{"id":67091,"date":"2026-08-20T18:28:58","date_gmt":"2026-08-20T18:28:58","guid":{"rendered":"https:\/\/www.sogolytics.com\/help\/?page_id=67091"},"modified":"2026-08-20T23:51:28","modified_gmt":"2026-08-20T23:51:28","slug":"entity-analysis-metadata","status":"publish","type":"page","link":"https:\/\/www.sogolytics.com\/help\/entity-analysis-metadata\/","title":{"rendered":"CX Dashboard: Metadata Analysis"},"content":{"rendered":"<style>\n\/* Offset anchor jumps so headings aren't hidden behind the fixed help header *\/\nh2[id], h3[id] {\n  scroll-margin-top: 100px;\n}\n<\/style>\n<p>The <strong>Metadata<\/strong> tab in Text Analysis, powered by <b>OpenAI<\/b>, identifies the specific people, products, companies, and places your customers name in their open-ended responses, and shows how often each one is mentioned and how customers feel about it. It turns unstructured, qualitative feedback into structured metadata you can rank, track, filter, and drill into \u2014 so you can analyze at scale instead of reading every comment by hand.<\/p>\n<p>Themes tell you what topics customers are discussing. Metadata tells you who and what they are naming inside those discussions. If a customer writes \u201cthe billing team took three days to respond,\u201d Theme Analysis identifies the theme Response Time; Metadata identifies the entity Billing Team and records the sentiment attached to it. Over hundreds of responses, this lets you rank the entities driving your feedback and drill into the verbatims behind each one.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity1.png\" \/><\/p>\n<div class=\"sogo-help-info-box\">\n<i class=\"fal fa-info-circle\"><\/i><\/p>\n<p class=\"sogo-info-text\">Entities are extracted from <strong>follow-up (open-ended) questions<\/strong> only. The sentiment shown for an entity reflects the sentiment of the surrounding response in which it was mentioned, not a direct rating of the entity itself.<\/p>\n<\/div>\n<p class=\"sogo-help-sub-head\">Where to Find<\/p>\n<p>Go to <strong>CX Dashboard &gt; Report &gt; Text Analysis &gt; Metadata<\/strong>. Metadata is the fourth tab in Text Analysis, alongside Open-Ended Responses, Theme Analysis, and Text Classification.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity2.png\" \/><\/p>\n<hr \/>\n<p class=\"sogo-help-sub-head\">Before you begin<\/p>\n<ul>\n<li>Metadata is available on <strong>Pro<\/strong> packages and above, under the same Text Analytics entitlement that covers Open-Ended Responses and Theme Analysis. There is no separate packaging change for this tab.<\/li>\n<li>Your account must have accepted the <b>OpenAI<\/b> Terms. If consent has not been given, the tab displays a consent prompt in place of the widgets. Only an <strong>Admin User<\/strong> can accept the terms.<\/li>\n<li>Your project must contain at least one open-ended question marked as a <strong>follow-up<\/strong>, with responses collected.<\/li>\n<li>If your sub-user account has a <strong>Data Visibility Rule<\/strong> applied, every widget, drill-down, and export on this tab is limited to the records your rule permits. An administrator also controls who can access the Metadata tab.<\/li>\n<\/ul>\n<hr \/>\n<p class=\"sogo-help-sub-head\">Navigate to<\/p>\n<ul class=\"sogo-help-sub-list\">\n<li><a href=\"#section-1\">Entity types<\/a><\/li>\n<li><a href=\"#section-2\">Getting oriented: the top controls<\/a><\/li>\n<li><a href=\"#section-3\">Available widgets in this tab<\/a><\/li>\n<li><a href=\"#section-4\">Entity Details<\/a><\/li>\n<li><a href=\"#section-5\">Filters<\/a><\/li>\n<li><a href=\"#section-6\">Good to Know<\/a><\/li>\n<\/ul>\n<h2 id=\"section-1\">Entity types<\/h2>\n<p>Every entity the system extracts is classified into exactly one of six types. Understanding the distinction between them \u2014 particularly Place and Location \u2014 makes the filters easier to use.<\/p>\n<div class=\"package-table-container no-scroll\">\n<table class=\"package-limit-table\" border=\"0\" cellpadding=\"0\">\n<thead>\n<tr>\n<td style=\"width: 20%;\">Type<\/td>\n<td style=\"width: 45%;\">What it identifies<\/td>\n<td style=\"width: 35%;\">Examples<\/td>\n<\/tr>\n<\/thead>\n<tbody>\n<tr class=\"column-bg-white-effect\">\n<td style=\"width: 20%;\">Person<\/td>\n<td style=\"width: 45%;\">A named individual \u2014 staff, contacts, or public figures.<\/td>\n<td style=\"width: 35%;\">Dr. Arjun Malhotra, Priya Nair<\/td>\n<\/tr>\n<tr class=\"column-bg-effect\">\n<td style=\"width: 20%;\">Product<\/td>\n<td style=\"width: 45%;\">A product, device, feature, or service named by respondents.<\/td>\n<td style=\"width: 35%;\">Philips patient monitor, Boston Scientific stent<\/td>\n<\/tr>\n<tr class=\"column-bg-white-effect\">\n<td style=\"width: 20%;\">Company<\/td>\n<td style=\"width: 45%;\">An organization, brand, or business.<\/td>\n<td style=\"width: 35%;\">Meridian Health Network, Cigna<\/td>\n<\/tr>\n<tr class=\"column-bg-effect\">\n<td style=\"width: 20%;\">Place<\/td>\n<td style=\"width: 45%;\">A store, branch, office, or identifiable establishment.<\/td>\n<td style=\"width: 35%;\">Noida Office, Cardiology wing exam room 4<\/td>\n<\/tr>\n<tr class=\"column-bg-white-effect\">\n<td style=\"width: 20%;\">Location<\/td>\n<td style=\"width: 45%;\">A geographical or map-based location \u2014 city, region, or country.<\/td>\n<td style=\"width: 35%;\">London Branch, APAC Region<\/td>\n<\/tr>\n<tr class=\"column-bg-effect\">\n<td style=\"width: 20%;\">Profession<\/td>\n<td style=\"width: 45%;\">A role, job title, or occupation.