Brainspace - 6.0-6.4 Release Notes
Release Notes
Version 6.4 (Release Date: 12/23/2020)
Version 6.3.3 (Release Date: 9/1/2020)
Version 6.3.2 GA (Release Date: 8/21/2020)
Version 6.3.1 GA (Release Date: 7/17/2020)
Version 6.3.0 GA (Release Date: 6/8/2020)
Version 6.2.14 GA Patch (Release Date: 9/17/2020)
Version 6.2.13 GA Patch (Release Date: 6/29/2020)
Version 6.2.12 GA Patch (Release Date: 6/8/2020)
Version 6.2.11 GA (Release Date: 1/31/2020)
Version 6.2.10 GA (Release Date: 11/25/2019)
Version 6.2.9 (Unreleased)
Version 6.2.8 Hot Fix (Release Date: 10/29/2019)
Version 6.2.8 GA (Release Date: 9/26/2019)
Version 6.2.7 GA (Release Date: 8/24/2019)
Version 6.2.6 GA (Release Date: 7/30/2019)
Version 6.2.5 GA (Release Date: 7/16/2019)
Version 6.2.4 GA (Release Date: 7/02/2019)
Version 6.2.3 GA (Release Date: 6/12/2019)
Version 6.2.2 GA (Release Date: 5/24/2019)
Version 6.2.1 GA (Release Date: 4/24/2019)
Version 6.2.0 GA (Release Date: 4/18/2019)
Version 6.1.6 GA (Release Date: 3/26/2019)
Version 6.1.5 GA (Release Date: 2/11/2019)
Version 6.1.4 GA (Release Date: 11/12/2018)
Version 6.1.3 GA (Release Date: 8/29/2018)
Version 6.1.2 GA (Release Date: 8/21/2018)
Version 6.1.1 GA (Release Date: 7/5/2018)
Version 6.1.0 GA (Release Date: 5/19/2018)
Version 6.0.8 GA (Release Date: 4/6/2018)
Version 6.0.7 GA (Release Date: 2/22/2018)
Version 6.0.6 GA (Release Date: 1/15/2018)
Version 6.0.5 GA (Release Date: 1/9/2018)
Version 6.0.4 GA (Release Date: 11/9/2017)
Version 6.0.3 GA (Release Date: 10/20/2017)
Version 6.0.2 GA (Release Date: 10/19/2017)
Version 6.0.1 GA (Release Date: 10/9/2017)
Version 6.0.0 GA (Release Date: 9/29/2017)
=====================================
Version 6.4 (Release Date: 12/23/2020)
Type | Category | Item | Description |
New | Connector | Nuix | Support for 8.4, 8.6, and 8.8 added. |
New | Analytics | Emoji’s | Emoji’s are now consumed by analytics as language, and are searchable. |
New | Search | Keyword Search | Added ability to search for documents by entity type. |
New | CMML | Probability Score Distributions | Added visualization to show distribution probability scores across a dataset over time. |
Change | Analytics | Logistic Regression | Replaced Vopal Wabbits SGD algorithm with Brainspace's own implementation of Bayesian Binary Regression to make portable models more readable, and more effectively consume content for training. |
Change | CMML | Classifier | Added indicator to classifier card, to show if it was created with the “include metadata” option enabled. |
Change | Analytics | Entity Extraction | Improved detection of Credit Card numbers. |
Change | Analytics | Korean Language Data | Improved handling of ending characters on some Korean words. |
Fix | Analytics | Language Detection | Improved language detection when only small text segments are available |
Version 6.3.3 (Release Date: 9/1/2020)
Type | Category | Item | Description |
Fix | Connectors | Relativity® Plus | Corrected some cases where unsupported characters in field names could cause syncs with Relativity® to fail. |
Version 6.3.2 GA (Release Date: 8/21/2020)
Type | Category | Item | Description |
Fix | Ingestion | Connectors | Ingestion wizard modals are now scrollable to ensure that action buttons are reachable |
Fix | CMML | Training Rounds | Classifiers created in previous versions can now be further trained with new rounds |
Fix | Communications Analysis | View Position | The Communications graph will now return to default zoom and position when reloading with different data |
Fix | Analytics | Results Sorting | Sorting the results panel by unsupported fields will no longer cause the results UI to resize inappropriately |
Fix | Administration | Datasets | Corrected some inconsistent text indentation in the dataset Administration interface |
Fix | Archive | Archive utility | Archives now export reliably and no longer include a copy of the archive script |
Fix | Installer | Upgrade | pg_hba.conf will no longer be overwritten when upgrading from 6.1.x to 6.3.x |
Fix | Dataset | Focus | Corrected a file sync inconsistency with On-Demand Analytics servers that could cause a Focus build to fail |
Fix | Connector | Relativity® Plus | Syncing from Relativity® will no longer fail when data fields contain single quotes |
Version 6.3.1 GA (Release Date: 7/17/2020)
Type | Category | Item | Description |
Fix | Analytics | Incremental Build | Expanded Incremental build functionality to include entity extraction and field population |
Fix | Dataset Build | Duplication Detection | Duplicate detection is now consistent between full and incremental builds |
Fix | Concept Search | Concept Weights | Concept weights no longer move when mousing over them in Chrome |
Fix | PSQL Database | SSL Settings | New installs will be configured for SSL by default, upgrades will honor existing settings |
Fix | Heat Map | Tool Tip | Heat Map tool tips now correctly account for window resizing |
Fix | Administration | Installer | Spaces in the directory path are now properly handled by the installer |
Fix | Notebooks | CMML Screen | “Add to Notebook” dialog moved to the top of the results window to be more consistent with other screens |
Fix | Administration | Ram Estimate Script | dataset-ram.sh script will now properly handle datasets built in 6.3 |
Fix | Administration | Installer | Corrected description for application-only option |
Fix | Search | Saved Search | Editing the name of a saved search will no longer crash the UI |
Fix | Connector | DAT | UI text in the DAT connector will no longer overlap |
Version 6.3.0 GA (Release Date: 6/8/2020)
Type | Category | Item | Description |
New | Administration | Relativity® and Relativity® Plus Connectors | An option is provided for a Relativity® Plus connector to support a legacy Brainspace dataset that was connected to a Relativity® connector |
New | Administration | Relativity® Connectors | Added the ability to overlay Brainspace fields from multiple datasets into a single Relativity® workspace |
New | Administration | Relativity® Connectors | Added configuration option to not clear fields prior to overlay |
New | Administration | Metadata Fields | Added new metadata fields |
New | Administration | Relativity® Overlay Fields | Added new Relativity® overlay fields |
New | Administration | Logging | Downloaded logs will now identify the server or dataset for the logs |
New | Analytics | Conversations Page | Updated look and feel of the Conversations page |
New | Analytics | Document Viewer | Added the ability to pin metada-ta fields on the Metadata tab |
New | Analytics | Dashboard Heatmap | Heatmap is now disabled by default to improve processing speed; an option is available to view as needed or to change the setting to enable by default |
New | Analytics | Dashboard Timeline Chart | Timeline chart now defaults to logarithmic scale view; an option is provided to view in linear scale |
New | Analytics | Dashboard Heatmap and Timeline Chart | For datasets that have a gap in documents for period of time, the Heatmap and Timeline chart display a chevron icon to show that no documents exists for the time period |
Type | Category | Item | Description |
New | Analytics | Results Pane | Added an option to sort document cards by metadata fields |
