Glossary
A reference guide to the key concepts and terms used throughout Text Explorer.
Cause & Effect
Section titled “Cause & Effect”A Text Auto Report tab (Beta) that shows which sub topics drive your outcome measure, using Driver Analysis, and how topics affect each other, using the Impact Map. See Cause & Effect.
Classification model
Section titled “Classification model”A model built during preprocessing that maps text responses to topics and sub topics. Classification models can be created automatically using AI (Auto-Classification), defined manually, or reused from an existing model in the library.
Co-Occurrence
Section titled “Co-Occurrence”An analysis type that identifies which themes and terms appear together in the same responses. Co-Occurrence reveals connections between topics rather than just which topics exist. See Co-Occurrence.
Combination
Section titled “Combination”A unique pairing of values produced by applying two Data Splits variables simultaneously. For example, splitting by Region (5 values) and Customer Type (4 values) produces 20 combinations. Text Explorer supports up to 20 combinations per project.
Comment labels
Section titled “Comment labels”Variables from the project’s data store configured to display as contextual metadata alongside each comment in the Text Auto Report. For example, a Region label shows each respondent’s region beneath their comment, giving reviewers immediate context without cross-referencing another dataset. Up to 10 label variables can be configured per project.
Comment Search
Section titled “Comment Search”A Text Auto Report tab (Beta) for searching and reading every comment behind a report page, colour-coded by sentiment or a rating from your data. See Comment Search.
Data Split
Section titled “Data Split”A configuration that segments analysis results by one or two supporting variables. Instead of seeing combined results, each split category gets its own chart series. See Data Splits.
Data store
Section titled “Data store”The source of data connected to a Text Explorer project — typically a survey dataset, data upload, or integrated data feed. A project draws its text variables and supporting variables from the data store.
Diff Terms (Differentiating Terms)
Section titled “Diff Terms (Differentiating Terms)”An analysis type that identifies words and phrases disproportionately associated with a specific group compared to others. Where Top Terms shows what is mentioned most, Diff Terms shows what makes one group’s language distinctive. See Diff Terms.
Driver Analysis
Section titled “Driver Analysis”Part of the Cause & Effect tab. A model that ranks the sub topics that best predict whether a respondent is in your outcome group — for example, what separates Promoters from Detractors. See Drivers.
EVE variable
Section titled “EVE variable”A variable extracted from responses collected via an EVE conversation. EVE variables reflect custom probes and prompts configured in EVE Builder and are available in Text Explorer as supporting variables for filtering and segmentation.
Ignore list
Section titled “Ignore list”A list of terms excluded from a specific analysis chart. Used to remove words that are too generic or irrelevant to add analytical value — for example, product names or survey instructions. Available in Top Terms and Diff Terms analyses.
Lemmatisation
Section titled “Lemmatisation”A preprocessing step that reduces words to their base form so that variations of the same word are treated as one term. For example, “running”, “runs”, and “ran” are all reduced to “run”.
Measure
Section titled “Measure”The outcome a Text Auto Report tracks over time or explains — overall sentiment, or a data variable such as NPS or OSAT reported as a mean or a range. Set in the Define Your Measure step of project setup, and used by the Time Series and Cause & Effect tabs. See Define Your Measure.
Metric property
Section titled “Metric property”A setting in the Top Terms analysis that determines how term frequency is counted. Options are Count (total appearances) and Unique mentions (number of distinct respondents who mentioned the term).
Minimum mentions
Section titled “Minimum mentions”A threshold setting that filters out terms appearing fewer times than the specified number. Increasing the minimum mentions threshold removes noise and focuses analysis on the most significant themes.
Outcome group
Section titled “Outcome group”In Cause & Effect, the group of respondents your measure defines as the outcome — for example, Promoters (your In Group). Driver Analysis compares the outcome group with a comparison group, such as Detractors or everyone else.
Preprocessing
Section titled “Preprocessing”The step that prepares raw text for analysis. During preprocessing, Text Explorer cleans the text, removes stop words, applies lemmatisation, extracts terms, and builds the models that power each analysis type.
Project
Section titled “Project”A container for a Text Explorer analysis. A project connects to one data store, holds a variable selection, stores analysis configuration, and accumulates a library of saved chart outputs.
Report page
Section titled “Report page”A page of the Text Auto Report, built from one text variable and its settings in the Configure Report Pages step of project setup. Each page appears as an option in the Analysis Title filter at the top of the report. See Configure Report Pages.
Report type
Section titled “Report type”One of the tabs a Text Auto Report can include — Snapshot, Time Series, Cause & Effect, or Comment Search. Chosen in the Pick Your Report Type step of project setup. See Pick Your Report Type.
Sentiment class
Section titled “Sentiment class”One of four emotional tone categories assigned to text responses: Positive, Negative, Neutral, or Mixed. Used in the Sentiment analysis type. See Sentiment.
Snapshot
Section titled “Snapshot”The main Text Auto Report tab, showing what people raised and how they felt about it — summaries, sentiment charts, sub topic deep dives, and optional Emotional Insights. See Text Auto Report.
Stop words
Section titled “Stop words”Common words removed during preprocessing because they carry little analytical meaning — for example, “the”, “and”, “is”. Stop words are excluded before term frequency and analysis models are built.
Sub topic
Section titled “Sub topic”A specific theme or idea within a broader topic, identified during preprocessing. Sub topics are the primary unit of analysis in the Text Auto Report — each chart section, deep dive, and filter operates at the sub topic level.
Sub topic report
Section titled “Sub topic report”A separate Text Auto Report built from only the responses that mention one sub topic of another report. It finds its own topics and sub topics within that sub topic. Created from the Sub Topic Deep Dive or the Time Series tab. See Creating a separate report for a sub topic.
Supporting variable
Section titled “Supporting variable”A categorical or numeric variable from the data store used for filtering, segmentation, and data splits. Supporting variables are not analysed as text — they provide context and comparison dimensions for the Text Auto Report.
Text Auto Report
Section titled “Text Auto Report”An automatically generated, interactive dashboard built from a Text Explorer project’s analysis results. It presents sentiment breakdowns, key themes, topic insights, AI-generated summaries, and more in a structured, filterable format. Owners and editors can update settings, append data, and manage the report layout from within it. See Text Auto Report.
Text Drivers
Section titled “Text Drivers”An analysis type that identifies which text themes are most strongly associated with a quantitative outcome — for example, which topics most predict high or low satisfaction scores. See Text Drivers.
Text variable
Section titled “Text variable”An open-ended response field selected for text analysis — the verbatims, comments, or free-text answers in the data store. At least one text variable is required per project.
A high-level grouping of related sub topics, identified during preprocessing. Topics organise the classification structure and represent the broadest categories of themes in the data. The full topic and sub topic tree can be explored in the Appendix section of the Text Auto Report.
Time Series
Section titled “Time Series”A Text Auto Report tab that tracks how topics, sentiment, and your measure change over time. Needs a date variable. See Time Series.
Top Terms
Section titled “Top Terms”An analysis type that identifies the words and phrases appearing most frequently in text responses. See Top Terms.
Workbench
Section titled “Workbench”A Text Explorer environment for configuring analyses and building custom chart outputs. Accessible once a project has a Success status. See Workbench.