<\/td>\n<td style=\"width: 35%;\">Cardiologist, Respiratory therapist<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p>Responses that contain no recognizable entity, and text the system cannot confidently classify into one of these six types, are not shown on this tab.<\/p>\n<p><strong>Mentions<\/strong> counts every time an entity appears across your responses. <strong>Unique Responses<\/strong> counts how many separate responses an entity appears in. A single response that names SogoAssist three times contributes three mentions but one unique response.<\/p>\n<p>Each entity also carries a <strong>sentiment score<\/strong> between <strong>\u22121.00<\/strong> and <strong>+1.00<\/strong>, drawn from the sentiment of the text surrounding the mention:<\/p>\n<ul>\n<li><strong>Positive<\/strong> \u2014 score of +0.20 or higher<\/li>\n<li><strong>Negative<\/strong> \u2014 score of \u22120.20 or lower<\/li>\n<li><strong>Neutral<\/strong> \u2014 any score between the two<\/li>\n<\/ul>\n<hr \/>\n<h2 id=\"section-2\">Getting oriented: the top controls<\/h2>\n<p>A set of controls runs across the top of the tab and governs everything below it. Set these first, then read the widgets.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity3.png\" \/><\/p>\n<div class=\"package-table-container no-scroll\">\n<table class=\"package-limit-table\" border=\"0\" cellpadding=\"0\">\n<thead>\n<tr>\n<td style=\"width: 30%;\">Control<\/td>\n<td style=\"width: 70%;\">What it does<\/td>\n<\/tr>\n<\/thead>\n<tbody>\n<tr class=\"column-bg-white-effect\">\n<td style=\"width: 30%;\">Time filter<\/td>\n<td style=\"width: 70%;\">Sets the reporting period for the entire tab (for example, All Time or a custom range). Every widget, delta, and trend recalculates to the selected period.<br \/>\n<img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity4.png\" \/><\/td>\n<\/tr>\n<tr class=\"column-bg-effect\">\n<td style=\"width: 30%;\">All Metric selector<\/td>\n<td style=\"width: 70%;\">Focuses the analysis on entities extracted from specific metric (follow-up) questions in your project.<br \/>\n<img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity5.png\" \/><\/td>\n<\/tr>\n<tr class=\"column-bg-white-effect\">\n<td style=\"width: 30%;\">Entity type chips<\/td>\n<td style=\"width: 70%;\">Quick filters for the six entity types (see below), each showing a live count.<br \/>\n<img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity6.png\" \/><\/td>\n<\/tr>\n<tr class=\"column-bg-effect\">\n<td style=\"width: 30%;\">Filter (funnel icon)<\/td>\n<td style=\"width: 70%;\">Opens the <strong>Create Filters<\/strong> panel for more detailed conditions, including the new <strong>Entity<\/strong> NLP criteria.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<h3 id=\"entity-type-chips\">Entity type filter (chips)<\/h3>\n<p>A row of filter chips sits at the top of the tab: <strong>All<\/strong>, <strong>Person<\/strong>, <strong>Product<\/strong>, <strong>Company<\/strong>, <strong>Place<\/strong>, <strong>Location<\/strong>, and <strong>Profession<\/strong>, each showing the number of entities of that type in the current scope (for example, All 1664, Person 1068, Product 96, Company 303, Place 98, Location 8, Profession 91).<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity7.gif\" \/><\/p>\n<p>These chips act as a filter for the entire page. Clicking a chip re-draws every widget on the tab to show only that entity type \u2014 the metric cards recalculate, the trend charts switch from type-level lines to the top five individual entities of the type you selected, and the word cloud and table filter to match. <strong>All<\/strong> is selected by default. Clicking the active chip a second time returns you to <strong>All<\/strong>. The chip row stays in place as you scroll, so the current scope is always visible.<\/p>\n<div class=\"example-block-container blue-shade\">\n        <span class=\"example-block-content\"><span class=\"example-block-header\">For example:<\/span>Click the <strong>Product<\/strong> chip and the whole tab reshapes around products only: the metric cards recalculate (23 distinct products, 96 total mentions), the charts show product trends, and the table lists only products such as Philips patient monitor and GE Vivid ultrasound machine. Click <strong>All<\/strong> to bring every entity type back into view.<\/span>\n  <\/div>\n<hr \/>\n<h2 id=\"section-3\">Available widgets in this tab<\/h2>\n<p>The Metadata tab is made up of the following widgets:<\/p>\n<h3>Available Widgets in this tab<\/h3>\n<ul>\n<li><a href=\"#md-unique-entities\">Unique Entities<\/a><\/li>\n<li><a href=\"#md-responses-with-entities\">Responses with Entities<\/a><\/li>\n<li><a href=\"#md-total-mentions\">Total Mentions<\/a><\/li>\n<li><a href=\"#md-distribution-volume\">Distribution &amp; Volume<\/a><\/li>\n<li><a href=\"#md-entity-mentions-trend\">Entity Mentions Trend<\/a><\/li>\n<li><a href=\"#md-net-sentiment\">Net Sentiment<\/a><\/li>\n<li><a href=\"#md-entity-sentiment-trend\">Entity Sentiment Trend<\/a><\/li>\n<li><a href=\"#md-top-entities\">Top Entities by Mention Volume<\/a><\/li>\n<li><a href=\"#md-entity-word-cloud\">Entity Word Cloud<\/a><\/li>\n<li><a href=\"#md-entities\">Entities<\/a><\/li>\n<\/ul>\n<h3 id=\"md-widget-controls\">Widget controls common to every card<\/h3>\n<p>Before looking at each widget individually, it helps to know the controls that appear on every card:<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity8.gif\" \/><\/p>\n<ul>\n<li><strong>Expand (full-screen) icon<\/strong> \u2014 the diagonal-arrows icon in the top-right of a card opens that widget in a larger, full-screen view.