Figure 1.
Framework of the biomedical text summarization system.
Figure 2.
An example of semantic relation extraction.
Table 1.
Direct relations with “Angina Pectoris”.
Figure 3.
The core relation set for “Angina Pectoris”.
Figure 4.
Semantic relation network for “Angina Pectoris” after relation retrieval.
Figure 5.
An example of relation and sentence retrieval.
Table 2.
Location scores used in our experiment.
Table 3.
Diseases use in our experiment.
Figure 6.
Comparison of summarization performance on ROUGE-1.
Table 4.
Performance of summarization for 24 diseases.
Table 5.
The impact of relation expansion, noise filtering and redundant removal.
Figure 7.
Relationship between ROUGE-1 and concept depth in MeSH based filtering.
Figure 8.
Relationship between ROUGE-1 and the trade-off parameter
Figure 9.
Relationship between ROUGE-1 and the trade-off parameter ω.