version history attached----- Original Message ----- From: "Ilian Kitchukov" <[EMAIL PROTECTED]>
To: "[KIM-discussion]" <[email protected]> Sent: Tuesday, August 01, 2006 14:02 PM Subject: [KIM-discussion] kim 1.6.7.20
Title: KIM Platform Version Historyhttp://ontodemo.sirma.bg/KIM/download
KIM Platform Version History
[version 1.6.7.20]
- Changed KIMAnnotation and KIMAnnotationSet interfaces,
annotations are now allowed to be added and removed to/from the KIMDocument annotation sets. - Fixed custom document-level features storage and indexing in Lucene.
- Fixed showing of entities in a document in the DocumentDetail screen of KIM WebUI..
- Changes to KIMDocument and KIMLimitedDocument:
- extended KIMLimitedAnnotation features
- added respectively toKIMLimitedDocument() and toKIMDocument() conversion methods
- added toXML() methods to both KIMDocument and KIMLimitedDocument - Exposed getMaxDocID() in the IndexAPI.
- Changed gate.creole.morph.Morph with com.ontotext.gate.morph.Morph in the IE.gapp.
- Added new important sections to the sys-doc:
- Extending the KIM Platform
- kim-cluster configuration - Renamed ./kim-cluster/bin/swan.bat(*.sh) scripts to cluster.bat(*.sh).
- Introduces a new PROTON specific EntityAPI, javadoc updated.
- Added and configured KIM_CONTEXT/kb/ignorelist.nt for use with KIM Gazetteer.
Aliases to be ignored during the IE process should be placed there (see the example in the file). - Fixed client application to KIM Server dependency.
- owlim 2.8.3 + trree 2.8.3 (Triple Reasoning and Rule Entailment Engine) integrated.
- Updated multiple grammars.
- Fixes in InstanceGenerator and MasterInstanceGenerator.
- System Documentation updated - configuration, examples, API javadoc.
- GATE 3.1 integrated, support for doc/pdf files is now available.
- Added a new hasAdjective property in KIM Lexical Ontology.
- New extraction features:
- Key Phrases and Key Entities extraction
- Relation Extraction for People's Positions (Person hasPosition Position withinOrganization Organization)
- Relation Extraction for activity of Organizations in a location (Organization locatedIn Location) - Updated multiple grammars.
- Added CoreDb connection timeout check for Oracle implementation.
- Added an option to show all entities for a document shown in the WebUI.
- KIM version is now displayed on the WebUI home page.
- Some parameters for configuring WebUI added to TOMCAT/webapps/kwebui.config :
<kim-rmi-host> and <kim-rmi-port>
for connection to a KIM Server (on a remote machine)
<show-document-result-urls>, <show-document-result-original>, <show-document-result-new-window>
these parameters allow the Documents Result screen to be configured to change its show behaviour - Fixed kim-cluster regarding context folder redesign.
- Fixed kim-cluster starting of components with parameters containing spaces.
- Added support for CoreDb to kim-cluster.
- Renamed NercAPI to SemanticAnnotationAPI.
- CoreDbAPI is officially released now.
- CoreDb table structure refactored
- CoreDb API refactored, introduces CoreDbQuery
- Added separate TimelinesAPI, introduces TimelineQuery
- Separated client and server classes
- Automatic entity ranking
- Solved the 32k CLOB limit issue for Oracle
- Introduces KIM CoreDb Cleaner Tool for cleaning and initalizing of the CoreDb repository - Introduces a context folder :
- Context folder holds the former kb and resources folders that specify an IE context
- Switching a context folder is as easy as changing KIM_CONTEXT parameter in bin/config_kim.bat(*.sh) file
- The default context folder is KIM_HOME/context/default
- Any new context folder must reside in KIM_HOME/context
- Each context folder may contain a file context.properties that can override any parameter from the common configs
- The KIM_CONTEXT/populated subfolder :
= holds former folders cache and index-and-store
= holds rkb.nt for newly recognized entities
= all recognized entities, documents and cache for the context can be easily removed at once - Added support for multiple GATE applications, runtime switching of the IE modules is now possible.