<\/li>\n<li><strong>More options (three-dot) menu<\/strong> \u2014 opens the actions for that widget. Common options are <b>Share<\/b> (share the widget&#8217;s real-time data via a link), <b>Download<\/b> (save the widget), and <b>Refresh Data<\/b> (re-pull the latest data). Charts download as <strong>PDF, PNG, or JPEG<\/strong>; tables download as <strong>Excel (XLSX) or PDF<\/strong>.<\/li>\n<\/ul>\n<p>Some widgets add their own options to this menu (such as Graph Type, Graph Colors, Calculation, Frequency, Scale Range, or Word Count). Those are called out under each widget below.<\/p>\n<h3 id=\"md-unique-entities\">Unique Entities<\/h3>\n<p>Displays the number of distinct entities found in the current scope, along with the percentage change compared to the previous period. The subtitle shows how many entity types those entities span.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity9.png\" \/><\/p>\n<div class=\"example-block-container blue-shade\">\n        <span class=\"example-block-content\"><span class=\"example-block-header\">For example:<\/span>241 distinct entities across 6 types, +102.5% vs prior period tells you your feedback is referencing far more distinct entities than before \u2014 a signal that customers are naming more specifics.<\/span>\n  <\/div>\n<h3 id=\"md-responses-with-entities\">Responses with Entities<\/h3>\n<p>Displays the share of your open-ended responses that contained at least one entity, shown as a percentage with the underlying counts beneath it (for example, 83% \u2014 1,020 of 1,230 open-ended responses). Responses with no entity count toward the total but not the share.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity10.png\" \/><\/p>\n<h3 id=\"md-total-mentions\">Total Mentions<\/h3>\n<p>Displays the total number of times entities were mentioned across all six types, with the change versus the prior period. Where Unique Entities measures variety and Responses with Entities measures coverage, Total Mentions measures raw volume \u2014 the overall amount of entity talk in your feedback (for example, 1,664 mentions, +169.7% vs prior period).<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity11.png\" \/><\/p>\n<p>Each of these three metric cards carries the standard expand icon and three-dot menu (<b>Share<\/b>, <b>Download<\/b>, <b>Refresh Data<\/b>).<\/p>\n<h3 id=\"md-distribution-volume\">Distribution &amp; Volume<\/h3>\n<p>Shows how your entity mentions are divided across the six entity types using a Donut chart. The center of the donut displays your <strong>total mentions<\/strong>; when an entity type filter is applied, the center shows that type&#8217;s mention count and its percentage share instead.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity12.png\" \/><\/p>\n<p>Hovering over a slice reveals the entity type, its mention count, and its percentage share. Clicking a slice applies that entity type as the page filter \u2014 the same as clicking the matching chip.<\/p>\n<p>The three-dot menu adds the following options:<\/p>\n<ul>\n<li><strong>Graph Type<\/strong> \u2014 change the visual style. Available types are Donut, Pie, Bar, Column, Line, and Area.<\/li>\n<li><strong>Graph Colors<\/strong> \u2014 recolor the chart by choosing from a range of color scales.<\/li>\n<li><b>Share<\/b>, <b>Download<\/b> (PDF, PNG, JPEG), and <b>Refresh Data<\/b>.<\/li>\n<\/ul>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity13.png\" \/><\/p>\n<div class=\"example-block-container blue-shade\">\n        <span class=\"example-block-content\"><span class=\"example-block-header\">For example:<\/span>Hovering the largest segment shows Person \u2014 64% (1,068), telling you that people (in this healthcare example, mostly doctors and staff) dominate the conversation. If you expected products to lead, that&#8217;s an immediate cue to look closer at how respondents talk about your team.<\/span>\n  <\/div>\n<h3 id=\"md-entity-mentions-trend\">Entity Mentions Trend<\/h3>\n<p>Displays how mention volume changes over time using a Line chart. What the lines represent depends on your entity type filter:<\/p>\n<ul>\n<li>With <strong>All<\/strong> selected, the chart plots one line per entity type, so you can see which types are being discussed more or less over time.<\/li>\n<li>With a single type selected, the chart plots the top five individual entities of that type.<\/li>\n<\/ul>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity14.png\" \/><\/p>\n<p>Hovering over any data point shows the time period, the series name, and the mention count.<\/p>\n<p>The three-dot menu adds the following options:<\/p>\n<ul>\n<li><strong>Calculation<\/strong> \u2014 <strong>Non-Cumulative<\/strong> (each point represents mentions within that specific time range only; this is the default) or <strong>Cumulative<\/strong> (each point includes all previous data up to that point).<\/li>\n<\/ul>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity15.png\" \/><\/p>\n<ul>\n<li><strong>Graph Type<\/strong> \u2014 Line Chart, Area Chart, or Vertical Stacked Bar.<\/li>\n<\/ul>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity16.png\" \/><\/p>\n<ul>\n<li><strong>Frequency<\/strong> \u2014 set the X-axis to Daily, Weekly, Monthly, Quarterly, or Yearly.