New | Datasets Card Page | Datasets Organizer | Added ability to pin dataset cards |
New | Infrastructure | Shared Data Directories | Added requirement for data directories on Application server to be shared with Analyt-ics On-Demand servers |
New | Notebooks | CMML Classifier | Added the ability to create batch notebooks based on classifier scoring |
New | Notebooks | Notebooks Screen | Added a list view for notebooks |
New | Supervised Learning | CMML Control Set | Added ability to generate control set statistical data for CMML classifiers |
New | Supervised Learning | CMML Automode | Added the ability to create training rounds automatically |
New | Supervised Learning | CMML Connect Tag | Added the ability to connect Relativity® tags to CMML classifiers |
New | Supervised Learning | CMML Top Scores Training Round | Added the ability to create a CMML training round using documents with the top scores in a dataset or in a subset of documents in a dataset |
New | Tags | Tagging uses +/- buttons | Applying/Selecting tags is now accomplished using a plus (+) button and a minus (-) button to remove a tag |
Version 6.2.14 GA Patch (Release Date: 9/17/2020)
Type | Category | Item | Description |
Fix | Analytics | Build | Expanded memory access for duplicate document detection. |
Installer | RPM Version Numbers | The BasisTech RPM now has proper linux version numbers, consistent with other Brain-space RPM’s |
Version 6.2.13 GA Patch (Release Date: 6/29/2020)
Type | Category | Item | Description |
Fix | Analytics | Incremental Build | Expanded Incremental build functionality to include entity extraction and field population |
Fix | Tagging | Edit menu | Canceling an edit to a tag will no longer delete the tag |
Fix | Datasets | Filter | The overlay will no longer cause the dataset filter to stop working |
Version 6.2.12 GA Patch (Release Date: 6/8/2020)
Type | Category | Item | Description |
Fix | Security | User Admin | Group Admins are restricted to only viewing users in their groups via the Quick Search |
Version 6.2.11 GA (Release Date: 1/31/2020)
Type | Category | Item | Description |
New | Datasets | Error Reporting | Admins may now download, if present, consolidated error reports for entity extraction and other spurious errors |
New | Installer | Permissions check and cor- rection | Installer will no longer check and update permissions of all files in /data and /localdata |
Fix | Notebooks | History | Deleting document(s) from a notebook will now properly clear query history |
Fix | CMML | Text cleanup | The text in a CMML modal post increment build has been edited for clarity and correctness |
Fix | Notebooks | History | Correct query history will be displayed on new notebook when documents were surfaced using search on another notebook |
Fix | Connectors | Relativity® Plus | Multi-choice fields are now properly concatenated on sync |
Fix | Connectors | Relativity® Plus | UI will no longer crash if sync in progress and connection is removed |
Fix | Datasets | Portable Models | UI will no longer crash upon disabling group if admin is not part of default group. |
Fix | Datasets | Heat Map | UI will now properly handle Heat Maps with identical names |
Fix | Connectors | JSON | Field Mapping dialog will now properly handle fields whose content is ‘NaN’ |
Version 6.2.10 GA (Release Date: 11/25/2019)
Type | Category | Item | Description |
New | Connector | Nuix Connector | Support for Nuix 8.0 added |
Fix | Predictive Coding | Autocoding | System will improve handling of document autocoding when:
|
Fix | Connector | Relativity® Plus | System will now overlay results correctly when a dataset is extended multiple times using an incremental build, which leverages a Saved Search that returns different results |
Fix | Connector | Relativity® Plus | System will now correctly handle all end-of-document list API results, and ingest will succeed |
Fix | Connector | Native File Analysis | System will now correctly populate brs_source label for *.pst files |
Fix | Connector | Native File Analysis | System will now generate Conversation Analysis data |
Fix | Connector | Relativity® Plus | System will now properly pull tags on multimillion document datasets |
Fix | Installer | Directories | New version of BatchTools will correctly deploy if previous version was deployed using a symlink |
Fix | Advanced Search | Numeric input | Search will now require numeric data for numeric Advanced Search |
Type | Category | Item | Description |
Fix | Connector | Native File Analysis | System will now properly log failure to ingest corrupt *.pst files |
Fix | Connector | Native File Analysis | System will send all documents, even errored, to analyt-ics |
Fix | Connector | Native File Analysis | System will now default to using SMTP format email addresses for data ingested via Native connector |
Fix | Search | Cluster Wheel | Cluster Wheel will no longer shrink upon right-click |
Fix | Analytics | Incremental Build | Memory management during incremental builds of multimillion document datasets has been improved |
Fix | Connector | Relativity® Plus | System will now preview documents correctly prior to ingest with the Relativity® Plus plugin |
Fix | Analytics | Reports | CMML classifier based on a Focus will now download results constrained by the Focus and not the entire dataset |
Fix | Datasets | Datasets Screen | User interface will no longer crash if viewing the connected tag modal and the build queue refreshes |
Fix | Datasets | Datasets Screen | User interface will now properly handle the deletion of a dataset that was present in the build queue |
Fix | Datasets | Datasets Screen | User interface will now correctly display the assignment of custom settings when there are many datasets present |
Version 6.2.9 (Unreleased)
Version 6.2.8 Hot Fix (Release Date: 10/29/2019)
Type | Category | Item | Description |
Fix | Connector | Relativity® Plus | Relativity® Plus connector hot fix for v6.2.8 that handles formatting change coming from RelativityOne following their 10.3.1 upgrade on 10/26/19 |
Version 6.2.8 GA (Release Date: 9/26/2019)
Type | Category | Item | Description |
New | Connector | Native File Analysis Connec- tor | Native File Analysis now capable of OCR functionality |
New | Connector | Native File Analysis Connec- tor | Native File Analysis now sup-ports filtering based upon NIST lists |
New | Connector | Native File Analysis Connec- tor | Format of Native File Analysis specific exception reports has been enhanced |
Fix | Ingestion | API Developer | API developers will no longer encounter data import errors; fixed issue where REST ‘/archive/*’ endpoints fail to ingest data |
Fix | Connector | Relativity® Plus | Systems upgraded after a failure to properly create a Relativity connector will now start up properly |
Fix | CMML | Scheduled Sync | Systems with multiple datasets configured to sync CMML sessions at the same time will now restart tomcat properly |
Fix | Notebooks | Scheduled Sync | Systems with multiple datasets configured to sync notebooks at the same time will now restart tomcat properly |
Fix | Supervised Learning | CMML/Predictive Coding In- teraction | Deletion of a Predictive Coding session will no longer delete CMML coding session data |
