Mentions
Hashtags, cashtags, linked sites, and handles found in matching posts, ranked on the Mentions tab.
The Mentions tab of the Social Intelligence dashboard ranks hashtags, cashtags, link domains, and account handles found in matching posts. The query, time range, and topic filter determine which posts contribute. Getting Started explains those controls.
Each count is the number of matching posts containing an entry. A post that repeats a hashtag five times contributes one to that hashtag's count. A post containing several different hashtags contributes to each of their counts, so adding the rows can produce a total larger than the number of posts. Several posts by the same author count separately.
Mention categories
The four buttons above the results change the category shown in both the visual card and the table.
| Category | What the entries represent |
|---|---|
| Hashtags | Tags carried by the posts, such as #climate or #energy. |
| Cashtags | Symbols prefixed with a dollar sign, such as $AAPL or $BTC. |
| Links | Links in the posts, grouped by site, such as youtube.com. Several pages on the same site count as one entry. |
| Handles | Accounts explicitly mentioned in post text, such as @nytimes.com. These counts describe references to an account; posts it writes contribute only if they also contain the mention. |
These categories identify specific parts of a post. Ordinary names and phrases do not receive a separate ranking here. The Recent posts feed below the results supplies the surrounding text and links, which can help explain why an entry appears. A count alone does not establish whether the author endorses what they mention or link to.
The category choice changes what you can examine within the same discussion. Hashtags show how posts are tagged, while cashtags identify the stock and crypto symbols they carry. The Links ranking can help identify sites supplying articles or videos to a discussion. Reading the posts behind a Handles entry shows whether an account is being addressed, credited, or discussed.
Single-query rankings
With one submitted query, Mentions shows an At a glance card followed by a Top table for the selected category. For example, choosing Hashtags displays Hashtags at a glance and Top hashtags. With several queries, Select opens this same layout for the query chosen from its dropdown.
Word cloud and treemap
Word cloud gives entries with higher post counts larger text. The sizes are relative to the other entries in that cloud, so a large word in one query's cloud cannot be compared directly with a large word in another. Colors distinguish entries visually and carry no sentiment meaning. Some entries may be omitted when there is not enough room to place them.
Treemap gives each entry a rectangle whose area is proportional to its post count. Hovering over a tile shows the count. The full area represents the sum of the displayed entries' counts. Because one post can contain several entries, a tile occupying half the treemap does not mean that half the matching posts contain that entry.
Both views use the leading entries in the selected category. Less frequent entries can fall outside the ranking, and posts containing none of the selected category contribute no entries. The table below shows the displayed ranking's counts in its Total column, from highest to lowest.
Clicking a word or tile adds its text to the query bar. You can review the revised query in Write queries before using Analyze to update the results. The table provides further actions for refining or comparing queries.
Refining queries with an entry
Each row's actions menu offers three ways to use an entry when viewing a single query. The following example starts with the query climate and uses the hashtag #energy.
| Action | Change to the query bar |
|---|---|
| Append to query | Changes climate to climate #energy. |
| Add as new query | Keeps climate and adds #energy as a separate query. |
| Append & add as new query | Keeps climate and adds climate #energy as a separate query. |
After Append to query, the next analysis requires posts to match the original query and the added term. In the example, every result must contain both “climate” and the hashtag #energy.
Add as new query lets the entry stand on its own. Its results can include posts unrelated to the original subject. Append & add as new query keeps the original results available for comparison with the narrower set. With climate and climate #energy submitted together, the second query's posts are also part of the first query's results. Their counts overlap.
These actions change the draft in the query bar. You can review or edit the terms in Write queries and submit them with Analyze. The results continue to describe the submitted queries until then. In Select, the action uses the selected query. The two actions that add a query are disabled when the query bar has reached its comparison limit.
The added entry becomes a text condition using the same syntax as a typed query. Adding a domain such as youtube.com matches that domain in post text, which can include a link or an ordinary textual reference. Adding a handle requires a mention in the text. The Query Syntax page explains these matches and how to combine alternatives, required terms, and exclusions.
Following a narrower discussion
Refinements can build on one another. Suppose a query for electric vehicles returns #tesla among its leading hashtags. Appending the hashtag and submitting electric vehicles #tesla recalculates the rankings from posts that also carry that tag. Choosing Links then shows the domains shared within that narrower result, and Handles shows the accounts mentioned there.
The revised query applies across the dashboard. Switching to Sentiment or Emotions shows the classifications for those same posts, while Volume shows their activity over time. Append & add as new query preserves the broader query alongside the refinement when the difference between them is the subject of the analysis.
Comparing rankings across queries
With several submitted queries, Compare shows a combined word cloud and a table with one count column per query. Each query retains its display name and color. Treemap is available in Select and Aggregate; Compare uses the word cloud.
The cloud takes leading entries from each query and colors them by their source query. A shared term can appear more than once, with a separate size and color for each query. The cloud has limited space and uses a shorter selection than the table. Seeing a term in only one color does not establish that it occurs only in that query's posts.
Comparison table
Rows are ordered by the sum of their counts across the query columns shown. A post matching two queries contributes to both columns, so that sum can count the same post more than once.
A dash means the entry was absent from that query's returned ranking. It may have no matches, or it may fall below the leading entries selected for that query. A dash therefore cannot establish that a term is exclusive to another query.
The selector above the table controls which queries appear as columns. Changing the selection also changes the sums used to order the rows. On a narrow screen, the table initially shows the first two queries; the selector lets you change that selection. At least two selected queries are needed to display the comparison table. Select above the results opens an individual query's ranking.
The All and Shared controls apply to the columns currently shown:
- All shows every entry present in at least one selected query's ranking.
- Shared keeps entries present in at least two selected queries' rankings. An entry can qualify without appearing in every selected query.
Changing the table's selected columns leaves the word cloud and submitted queries unchanged. Clicking a word in the cloud or a row's Append to all queries button adds the entry to every query in the bar, including queries whose table columns are hidden. The results update when you submit the edited queries with Analyze in Write queries.
Combining queries with Aggregate
Aggregate produces one ranking across posts matching any submitted query. A shared post is counted once for each entry it contains, even if the post matches several queries. This is useful when the queries represent alternative spellings or related subjects and you want their combined post counts.
Aggregate uses the same word cloud, treemap, and Total table as the single-query view. Clicking a visual entry or a row's Append to all queries button adds the entry to every query in the bar. When you submit those edits with Analyze in Write queries, the combined result includes posts matching at least one edited query.