<\/li>\n<\/ul>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity17.png\" \/><\/p>\n<ul>\n<li><b>Share<\/b>, <b>Download<\/b> (PDF, PNG, JPEG), and <b>Refresh Data<\/b>.<\/li>\n<\/ul>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity18.png\" \/><\/p>\n<div class=\"example-block-container blue-shade\">\n        <span class=\"example-block-content\"><span class=\"example-block-header\">For example:<\/span>Set <strong>Frequency<\/strong> to Monthly and hover the peak to read Person: 1,068 Mentions, Product: 96, Company: 303, Place: 98, Location: 8, Profession: 91 for that period. If the Person line climbs steadily while others stay flat, staff and individuals are drawing a growing share of attention \u2014 worth investigating in the drill-downs.<\/span>\n  <\/div>\n<h3 id=\"md-net-sentiment\">Net Sentiment<\/h3>\n<p>Summarizes how customers feel about each entity type. With <strong>All<\/strong> selected, the widget lists all six types ranked from the most positive net score to the most negative. Each row shows the entity type, a stacked bar dividing its mentions into Positive, Neutral, and Negative, and its net sentiment score. When an entity type filter is applied, only that type&#8217;s row is shown.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity19.png\" \/><\/p>\n<p>Hovering over a row displays the sentiment score along with the count and percentage of positive, neutral, and negative mentions. A legend below the list identifies the three sentiment colors. This widget&#8217;s three-dot menu offers <b>Share<\/b>, <b>Download<\/b>, and <b>Refresh Data<\/b>.<\/p>\n<div class=\"example-block-container blue-shade\">\n        <span class=\"example-block-content\"><span class=\"example-block-header\">For example:<\/span>Location shows a net score of \u22120.13 while every other type is positive. That single red number tells you geographic mentions carry the most dissatisfaction \u2014 a cue to filter to Location and read what respondents are saying about specific regions or branches.<\/span>\n  <\/div>\n<h3 id=\"md-entity-sentiment-trend\">Entity Sentiment Trend<\/h3>\n<p>Displays how sentiment changes over time. The Y-axis runs from <strong>+1.00<\/strong> at the top to <strong>\u22121.00<\/strong> at the bottom, with the zero line emphasized so you can see when a series crosses from positive to negative.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity20.png\" \/><\/p>\n<p>As with Entity Mentions Trend, the series depend on your filter: <strong>All<\/strong> plots the average sentiment of each of the six entity types, while a single type plots the top five entities of that type.<\/p>\n<p>Hovering over a data point shows the time period, the series name, the sentiment score, and the distribution of positive, neutral, and negative mentions behind it.<\/p>\n<p>The three-dot menu adds <strong>Calculation<\/strong> (Cumulative \/ Non-Cumulative), <strong>Graph Type<\/strong> (Line Chart, Area Chart, Vertical Stacked Bar), <strong>Frequency<\/strong> (Daily, Weekly, Monthly, Quarterly, Yearly), and <strong>Scale Range<\/strong> (Original Range \/ Best Fit), along with <b>Share<\/b>, <b>Download<\/b>, and <b>Refresh Data<\/b>.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity21.png\" \/><\/p>\n<div class=\"example-block-container blue-shade\">\n        <span class=\"example-block-content\"><span class=\"example-block-header\">For example:<\/span>Hovering the week of 03\u201310 Jun 26 reveals Person: 0.25, Product: 0.23, Company: 0.09, Place: 0.03, Location: 0.00, Profession: 0.35. Watching these lines week over week shows whether a dip in sentiment is a one-off or a developing trend.<\/span>\n  <\/div>\n<h3 id=\"md-top-entities\">Top Entities by Mention Volume<\/h3>\n<p>Ranks your 15 most-mentioned entities as a horizontal bar list. Each row shows the entity name with a colored pill indicating its type, a bar showing its Positive \/ Neutral \/ Negative sentiment split, and its mention count. An entity that is both frequently mentioned and negative shows a long bar with a large red segment. A footer beneath the list shows how many entities the top 15 was drawn from (for example, Top 15 of 241 entities shown).<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity22.png\" \/><\/p>\n<p>Hovering a bar shows a sentiment detail such as the negative share and count. Applying an entity type filter re-ranks the list to the top 15 entities of that type. Clicking any row opens the Entity Details view for that entity. The three-dot menu offers <b>Share<\/b>, <b>Download<\/b>, and <b>Refresh Data<\/b>.<\/p>\n<div class=\"example-block-container blue-shade\">\n        <span class=\"example-block-content\"><span class=\"example-block-header\">For example:<\/span>Meridian Health Network tops the list with 79 mentions but carries a visibly large red segment; hovering Dr. Simone Beaumont shows Negative: 32% (10), 31 Mentions. High volume plus high negativity is exactly the combination worth drilling into first.<\/span>\n  <\/div>\n<h3 id=\"md-entity-word-cloud\">Entity Word Cloud<\/h3>\n<p>Displays your most-mentioned entities in a word cloud. The <strong>size<\/strong> of each term corresponds to how often that entity is mentioned, and the <strong>color<\/strong> corresponds to its sentiment. A large red term is both prominent and poorly regarded.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity23.png\" \/><\/p>\n<p>Hovering over a term shows the entity name, its type, its mention count, and its sentiment score. Clicking a term opens the Entity Details view.