Using a Combination Method of MDS and SOM to Visually Analyze Postpartum Depression Domain
- Resource Type
- Conference
- Authors
- Meng, Xi; Shen, Ruifang; Li, Jianqiang; Yang, Jijiang
- Source
- 2015 IEEE 39th Annual Computer Software and Applications Conference Computer Software and Applications Conference (COMPSAC), 2015 IEEE 39th Annual. 3:95-95 Jul, 2015
- Subject
- Computing and Processing
Data visualization
Pediatrics
XML
Text mining
Visualization
Artificial neural networks
Web mining
multidimensional scaling (MDS); Sammons mapping; self-organizing map(SOM); postpartum depression (PPD); visual analysis; visualization method
- Language
- ISSN
- 0730-3157
This study focused on a combination method that included MDS and SOM to visually reveal research topics in postpartum depression research domain from the bibliographic point of view. 391 documents which were under postpartum depression domain were used to be analyzed in this study. Considering that MDS and SOM each dimension reduction method has its own disadvantages, we used the two visualization techniques. The nonmetric MDS method (Sammon's mapping) combined with self-organizing map was employed to separate documents based on their topics, and a cluster analysis was conducted. The experiment result proved that Sammon's mapping combined with self-organizing map techniques yielded the better visualization effect in revealing postpartum depression domain.