Main Page
From pubcurator
Contents |
PubCurator - A Biomedical Text Mining Platform
Welcome to the PubCurator homepage.
PubCurator is a highly extensible text mining platform for the biomedical field. PubCurator allows to automatically evaluate given texts and supports the manual validation by the user. For that task the whole architecture is built upon the Apache UIMA analysis framework and [Eclipse RCP http://wiki.eclipse.org/index.php/Rich_Client_Platform]. For text evaluation various open source text mining algorithms are integrated. New algorithms can be integrated by using Plug-Ins. The graphical user interface can also be extended by third party Plug-ins. That allows to integrate in PubCurator further functionality, e.g. new visualizations of evaluation results. PubCurator is very graph centric. It understands a text as text terms and relations between them. That is a very common concept in the biomedical field, e.g. protein protein interactions, gene disease assocations or drug side effects. Building up networks over multiple texts also allows to extract more knowledge than a single text provides.
Features
- Built upon Eclipse RCP. Multi-platform (Linux and Windows, MacOS comming soon) and native GUI widgets.
- Feature rich document manager that can organize thousands of documents (with sorting, tagging, filtering, paging).
- Document import from PubMed, PubMed Central, OMIM and batch files (in JSON format). (PDF and PubGet import is comming soon.)
- Integrates the Apache UIMA framework for text analysis.
- Already integrated 3rd party ananlyzers: MetaMap (account required), MaxMatcher, ABNER.
- Custom analzyers: String (regular expression search feature), Range (sentence range term checker).
- Design the search by defining a graph.
- Automatic result form filling.
- Links to external databases (if supported by analyzer).
- Export results to comma separated file (CSV).
- more
Downloads
Download PubCurator from Sourceforge.
Screenshots
 |
 |
 |
 |
