Name
Abstract | ||
|---|---|---|
Literature-based discovery, the process of choosing bridge terms to build plausible connections between “unrelated” terms, helps the biomedical research by generating potentially useful hypotheses. The basic idea of previous work is conducting two or higher order association rules searching and ranking, either according to the semantics or the statistical measurements. However, the high order searching nature makes current approaches computationally complex. We address this problem by proposing a generative model assuming all the connection words as a rank of words. We further present an inference method for the generative model to learn the rank. Our model avoids the process of high order searching. Experiments show that our approach addresses the problem both effectively and efficiently. |
Year | DOI | Venue |
|---|---|---|
2019 | 10.1109/BIBM47256.2019.8983400 | 2019 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) |
Keywords | Field | DocType |
text mining,literature-base discovery,generative model,topic model | Ranking,Computer science,Inference,Association rule learning,Artificial intelligence,Literature-based discovery,Topic model,Generative grammar,Machine learning,Semantics,Generative model | Conference |
ISSN | ISBN | Citations |
2156-1125 | 978-1-7281-1868-0 | 0 |
PageRank | References | Authors |
0.34 | 0 | 2 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Juncheng Ding | 1 | 0 | 0.68 |
| Wei Jin | 2 | 83 | 25.25 |