<\/p>\n<p>The three-dot menu adds a <strong>Word Count<\/strong> option \u2014 set how many entities appear in the cloud by entering a number between <strong>1 and 70<\/strong> \u2014 along with <b>Share<\/b> and <b>Refresh Data<\/b>.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity24.png\" \/><\/p>\n<div class=\"sogo-help-info-box\">\n<i class=\"fal fa-info-circle\"><\/i><\/p>\n<p class=\"sogo-info-text\">An Admin User can also set the default and maximum number of terms in NLP Settings.<\/p>\n<\/div>\n<div class=\"example-block-container blue-shade\">\n        <span class=\"example-block-content\"><span class=\"example-block-header\">For example:<\/span>Meridian Downtown Medical Center appears large and red \u2014 frequently mentioned and negative. Lowering the Word Count to the top 20 strips away the long tail so the few entities that truly drive your feedback stand out.<\/span>\n  <\/div>\n<h3 id=\"md-entities\">Entities<\/h3>\n<p>Lists every entity in the current scope in a table. Use the <strong>Search<\/strong> box above the table to find a specific entity by name.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity25.png\" \/><\/p>\n<div class=\"package-table-container no-scroll\">\n<table class=\"package-limit-table\" border=\"0\" cellpadding=\"0\">\n<thead>\n<tr>\n<td style=\"width: 28%;\">Column<\/td>\n<td style=\"width: 72%;\">Description<\/td>\n<\/tr>\n<\/thead>\n<tbody>\n<tr class=\"column-bg-white-effect\">\n<td style=\"width: 28%;\">Sentiment<\/td>\n<td style=\"width: 72%;\">The entity&#8217;s sentiment score, with an icon and color indicating its bucket (for example, +1.00, \u22120.60).<\/td>\n<\/tr>\n<tr class=\"column-bg-effect\">\n<td style=\"width: 28%;\">Entity<\/td>\n<td style=\"width: 72%;\">The entity name.<\/td>\n<\/tr>\n<tr class=\"column-bg-white-effect\">\n<td style=\"width: 28%;\">Type<\/td>\n<td style=\"width: 72%;\">The entity type, shown as a colored pill.<\/td>\n<\/tr>\n<tr class=\"column-bg-effect\">\n<td style=\"width: 28%;\">Mentions<\/td>\n<td style=\"width: 72%;\">The total number of times the entity was mentioned.<\/td>\n<\/tr>\n<tr class=\"column-bg-white-effect\">\n<td style=\"width: 28%;\">Unique<\/td>\n<td style=\"width: 72%;\">The number of separate responses the entity appears in.<\/td>\n<\/tr>\n<tr class=\"column-bg-effect\">\n<td style=\"width: 28%;\">Negative Shares<\/td>\n<td style=\"width: 72%;\">The share of the entity&#8217;s mentions that were negative, shown as a bar and a percentage.<\/td>\n<\/tr>\n<tr class=\"column-bg-white-effect\">\n<td style=\"width: 28%;\">Trend vs Prior<\/td>\n<td style=\"width: 72%;\">Whether mentions rose, fell, or held steady against the previous period, with the percentage change and a direction arrow.<\/td>\n<\/tr>\n<tr class=\"column-bg-effect\">\n<td style=\"width: 28%;\">Last Mentioned<\/td>\n<td style=\"width: 72%;\">How long ago the entity was most recently mentioned (for example, 24 days ago).<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><strong>Actions available on the table:<\/strong><\/p>\n<ul>\n<li><strong>Search entities<\/strong> \u2014 find a specific entity by name.<\/li>\n<li><strong>Sort<\/strong> \u2014 every column is sortable; click a column header to sort ascending or descending.<\/li>\n<li><strong>Expand a row<\/strong> \u2014 the chevron at the start of each row expands it to reveal up to three representative phrases: the verbatim snippets where that entity appears.<\/li>\n<li><strong>Column Options<\/strong> \u2014 the columns icon lets you choose which columns are displayed.<\/li>\n<li><strong>Paging<\/strong> \u2014 the table displays up to 100 entities per page; use the paging control at the bottom-right to move through the list.<\/li>\n<li><strong>Download<\/strong> \u2014 the three-dot menu offers <b>Download \u2192 As PDF<\/b> or <b>As XLSX<\/b> (Excel), plus <b>Share<\/b> and <b>Refresh Data<\/b>.<\/li>\n<li><strong>Drill down<\/strong> \u2014 clicking the entity name opens the Entity Details view.<\/li>\n<\/ul>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity26.gif\" \/><\/p>\n<hr \/>\n<h2 id=\"section-4\">Entity Details<\/h2>\n<p>Clicking an entity anywhere on the tab \u2014 a bar in Top Entities by Mention Volume, a term in the Entity Word Cloud, or an entity name in the Entities table \u2014 opens a full-screen view for that entity. The header states how many times the entity was mentioned in the current date range, along with its sentiment breakdown.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity27.png\" \/><\/p>\n<p>Below the header are three sections:<\/p>\n<ul>\n<li><strong>Summary<\/strong> \u2014 a chart dividing the entity&#8217;s mentions into Positive, Neutral, and Negative, alongside its Total Mentions, Unique Responses, and Average Sentiment. Hovering shows the percentage and count for each segment. The three-dot menu offers <strong>Graph Type<\/strong> (Donut, Pie, Bar, Column, Line, Area), <b>Share<\/b>, <b>Download<\/b> (PDF, PNG, JPEG), and <b>Refresh Data<\/b>.<\/li>\n<li><strong>Sentiment Trend<\/strong> \u2014 a dual-axis graph plotting <strong>mentions<\/strong> (left axis) against <strong>average sentiment<\/strong> (right axis) over time. The three-dot menu offers <strong>Frequency<\/strong>, <strong>Calculation<\/strong>, <strong>Graph Type<\/strong>, <b>Share<\/b>, <b>Download<\/b>, and <b>Refresh Data<\/b>.