 | Information Resources Management Association. International Conference - 2000 - 1244 lapas
...feature vectors B, B, Color Index Sjiaj» ftatwejjjflor < AAA content-based retrieval efficiently. Figure 9 shows both color and shape index structures...retrieval, there is little research on retrieval models for both structured- and content-based multimedia information retrieval. In this section, we can specially... | |
 | Qiyang Chen - 2001 - 272 lapas
...|0id| 1^ DID 2 EF2 ... | >--». | \ Figure 10. Attribute index structure Posting File Attribute Index Unified Retrieval Even though there have been various retrieval models in the traditional text- based information retrieval, there is little research on retrieval models for structured- and... | |
 | Rahman, Syed M. - 2001 - 316 lapas
...dimension, while it uses a linear directory structure that can save a memory space and lead to fast accesses to nodes in a high dimension. Thus, in this chapter,...that of node t, as shown in the following equation (Saltón and McGill, 1 983). Sw = COSINE(NODE,,NODE.) = J £ (7EKA/,*)2 ¿ i*=i *=i When results for... | |
 | Hiroshi Arisawa - 2002 - 515 lapas
...high-dimensional feature vectors of both color and shape, we are doing research on a parallel CBF method. 3.4 Unified Retrieval Even though there have been various...retrieval models for structured- and content-based XML information retrieval. To answer a document structure query, a similarity (SJ between an element... | |
 | Osvaldo Gervasi - 2005 - 1297 lapas
...for retrieval based on document structure, a similarity measure (Sw) between two elements, say q and t, is computed as the similarity between the term vector of node q and that of node t by using a cosine measure [15]. Supposed that a document can be represented as D = { Eo, E 1 , E¿... | |
 | Edward A. Fox - 2005 - 545 lapas
...for retrieval based on document structure, a similarity measure (Sw) between two elements, say q and t, is computed as the similarity between the term vector of node q and that of node t by using a cosine measure. Supposed that a document can be represented as D = { EO, E\, ..... En.|... | |
 | Heng Tao Shen, Jinbao Li, Minglu Li, Jun Ni, Wei Wang - 2006 - 1083 lapas
...retrieval based on document structure, a similarity measure (S w ) between two elements, say q and /, is computed as the similarity between the term vector of node q and that of node t by using a cosine measure [11]. Supposed that a document can be represented as D = { EQ, E\, ....,... | |
| |
Figure 9 shows both color and shape index structures constructed using X-tree. Unified Retrieval Even though there have been various retrieval models in the traditional text-based information retrieval, there is little research on retrieval models for... 