Download RELATIVITY ANALYTICS SPECIALIST EXAM SCRIPT 2026 PRACTICE SOLUTION BUNDLED and more Exams Reasoning in PDF only on Docsity!
RELATIVITY ANALYTICS SPECIALIST EXAM
SCRIPT 2026 PRACTICE SOLUTION BUNDLED
◉ What best describes coverage mode in Review Center? a. Serves up documents that are most likely to be relevant. b. Serves up documents that are optimal for training the model. c. Focuses on documents least likely to be responsive to train the model. d. Gets a large percentage of relevant documents to the reviewers. Answer: Serves up documents that are optimal for training the model ◉ What documents are listed in the Documents panel in the viewer when in a Review Center queue? a. Documents the queue will serve next to this logged in user. b. Documents the queue will serve next to one of the reviewers in the queue. c. Documents the logged in user has reviewed in the
queue. d. Documents the logged in user has coded inside or outside the queue. Answer: Documents the logged in user has reviewed in the queue. ◉ What additional validation metric is calculated if you set a cutoff? a. Precision b. Recall c. Elusion Rate d. Richness Answer: Precision ◉ What's the difference between conceptual and structured analytics? Answer: Conceptual analytics focuses on the related concepts whereas structured relies on the specific structure of the content. ◉ When you are creating a new conceptual analytics index, what few steps need to be performed? Answer: 1 Creating a saved search to define the training and searchable sets. 2 Running a structured analytics repeated content set. 3 Creating the index and linking the repeated content ID.
◉ How can Clustering assist in your review? Answer: Investigate a large unknown data set Deprioritize non-relevant documents quickly Speed up a linear review by batching conceptually related documents together ◉ How can Keyword Expansion help your review? Answer: Expand original keyword terms list to uncover unknown related terms Overcome obfuscatory language, uncover code words, or help with regional variation language ◉ When you install Review Center app, what tabs are created? Answer: Review Center Review Queue Review Library ◉ What field must be created when you start a Prioritized Review queue? Answer: Rank Output Field - Decimal ◉ How does Review Queues serve up documents in a saved search queue? Answer: By the saved search sort order ◉ What categories are tracked on the Review Center charts? Answer: Coding Progress
Relevance Rate Review Speed Rank Distribution Queue History ◉ Some of Review Center's key features include: Answer: Templatization Customized queues Powerful AI classifier Clear progress reporting ◉ You are reviewing documents in Review Center using a prioritized review queue and notice documents are grouped by family. How should you code groups of documents in the same family? a. You should code each document in the family using the "Four Corners" test, and code each document independently of other documents in the family. b. You should code each document in the family based on the custodian and date range of the document. c. You should code all other documents in the family Responsive if at least one document in the family is Responsive. d. You should code all documents in that family Responsive if at least one document in the family comes from a custodian whose documents are likely responsive Answer: a. You should code each
◉ If enabled for a queue, where can you see the total count of documents you coded in a Review Center queue? a. Queue card. b. Saved search. c. Field Tree. d. Reviewed Documents table. Answer: a. Queue card. ◉ How do you access documents you previously coded in Review Center? a. Go to the Documents panel within the review queue document viewer. b. Use the Review Queues view on the Documents tab. c. Go to the Review Center queue page. d. It is not possible to access previously coded documents. Answer: a. Go to the Documents panel within the review queue document viewer. ◉ The Start Review button is missing from the Review Queues card. Why might this happen? a. The queue is paused. b. You do not have permissions to that queue. c. Too many reviewers are already in the queue. d. Another queue is active. Answer: a. The queue is paused.
◉ What Review Center queue serves up documents that are the most likely to be responsive? a. Prioritized review queue. b. Responsive review queue. c. Saved search queue. d. Family-based review queue. Answer: a. Prioritized review queue. ◉ You are reviewing a document and would like to find documents that are conceptually similar to the entire document you are currently reviewing. How can you do this? a. Click the View Similar Documents button in the viewer. b. Highlight a paragraph of text in the document and use the conceptual analytics index to search this text. c. Choose the most relevant phrase from the document and run a dtSearch on this phrase. d. Use the Related Items pane to look for textual near duplicate groups. Answer: a. Click the View Similar Documents button in the viewer. ◉ You are told a paragraph is highly relevant. You would like to use a search to return other documents that may be closely related to the paragraph above. What should you do to return the best results?