Fix | Notebooks | Tag Status Visibility | User interface will now properly display more than 20 tags in the Tag Status associated with the notebook |
Fix | Search | Keyword Search | User can now properly edit Keyword search pill |
Fix | Administration | Connectors | A system admin can now properly filter upon connector name |
Type | Category | Item | Description |
Fix | Administration | Scheduled Builds | System will now properly remove icon indicating a scheduled build for a non- recurring build after the build has completed |
Fix | Installer | RPM version numbers | Brainspace Installer RPM files now have a proper Linux version number |
Fix | Ingestion | Chinese and Japanese Lan- guage Data | Performance improvements for Chinese and Japanese language parsing and entity extraction |
Version 6.2.7 GA (Release Date: 8/24/2019)
Type | Category | Item | Description |
New | Datasets | Analytics | Admin may now select the EMT Enhanced BCC Handling option for analytics to mark emails with a BCC field populated as IsUnique |
Fix | Upgrades | Backups | During upgrades system backups will now be written to the / data/backups directory |
Fix | Datasets | Incremental Builds | Incremental Builds on multi million document datasets will no longer encounter an out of memory error |
Fix | Datasets | Analytics | When conversation index is not present, rarely email threads would have emails change from not unique to unique. This will no longer occur |
Fix | Ingestion | API Developer | API developers will no longer encounter data import errors. |
Fix | Ingestion | NFA Connector | Ingestion process will now handle more diverse formats of mbox |
Fix | CMML | Portable Models | Portable Models imported from a csv file will now properly code negative documents |
Fix | Ingestion | Relativity® Plus Connector | Data stored in Relativity® in Multi choice and Multi Object fields will now be ingested properly |
Fix | Notebooks | Relativity® Plus Connector | Notebook Sync will now properly push data to Yes/No fields in Relativity® |
Fix | Ingestion | Chinese Language Data | Performance improvements for Chinese language parsing and entity extraction |
Fix | Administration | Server Communication | Resiliency of communication between Application and Analytics servers has been improved |
Version 6.2.6 GA (Release Date: 7/30/2019)
Type | Category | Item | Description |
New | Datasets | Logging | Archive previous build reports prior to rebuild or incremental build |
New | Upgrades | Logging | Logging is now more verbose during database operations |
Fix | Supervised Learning | Predictive Coding | UI will no longer timeout when user attempts to expand control set |
Fix | Supervised Learning | Predictive Coding | Ensures that user is prompted for control set expansion when new documents are added to a dataset used for PC |
Fix | Supervised Learning | CMML | UI will no longer fail if a deleted CMML classifier was in the user’s recent search history |
Fix | Supervised Learning | CMML | UI will no longer fail when user attempts to download a result set when unchecking the Include Headers, Include Family, and Include Score check boxes |
Fix | Datasets | Logging | Allow for larger size build log |
Fix | Datasets | Reports | Dataset build process will again generate document-counts.txt |
Fix | Supervised Learning | CMML | New classifiers built from a portable model will no longer fail if the first round utilized Fast Active |
Version 6.2.5 GA (Release Date: 7/16/2019)
Type | Category | Item | Description |
New | Connector | Native File Analysis | Several new hash fields introduced to facilitate review |
Fix | Build Process | Cancelling a build | System will properly stop the running process on the analytics server when an administrator cancels a build |
Fix | Ingestion | Incremental Ingest on large datasets | Incremental ingests on very large datasets will no longer result in a BatchWriter timeout failure |
Fix | Search | Advanced Search | Users may now edit the keyword search pill |
Version 6.2.4 GA (Release Date: 7/02/2019)
Type | Category | Item | Description |
Fix | Communications Analysis | Recipient Count Queries | Searches in communication analysis based upon number of recipients will no longer include loose documents that have recipient count of 0 |
Fix | Search | Recipient Count Queries | Keyword/Lucene searches in the dashboard for recipient count greater than 14 will now return the correct number of results |
Fix | CMML | CMML/Portable Model interac- tion | CMML Sessions initiated from a portable model will no longer error if second training round is diverse active |
Fix | PSQL/Database | Sequence numbers can exceed maxint (2,147,483,647) | Systems that have processed many datasets and undergone numerous application restarts will no longer overflow table sequence ID counter |
Fix | Notebooks | Inclusivity score overlap | When documents are assigned to a notebook based classifier score the start/stop ranges will no longer overlap |
Fix | Connector | Relativity® Plus | Meta data field URL will again be assigned to each document as a link back to Relativity®’s Native Document viewer |
Fix | Installer | Certificate Generation | Installer will no longer fail when there is invalid information in the file /etc/hosts |
Fix | Archive | Archive utility | Archive script will now properly check to ensure all Supervised Learning is complete prior to archive |
Fix | Archive | Archive utility | Archive script will now provide better user feedback when data is not retrieved from database |
Type | Category | Item | Description |
Fix | Connector | Native File Analysis | Error report will now correctly use nativeUri when capturing NOT_SENT_TO_ANALYTICS errors |
Fix | Connector | Native File Analysis | Error report will now correctly capture the number of 0 byte files |
Fix | Connector | Native File Analysis | Error report will now correctly capture the individual errors of a file that has more than one error |
Fix | UI | Communications Analysis | Mouse wheel events will no longer result in improper page scrolls |
Fix | Installer | Logging/Reporting | Installer will no longer generate spurious “ignored package” log |
Fix | Concept Search | Top Concepts | When adding top concepts to a search a user can no longer add the same concept multiple times if the text case is different |
Version 6.2.3 GA (Release Date: 6/12/2019)
Type | Category | Item | Description |
New | Datasets | Optional Analytics | If a dataset has been created with optional analytics disabling a feature (e.g. EMT), then Incremental builds are not available through the UI. Full Report is now available even if Optional Analytics only has Email Threading |
Fix | Datasets | Search Results | Exporting body text as separate files fails if dataset has no data source |
Fix | Installer | chown | Performance on post-upgrade operation to validate file ownership |
Fix | Installer | Upgrade | systemd unit files and init scripts are not removed on upgrade or legacy uninstall |
Fix | Installer | OS Support | Brainspace installer now supports Redhat 7.6 |
Fix | NFA | MSG File | When dealing with null bodies in MSG file, no longer throws NPE exception. |
