| Removed eclipse settings file. |
|
More...
|
about 15 years ago
|
| Removed eclipse settings. |
|
More...
|
about 15 years ago
|
| Fixed out of bounds error in a log statement. |
|
More...
|
about 15 years ago
|
| Added a space between the command name and the first argument. |
|
More...
|
about 15 years ago
|
| Updated pom to depend on 1.5.1-SNAPSHOT version. |
|
More...
|
about 15 years ago
|
| [ opennlp-Bugs-3101562 ] uima name finder trainer needs cutoff and iterations params |
|
More...
|
about 15 years ago
|
| [ opennlp-Bugs-3101562 ] Added a new method to retrieve an optional int parameter, if the param is not defined a provided default value will be returned instead. Additionally types in the javadoc have been fixed. |
|
More...
|
about 15 years ago
|
| Added a util method to support the merging of multiple streams. |
|
More...
|
about 15 years ago
|
| Removed a comment which says to add support for the prev map. The prev map is handled by the corresponding feature generator. And the clearAdaptiveData method is called correctly. So there is nothing wrong with the handling of the prev map, or other adaptive feature generators. |
|
More...
|
about 15 years ago
|
| [3101555] Only use the default feature generation |
|
More...
|
about 15 years ago
|
| [3101555] Only use the default feature generation and added an option to supplement the training data provided trough CASes with a file which contains training data in the native opennlp format. |
|
More...
|
about 15 years ago
|
| [ opennlp-Feature Requests-3081785 ] CoNLL 03 English Support Incorporated the logic for the German corpus data. I don't have my hands on the data to validate this myself. Marked as a todo item in the code! |
|
More...
|
about 15 years ago
|
| [ opennlp-Feature Requests-3081785 ] CoNLL 03 English Support Fixed the -types parsing to allow multiple types. Someone may want to train for multiple types in the future. |
|
More...
|
about 15 years ago
|
| [ opennlp-Feature Requests-3081785 ] CoNLL 03 English Support Fixed the todo item on multiple I- values after each other without a breaking O. This shouldn't really happen in English; however, it may happen with mutiple name types being parsed. |
|
More...
|
about 15 years ago
|
| [ opennlp-Feature Requests-3081785 ] CoNLL 03 English Support Added German to the factory, we should really support this soon. |
|
More...
|
about 15 years ago
|
| [ opennlp-Feature Requests-3081785 ] CoNLL 03 English Support Fixed a bug with missing support for the 'B-' prefix. Added a TODO item to allow multiple types training, or in this case parsing. |
|
More...
|
about 15 years ago
|
| Added the CoNLL 03 converter for the english data set for the Reuters data. |
|
More...
|
about 15 years ago
|
| Added test to also serialize a POS Tagger Perceptron model. |
|
More...
|
about 15 years ago
|
| Bumped version for build in CVS to 1.5.1-SNAPSHOT. |
|
More...
|
over 15 years ago
|
| [ opennlp-Patches-3074239 ] Add NER for Portuguese |
|
More...
|
over 15 years ago
|
| [ opennlp-Patches-3074239 ] Add NER for Portuguese |
|
More...
|
over 15 years ago
|
| Now also generates a javadoc jar file. |
|
More...
|
over 15 years ago
|
| Now also generates a source jar file. |
|
More...
|
over 15 years ago
|
| Now depends on opennlp tools 1.5.0 (without SNAPSHOT) |
|
More...
|
over 15 years ago
|
| Incremented version to 1.5.1-SNAPSHOT |
|
More...
|
over 15 years ago
|
| Added internal usage only comment. |
|
More...
|
over 15 years ago
|
| Initial check in |
|
More...
|
over 15 years ago
|
| Added internal usage only comment. |
|
More...
|
over 15 years ago
|
| Added internal usage only comment to all factories. |
|
More...
|
over 15 years ago
|
| Added link for 1.5 models. |
|
More...
|
over 15 years ago
|