◉ What can you do from the Review Center dashboard? Answer: prepare, start, stop, edit, and refresh your Review Center queues. observe the progress of the review using relevant project metrics. monitor the overall projects via the charts and tables that are displayed including: Relevance Rate, Coding Progress, Review Speed, Reviewer Detail, and Rank Distribution (for prioritized reviews). ◉ What factors are taken in when you should validate a prioritized review queue? Answer: rank distribution - your rank distribution has few or no unreviewed documents with a rank of 50 or higher relevance rate - you should notice a decline in the progress line trending towards zero. This indicates that very few responsive documents are still being found in your project. ◉ Prior to validation you must... Answer: Pause review queue Disable auto refresh ◉ Validations setting include: Answer: Review Groups Sample Size Margin of Error (Elusion) Margin of Error (Recall)
◉ Margin of Error (Elusion) Answer: Calculates a size for the sample based on how accurate the Elusion statistics will be. A lower margin will increase the sample size ◉ Margin of Error (Recall) Answer: Calculates a size for the sample based on how accurate the Recall statistics will be. A lower margin will increase the sample size. ◉ All validation statistics aim for a 95% confidence interval alongside the margin of error (MOE): True or False Answer: True ◉ Scenarios that can nullify and require recalculation of your validation: Answer: Coding skipped documents or documents coded with a neutral choice that are part of the validation sample Changing already coded documents outside the sample Adding already coded documents to the queue after validation starts ◉ Scenarios that require a new validation queue: Answer: Adding uncoded or neutral documents to the queue after validation starts Changing positive or negative coded documents outside the sample to skipped or neutral ◉ Unlike traditional searching methods like dtSearch, how is Analytics different? a) It uses outside word lists
c) The searchable area within the index d) The physical storage space required for the index Answer: b) A mathematical model where documents are mapped ◉ Concept Rank in the context of Analytics indexing is a measure of: a) Conceptual relevance of the document b) Distance between two items c) Number of concepts in the document d) Percentage of shared terms in a document Answer: b) Distance between two items ◉ What is the purpose of 'Optimize Training Set' feature in an Analytics index? a) To include all documents from the saved search b) To include only conceptually valuable documents from the training data source c) To increase the number of dimensions in the concept space d) To increase the size of the index Answer: b) To include only conceptually valuable documents from the training data source ◉ The 'Enable Email Header Filter' gets activated when: a) 'Remove English Email Signatures and Footers' is set to 'Yes' b) It is always activated
c) 'Remove English Email Signatures and Footers' is set to 'No' d) 'Optimize Training Set' is set to 'Yes' Answer: c) 'Remove English Email Signatures and Footers' is set to 'No' ◉ What does an 'Exception' mean in the context of an Analytics index? a) It refers to the documents that have been excluded from indexing b) It refers to any errors encountered during the indexing process c) It refers to the documents that are excluded from the training data source d) It refers to the unique words that have been removed from the index Answer: b) It refers to any errors encountered during the indexing process ◉ What does a classification index use for discovering concepts between documents? a) Support Vector Machine (SVM) b) Latent Semantic Indexing (LSI) c) Conceptual Semantic Indexing (CSI) d) Latent Vector Indexing (LVI) Answer: a) Support Vector Machine (SVM) ◉ What is the significance of the 'Dimensions' field under 'Advanced Settings'? a) It defines the physical dimensions of the server storing the index
◉ The 'Rank' in Support Vector Machine learning (SVM) measures: a) The strength or confidence the model has in a document being relevant or not b) The number of dimensions in the SVM c) The confidence level of the SVM model d) The speed of the SVM model Answer: a) The strength or confidence the model has in a document being relevant or not ◉ In the context of Analytics indexing, what does Latent Semantic Indexing (LSI) enable? a) It enables the analysis of document syntax b) It enables the extraction of keyphrases from the document c) It enables the analysis of document metadata d) It enables Relativity Analytics to learn the language and the conceptuality of each document Answer: d) It enables Relativity Analytics to learn the language and the conceptuality of each document ◉ What does the 'Dimensions' field determine in creating a conceptual index? a) The dimensions of the physical storage space b) The dimensions of the hyperplane used for differentiating documents
c) The dimensions of the concept space into which documents are mapped when the index is built d) The dimensions of the arrays used for storing the index Answer: c) The dimensions of the concept space into which documents are mapped when the index is built ◉ Why might you need to run a 'Full Population' while updating an index? a) When there are new documents that introduce new concepts b) When you want to update the noise words c) When the extracted text is updated d) All of the above Answer: d) All of the above ◉ When does an index automatically suppress documents? a) When the documents are larger than 30 MB b) When the documents contain system files c) When the documents contain image files d) When the documents are compressed files like ZIP or RAR files Answer: a) When the documents are larger than 30 MB ◉ When creating an Analytics Index, what unique function does the 'Cluster Documents' field serve? a) It organizes documents based on their content.
◉ In 'Analytics and Support Vector Machine learning (SVM)', what does SVM predict? a) The relevance of uncoded documents based on their distance to the hyperplane. b) The number of languages used in the documents. c) The size of the document. d) The complexity of the text within the documents. Answer: a) The relevance of uncoded documents based on their distance to the hyperplane. ◉ Which of the following is not a type of review queue in Review Center? a) Saved search queues b) Random queues c) Prioritized review queues d) Custom-sorted document queues Answer: b) Random queues ◉ What is the AI classifier's purpose in Review Center? a) To sort documents into folders b) To serve up documents it predicts as relevant c) To delete irrelevant documents d) To classify documents into types Answer: b) To serve up documents it predicts as relevant
◉ What happens when Coverage Mode is turned on for a prioritized review queue? a) It serves up the highest-ranking documents b) It serves up documents that are better for training the model c) It stops the queue from functioning d) None of the above Answer: b) It serves up documents that are better for training the model ◉ How many documents are checked out to each active reviewer at a time by default? a) Three b) Five c) Ten d) Fifteen Answer: b) Five ◉ What does setting a relational field in a Review Center queue do? a) It clusters similar documents together b) It separates unrelated documents c) It assigns the whole relational group of documents present in the queue to the same reviewer