Fix | Predictive Coding | Reports | Relevance Report and Training Report are not downloadable. |
Fix | Reports | Consistency Report | While a model is being created from training examples and used to score documents, other columns are being filled in as if there are no training documents. |
Version 6.2.2 GA (Release Date: 5/24/2019)
Type | Category | Item | Description |
New | Administration | RAM Estimate Script | Updated script and added it to the scripts directory in /var/lib/ brains |
New | Cluster Wheel | API Developer | Brainspace Cluster Wheel is available as a stand-alone product that can be licensed for use with third-party applications |
New | Lucene | Lucene Search from Storage | Separate Search index from Values index. Improves performance while reducing memory requirements for larger datasets |
Change | Brainspace Installer | Interface and Functionality Improvements | Improved user interaction for install or upgrade. Installer alerts for bad /data mount status or unsuccessful backup |
Change | Data Ingestion | Language Detection | More efficient language detection (smaller indexes, less CPU overhead) |
Change | Native Documents | Bulk-Download Optimization | If the user has not chosen the nativePath field when requesting source files on bulk-download, that field will be added to the request and will show in metadata-report.csv |
Change | Data Ingestion | Native File Analysis | Location in ingest-error-report.csv changed from 'human readable' form of native-path URI to the 'real' native-path URI NFA does not support skip-Count or skipValue. This means that pausing and resuming does not work as expected and should not be used |
Change | Supervised Learning | CMML | Improved storage efficiency for a CMML session that is run on a Focus |
Change | Brainspace Installer | RPM-Based Installer | Installer provides a feedback message when upgrade completes successfully. |
Type | Category | Item | Description |
Fix | DocumentWriter | API | DocumentWriter will now appropriately close for further writing |
Fix | Supervised Learning | Predictive Coding | Fixed bug that deleted a CMML session when deleting a PC session. |
Fix | Supervised Learning | Predictive Coding | Fixed timeout that prevented training rounds from completing successfully |
Fix | Data Ingestion | Native File Analysis | PluginWorkspace has a new method for obtaining reference to a durable storage area |
Fix | Supervised Learning | CMML | There are 2 cases where documents can have a CMML classifier score of -1:
Users can find the documents with a -1 score using advanced search range query for classi-fier_X_score -1 to -1 |
Fix | Data Ingestion | Native File Plugin | Native file analysis ingestion will no longer fail with silent errors |
Type | Category | Item | Description |
Fix | Dataset Manage- ment and Brain- space System Management | Archive and Import Datasets | Archive breaks on multiple hits for brains in /etc/password |
Fix | Dataset Manage- ment and Brain- space System Man- agement | Archive and Import Datasets | Dataset failed to archive when FQDN of the localhost was present in DBConfig file |
Fix | Connector | Relativity® Plus Plugin V2 | Relativity® Plus overlays data incorrectly |
Fix | Supervised Learning | CMML and Predictive Coding | CMML classifiers and Predictive Coding classifiers use separate auto-increment, and subsequently the same directory |
Fix | Dataset Mode | Zen Mode | Screen Freezes when we turn off Zen mode on large dataset |
Fix | Groups | Datasets and Connectors | Error when creating new group with dataset or connector |
Fix | Search | Endpoint | /search endpoint spending most of its time computing family counts |
Fix | Visualization | Heatmap | Heatmap does not change after switching from Document Volume to Anomaly Detection |
Fix | Natural Language Processing | Basis Technology | Upgrade of BasisTech |
Fix | Native File Analysis | Silent Errors | Silent errors in NFA ingestion |
Fix | Installer | RPM Save Files | Installer leaves rpmsave files |
Fix | Supervised Learning | CMML and Predictive Coding | Potential to write predictive coding (PC) and continuous multi-model learning (CMML) classifier results to the same fields, resulting in overwritten and corrupt results |
Version 6.2.1 GA (Release Date: 4/24/2019)
Type | Category | Item | Description |
Fix | Focus | Trim Function | When a search is trimmed, the full document count is sent to the Focus instead of sending the reduced document count |
Fix | Predictive Coding | Reports | Relevance Report and Training Report are not download-able |
Change | Reports | Relevancy Report | Relevance Report renamed to Cutoff Report |
Version 6.2.0 GA (Release Date: 4/18/2019)
Type | Category | Item | Description |
New | Connector | Relativity® Plus Plugin | Relativity® Plus plugin can be used with Relativity® v9.7 and newer versions. User authentication is achieved using OAuth authentication. Relativity® credentials are no longer stored in Brainspace. API-only integration no longer requires dedicated access to SQL and is independent of the Relativity® backend. |
New | Connector | Venio Plugin | Venio connector is available as a connector option when creating datasets. |
New | Dataset Manage- ment and Brain- space System Man- agement | Archive and Import Datasets | Datasets and associated work products can be archived to free-up Brain-space system resources. Archived datasets and associated work products can then be imported back into Brainspace at a later date, including datasets migrated from another Brainspace v6.2 instance. |
New | Native Documents | Bulk-Download Native Docu- ments | Native documents can be downloaded in bulk for an entire dataset, a notebook, a focus, a concept search, or any other method of sorting and organizing native documents in a dataset. |
New | Supervised Learning | Download Classifier Insights Comparison | A *.csv file that compares two different classifier training rounds can be downloaded to a user’s local machine. |
Change | Data Sources | Custom Delimiter | Custom delimiter allow users to leverage Communications and Conversations graphs for datasets that do not include emails by specifying a split-Field to override the default behavior. |
Change | Concept Search | Concept Search Drawer | Removed slider bars from Concept Search drawer. |
Type | Category | Item | Description |
Change | Brainspace Installa- tion and Upgrades | RPM-Based Installer | Simplified and automated more of the installation and environment management process from the command-line interface and Brainspace UI. |
Change | System Manage- ment | Memory Management | Brainspace automatically detects and configures available memory resources when a Batch Tools (bt) job runs. |
Change | Dataset Manage- ment | Dataset Status | Dataset status has been up-dated from Running and Stopped to Active and Inactive. |
Change | Native Documents | Relativity® Plugin | View Native Doc button has been moved from the Meta-data tab to the View Source Document button at the bottom of the Document Viewer. |
Change | User Interface | Icons | Updated icons throughout the user interface. |
Change | Infrastructure | Java | Upgrade to Java 11. |
