Nuix NLP v1.3
1. New Features
  1. United Kingdom and Australian Compound Lexemes
  1. Investigations Compound Lexemes
  1. Investigations Skillsets
  1. Model Improvements
  1. User Interface
  1. Processing Improvement
  1. Graph Connector Improvements
1. Resolved Issues
  1. User Interface
  1. Processing
  1. Graph Connector
1. Models

Nuix NLP v1.3

The following new features and fixed issues are included in Nuix NLP.

New Features

United Kingdom and Australian Compound Lexemes

In this version, NLP introduces United Kingdom and Australian-specific Compound Lexemes. This is the first release with tailored features for these regions.

Australian Compound Lexemes include:

Australian Business Number

Australian Company Number

Bank State Branch

Immicard Number

Individual Healthcare Identifier

Medicare Number

Tax File Number

United Kingdom Compound Lexemes include:

Company Registration Number

National Insurance Number

NHS Number

Sort Code

UTR Number

Investigations Compound Lexemes

NLP has added the ability to search for sensitive financial information related to Crypto and XRP. Additionally, NLP provides insights into a person's online habits, including frequently referenced URLs and search terms.

Investigations Compound Lexemes include:

Crypto Value

Cryptocurrency Transaction

Search Term

Web URL

XRP Address

Investigations Skillsets

NLP has enhanced investigations skillsets to easily search for common scam emails, including 419 and phishing emails. Investigators can use the advanced capabilities to identify instances where the Fraud Triangle may be present, enabling a deeper understanding of potential fraudulent activities.

Added skillsets include the following:

Skillset: Scam Emails

Skill: 419 Scam Emails

Skill: Phishing Emails

Skillset: Fraud Indicators

Skill: Opportunity

Skill: Pressure

Skill: Rationalization

Model Improvements

Added backend functionality to Entity Risk rules to "require all".

Added Italian language support.

User Interface

Added sorting parameters to /listLabels for Compound Lexemes and Regexes.

Updated sorting for Labels in Compound Lexemes and Regexes.

Improved icon colors in Compound Lexemes for information, warning, alert, and error icons.

Added ‘collapse’ and ‘expand all’ functionality to Compound Lexemes.

Added a ‘require all’ component for entity risk.

Compound Lexeme blocks are now expanded by default.

Added a list of items referenced in the current Compound Lexeme in the references dialog.

Improvements to the validation visualization graph.

Added the ability to filter data based on values in the playbook.

Added a Playbook Rule for linking named items.

In Compound Lexemes, improved validation of invalid or incompatible rules.

A user can now specify a persistence instance when calling an NLP Job Control Service.

Added links for cross-app navigation.

Processing Improvement

Added the option to remove ‘Spacy Entities’ from Job Control results.

Graph Connector Improvements

In the playbook, the normalizedName field is used as the main name in the graph.

Added the ability to view relationships between item metadata by analyzing the nodes and edges of text-less items from the Nuix Engine in Linkurious.

Added the Playbook Rule for Linking Named Items.

Updated Playbook to remove spaces instead of replacing them with an underscore.

Resolved Issues

User Interface

Skills user interface filters don't display correct Counts. This is now fixed.

Changes to the Stopwords list required a page refresh to be visible to users. This is now fixed.

There was an issue with Umlaut German characters. This is now fixed.

Accented letters disappeared in certain validation documents and proximity issues. This is now fixed.

Dictionaries-Limit to Current Tier option was not working as expected. This is now fixed.

Misleading errors occurred when adding labels containing special characters to Compound Lexemes and Regex. This is now fixed.

No results were shown in Entity Risk Rules for Compound Lexeme labels. This is now fixed.

Accented letters in the text did not appear in the user interface. This is now fixed.

Processing

Specific text files returned an error when processed through the bulk uploader. This is now fixed.

Processing an image file caused a service crash. This is now fixed.

Feeds were not processing specific documents. This is now fixed.

Graph Connector

Errors and nodes were not created if ‘No Edge Rule’ was specified. This is now fixed.

In Playbook, Edges were not created in dictionaryProximities section. This is now fixed.

Added fixes for Memgraph Connector performance issues.

Models

The NHS Number compound lexeme was referencing the incorrect lexeme. This is now fixed.