<\/li>\n<li><strong>Responses<\/strong> \u2014 a table of every response in which the entity was mentioned, showing the <strong>Response sentiment<\/strong>, the <strong>Answer text<\/strong> with the entity occurrence highlighted, and the <strong>Entity<\/strong>, <strong>Theme<\/strong>, <strong>Emotion<\/strong>, <strong>Topic<\/strong>, and <strong>Received<\/strong> columns. All columns are sortable, and <strong>Column Options<\/strong> controls which are shown. The table displays up to 100 responses per page and can be downloaded as an <strong>Excel<\/strong> or a <strong>PPT<\/strong> file.<\/li>\n<\/ul>\n<p>Use the search box, date range, sentiment, and respondent controls above the Responses table to narrow the list further; these apply on top of the filters already active on the tab. Close the view using the <strong>X<\/strong> in the header, the <strong>Close<\/strong> button, the <strong>Esc<\/strong> key, or by clicking outside it.<\/p>\n<div class=\"example-block-container blue-shade\">\n        <span class=\"example-block-content\"><span class=\"example-block-header\">For example:<\/span>Opening Dr. Arjun Malhotra (52 mentions; 29% positive, 44% neutral, 27% negative) lets you read each of those 52 comments in context in the Responses table \u2014 turning a single sentiment score into specific, actionable feedback.<\/span>\n  <\/div>\n<hr \/>\n<h2 id=\"section-5\">Filters<\/h2>\n<p>The controls at the top of the tab let you refine the data shown across every Metadata widget. Filters applied to a tab affect only the widgets within that tab and do not carry over to other tabs.<\/p>\n<h3 id=\"md-all-time\">All Time Filter<\/h3>\n<p>The date range filter, in the top-left corner, controls the time period for which data is displayed. By default it is set to <strong>All Time<\/strong>, and the exact date range currently in effect is shown next to the dropdown. Click the dropdown to choose from preset ranges: All Time, Today, Yesterday, Last 24 hours, This Week, Last Week, Last 7 days, Last 14 days, This Month, Last Month, Last 30 days, This Quarter, Last Quarter, Last 90 days, Last 180 days, This Year, Last Year, Last 12 Months, and Custom (define your own start and end dates).<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity28.png\" \/><\/p>\n<div class=\"sogo-help-info-box\">\n<i class=\"fal fa-info-circle\"><\/i><\/p>\n<p class=\"sogo-info-text\">Entities are placed on the trend charts by the date the response was <strong>received<\/strong>, not the date it was analyzed. Reprocessing older responses will not shift your historical trends.<\/p>\n<\/div>\n<h3 id=\"md-metric-filter\">Metric Filter<\/h3>\n<p>The metric filter, labeled <strong>All Metric<\/strong> by default, lets you focus the analysis on entities extracted from specific metric (follow-up) questions in your project. Click the dropdown to see the full list of available metrics, each with a checkbox. Use the checkboxes to include or exclude individual metrics, or select <strong>Select All<\/strong> to choose every metric at once. <strong>Apply<\/strong> saves your selection and updates all widgets on the tab; <strong>Reset<\/strong> clears your current selection and returns the filter to its default state.<\/p>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity29.png\" \/><\/p>\n<h3 id=\"md-create-filters\">Create Filters (including the new Entity criteria)<\/h3>\n<p>The <strong>filter icon<\/strong> in the top-right corner opens the <strong>Create Filters<\/strong> panel, where you can build more detailed conditions to segment your feedback. You can filter by Projects, Activity Fields, Respondents, Feedback Recipients, Tags, and <strong>NLP Conditions<\/strong>. Applied filters remain active until you remove or change them.<\/p>\n<p>A new <strong>Entity<\/strong> criteria has been added under <strong>NLP Conditions<\/strong> (alongside Sentiment, Theme, Topic, Emotion, Language, and Metrics). To filter by entity:<\/p>\n<ul>\n<li>Open <strong>Create Filters<\/strong> and choose <strong>NLP Conditions<\/strong>.<\/li>\n<\/ul>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity30.gif\" \/><\/p>\n<ul>\n<li>Set the condition (for example, is one of the following) and select <strong>Entity<\/strong> as the criteria.<\/li>\n<\/ul>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity31.png\" \/><\/p>\n<ul>\n<li>Optionally choose the open-ended (follow-up) question to base the filter on, just as you would for other NLP filters.<\/li>\n<\/ul>\n<p><img decoding=\"async\" src=\"\/help\/wp-content\/uploads\/entity32.png\" \/><\/p>\n<ul>\n<li>Select the specific entities you want to filter by, then apply the filter to update the dashboard.<\/li>\n<\/ul>\n<div class=\"sogo-help-info-box\">\n<i class=\"fal fa-info-circle\"><\/i><\/p>\n<p class=\"sogo-info-text\">If a filter combination returns no entities, the widgets display an empty state rather than an error. Clear or widen the filter to bring the data back.<\/p>\n<\/div>\n<div class=\"example-block-container blue-shade\">\n        <span class=\"example-block-content\"><span class=\"example-block-header\">For example:<\/span>Filter by <strong>Entity is one of the following: \u201cMeridian Health Network\u201d<\/strong> to reshape every widget around just that company \u2014 its mention trend, its sentiment over time, and the individual comments that name it \u2014 so you can build a focused view of one organization from thousands of responses.<\/span>\n  <\/div>\n<hr \/>\n<h2 id=\"section-6\">Good to Know<\/h2>\n<ul>\n<li>Entities are extracted from <strong>follow-up questions only<\/strong>. Open-ended questions that are not marked as follow-up are not analyzed on the Dashboard.