Release Notes for Mindbreeze InSpire
Version 23.6

Innovations and new features

Optimized "Did you mean" suggestions for searches in natural language

Searching in natural language enables users to find content in a natural and intuitive way, without having to restrict themselves to specific terms. The Mindbreeze InSpire 23.6 release now further supports users by optimizing "Did You Mean" suggestions for natural language queries. With the optimization of this feature, the system can now ensure that the context between words is better preserved and that the user is offered helpful suggestions even for longer search queries.

This is made possible by the automatic exclusion of certain words like "is" and "for" – so called stop words. Stop words are words that do not make a significant difference to the meaning of a query. They are listed in stop word catalogues and considered during the index inversion. When a search query is entered, Mindbreeze InSpire uses this multi word lexicon to generate "Did you mean" suggestions. This keeps the focus on the really relevant words.

This function is available in German, English, French, Spanish and Italian.

Link to documentation

More specific highlighting of relevant terms and sections

Highlighting terms and sections helps users to find relevant information. Thanks to the visual highlighting in the document, users can more quickly find the content they are looking for. With the release of Mindbreeze InSpire 23.6, the applied highlighting is now more focused and more precisely targeted. The highlighting of words with little relevance is no longer executed in order to focus the user’s attention to relevant words and paragraphs. In addition, the highlighting of words and paragraphs is now also available directly within the PDF preview.

With this optimization, the highlighting of words which are part of stop word catalogues is now disabled. As these words do not have much relevance due to their frequency, they are excluded. This feature is available in German, English, French, Spanish and Italian.

Link to documentation

Improved availability of document specific filters in the Insight App Designer

Filters are an essential part of designing Insight Apps. They ensure that search results, dashboards and charts can be further narrowed down. This makes it possible to quickly find the desired content and information in huge amounts of data. With the Mindbreeze InSpire 23.6 release, additional filters can be applied in the Insight App Designer that refer to document specific metadata (dynamic aggregated metadata).

Administrators can now use filters within their Insight Apps that are tailored to the data source used. The permissions of the user also play a decisive role. Accordingly, only those filters for which the user has access rights are available. Furthermore, only those filters that are available as metadata in the search results can be selected. For more information on dynamic aggregated metadata, see the Index Operating Concepts - Aggregateable.

In addition, the performance of the search function in the Insight App Designer has been optimized. This means that users can design their Insight apps more efficiently and the editing mode provides a more accurate preview of the final version of the Insight app.

Link to documentation

New and Extended Connectors

Indexing of file attachments and increased crawler robustness in Microsoft Dynamics CRM

Users of Microsoft Dynamics CRM can write comments in the form of "Notes". Files can be attached to notes to add information and context. With the release of Mindbreeze InSpire 23.6, it is now possible to index and display file attachments in the client service. This means that users can access file attachments and download them via the Client Service.

Attachments and their associated notes share the same metadata and can therefore be easily found and displayed together in the Client Service. Please note that file attachments can only be downloaded if Microsoft Dynamics CRM On-Premises is used with Kerberos authentication.

Additionally, the number of log messages generated by crawling untitled documents has been reduced. These are now summarized into a single log message to maintain log clarity. The robustness of the crawler has also been increased so that fewer crawl runs are aborted due to errors.

Link to documentation

Security relevant changes

23.6.2.330

• Updated: CoreOS update to version 38.20230918.3.0 (CVE-2023-4016, CVE-2023-4563, CVE-2023-4244, CVE-2023-4911).
• Updated: app.telemetry to version 23.1.107 (XSS).

23.6.0.307

• Updated: Chromium to version 118.0.5993.18 (various CVEs).
• Updated: Java GRPC to version 1.5.7.2 (CVE-2023-33953).
• Updated: Dell Firmware (CVE-2022-40982, CVE-2022-43505)
• Updated: Keycloak to version 22.0.1. (CVE-2022-4361)

Additional changes

23.6.3.336

• Fix for: In the Basic Login Configuration setting the parameter „useKeyTabs“ was set to "true".

23.6.1.319

• In version 22.3, a change was introduced that proved to be problematic. This change was reversed due to the possible disclosure of information.

23.6.0.307

• Added: Filters for dynamically aggregated metadata are displayed in the Insight App Designer.
• Added: The application options accept a new value called “editDebounce”. This delays the “Edit View” initialization.
• Fix for: Windows setup did not detect existing installations in custom install paths.
• Fix for: Race condition in initial configuration.
• Fix for: Incorrect STX offset of the highlighting in the PDF preview (when using CSV Enricher or Hierarchical CSV Transformer).
• Fix for: ACL threads did not wake up and the first timeout was slowing down the performance.
• Fix for: Unable to overwrite "Did You Mean" to be computed when no results were found per query service.
• Fix for: List of filters is incomplete in the Insight App Designer if no filters have been added to the Insight App yet.
• Fix for: The servlet “/wait” is not reliable if only small updates are done in the index.
• Fix for: Heritrix will continue logging after a log backup.
• Fix for: Query Expression Transformation did not work as expected when a TERM's expression was nested inside an “AND” expression.
• Fix for: Confluence Sitemap Generator may not operate correctly after an update or a configuration change.
• Fix for: Issue with incorrectly exported properties in the export.
• Improved: Saving of disk space in case of frequent updates or deleted documents.
• Improved: The servlet „/wait“ sends responses as server-sent events (SSE) to support long running waits.
• Improved: Configurability of time zone and time server. In addition, achievement of CIS-L1 and default umask is 027.
• Improved: Did-You-Mean (e.g. ignore stop words).
• Improved: Sample texting has been optimized by skipping the processing of content if there is no hit in the content.
• Improved: Microsoft CRM Connector can now crawl attachments from notes (annotations).
• Improved: Performance of "Update app.telemetry dashboards" task.
• Changed: Index setting "Max Query Timeout ms for Did You Mean" default value from 0 to 500.
• Changed: The saved search encoded URLs won't be effective on versions older than 23.6.
• Removed: Webmin has been removed from the product.
• Removed: Obsolete cipher suites from MMC.
• Updated: App.telemetry to version 23.1.