conference · 2004
Un modelo de recuperación de información basado en SVMs
See where this sits in the topic map →Summary AI-generated
- TL;DR
- While Support Vector Machines (SVMs) have proven very efficient for document classification, their usefulness in Information Retrieval (IR) for ranking documents remains unverified.
- Problem
- The effectiveness of SVM classifiers in the document ranking phase of Information Retrieval has not yet been demonstrated.
- Method
- The authors propose a transformation that maps the IR process into a new vector space, allowing an SVM-based classifier to be trained to learn the concept of document similarity.
- Results
- Not specified in the abstract.
- Contributions
- A novel transformation method that adapts SVM classifiers to learn document similarity within a new vector space for Information Retrieval.
- Limitations
- Not specified in the abstract.
- Takeaways
- SVMs can be successfully adapted for Information Retrieval tasks by transforming the process into a new vector space to learn document similarity.
- Applications
- Not specified in the abstract.
- Topics
- Not specified in the abstract.
- For industry
- Not specified in the abstract.
- Why it matters
- Not specified in the abstract.
Abstract
Los clasificadores como los SVMs (Support Vector Machines) se usaron para la clasificacion de documentos de manera muy eficiente, pero su utilidad no ha sido comprobada para la recuperacion de informacion (RI) en el momento de jerarquerizar los documentos. En este articulo proponemos una transformacion que asocia el proceso de la RI a un nuevo espacio vectorial en el que un clasificador basado en SVMs se entrena para aprender el concepto de similitud frente a los documentos.