Métodos de acceso para bases de datos multimedia y sus paralelizaciones
Ver/ Abrir
Metadatos
Mostrar el registro completo del ítemcomunitat-uji-handle:10234/9
comunitat-uji-handle2:10234/7036
comunitat-uji-handle3:10234/27725
comunitat-uji-handle4:
INVESTIGACIONMetadatos
Título
Métodos de acceso para bases de datos multimedia y sus paralelizacionesFecha de publicación
2007-04Editor
Departament d' Enginyeria i Ciència dels Computadors, Universitat Jaume ITipo de documento
info:eu-repo/semantics/reportPalabras clave / Materias
Access methods | Multidimensional indexing | Similarity indexing | High-dimensional indexing | Paralellization | Text mining | Data mining | Métodos de acceso | Indexado multidimensional | Indexado por semejanza | Indexado en alta dimensionalidad | Paralelización | Minería de textos | Minería de datos
Resumen
Similarities queries are very important in data mining, and specially in text mining. The goal of this type of query is searching for all the objects in the database which are similar to a given object. Most similarity ... [+]
Similarities queries are very important in data mining, and specially in text mining. The goal of this type of query is searching for all the objects in the database which are similar to a given object. Most similarity search techniques map the data objects into some feature space. Particulary in text mining this feature space has thousands of dimensions. Then, the similarity search correspond toa search of nearest-neighbourobjects in the feature space.
On the other hand text databases are very large and so it is necessary to use indexing methods to organize the objects in the database in such a way that only a small portion of database must be explored to retrieve the target objects. This technical report reviews and compares the main tree indexing techniques, to determine their suitability to text mining in some stage of text processing. Most tree indexing methods perform badly in very-high-dimensional space, and so other techniques suitable for indexing in this type of space are also described. Our main conclusion is that only very few existing searching techniques (VA-File, IQ-Tree, Nene’s method) can be applied to very high-dimensional spaces such as the spaces arising in text mining.
This report also reviews and compares the main parallel indexing techniques that appear in the literature. [-]
Derechos de acceso
http://rightsstatements.org/vocab/CNE/1.0/
info:eu-repo/semantics/openAccess
info:eu-repo/semantics/openAccess
Aparece en las colecciones
- ICC_Reports [18]