Fix | Dataset Build | Email Threading | EMT email header parsing after blank line. |
Fix | Brainspace Addons | Relativity® Plugin | Cannot install the Brainspace addons to Relativity® V10 (relativitydevvm). |
Fix | Authentication | SAML | Fix password behavior for users/creation when SAML enabled. Do not show the password field. |
Fix | Reports | Facets | The two Facet reports are throwing 500 error when we have no results. |
Fix | Saved Search | Relativity® Plugin | Relativity® saved search not showing up in Brainspace UI. |
Fix | Conversations | Conversations | People in conv analysis not automatically enabled/dis-abled after adding/removing from graph. |
Type | Category | Item | Description |
Fix | Supervised Learning | CMML | Classifier not in Advanced Search (until log out/log in). |
Fix | Supervised Learning | Predictive Coding | Stop automatically deleting failed PC sessions. |
Fix | Concept Search | Advanced Search | Search drop-down values should be sorted by count in current search instead of by count in the dataset. |
Fix | Connector | Relativity® | Relativity® ingestion exceptions & logging |
Fix | Reports | Usage Report | Usage report endpoint is zero-based. |
Fix | Application Server | Tomcat | Runtime (potential) memory leak. |
Fix | Browser Compatibility | Edge | Missing Brainspace icon on Edge browser. |
Fix | Dataset Sort Order | Datasets Page | Dataset "Size" column sorts datasets incorrectly and flips back and forth between MB and GB in an inconsistent way. |
Fix | Connector | Relativity® | When pushing to Relativity® without Concept, we get error. |
Fix | Cluster Wheel | Visualization Rendering Time | Cluster Wheel slow to render for some data sets / portions of hierarchy. |
Fix | Connector | Relativity® | Brainspace enabled for Relativity connector with 2FA enabled. |
Fix | Dataset Build | bt increment | When an increment causes a new near duplicate cluster to be formed, original documents that existed prior to the increment are not being reindexed. This means that the brs_near_dup_set_id is not being populated on the original documents. |
Version 6.1.6 GA (Release Date: 3/26/2019)
Type | Category | Item | Description |
Change | Data Ingestion | Native File Analysis | Relative paths are now simplified. Reingesting the same data may ingest a second document with the simplified path. |
Fix | Cluster Wheel | Focusses | Switching Focuses when viewing the Cluster Wheel caused infinite loading. |
Fix | Security | Connectors | Paths in connector configuration were not validated to disallow relative paths. |
Fix | Native Document Viewer | Control | View Native document toggle switch was incorrectly styled. |
Version 6.1.5 GA (Release Date: 2/11/2019)
Type | Category | Item | Description |
New | Native Documents | Download | Added the ability to download native documents from the document viewer. |
Change | Languages | Stopwords | Improved parts of speech for Danish, Finnish, Icelandic, Norwegian, and Swedish stopwords. |
Change | Native Documents | Display | Improved error messages when native view of document is unavailable (missing file, unable to render, etc.). |
Fix | Clustering | Deduplication | Fixed an issue where certain types of data could cause clustering to fail during dedu-plication. |
Fix | Supervised Learning | Portable Models | Fixed an issue where attempting to save a portable model when no “default” group exists resulted in a white screen. |
Fix | Supervised Learning | Portable Models | Fixed an issue where only the first 20 available groups would be assignable when saving a portable model. |
Version 6.1.4 GA (Release Date: 11/12/2018)
Type | Category | Item | Description |
Change | Connector | Nuix Plugin | Improved performance, error handling, license type naming, and logging. Improved handling of issues regarding licenses and mismatched Nuix server versions. |
Change | Authentication | SAML | inactive.session.timeout property is now enforced if set in a SAML configuration. |
Fix | Memory Leak | Dataset Analytics Process | Fixed a memory leak involving redundant resources accessing underlying analytics. |
Fix | UI does not accept input. | Communications Analysis | Fixed an issue where the advanced settings sliders and the “Number of Recipients” text box would not accept input. |
Fix | Search by Notebook ID doesn’t return re- sults from /date- facets API | Advanced Search | Fixed a caching error that intermittently resulted in search errors for Notebooks when querying for date fields. |
Fix | CSRF Prevention | REST API | File download functionality has been standardized and includes additional headers to mitigate CSRF vulnerabilities. Note: Users must enable pop-ups to enable downloads. |
Fix | Notebook Creation | REST API | Notebook creation is now possible via API Key authentication. |
Fix | Document Viewer Scrolling | Document Viewer | Fixed an issue where the document viewer would scroll to the top of the document whenever the user clicked. |
Version 6.1.3 GA (Release Date: 8/29/2018)
Type | Category | Item | Description |
New | Datasets | Administration | Added ability to configure boilerplate settings via the UI |
New | Analysis | Languages | Added support for
|
New | REST API | Documentation | Expanded documentation and an SDK for common operations is now available for customers with API access |
Change | Reporting | Documents | Added inactive document statistics to the monthly report |
Change | Application Server | Security | Improved configuration and installation checks related to HTTPS on the application server |
Change | Login | Documentation | Added documentation of login message/disclaimer parameter in brainspace.defaults.properties |
Change | Dataset | Performance | Improved performance of deleting a dataset |
Change | Predictive Coding | Autocoding | All scored documents will now get a value during au-tocoding |
Change | Analysis | Email Addresses | Added support for a nonstandard (non-RFC compliant) email format that could result from errors in source data, OCR, etc. Emails in the form of “(email@example.com)” will now parse as “email@example.com ( |
Fix | Focus | Cluster Wheel | Fixed a caching issue which could result in incorrect document counts in the cluster wheel in some scenarios |
Type | Category | Item | Description |
Fix | Upgrade | Supervised Learning | Fixed an issue where Discovery 5 rounds were not properly upgraded, resulting in error messages in new sessions. |
Fix | Upgrade | Servers | Fixed an issue where servers would not be displayed properly after upgrade from Discovery 5 |
Fix | Upgrade | Database | Fixed an issue where the database could fail to properly start during an upgrade from Discovery 5 |
Fix | Installation | OS Version | Fixed an issue where the version of the OS was unable to be detected by the installer |
Fix | Installation | Tomcat Version | Fixed obscuring of Tomcat version during install |
Fix | Nuix Connector | Case Access | Fixed UI issues when a user’s credentials did not grant them access to a case in Nuix |
Fix | XML Connector | Error Handling | Fixed XML Ingest failing when encountering 0-byte files |
Fix | XML Connector | Ingestion | Fixed an issue where an incremental ingest could become unresponsive |