All the applications can now be configured for a single or all contexts using the parameter
com.ontotext.kim.KIMConstants.IE_APP ( config/nerc.properties or CONTEXT/context.properties ) - Added support for KeyPhrases:
- GeneralTerms included in context/default/kb/visibility.nt
- Document detailed view in WebUI now displays KeyPhrases and KeyEntities for a document - Adding new entities with aliases is now entirely centralized in InstanceGenerator.
- KIM WebUI changes:
- Updated the TOMCAT/webapps/kwebui.config regarding context folder redesign
- Updated styles and menus
- Fixed WebUI to limit the results for documents and entities using
INDEX_MAX_RESULT_COUNT and SESAME_MAX_RESULT_COUNT
- Fixed the comboboxes on the EntityLookup page to remember their values
- Integrated CoreSearch in WebUI (requires a properly configured CoreDb)
- Integrated Timelines for entities/docs in WebUI (requires a properly configured CoreDb) - Applied various fixes in bat/sh scripts - Unix/PC LF, quotes, paths, classpaths, java options.
- Updated Linux shortcuts.
- Updated kim-cluster libs.
[version 1.5.11.4]
- Owlim version 2.8.0 breaks into two libraries owlim 2.8.0 + irre 2.8.0, thus separating the Inductive Reasoning Rule Engine as a standalone module.
- Updated config/sesame.inmem.conf to match Owlim changes.
- Added a new parameter in config/persist.properties:
DB_STORE_LIMITED_DOC
that is used to force the PersistAPI to store/load limited (smaller) documents. Please, read the HowTo... section of the system documentation FAQ for further details and instructions. - Updated sys-doc APIs and examples that concern limited document usage.
- KIM WebUI changed to support both types of documents KIMDocument / KIMLimitedDocument.
- Owlim version 2.8.0 integrated.
- Added a new property that indicates if Sesame is used remotely or in a built-in fashion.
Added a coresponding parameter in config/semantic.repository.properties:
SESAME_IS_REMOTE
See the Getting Started system documentation section for details on setting it and starting KIM with Sesame in remote ot built-in mode. - Updated KIM Plugin v.1.6 bundled with the install.
- A new separate component EntitiesCache implemented for:
- holding Trusted and Recognized entities
- speeding up entity generation and lookup
- caching entities on disk for faster loading - Major changes in the automatic named entity recognition:
- new aliases can be added to already existing entities
- entity class can be changed from more general to more specific
- multiple bug fixes - Fixed generating escaped entity URIs. Fixes applied also for :
- WebUI
- KimPlugin communicator - Document Store Population tool improvements:
- two status modes Full/CurrentFile now available
- faster and correct parsing for document features in the XML files - Lucene fixes for generating doc IDs and batch storing of documents.
- System documentation and javadoc updated to match all the changes, new examples that concern these changes are produced.
[version 1.4.8.12]
- Sesame version 1.2.1 integrated.
- Sesame WebUI added.
- Entity cache improvements.
- Grammar fixes for dates and time.
- Refactored KIM classes so that client and server sides are separated.
There are two archives available for each side now that come within the installation:
- server-side classes in KIM.jar
- client-side classes in KIM-api.jar - Example project and sys-doc examples section updated to match changes.
- Added a new CoreDbAPI but it will be introduced and documented in some subsequent release.
- also added config for CoreDbAPI
- also added a population tool for CoreDbAPI
- CoreDbAPI is by default disabled (not used in this version) - Javadoc updated to match all changes.
- Fixed the bug that prevented from using KIM within a GATE processing resource.
- Escaping entity URI fixes for :
- WebUI
- KimPlugin communicator - Some fixes in the QueryAPI about getting lucene queries and documents.