<\/li>\n<li>Entity sentiment reflects the sentiment of the text surrounding the mention, not a judgment of the entity itself. An entity named in a negative response carries a negative score even when the complaint is not about that entity.<\/li>\n<li>Different spellings of the same entity \u2014 for example, SogoAssist, Sogo Assist, and Sogoassist \u2014 are recognized as the same entity and counted together.<\/li>\n<li>An Admin User can <strong>merge<\/strong> entities under a single display name, <strong>hide<\/strong> (exclude) entities from all reporting, and set the word cloud&#8217;s term limit in NLP Settings.<\/li>\n<li>Person entities are subject to your account&#8217;s existing <strong>PII redaction<\/strong> setting. Redaction is applied before analysis, so any name your account redacts is not surfaced here.<\/li>\n<li>Entities now also appear in <strong>response tables<\/strong> (alongside Emotions and Themes) across the areas where NLP filtering works \u2014 CX Dashboard, EX Dashboard, and NLP Report. You can show or hide the entity column using Column Options. (Entities are not added to the Text Classification tab.)<\/li>\n<li>A common <strong>\u201cfilter by entity\u201d<\/strong> option is available across all modules where the NLP filter works.<\/li>\n<li>The Metadata tab is included in <strong>Share<\/strong>, so it can be shared like other dashboard tabs. Navigation and chatbot functionality are supported.<\/li>\n<li><strong>Exports:<\/strong> charts download as PNG, JPEG, or PDF; the Entities table downloads as Excel (XLSX) or PDF; the Entity Details Responses table downloads as Excel or PPT.<\/li>\n<li>There is <strong>no change to packaging<\/strong> for this feature.<\/li>\n<\/ul>\n<hr \/>\n<h2 id=\"section-8\">See also<\/h2>\n<ul>\n<li><a href=\"\/help\/cx-dashboard-text-analysis\/\" rel=\"noopener\">CX Dashboard: Text Analysis<\/a><\/li>\n<li><a href=\"\/help\/nlp_settings_in_cx\/\" rel=\"noopener\">NLP Settings in CX Dashboard<\/a><\/li>\n<li><a href=\"\/help\/ex-dashboard-metadata\/\" rel=\"noopener\">EX Dashboard: Metadata<\/a><\/li>\n<li><a href=\"\/help\/data-visibility-permissions\/\" rel=\"noopener\">Data Visibility Permissions<\/a><\/li>\n<\/ul>\n<p><!-- =========================================================\n     Definition of right side asset:\n     ========================================================= --><\/p>\n","protected":false},"excerpt":{"rendered":"<p>The Metadata tab in Text Analysis, powered by OpenAI, identifies the specific people, products, companies, and places your customers name in their open-ended responses, and shows how often each one is mentioned and how customers feel about it. It turns unstructured, qualitative feedback into structured metadata you can rank, track, filter, and drill into \u2014 so you can analyze at scale instead of reading every comment by hand. Themes tell you what topics customers are discussing. Metadata tells you who and what they are naming inside those discussions. If a customer writes \u201cthe billing team took three days to respond,\u201d Theme Analysis identifies the theme Response Time; Metadata identifies the entity Billing Team and records the sentiment attached to it. Over hundreds of responses, this lets you rank the entities driving your feedback and drill into the verbatims behind each one. Entities are extracted from follow-up (open-ended) questions only. The sentiment shown for an entity reflects the sentiment of the surrounding response in which it was mentioned, not a direct rating of the entity itself. Where to Find Go to CX Dashboard &gt; Report &gt; Text Analysis &gt; Metadata. Metadata is the fourth tab in Text Analysis, alongside Open-Ended Responses, Theme Analysis, and Text Classification. Before you begin Metadata is available on Pro packages and above, under the same Text Analytics entitlement that covers Open-Ended Responses and Theme Analysis. There is no separate packaging change for this tab. Your account must have accepted the OpenAI Terms. If consent has not [&hellip;]<\/p>\n","protected":false},"author":30,"featured_media":0,"parent":0,"menu_order":680,"comment_status":"closed","ping_status":"closed","template":"faq_php.php","meta":{"_acf_changed":false,"footnotes":""},"categories":[530],"tags":[531,532,533,515,510,482,388],"class_list":["post-67091","page","type-page","status-publish","hentry","category-cx-dashboard-text-analysis-new","tag-metadata","tag-entity","tag-analysis","tag-open-ended","tag-text-analysis","tag-nlp","tag-cx"],"acf":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v28.1 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Metadata Analysis on the CX Dashboard - Sogolytics<\/title>\n<meta name=\"description\" content=\"The Metadata tab identifies the people, products, companies, and places customers name in open-ended responses, with mention counts, sentiment scores, and drill-downs.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.sogolytics.com\/help\/entity-analysis-metadata\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Metadata Analysis on the CX Dashboard - Sogolytics\" \/>\n<meta property=\"og:description\" content=\"The Metadata tab identifies the people, products, companies, and places customers name in open-ended responses, with mention counts, sentiment scores, and drill-downs.