Fix | XML Connector | Security | Mitigated an XXE injection vulnerability in XML ingestion |
Fix | Nuix Connector | Ingest | Fixed date fields not being properly ingested via the Nuix Connector |
Fix | Relativity® Connector | Overlay | Fixed an issue where manually overlaying after an incremental build would only overlay newly ingested documents. Manual overlays will now always overlay for all documents. |
Fix | Native Connector | Ingestion | Fixed ability to pause ingestions and perform incremental ingestions |
Type | Category | Item | Description |
Fix | Upload | Security | Improved data validation of uploaded Person and Profile images |
Fix | CMML | Insights | Fixed an issue where rounds without portable models could appear to be available for model comparison |
Fix | CMML | Document List | Fixed an issue where the document list could not be reordered properly when refreshing |
Fix | REST API | API Key | Fixed ability to create a Focus via API Key authentication |
Fix | REST API | Datasets | API now returns dataset information correctly while a dataset is starting |
Fix | REST API | Datasets | Datasets created via API will no longer attempt to overlay by default |
Version 6.1.2 GA (Release Date: 8/21/2018)
Type | Category | Item | Description |
Fix | Analysis | Email Threading | Fixed display of X.400 email aliases in Conversation and Communication Analysis |
Version 6.1.1 GA (Release Date: 7/5/2018)
Type | Category | Item | Description |
New | Dashboard | CMML Context | The Dashboard can now be navigated in the context of a CMML Classifier - Document Results will be sorted by Predictive Rank and viewed in Zen Mode for rapid tagging |
New | Notebooks | CMML Context | Notebooks can now be navigated in the context of a CMML Classifier - Documents can be filtered by tagged status and viewed in Zen Mode for rapid tagging |
New | Notebooks | Batch Creation | Notebooks can now be created in batches by ranges of Predictive Rank when in a CMML Classifier Context |
New | Notebooks | Scheduled Sync | Notebook Sync Pushes and Pulls are now able to be scheduled |
New | Dashboard | Heatmap | Heatmap cells may be multi-selected via a toggle or by clicking and holding a cell |
New | Integration | Connected Tags | Connected Tags can now be used to push Brainspace tagging data into the connected data source |
New | Integration | Native Plugin | Attachment and Calendar documents are now indicated by their icon in the Search Results Panel |
New | Login | Messaging | A login message or disclaimer can now be presented when a user logs in via a configuration in brainspace.properties |
New | Datasets | Email Threading | Chinese email headers are now supported |
Change | Search | Export | Exported searches are now exported in a format that retains queries that can be copied and pasted back into the Advanced Search bar |
Type | Category | Item | Description |
Change | Search | Export | Exported searches now include CMML Context used, creation date, and user name |
Change | Installation | Dependencies | Certain font dependencies required for native viewing are now packaged in the installer, enabling non-internet connected installs |
Change | CMML | Predictive Ranking | Improved sorting of documents within ‘plateaus’ of equal Predictive Rank in the Search Results Panel |
Change | CMML | Insights | Access to Insights made more contextual based on entering from a Round or Supervised Learning |
Change | Administration | Services | Improved display of disk and memory usage |
Change | Datasets | Build Process | Schema and Boilerplate configuration now logged when building to improve ability to retrieve and troubleshoot |
Change | Integration | Native Plugin | `application/xml` added to mime type whitelist |
Change | Integration | Nuix Plugin | Nuix Plugin now requires the use of a Metadata Profile to select fields for import |
Change | Predictive Coding | Progress Indicator | Tooltip added to the PC progress bar to direct users to train documents to make progress |
Change | Login | Logout | Added a logout confirmation landing page |
Change | Tomcat | Configuration | Tomcat version information is now obscured to reduce data available to attackers |
Change | Tomcat | Version Upgrade | Tomcat version upgraded to 8.5.31 |
Type | Category | Item | Description |
Change | HAProxy | Configuration | Updated configuration to make timeout values consistent with the rest of the application |
Change | Datasets | Entity Extraction | Improved quality and speed of Entity Extraction |
Change | Datasets | Deploy | Datasets with corrupt Graph Data will now be able to be deployed without functional Communication and Conversation Analysis |
Change | CMML | Scheduling | Added further validation and restrictions to `Twice Daily` schedule option |
Change | Administration | Logging | Changed a number of log statements to log names instead of IDs where available |
Change | Dataset | Build Process | Email address parsing has been made more robust, with limited support for multiple `From` values, i.e. a sender on behalf of another sender |
Change | CMML | Insights | Improved sorting and comparing of values across models |
Fix | Integration | Relativity® Plugin | Fixed an issue where a long Connected Tag name could result in a Predictive Rank column name that was too long for Relativity® |
Fix | Security | Cross-Site Scripting | Added protection against a cross-site scripting vulnerability |
Fix | Dataset | Ingest | Improved ingest speed by narrowing the amount of data transferred by the Nuix, XML, and DAT plugins |
Fix | Notebooks | Adding Documents | Fixed `Add to Notebook` action not respecting Trim on the Search Results Panel Graph |
Type | Category | Item | Description |
Fix | Dashboard | Conversation Analysis | Fixed a crash when using Edge caused by a browser update |
Fix | Conversations | Conversations | People in conv analysis not automatically enabled/dis-abled after adding/removing from graph. |
Fix | Supervised Learning | CMML | Classifier Not in Advanced Search (until log out/log in) |
Version 6.1.0 GA (Release Date: 5/19/2018)
Type | Category | Item | Description |
New | Integration | Native Plugin | New Plugin to enable loading loose files, PSTs and more into Brainspace. Native representations of documents available in the Document Reader |
New | CMML | Portable Models | Exportable, readable, editable models can now be saved or downloaded from any CMML Round. New CMML Classifiers can be created using a Portable Model to apply previous learning to a new set of data |
New | CMML | Classifier Insights | New UI to see the effects of training on the model - see how relevant features and metadata change over time |
New | Ingest Wizard | Advanced Config | Email Threading Related and Containment values now configurable |
New | Ingest Wizard | Optional Analytics | Ability to turn off parts of analysis with corresponding build time reduction and related analytics unavailable |
New | Document Reader | Navigation | Hotkeys for navigation to previous and next doc |
New | CMML | Focus Scope | Ability to create a CMML Session based on a Focus |
New | Datasets | Language Support | Enhanced support for more languages, including Czech and Romanian |