[version 1.3.5.25]
- System documentation updated to match all the following changes, a new example that concerns the new functionality in the QueryAPI is produced.
- SemanticRepositoryAPI updated to use limit clause for Sesame queries.
- Fixed a bug in SeRQLQueryConstructor when constructing queries that have label restrictions.
- Enhanced SemanticQuery and QueryAPI to support mixed {semantic + keyword} search (a single default and comparably slow execution plan is implemented for the moment).
- Added methods to SemanticQuery (details in sys-doc):
- for sorting
- for grouping
- for binding variables to document fields
- for setting keyword restrictions
- toString() method changed to print in a more convenient way - Added methods to QueryAPI (details in sys-doc):
- for getting documents matching entities from semantic/composite queries or from EntityQueryResult, with optional limitation of the results and an option to set an execution mode (a single default mode in this release)
- for changing the default value for the parameter from config/semantic.repository.properties:
com.ontotext.kim.KIMConstants.SESAME_MAX_RESULT_COUNT
- all getLuceneQuery(*) methods have changed to return a LuceneQuery object instead of String
- automatical internal usage of the default Index/IndexAndPersist API for combined {semantic + keyword} searches - Added sorting and grouping to EntityQueryResult, methods for browsing the groupped result.
- Added QueryResultFlags to EntityQueryResult and DocumentQueryResult to mark what limitations were applied during getting these results.
- Exposed a parameter in config/install.properties that determines the way the semantic repository keeps entity descriptions (by Aliases or by Labels):
com.ontotext.kim.KIMConstants.ENTITY_DESCR - Namespaces in the ontology, KB and the JAPE grammars updated. Resources updated.
- Added a label part in the URIs for the recognized entities.
- Fixed getting of IndexAndPersistAPI, getting the default one allowed by using null parameter.
- Optimized IndexAndPersistAPI to execute sequental read/write/search operations without opening an input/output stream for each of them.
- Added methods to IndexAPI and IndexAndStoreAPI (details in sys-doc):
- for getting limited document results
- for getting only the count of docs for a certain Lucene query
- for checking the validity of a Lucene query
- for changing the defaults for the parameters in config/indexing.properties:
INDEX_MAX_RESULT_COUNT
INDEX_MAX_BOOLEAN_CLAUSES
INDEX_MAX_SEQUENTIAL_OPERATIONS
- deleteDocument() method now returns the count of affected docs
- GATE KIMClient PR extended to allow storage and indexing of the docs annotated.
- Added cache version check - prevents using cached data with different KB or KIM Server versions.
- Console logging updated, also showing correctly current KB version and KIM Server version.
[version 1.2.14.24]
- Added new Proton Knowledge Management (PROTONKM = ex-SKULO) layer in the ontology.
- Implemented cache files for KIMGazeteer and InstanceGenerator, initial load boosted.
- Lucene version 1.4.3 integrated.
- IndexAndPersistAPI - load/store document synchronization issues fixed.
- IndexAndPersistAPI - generation/setting/getting of documents' persistance IDs fixed.
- Local name vs full URI map management fix.
- Fixed shell scripts for Linux after the GATE 3 integration.
- Fixed generating of config/install.properties file during install.
- Configuration of logging updated.
- GATE version 3.0 integrated.
- GATE plugins added.
- Resources update.
- Lucene version 1.3 Final integrated.
- Sesame configuration fix.
- WebUI fixes for OWL version.
- Transition from RDFS to OWL Lite.
- Alignment with the latest official release of SESAME and its OWL Lite extension provided by Ontotext.
- Transition from KIMO (the KIM Ontology) to PROTON (find more in the system documentation).
- support for ontological modules - providing for human readabiliy and easy extension.
