and evaluation of the results. As data continues to grow in volume, variety, and complexity, there is an increasing need for advanced but user-friendly tools that facilitate the user's interaction with them. This is relevant for end users who may not have technical knowledge but still need to take advantage of these powerful tools to make informed decisions. To this end, the primary goal of this Master's Thesis is to investigate and develop methods for capturing and anticipating user intents within complex analytics workflows, focusing on making these systems more intuitive for inexperienced users. This is, the aim is to assist end users when they express their preferences, constraints and/or intentions over analytics workflows, in hopes to enhance their ability to effectively engage with these systems. To achieve this objective, this thesis explores the use of Knowledge Graphs as the basic framework for capturing and anticipating user intentions. First, a Knowledge Graph capable of capturing data mining processes as a whole, from the different workflow steps to the user's inputs and feedback, has been designed by creating new concepts to extend and link other existing ontologies. Then, the graph has been leveraged for anticipation by knowledge extraction through queries. Additionally, the usage of Knowledge Graph Embeddings for recommendation tasks with Link Prediction has been explored. |
|---|