What's Changed
* Removed: Top level WIP warnings in information extraction and document categorization files by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/327
* Fix: Downgrade scikit-learn dependency by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/329
* Changed: optional installation of large dependency packages by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/287
* Changed: Documentation about custom AI creation, saving, and uploading by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/312
* Changed: Paragraph and Sentence Tokenizer Documentation by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/296
* Changed: Replaced document ID image in DVUI docs by mexicat in https://github.com/konfuzio-ai/konfuzio-sdk/pull/328
* Added: Information on CategoryAnnotation into the Data layer concept overview by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/330
* Changed: Obsolete encoding method and indirect dependencies' versions by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/331
* Added: Documentation on preparing the Documents for the custom AI's training by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/324
* Fix: No Category categorization Vocab representation by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/334
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.23...0.2.24
v.0.2.23
What's Changed
* Added: Use of relative position and first word features in Extraction AI by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/321
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.22...v.0.2.23
v.0.2.22
What's Changed
* Add detailed error messages to bbox validation by genego-io in https://github.com/konfuzio-ai/konfuzio-sdk/pull/304
* Fixed: spacy version suitable for installing the SDK in the Colab environment by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/317
* Fix: Delete boilerplate test Document by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/322
* Added: Paragraph Extraction AI code example by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/323
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.21...v.0.2.22
v.0.2.21
What's Changed
* Fixed: Retain category annotations in subdocuments by mexicat in https://github.com/konfuzio-ai/konfuzio-sdk/pull/316
* Fixed: Retro Compatibility with older Extraction AI Models by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/319
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.20...v.0.2.21
v.0.2.20
What's Changed
* Tutorial how to retrieve Bounding Boxes on Word Level of a Page by atraining in https://github.com/konfuzio-ai/konfuzio-sdk/pull/294
* Fixed: Fetching all Categories for all Label Sets in the Project by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/308
* Added: A method to get a Page by ID from the Document, a method to get a Page's original from its copy by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/285
* Fix: Tutorial title and explanation by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/311
* Fixed: File Splitting Evaluation's ZeroDivisionError handling by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/313
* Changed: Code examples in the Documentation contain relative references to the .py scripts by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/309
* Changed: BERT and image processing architecture used for Multimodal File Splitting model; lz4 compression for pickle files by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/314
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.19...v.0.2.20
v.0.2.19
What's Changed
* Added: deepcopy Page.category by mexicat in https://github.com/konfuzio-ai/konfuzio-sdk/pull/293
* Fix: get_segmentation call from Virtual Document by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/292
* Changed: Use images instead of image_paths by mexicat in https://github.com/konfuzio-ai/konfuzio-sdk/pull/298
* Fix: Normalize to float edge case by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/297
* Changed: Additional data and `reduce_weight` in saved categorization pickle by mexicat in https://github.com/konfuzio-ai/konfuzio-sdk/pull/300
* Fixed: Categorization compatibility issue with Server's torch version by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/299
* Fixed: spacy's version conflict with Python 3.10 & 3.11 by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/305
* Changed: Default input parameters for the Categorization AI's fitting by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/306
* Fixed: Categorization AI predicting Document as having a NO_CATEGORY by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/307
* Fix: ValueError Exception with non-strict Evaluation and Empty Extracted Document by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/303
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.18...v.0.2.19
v.0.2.18
What's Changed
* Added: `Document.view_annotations` shows only top annotations for Labels with `has_multiple_top_candidates=False` by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/272
* Fixed: Automatic tests of SDK code examples in the Documentation by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/275
* Added: Information Required to Use AIs in a Konfuzio Server instance by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/277
* Rename Default Evaluation to Non-Strict Evaluation by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/283
* Fix: Deepcopied Document Pages retain image_bytes by mexicat in https://github.com/konfuzio-ai/konfuzio-sdk/pull/291
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.17...v.0.2.18
v.0.2.17
What's Changed
* Added: Paragraph and Sentence Tokenizer by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/209
* Fixed: Default Evaluation filters out the case where a Label should only appear once per AnnotationSet by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/273
* Changed: `Document.from_file` method to upload files to a Konfuzio instance by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/274
* Changed: Pages of sub-Documents after file splitting inherit Categories of Pages in the original Document by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/279
* Changed: Trainer class renamed to AbstractExtractionAI class by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/268
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.16...v.0.2.17
v.0.2.16
What's Changed
* Changed: pandas version requirement updated to `pandas>=1.3.5,<2.0.0` by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/267
* Fixed: Cross-referencing of headlines as SDK documentation URLS by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/266
