Incremental Support Vector Machine Construction
ICDM '01 Proceedings of the 2001 IEEE International Conference on Data Mining
AllInOneNews: development and evaluation of a large-scale news metasearch engine
Proceedings of the 2007 ACM SIGMOD international conference on Management of data
Incremental Support Vector Learning: Analysis, Implementation and Applications
The Journal of Machine Learning Research
Author name disambiguation in MEDLINE
ACM Transactions on Knowledge Discovery from Data (TKDD)
Solving multiclass learning problems via error-correcting output codes
Journal of Artificial Intelligence Research
Evaluation of result merging strategies for metasearch engines
WISE'05 Proceedings of the 6th international conference on Web Information Systems Engineering
Screening nonrandomized studies for medical systematic reviews: A comparative study of classifiers
Artificial Intelligence in Medicine
| Hi-index | 0.00 |
High quality, cost-effective medical care requires consideration of the best available, most appropriate evidence in the care of each patient, a practice known as Evidence-based Medicine (EBM). EBM is dependent upon the wide availability and coverage of accurate, objective syntheses called evidence reports (also called systematic reviews). These are compiled by a time and resource-intensive process that is largely manual, and that has not taken advantage of many of the advances in information processing technologies that have assisted other textual domains. We propose a specific text-mining based pipeline to support the creation and updating of evidence reports that provides support for the literature collection, collation, and triage steps of the systematic review process. The pipeline includes a metasearch engine that covers both bibliographic databases and selected "grey" literature; a module that classifies articles according to study type; a module for grouping studies that are closely related (e.g. that derive from the same underlying clinical trial or same study cohort); and an automated system that ranks publications according to the likelihood that they will meet inclusion criteria for the report. The proposed pipeline will also enable groups performing systematic review to reuse tools and models created by other groups, and will provide a test-bed for further informatics research to develop improved tools in the future. Ultimately, this should increase the rate that high-quality systematic reviews and meta-analyses can be generated, accessed and utilized by clinicians, patients, care-givers, and policymakers, resulting in better and more cost-effective care.