Perform OCR

Option

Action

OCR Profile

Leave the Default selected or click Configure to do either of the following:

Select a pre-customized OCR Profile.

Edit the Default to customize a new OCR Profile.

Evidence Settings

Select Store binary of data items to store a copy of the native file in the databases in the case directory while you OCR items, up to the maximum size of one gigabyte. The default maximum is 250 megabytes.

This allows you to directly review the PDF files, that are generated after OCRing, as images.

Note: This option can reduce indexing speed by 15-20%, and increase the amount of storage required for the case folder from 20-50% of the original data set to 220-250%.

Family Text Settings

Select either or both of the following options:

Create family search fields for top level items

To create an extra field with all child items' text added to their parent's text, allowing for the use of "family" searches for faster searching. (A top level item is a material item with no material ancestors, that is, the first material item in an item path.)

See the Nuix Search Guide for syntax when using this field or from the Help menu select Help Topics > References > File Types.

Hide immaterial items (text rolled up to parent)

To automatically hide immaterial items in the Results pane when processed and remove clutter from the results set, yet append the extracted text from hidden immaterial items to its parent item so it is still available to search on.

Text Indexing Settings

In Analysis language, select English or Japanese.

You can only use one language for each evidence store and you cannot change it when you select the Reuse Evidence Store option.

Then select any of the following options:

Use stop words: To not index all stop words, or noise words, for your selected language. English stop words are: a, an, and, are, as, at, be, but, by, for, if, in, into, is, it, no, not, of, on, or, such, that, the, their, then, there, these, they, this, to, was, will and with.

Use stemming: To create a text index that only stores stemmed components of words for an entire case.

For example, if your search word is "control", enabling this option returns documents containing "control", and "controlling", "controller", and "controls". However, be aware of this warning:

Enable exact queries: To store the text content of items with punctuation and capitalization, so that when searching, you perform an exact string match. (For such a search, remember to use single quotes (‘ ‘) around your search term to invoke exact queries.)

Hover over this field to see this tooltip:

“Create a separate text index with characters indexed one by one to allow searching for punctuation and other symbols which are usually not indexed. This will increase processing time and significantly increase case size.”

Worker Settings

Reconfigure any of the following settings to process your OCR job more efficiently:

Number of local workers Memory per worker (MB)

In general, 4GB of RAM is sufficient for most operations. If working with very large datasets and performing operations like finding top- level items or deduplicating, it is not uncommon to increase the Memory to 30GB or more depending on your hardware. It is important to balance the amount of memory you allocate to Nuix Workstation and to Nuix single Workers for processing operations. For information on allocating application memory, in the Nuix Workstation Installation Guide see the Hardware configuration page, and Recommended memory (RAM and CPU) allocation section.

Worker temp directory

Option

Action

Worker broker IP address and port

Item Content Settings

Select any of the following to use these features:

Process Text: To extract the text content of evidence to enable searching. If you disable this option, you can only search across an evidence item’s metadata.

Enable near-duplicates: To identify word shingles for Near-Duplicate detection and clustering in the case. This option automatically selects the Process Text option, if not pre-selected.

Enable text summarisation: To calculate and store text summaries from document data that is ingested. This option automatically selects the Process Text option, if not pre-selected.

Extract named entities from text: To extract named entities from text; and select Include text stripped items to include text stripped items while extracting named entities from text.

Nuix Workstation uses Text Stripping to identify an item’s file type but it cannot cleanly extract all text and metadata in accordance with the file type’s API. The result is a searchable data item, but the text may be garbled or not properly formatted. Text Stripping scans the bytes of the data item and looks for runs of string data at least three characters long.

Extract named entities from properties: To search in the metadata properties associated with an item.

Extract named entities from communications: To extract named entities such as phone numbers and email addresses from the communication metadata using Nuix Workstation standard or custom named entities. When selected, it enables the Use custom named entity profile option.

Use custom named entity profile: To enable selection of a Custom Named Entity Profile for processing from the menu if you selected the Extract named entities from text or Extract named entities from properties option.