| H01-1030 | attempt to explain why certain | data fusion | experiments succeed where others |
| E95-1010 | combined with a direct approach to | data fusion | for multiple sources of information |
| H01-1030 | CONCLUSIONS A variety of techniques for | data fusion | have been proposed in IR literature |
| H01-1030 | . There are many papers in the | data fusion | literature , which attempt to |
| H01-1030 | in a news stream . Instead the | data fusion | element of our system would involve |
| H01-1030 | paper , we explore the effects of | data fusion | on First Story Detection -LSB- |
| H01-1030 | give better performance . In our | data fusion | experiment in Section 5 we observed |
| H01-1030 | general preconditions for successful | data fusion | involving non-specific sources |
| H01-1030 | evidence . 1 . Dissimilarity : | Data fusion | between operationally very similar |
| H01-1030 | a broadcast news domain . The | data fusion | element of this experiment involves |
| H01-1030 | any concrete reasoning as to why | data fusion | under these particular conditions |
| H01-1030 | performance were observed in our | data fusion | experiment . The second criteria |
| H01-1030 | before they are combined in the | data fusion | process . 2 . Efficacy : Data |
| H01-1030 | strategies and propose reasons why our | data fusion | experiment shows performance |
| D15-1303 | parallelizable decision-level | data fusion | method , which is much faster |
| H01-1030 | general criteria for successful | data fusion | . 2 . LEXICAL CHAINING A lexical |
| H01-1030 | fusion process . 2 . Efficacy : | Data fusion | between a capable IR system and |
| H01-1030 | between two weighted vectors ) . The | data fusion | element of this experiment involves |
| H01-1030 | FSD classification based on our | data fusion | strategy in Section 4 . The remaining |
| H01-1030 | criteria defined for successful | data fusion | regards efficacy or the quality |