- Semantic Repository API changed: model changed to the new SESAME model;
methods: changed names, some methods removed (e.g. getStatements - SeRQL queries need to be executed instead);
returned iterators are simplified (StatementIterators are no longer around - now CloseableIterators that iterate over Resources or Values/Literals). - Semantic Repository configuration changed (refer to sys-doc).
- Examples in sys-doc refactored according the changes.
[version 1.1.12.13]
- KIM-cluster fixed dependency from JDK version.
- KIM-cluster WebConsole enabled.
- KIM-cluster running Sesame and MIG on separate machines fix for Linux.
- TouchGraph applet port fix.
- EditQuery and NewQuery buttons fix in WebUI.
- WebUI storage statistics fix - counting all documents.
- KIM Document Store Population tool URL document feature fix.
- DeleteDoc() and sync() method fixes for Lucene store implementation.
- Create and Manage Documents Example from sys-doc fixed for correctly deleting documents.
- Fixed commiting statements to the semantic repository to be syncronized.
- Documentation updated for v.2.0 examples, FAQ updated.
- WebUI browse ontology, document search and storage statistics fixes.
- Installation creating and cleaning folders fixes.
- KIM Server default configuration update.
- KIM-cluster RMI fixes for Linux.
- WebUI dynamic fixes.
- KB dynamic fixes.
- IE resources fixes.
- KIM-cluster included.
- Date recognition fixed in IE.
- New axiom added to Sesame.
- New Ontology API added.
- Examples in sys-doc refactored, v.2.0 released.
NOTE: system documentation is not updated to match new examples.
- Browse Ontology script fixed for most browsers, initial state set to 3 levels open.
- Dump file script.inc that caches ontotlogy tree moved to TOMCAT/webapps/KIM/screen.
- KIM Document Store Population tool now capable of setting documents' features (see documentation).
- Added install check for Tomcat folder with inavalid characters or spaces.
- Critical install checks now prompt to quit installation, instructions come along.
- Install version showed during install.
- This version history file included into install (see doc folder).
- Startup desktop icons and shortcuts created for Linux.
- KIM_OPTS variable added to all bat/sh files and set to -Dkim.home.dir=KIM_HOME.
- Tomcat catalina bat/sh modified to include KIM_OPTS.
- Log4J debug mode disabled for KIM log files.
- KIM install checks and warns if not root user under Linux.
- KIM install checks and removes previously cached Tomcat files (if any).
- Fixed URLs in config/indexing.properties and config/persist.properties files.
- Fixed kwebui.config file, removed from KIM.war and placed in TOMCAT/webapps folder.
- Added ENTITY_DESCR parameter in config/install.properties that determines the way the semantic repository keeps entity descriptions.
- Added IE_APP parameter in config/nerc.properties that points to an external IE application, default is KimNerc.gapp.
- Fixed the difference loading config/creole.xml under Windows and Linux.
- Fixed keyword search in WEB UI, interface additions in WEB UI.
- KIM Document Store Population tool added (with GUI), replacing previous datastore population tools (see bin folder).
- System documentation updated for Linux, new populate tool documentation, FAQ section formed.
- Windows, Linux, Mac OS X, Web Installer versions available.
- Renamed startup folder to bin.
- Linux (Mac OS X) sh scripts debugged and fixed.
- Tomcat startup fix by passing -Dkim.home.dir=KIM_HOME to catalina bat/sh.
- Install application design changed, Choose Shortcut Folder pannel added.
- Shortcut folders rearranged, icons updated.
- Folder startup created, bat/sh scripts moved there.
- Tomcat configuration changed, added corresponding bat/sh files in startup.
- Folder docs changed to doc, obsolete docs removed, system documentation added.
- Test corpus and TestPopulate tool added.
- TouchGraph applet added to WEB UI.
_______________________________________________ NOTE: Please REPLY TO ALL to ensure that your reply reaches all members of this mailing list. KIM-discussion mailing list [email protected] http://ontotext.com/mailman/listinfo/kim-discussion_ontotext.com