* Added: `Page.image_bytes` to load a Page image from bytes by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/270
* Fixed: Image preprocessing in Multimodal File Splitting Model by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/271
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.15...v.0.2.16
v.0.2.15
What's Changed
* Added: Train custom Page-based Categorization AI by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/136
* Removed: Catchphrase and substring features by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/253
* Changed: Extraction AI can extract Documents from a copied Category by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/256
* Fix: Documents downloaded again when already existing by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/251
* Fix: Duplicate Annotation error when extracting Document by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/260
* Add top attribute to Bbox by zypriafl in https://github.com/konfuzio-ai/konfuzio-sdk/pull/258
* Added: A Label method to detect outliers in the Annotations by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/248
* Fixed: SDK documentation broken URLs, images, and other typos by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/254
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.14...v.0.2.15
v.0.2.14
What's Changed
* Changed: Number of units in Multimodal File Splitting Model's layers is scalable via the `scale` argument by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/241
* Changed: Allow File Splitting AI to be trained on Projects with where not all Categories have training data by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/244
* Added: Decimal rounding for Bounding Box coordinate validation by genego-io in https://github.com/konfuzio-ai/konfuzio-sdk/pull/247
* Fix: Error during LabelSet extraction when no Span above detection threshold by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/245
* Fix: Calculate AnnotationSet with all Annotations above threshold; allow to sort AnnotationSets by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/246
* Fixed: A minor bug that caused extracted Documents to have two different default AnnotationSets by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/249
* Fix: Restore `n_nearest_across_lines` Extraction AI option during feature calculation by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/237
* Fix: Various Documentation issues by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/231
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.13...v.0.2.14
v.0.2.13
What's Changed
* Added: Multimodal File Splitting Model for splitting multi-Document files by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/139
* Fix: Handling of overlapping spans during horizontal merging by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/230
* Fixed: A bug when validating the interface of a loaded AI model by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/228
* Added: Option to initialize a Page without specifying offsets by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/233
* Fixed: Runtime of Multimodal File Splitting Model testcases which was longer than necessary by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/232
* Changed: feature_function retokenize setting takes on a smart default value by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/235
* Fixed: A minor bug in `Document.category`, and renamed "Fallback Categorization" to "Name-based Categorization" by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/236
* Added: GitHub tests now support Python 3.11 by zypriafl in https://github.com/konfuzio-ai/konfuzio-sdk/pull/225
* Changed: PyTorch version requirements to be compatible with Python 3.8 through 3.11 by zypriafl in https://github.com/konfuzio-ai/konfuzio-sdk/pull/243
* Added: GitHub workflows to automatically test the Code Examples in the SDK documentation by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/229
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.12...v.0.2.13
v.0.2.12
What's Changed
* Fix: Annotation bboxes is calculated for each Span by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/216
* Added: Extended Exception information to data validation errors by genego-io in https://github.com/konfuzio-ai/konfuzio-sdk/pull/218
* Changed: Categorization can be evaluated and calculated for each Page of a Document by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/163
* Added: normalize DD.M.YYYY into YYYY-MM-DD date format by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/217
* Added: GitHub workflow now runs testcases using Python 3.8, 3.9 and 3.10 by zypriafl in https://github.com/konfuzio-ai/konfuzio-sdk/pull/213
* Bug fix: prediction of FileSplittingModel when some Categories don't have exclusive first-page strings by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/211
* Changed: A virtual Document with NO_CATEGORY can now have Annotations with NO_LABEL by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/203
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.11...v.0.2.12
v.0.2.11
What's Changed
* Added: Monitoring of RAM usage by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/176
* Added: Methods to create and delete Documents on Konfuzio Server, and a method to patch their metadata by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/205
* Added: Rule based logic for multi-file Document splitting by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/150
* Edited: a diagram for SDK and Server integration by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/145
* Added: Option to disable data validations by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/195
* Changed: Code snippets about authorization with API v3 by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/210
* Added: FileSplittingAI automatically-generated documentation by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/212
* Fix: Vertical merge sometimes not working for loaded RFExtractionAI models by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/204
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.10...v.0.2.11
v.0.2.10
What's Changed
* Fix: Reload Project documents from memory instead of disk after RFExtractionAI model save by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/198
* Changed: Extraction AI flag for the `use_separate_labels` option now defaults to True by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/194