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.sogolytics.com\/help\/entity-analysis-metadata\/\" \/>\n<meta property=\"og:site_name\" content=\"Sogolytics\" \/>\n<meta property=\"article:modified_time\" content=\"2026-08-20T23:51:28+00:00\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data1\" content=\"22 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/www.sogolytics.com\\\/help\\\/entity-analysis-metadata\\\/\",\"url\":\"https:\\\/\\\/www.sogolytics.com\\\/help\\\/entity-analysis-metadata\\\/\",\"name\":\"Metadata Analysis on the CX Dashboard - Sogolytics\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.sogolytics.com\\\/help\\\/#website\"},\"datePublished\":\"2026-08-20T18:28:58+00:00\",\"dateModified\":\"2026-08-20T23:51:28+00:00\",\"description\":\"The Metadata tab identifies the people, products, companies, and places customers name in open-ended responses, with mention counts, sentiment scores, and drill-downs.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/www.sogolytics.com\\\/help\\\/entity-analysis-metadata\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/www.sogolytics.com\\\/help\\\/entity-analysis-metadata\\\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/www.sogolytics.com\\\/help\\\/entity-analysis-metadata\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/www.sogolytics.com\\\/help\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"CX Dashboard: Metadata Analysis\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/www.sogolytics.com\\\/help\\\/#website\",\"url\":\"https:\\\/\\\/www.sogolytics.com\\\/help\\\/\",\"name\":\"Sogolytics\",\"description\":\"Sogolytics - Online Survey Tool\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/www.sogolytics.com\\\/help\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Metadata Analysis on the CX Dashboard - Sogolytics","description":"The Metadata tab identifies the people, products, companies, and places customers name in open-ended responses, with mention counts, sentiment scores, and drill-downs.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.sogolytics.com\/help\/entity-analysis-metadata\/","og_locale":"en_US","og_type":"article","og_title":"Metadata Analysis on the CX Dashboard - Sogolytics","og_description":"The Metadata tab identifies the people, products, companies, and places customers name in open-ended responses, with mention counts, sentiment scores, and drill-downs.","og_url":"https:\/\/www.sogolytics.com\/help\/entity-analysis-metadata\/","og_site_name":"Sogolytics","article_modified_time":"2026-08-20T23:51:28+00:00","twitter_card":"summary_large_image","twitter_misc":{"Est. reading time":"22 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.sogolytics.com\/help\/entity-analysis-metadata\/","url":"https:\/\/www.sogolytics.com\/help\/entity-analysis-metadata\/","name":"Metadata Analysis on the CX Dashboard - Sogolytics","isPartOf":{"@id":"https:\/\/www.sogolytics.com\/help\/#website"},"datePublished":"2026-08-20T18:28:58+00:00","dateModified":"2026-08-20T23:51:28+00:00","description":"The Metadata tab identifies the people, products, companies, and places customers name in open-ended responses, with mention counts, sentiment scores, and drill-downs.","breadcrumb":{"@id":"https:\/\/www.sogolytics.com\/help\/entity-analysis-metadata\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.sogolytics.com\/help\/entity-analysis-metadata\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.sogolytics.com\/help\/entity-analysis-metadata\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.sogolytics.com\/help\/"},{"@type":"ListItem","position":2,"name":"CX Dashboard: Metadata Analysis"}]},{"@type":"WebSite","@id":"https:\/\/www.sogolytics.com\/help\/#website","url":"https:\/\/www.sogolytics.com\/help\/","name":"Sogolytics","description":"Sogolytics - Online Survey Tool","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.sogolytics.com\/help\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"}]}},"seo_meta_info":{"yoast_noindex":"","rankmath_noindex":"","yoast_title":"Metadata Analysis on the CX Dashboard - Sogolytics","yoast_desc":"The Metadata tab identifies the people, products, companies, and places customers name in open-ended responses, with mention counts, sentiment scores, and drill-downs.","rankmath_title":"CX Dashboard: Metadata Analysis","rankmath_desc":"\/* Offset anchor jumps so headings aren't hidden behind the fixed help header *\/ h2[id], h3[id] { scroll-margin-top: 100px; } The Metadata tab in Text&hellip;"},"_links":{"self":[{"href":"https:\/\/www.sogolytics.com\/help\/wp-json\/wp\/v2\/pages\/67091","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.sogolytics.com\/help\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.sogolytics.com\/help\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.sogolytics.com\/help\/wp-json\/wp\/v2\/users\/30"}],"replies":[{"embeddable":true,"href":"https:\/\/www.sogolytics.com\/help\/wp-json\/wp\/v2\/comments?post=67091"}],"version-history":[{"count":6,"href":"https:\/\/www.sogolytics.com\/help\/wp-json\/wp\/v2\/pages\/67091\/revisions"}],"predecessor-version":[{"id":67133,"href":"https:\/\/www.sogolytics.com\/help\/wp-json\/wp\/v2\/pages\/67091\/revisions\/67133"}],"wp:attachment":[{"href":"https:\/\/www.sogolytics.com\/help\/wp-json\/wp\/v2\/media?parent=67091"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.sogolytics.com\/help\/wp-json\/wp\/v2\/categories?post=67091"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.sogolytics.com\/help\/wp-json\/wp\/v2\/tags?post=67091"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}