The correct answer is B. dog not cat. Relativity’s dtSearch syntax guidance explains that the NOT operator can be used by itself at the beginning of an expression, or in the middle of an expression only when combined with either AND or OR. Relativity gives examples such as apple OR NOT pear and applesauce and NOT pear as valid syntax. That means expressions like dog OR NOT cat and dog AND NOT cat are valid, while dog NOT cat is missing the required connector and therefore causes a syntax problem.

Relativity’s Search Terms Report error documentation further explains that STRs surface syntax errors when the search index cannot process the term because of invalid syntax. In this case, dog not cat violates the dtSearch operator rule, so it is the option most clearly expected to fail with a syntax error in a search terms report. Option A is a standard Boolean expression and is valid. Options C and D both include the required connector before NOT, which matches Relativity’s documented dtSearch usage. Therefore, among the listed queries, dog not cat is the one that would cause a syntax error on an STR.

The correct answer is A. Ignore Numbers. Relativity’s near duplicate documentation explains that when Ignore Numbers is set to true, the similarity calculation considers only tokens beginning with letters, which means numbers are excluded from the similarity percentage calculation. That is directly relevant for invoice populations, because many invoices share nearly identical wording while differing mainly by dates, invoice numbers, quantities, or amounts. In a set like the one described, toggling Ignore Numbers is the setting most likely to change how those documents are grouped and help you assess whether the current near duplicate results can be improved.

The other options do not fit this use case. MD5Hash Verifier is related to hash-based duplicate validation rather than textual near duplicate tuning. Relativity Compare is a review tool for examining differences after grouping, not a setting that changes how the grouping is calculated. Auto-recognize Dates is not the structured analytics tuning option documented for this scenario. Therefore, for invoice-heavy data where numerical differences are driving questionable textual near duplicate results, the correct setting to toggle is Ignore Numbers.

The correct answer is B. Runs the saved search and trains the classifier. Relativity’s Review Center documentation states that during the Preparing state, the queue is refreshing the saved search for the first time , and for a prioritized review queue , this also trains the classifier . The Review Center admin quick reference guide further states that Prepare Only re-runs the saved search for all queues and retrains the AI classifier for prioritized review queues , and it does not start the queue.

That makes option B the direct documented effect of Prepare Only. It does not grant reviewer access by itself, it does not release checked-out documents, and it is not a statistical sampling action. From a case administration perspective, Prepare Only is used to get the queue ready and refresh the machine-learning state before reviewers begin or resume work. Therefore, the correct answer is B .

The correct answer is A. Configure the index to be accent sensitive. Relativity’s official dtSearch index documentation lists Create accent sensitive as a configurable Yes/No setting on a dtSearch index. When enabled, the index becomes sensitive to accents and other language-specific characters. This is an example of a valid dtSearch index modification that an administrator can configure.

The other options are incorrect. Relativity does not document a dtSearch feature that translates search terms into multiple languages, so option B is not a valid dtSearch index modification. A repeated content filter is associated with Analytics workflows, not dtSearch index configuration, so option C is also wrong. Option D is clearly incorrect because dtSearch indexes are not simply automatic and fixed upon import; Relativity provides a dedicated dtSearch object with editable settings such as searchable set, sub-index sizing behavior, skip malicious files, and accent sensitivity.

From an RCA perspective, this matters because dtSearch is a configurable search technology in Relativity, and administrators are responsible for tuning it for the needs of a workspace. Accent sensitivity can materially affect search precision in multilingual or language-specific datasets. Therefore, the correct answer is A .

The correct answer is D. Regular expression filter. Relativity’s Structured Analytics documentation states that a Regular expression filter can be linked to a structured analytics set as an optional setting, and that it is used to clean up extracted text before analysis . The documentation further explains that this setting is implemented through a repeated content filter whose type must be Regular Expression . This directly matches the question’s wording about excluding document text before analytics processing begins.

This capability is especially useful when documents contain extraneous content such as dates, URLs, page numbers, Bates numbers, standard disclaimers, or other repeated text that can interfere with email threading or textual near duplicate results. Relativity notes that the filter applies only to the field being analyzed, which is usually Extracted Text unless another field is selected. By removing this noise before the algorithm runs, administrators can improve the quality of structured analytics outcomes.

The other options are not the correct named filter for this function in Relativity’s structured analytics workflow. The official setting is specifically the Regular expression filter . Therefore, the correct answer is D .

The correct answer is D. Check the box to export multiple-choice fields as nested. Relativity’s official export documentation states that Export Multiple Choice Fields as Nested is the option used to maintain the hierarchy of Relativity multiple-choice lists during export. It further explains that the nested value delimiter is a backslash, which preserves the parent-child structure of those choices in the exported data.

This means the hierarchy is not lost by default if the appropriate export option is selected, so option A is incorrect. Sorting on the field does not preserve the nested hierarchy, making option B wrong. Exporting as CSV for Excel recognition is also not the documented mechanism, so option C is incorrect. From an RCA and data loading standpoint, this setting matters because exported hierarchical values often need to be re-used downstream, re-imported, or validated in a way that preserves their original structure. Without the nested export option, the relationship between parent and child values may not be represented correctly. Therefore, the proper way to maintain multiple-choice hierarchy when exporting a saved search through Import/Export is to check the box to export multiple-choice fields as nested .

The correct answer is C. Possible cut-off content on imaged MSG or EML files. Relativity’s official Imaging Warnings documentation states that the Imaging Warnings object identifies possible cut-off content on imaged MSG or EML files and stores that information. It further explains that users can review these items on the Imaging Warnings tab after imaging completes.

The other choices describe imaging failures or limitations, but they are not what the Imaging Warnings tab is specifically designed to identify. Password-protected failures and general non-imageable content are handled through other imaging or processing error mechanisms. The Imaging Warnings feature is much narrower and is specifically meant to flag the risk that content in imaged MSG or EML items may have been cut off in the rendered output. Therefore, the correct answer is C .