New | Datasets | Language Support | Expanded support for Entity Extraction to more languages |
New | Datasets | Build Status | New status screen to see progress of build and time taken for each step |
New | Saved Searches | Export | Saved Searches can now be exported |
New | Search | ID Upload | Ability to upload a list of Doc IDs to search |
Type | Category | Item | Description |
New | Notebooks | ID Upload | Ability to create a notebook via a list of uploaded Doc IDs |
New | Integration | Relativity® | Support added for Relativity® 9.6 |
Change | Predictive Coding | Control Set | Can now reanalyze Control Set and Training Rounds in the event that coding of Control Set docs has changed |
Change | Security | Encryption | TLS 1.2 now enforced for all server communication |
Change | Security | Login | Autocomplete now disabled on login page (not honored by Chrome) |
Change | Ingest Wizard | Document Count | Document counting can now be skipped if desired |
Change | Datasets | Build Process | Documents previously rejected due to no analyzed text will now be included in Email Threading analysis |
Change | Datasets | Build Process | Documents previously rejected due to fields too large will now be analyzed |
Change | Ingest Wizard | Advanced Config | Max Index Size and Max Index Field Size options removed |
Fix | Reporting | Metadata Report | Fixed metadata report only including one value for each field, all values are now included |
Fix | Dashboard | Focuses | Fixed focus list being unable to display more than 40 focuses |
Fix | Search | Top Terms | Fixed a vocabulary score calculation that sometimes changed the order of term results |
Fix | Search | Communications Analysis | Fixed the recipient count not being added to the search query |
Type | Category | Item | Description |
Fix | Ingest Wizard | Build State | Fixed an issue that prevented ingestion after canceling ingestion |
Fix | REST API | API Key | Fixed endpoints that failed when authenticating via an API Key |
Version 6.0.8 GA (Release Date: 4/6/2018)
Type | Category | Item | Description |
Change | Integration | Relativity® Connector | Greater efficiency in synchronizing data during Predictive Coding, CMML and Analytics Overlay. Family lookup performance improvements |
Fix | Analytics | Speed and efficiency | Slowdowns, and stalls observed during ndf no longer occur |
Fix | Analytics | ndf | Several failures during ndf due to text or corrupt Con-versationindex will no longer occur |
Fix | Security | Session timeout | System now properly respects inactive.session.time-out setting when it is in excess of 30 minutes |
Fix | Analytics | Incremental builds | Incremental builds now supported for datasets created prior to our 6.0 Release |
Fix | Classification | User Interface | Microsoft Edge will now properly present relevancy distribution graph |
Fix | Analytics | Reports | Corrupt installations of SW will no longer result in documents being duplicated in full report |
Fix | Installer | Software verification | Self installer will perform validation of bundled software |
Version 6.0.7 GA (Release Date: 2/22/2018)
Type | Category | Item | Description |
New | Integration | Relativity® Connector | Added Encryption option for SQL Connection |
New | Integration | Overlay | Added EMT_Attachment-Count, EMT_WasUnique-Reason, EMT_WasUnique and EMT_ThreadHasMiss-ingMessage fields |
New | Administration | Reports | Dataset report now includes the size of the dataset in bytes |
New | Document Reader | Navigation | Added keyboard navigation and legend to Document Reader |
New | Thread Analysis | Navigation | Added keyboard navigation and legend to Thread Analysis |
Change | Integration | Relativity® Connector | Reduced number of columns in the EDDS database that the configured user requires access to |
Change | Integration | Relativity® Connector | Modified overlay of decimal data into Relativity® to mitigate issues in display and rounding |
Change | Integration | Relativity® Connector | Increased timeout for Overlay to mitigate network bandwidth issues |
Change | Thread Analysis | Thread Selection | Leading zeros will now be disregarded to enable copying/pasting from overlayed systems |
Change | Integration | Overlay | Improved logging during overlay to better surface issues with missing or invalid credentials |
Change | Thread Analysis | Required Fields | Thread analysis no longer requires CC and BCC fields to be mapped |
Fix | Administration | Build Logs | Fixed an issue where the build logs could become unavailable |
Fix | Datasets | Deploy | Fixed an issue where a dataset could fail to deploy if a CMML Classifier had previously been created and deleted |
Fix | Administration | Errors Tab | Fixed UI artifact resulting in checkboxes always appearing unchecked |
Fix | Administration | Reports | Fixed an issue where reports could appear unavailable in some network configurations |
Fix | Logging | Push Notifications | Fixed an issue where log messages related to push notifications were being logged to the incorrect location |
Fix | Integration | Connector API | Improved interface with connectors supporting Predictive Coding and CMML |
Fix | REST API | Datasets | Improved functionality for building datasets via the REST API |
Version 6.0.6 GA (Release Date: 1/15/2018)
Type | Category | Item | Description |
Change | Integration | Relativity® Connector | Renamed BD Collection field to BD Notebook |
Fix | Classification | Predictive Coding | Fixed an issue where Excluded documents could be included in a Notebook based training round |
Fix | Authentication | SAML | Fixed an issue where a SAML signature could not be validated if multiple signing keys were present in the federation metadata |
Fix | Datasets | Scheduled builds | Fixed an error that prevented scheduled builds from executing |
Fix | Datasets | Build Configuration | Fixed an issue where Auto-Overlay did not have the correct default value based on the configured Connector |
Fix | Datasets | Build Process | Fixed a build failure that could occur when creating the dup report |
Version 6.0.5 GA (Release Date: 1/9/2018)
Type | Category | Item | Description |
New | REST API | Document Ingestion | New endpoints for ingesting documents via the REST API |
New | Integration | XML Connector | Added support for ingesting a directory containing multiple XML files |
New | Datasets | Entity Extraction | Added support for defining custom entities via Batch Tools configuration files |
Change | Integration | Relativity® Connector | Optimized sizing of fields created in Relativity® |
Change | Classification | Predictive Coding | Standardized language for classified documents as ‘Responsive’ and ‘Not Responsive’ |
Change | Security | User Session | Users are now logged out of Brainspace when their password is changed |
Change | Security | HAProxy Configuration | Weak TLS Ciphers are no longer supported |
Change | Security | Browser Caching | API responses now instruct the browser to not cache |
Fix | Integration | Relativity® Connector | Improved field naming convention and format for synced Notebook data |
Fix | Integration | Overlay | Fixed an issue where EMT_IsDuplicate field was not being populated in the full report |
Fix | Upgrade | Users and Groups | Improved migration of user dataset permissions from Discovery 5 |