* Add: Konfuzio and Python version metadata to Extraction AI pickle files by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/200
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.9...v.0.2.10
v.0.2.9
What's Changed
* Fixed: Missing raise exception when initializing an Annotation using legacy multiline support by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/175
* Fixed: "cannot pickle '_thread.RLock' object" error when saving an ExtractionAI with include_konfuzio by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/197
* Fixed: A minor issue with regex group name removal by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/192
* Fixed: Validate if an RFExtractionAI can be used for extraction rather than crashing by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/193
New Contributors
* vebarina made their first contribution in https://github.com/konfuzio-ai/konfuzio-sdk/pull/186
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.8...v.0.2.9
v.0.2.8
What's Changed
* Changed: GitHub actions use Ubuntu 22.04 by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/178
* Added: SDK Quickstart page by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/161
* Added: Option to set limit to AI model RAM usage before saving and loading. by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/180
* Added: Options to reduce RAM usage and file size of Extraction AIs by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/184
* Fixed: Bug in Automated Regex Generator that was producing duplicated capture group names by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/187
* Changed: Documentation for setting up the SDK with PyCharm updated to reflect SDK versions >=0.2 by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/171
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.7...v.0.2.8
v.0.2.7
What's Changed
* Fixed: Typo in documentation for Server/DVUI internal release process by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/165
* Added: Regex testcase for IBAN support by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/146
* Added: Improved postprocessing for extracting one line Annotations by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/155
* Changed: Intro code examples updated to reflect the latest version of the SDK by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/157
* Changed: Dependency for the `regex` package updated to use the latest version by zypriafl in https://github.com/konfuzio-ai/konfuzio-sdk/pull/169
* Fixed: Two minor bugs that were wrongly filtering out Annotations when iterating through `AnnotationSet.annotations` by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/172
* Added: Improved postprocessing for extracting multi line Annotations by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/148
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.6...v.0.2.7
v.0.2.6
What's Changed
* Changed: AnnotationSet method to list Annotations can now include the unrevised ones and provides a confidence filter by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/160
* Changed: The logger error message about Characters without Bbox now shows the ID of the Document by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/159
* Added: Normalization rule for negative numbers expressed with asterisk prefix before round brackets by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/152
* Added: Normalization rule for removing asterisk at the end of a numerical string by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/154
* Added: Documentation about the Konfuzio Release Process by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/153
* Changed: Log File creation can be disabled via an environment variable by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/156
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/v.0.2.5...v.0.2.6
v.0.2.5
What's Changed
* Added: Calculator class for precision, recall, and F1 by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/112
* Added: Update to Server documentation by genego-io in https://github.com/konfuzio-ai/konfuzio-sdk/pull/115
* Fix: Reduce the runtime of ListTokenizers by remove ListTokenizer duplicates by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/114
* Add: Inside–outside–beginning (tagging) when Annotations in one Document overlap by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/111
* Add: Filter for Spans in a Document by Label, region of Text and the Feedback provided to the related Annotation (use_correct) by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/113
* Changed: Allow setting the strict parameter when initializing Evaluation by mexicat in https://github.com/konfuzio-ai/konfuzio-sdk/pull/118
* Add detailed documentation about OCR process by genego-io in https://github.com/konfuzio-ai/konfuzio-sdk/pull/121
* Added: check negative coordinates for bboxes by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/120
* Added: Notebook for extracting the "total amount" label in payslip documents by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/119
* Cleanup API v3 docs; add bbox documentation by mexicat in https://github.com/konfuzio-ai/konfuzio-sdk/pull/117
* Added: Image operations on Konfuzio data when using OpenCV as a visual segmentation option by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/107
* Changed: Checks that disallow zero width/height bboxes are now optional by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/122
* Added: Integration between Extraction AIs and Konfuzio Documents by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/109
* Added: Methods to retrieve and delete online annotations by iftwigs in https://github.com/konfuzio-ai/konfuzio-sdk/pull/133
* Added: Categorization fallback logic to categorize a document when a trained model is not available by da-konfuzio in https://github.com/konfuzio-ai/konfuzio-sdk/pull/125
* Added: Loading of AI Model Function by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/131
* Added: Tokenizer Information to Spans by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/135
* Added: Evaluation method for all AI training stages by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/138
* Added: Unification of Random Forest Extraction AI Classes by samuel500 in https://github.com/konfuzio-ai/konfuzio-sdk/pull/124
New Contributors
* genego-io made their first contribution in https://github.com/konfuzio-ai/konfuzio-sdk/pull/115
* iftwigs made their first contribution in https://github.com/konfuzio-ai/konfuzio-sdk/pull/119
**Full Changelog**: https://github.com/konfuzio-ai/konfuzio-sdk/compare/0.2.4...v.0.2.5