The first real-time, plug-and-play AI entity resolution product for fraud detection.       Learn more >

Releases

Senzing App Version History

Version: 1.1.18199
Released: 19 July 2018

  1. New Icon, because Icons are cool
  2. Check for Updates:
    1. Opt-in to Automatically Check for Updates allows you to get the latest improvements to the Senzing experience
    2. NEW in Preferences: We believe in choice, you can turn off automatic updates too
  3. NEW Options: Accessible via the Gear Icon (Settings)
    1. Get Help: Submit Support Ticket
    2. About Senzing
    3. Access to Getting Started Page
    4. Access to Online Documentation
    5. Check for Updates
  1. Fixed drag-and-drop for adding one or more CSVs into the Data Sources screen
  2. Fixed copy/paste of in-app text on macOS
  3. Improved privacy messaging when enabling detailed application logs
  4. Improved status on and handling of license limits
  5. Various other bug fixes and improvements

Version: 1.1.18170
Released: 19 June 2018

  1. Initial GA version

Senzing API Version History

Version: 2.20
Released: 5 Feb 2018

  1. Improve some repeated data values being considered less important too soon
  2. Fix Audit Report could coming back with truncated records
  3. Significantly improve Possible Matches and Possibly Related results
  4. Automatically suppress 0/0/0 DATE and unknown GENDER values
  5. Substantial streamlining of MySQL integration
  6. Substantial streamlining of SQLite integration
  7. Allow for arbitrarily large JSON data in MySQL
  8. Searches will now return name only matches

Version: 2.15 (20180112)
Released: 12 Jan 2018

  1. Performance improvements
  2. Fix statistics calculation issue with SQL value out of range error on MySQL

Version: 2.14 (20180105)
Released: 5 Jan 2018

  1. Add G2Config API for adding datasources to the JSON configuration
  2. Make G2 Config JSON the primary configuration mechanism
  3. Add new GROUP ASSOCIATION feature for things such as EMPLOYER or group membership (Las Vegas Rotary)
  4. Database optimizations for MySQL and SQLite
  5. Address scoring performance improvements (2-3x)
  6. Export and Reporting APIs are much faster and the results are more rich
  7. Upgrade name scoring from v5 to v6

Version: 2.12.2 (20170925)
Released: 25 Sep 2017

  1. Fix issue with disclosed relationships

Version: 2.12.1 (20170901)
Released: 1 Sep 2017

  1. Fix issue with G2Loader python3 compatibility
  2. If G2 fails to load, add a reminder for Ubuntu/Debian about the required platform specific libraries
  3. Fix python3 issue with some JSON data and G2Command
  4. Fix issue with unicodecsv being loaded with python3

Version: 2.12 (20170828)
Released: 28 Aug 2017

  1. Completely revamped API for much more intuitive use
    1. Parameters are simplified with extra parameters removed
    2. RecordID is consistently used
  2. Made the JSON data and API better align with terms
    1. SOURCE_ID is now LOAD_ID
    2. ENTITY_ID is consistently used instead of RES_ENT_ID, RESOLVED_ID, and RELATED_ID
    3. Flattened CSV outputs RESOLVED_ENTITY_ID and RELATED_ENTITY_ID
    4. Confusion over ENTITY_KEY and RECORD_ID has been removed leaving just RECORD_ID being consistently used
  3. Complete Java/JNI interface
  4. G2Command shell for an interactive learning tool
  5. Adds Ubuntu and Debian support
  6. Now supports Python3
  7. Handle multiple values in simple fields (e.g. Phone)
  8. If G2 crashes it will automatically save a stack as g2.dump.#### with useful information for support

Version: v2 (20170727)
Released: 27 Jul 2017

  1. Automatically detect and downgrade the importance of invalid, partial, or misused IDs and SSNs
  2. Improve the best match algorithm when determining where to resolve a record when there are multiple options
  3. Support loading the G2 configuration via a JSON file making for easy promotion, history, and version control
  4. Improve memory footprint and performance while further reducing datastore access
  5. Automatically detect and downgrade the importance of invalid GENDER values
  6. Dramatically improve memory footprint for reporting
  7. Fix disclosed relations regression
  8. Improved date scoring
  9. Fix license display for Enterprise licenses

Version: v2 (20170624)
Released: 24 Jun 2017

  1. Increased performance with reduced repetitive calls to the data store
  2. Improve Full Address processing
  3. Enhance results when including name and address
  4. Improve Date format handling
  5. Increased flexibility with input data files
  6. License key support (required now for non-Eval uses)
  7. Fix memory growth in name processing

Version: v2 (20170512)
Released: 12 May 2017

  1. Speed speed speed... up to 60% faster
  2. New Quickstart Video
  3. Added DB2 support to the Developer package
  4. Many new high level, easy to consume APIs in Python to the Developer package
    1. Adding and Removing Record with Python
    2. Getting Resolved Entities with Python
    3. Searching by Entity Attributes with Python
    4. Real-time Searching with Python
    5. and more
    6. All APIs are available in a C interface also
  5. Upgrade Name and Address processing for improved quality
  6. Allow merging of records without names, in some cases
  7. Improved error handling and reporting
  8. Move data exporting and reporting into the core C engine
  9. Various bug fixes

Version: v2 (20170411)
Released: 11 Apr 2017

  1. Real-time search interface using JSON available via python
  2. Feature behavior overrides based on usage type and entity class. A business address (USAGE_TYPE of BUSINESS) can be considered exclusive where a residence is not.
  3. Added a match_score for improved sorting of results
  4. Added GEO location scoring and STB sliders for candidate list building
  5. Substantial performance improvements to G2Loader.py
  6. Improved address scoring
  7. Much of the G2Loader.py keyless entity processing moved into core libG2
  8. General quality improvements
  9. OFAC converter sample posted as well as sample Panama Papers data

Version: v2 (20170315)
Released: 15 Mar 2017

  1. Implement redo to re-evaluate past assertions made on now generic data
  2. libG2 now handles search requests (limited to XML for now)
  3. G2Report will limit results, if necessary, to prevent failures in the prototype visualization
  4. Visualization will display strongest possible matches and relationships first
  5. Support compressed input files to G2Loader
  6. Dramatically improve reloading of data without purging first
  7. Implement Name generation scoring
  8. Made workload statistics available in the libG2 C API (not yet available in the python G2Loader output)
  9. Improved single node scaling
  10. ID scoring automatically recognizes that people can have IDs in different states without it being a conflict (e.g. passports, drivers licenses)
  11. General performance, quality, and feature improvements

Version: v2 (20170221)
Released: 21 Feb 2017

August 8th, 2016 Senzing was launched with a source code license to IBM’s G2 Entity Resolution (ER) technology. Since August, we have been advancing this code base in extraordinary ways. Our new version, G2v2 is even more accurate, even easier to use, faster, and uses less storage space. Note: This is just the beginning as we have over 100 specific engineering tasks identified and prioritized on our internal project board. Our Entity Resolution 2.0 libraries for application builders are already well beyond the ER 1.0 offerings on the market today. Get a G2v2 subscription today and enjoy the journey with us as we continue to revolutionize the ER space.