Fix | Ingest Wizard | Field Mapping | Fixed an issue where the schema would be generated incorrectly when parsing user added delimiters |
Fix | Ingest Wizard | Build Process | Fixed an issue where clustering could fail due to extremely long tokens |
Version 6.0.4 GA (Release Date: 11/9/2017)
Type | Category | Item | Description |
New | Ingest Wizard | Field Mapping | Added support for autodetection of EDRM style date formats |
New | Analysis Tools | Cluster Wheel | Added zoom action for Chrome |
Change | Security | Server Configuration | The version of the webserver will no longer be reported to reduce the information available to an attacker |
Fix | Classification | Training Rounds | Help text has been standardized across CMML sessions and rounds |
Fix | Document Reader | Text Direction | Improved display of Right-to-Left languages |
Fix | Admin | Server Configuration | Improved URL resolution for certain types of DNS configuration |
Fix | Tag Sync | Tag Pull Speed | Significantly increased speed of pulling values for connected tags |
Fix | Search | Type Ahead | Fixed an error around type ahead searching being attempted for unfaceted fields |
Fix | Security | SQL Injection | Removed a vulnerability from the REST API |
Fix | Admin | User Creation | Fixed an error during user creation occurring in systems using SAML authentication |
Fix | Security | Errors | Removed stack traces from REST API responses |
Fix | Notebooks | UX | Improved visuals around deleting notebooks |
Fix | Datasets | Near Duplicates | Fixed an issue where Near Duplicate documents were sometimes being put in multiple groups instead of one |
Version 6.0.3 GA (Release Date: 10/20/2017)
Type | Category | Item | Description |
Fix | Dashboard | Heat Map | Fixed an issue with Heat Map showing inconsistent results. |
Fix | Administration | Datasets | Fixed an issue with dataset startup if the system admin user had been removed. |
Version 6.0.2 GA (Release Date: 10/19/2017)
Type | Category | Item | Description |
New | License | Usage Report | The value of the total documents analyzed by Brainspace was added the the Usage report. |
New | Classification | Reporting | For CMML training, the round report download has been added for completed validation rounds. |
Change | Classification | Training Rounds | The number of documents that can be added to an training rounds have been limited to 5000 for Random, Influential, and Fast Active and 250 for Diverse Active. |
Fix | Installation | haproxy | Fixed an issue upgrading to Brainspace 6 when haproxy was not present. |
Fix | Ingest | fieldMap.xml | Fixed issue with build failing in an unrecoverable state relating to changes to fieldmap.xml. |
Fix | Ingest | Euro Dates | Fixed issue matching euro dates that use period as delimiters. |
Fix | Search | Query Builder | Fixed issue where type-ahead was adding a new focus pill for each character typed once a single focus was matched. |
Fix | Analysis Tools | Communications Analysis | Changing the filters on the communications graph no longer lose the selected nodes and/or edges. |
Fix | Analysis | Configuration | The <minOccurences> setting in boilerplateConfig.xml will now be used when running analysis on a dataset. |
Fix | Dashboard | Document Viewer | Boiler plated subject and title will now show correctly in Document Viewer. |
Type | Category | Item | Description |
Fix | Classification | Training | On-demand Analytics service will now accept more than one classifier training request and queue up the requests if the service is busy. |
Version 6.0.1 GA (Release Date: 10/9/2017)
Type | Category | Item | Description |
New | Security | Session Hijacking | The user session can no-longer reused on a separate machine. |
Change | Ingest Wizard | Load Existing Dataset | Updated the action to load an existing dataset from disk to read “Load Existing Build” to clarify what the action would perform. |
Fix | Installation | Deployment Center | Installing in an environment that does not have internet access will not fail due to packages missing from the install file. Required packages are now all included. |
Fix | Ingest Wizard | Load Existing Dataset | Loading an existing dataset from disk no longer gets stuck while copying the files. |
Fix | Ingest Wizard | Field Mapping | Field map now resolves duplicate mappings. |
Fix | Ingest Wizard | XML Plug-in | Brainspace 6 XML plug-in was updated to version 2.0 and enhanced with Brain-space 6 plug-in features like a file picker. |
Fix | Ingest Wizard | XML Plug-in | Large XML ingestions will not fail due to memory usage or garbage collection errors. |
Fix | Upgrade | Datasets | Orphaned builds will not fail to be upgraded, a dataset will be created for them and user will be able to deploy the dataset. |
Fix | Upgrade | Field Map Template | Field map template will now be upgraded properly. |
Fix | Upgrade | Nuix Plug-in | Nuix connector configuration will now be upgraded properly. |
Fix | Upgrade | Postgres | Postgres 9.1 will be automatically removed on upgrade to Postgres 9.6.3. |
Fix | Analysis | Communications Analysis | Fixed display issue in the settings dialog. |
Fix | Search | Saved Searches | Saved search no-longer fails with a Communication Analysis search clause. |
Version 6.0.0 GA (Release Date: 9/29/2017)
Type | Category | Item | Description |
New | Integration | Plug-in SDK | Support for Incremental Ingestion |
New | Integration | Plug-in SDK | Support for Tag import and sync |
New | Integration | Plug-in SDK | Support for Continuous Multi-Modal Learning (CMML). |
New | Integration | Plug-in SDK | New methods on interfaces (some are optional depending on feature flags you enable): Plugin: 1 new method for configuring a CMML Review, 3 new feature flags Connector: 1 new method for ensuring the correct time between Brainspace and the connected system DataSource: 2 new methods for Field Mapping, 1 new method for Tag Syncing Ingest: 2 new methods for Incremental Ingestion |
Change | Integration | Plug-in SDK | Configuration Interfaces are now specific to each type of configuration. |
Change | Integration | Plug-in SDK | Exception classes moved into their own package. |
Change | Integration | Plug-in SDK | Annotation classes simplified and renamed. |
Change | Integration | Plug-in SDK | DataSpace was renamed to DataSource. |
Change | Integration | Plug-in SDK | Updated Field Mapping functionality and moved into a new package. |
Change | Integration | Plug-in SDK | DocumentWriter adds a delete method to remove documents already ingested - to be used during an incremental ingestion into a dataset |
Change | Integration | Plug-in SDK | Added additional fields to the Overlay Report to enable overlaying new Brain-space analytics. |
Change | Integration | Plug-in SDK | Choice fields are annotated and encoded in a new, simplified way. |
Change | Integration | Plug-in SDK | Example implementations have been enhanced and made more helpful to a developer. |
Change | Integration | Plug-in SDK | Added additional instructions and environment details for plugin developers. |
Change | Integration | Plug-in SDK | Added a new plugin.properties file for specifying version information. |