Author Analysis

The Authors page is your dedicated workspace for exploring and analyzing the social media accounts that are active within your project's dataset. Whether you are investigating bot-like behavior, profiling cohort membership, or identifying the most engaged voices in a narrative, this page provides the author-level granularity you need.

"Authors" refers to any unique account handle ingested into your project — across social media platforms — displayed in a unified, filterable table view.

• Monitor high-level author metrics at a glance using the top-level Stats Bar
• Explore author data across your project with a sortable, customizable data grid
• Search for specific authors by handle using the inline search bar
• Filter the dataset using the Segment Bar's boolean search and filter capabilities
• Customize your view by adding, removing, and reordering columns to suit your workflow
• Investigate cohort membership in detail by expanding the full cohorts list for any author
• Identify bot-like accounts using dedicated risk signal columns

1. Open your project in Constellation
2. From the left navigation menu, select Authors
3. The page will load with your project's author data, applying any active Segment Bar filters

The Authors page is organized into two main sections:

1. Detailed Stats Widget — Two side-by-side summary panels providing post, engagement, platform, and author metrics for the currently filtered dataset
2. Segment Overlay Search — A secondary search bar for applying an additional boolean query on top of the active segment
3. Authors Table — A searchable, sortable, and customizable grid of individual author records

The Detailed Stats Widget appears at the top of the Post Analysis page and provides instant, filter-aware metrics across two summary panels displayed side by side. All values update dynamically as you apply or remove filters.

For full details on the Detailed Stats Widget, including metric definitions and the breakdown dialog, see the Detailed Stats Widget user guide.

Pro Tip: Click either summary panel to open a detailed breakdown dialog with platform-by-platform and author-level metrics.

Note: All metrics in the Detailed Stats Widget reflect the currently filtered dataset, not the full project. Check for active filter pills in the Segment Bar if numbers appear lower than expected.

The Authors Table is the core of the Authors page. It displays individual author records in a structured, sortable, and customizable grid.

A search bar is located above the table. Enter any partial or full author handle to filter the table to matching results. The search performs a "contains" match — any author handle that includes your search term will be returned.

The following columns can be added to the Authors Table via the Column Selector:

To add or remove columns:

1. Click the Columns button located in the top-right area of the Authors Table
2. A panel will appear listing all available columns with checkboxes
3. Check or uncheck columns to show or hide them

To reorder columns:

1. Click and drag any column header to a new position in the table

Note: Column configurations are not persisted across sessions. You will need to reconfigure your column layout each time you return to the Authors page.

Click any column header to sort the table by that column in ascending order. Click the same header again to toggle to descending order. Sorting is available on all visible columns.

When an author belongs to more than 3 cohorts, the Cohorts column displays the first 3 along with a +N more link.

To view the full cohort list:

1. Click the +N more link in the Cohorts column for the relevant author
2. A modal will open displaying all cohorts associated with that author, sorted alphabetically

The Segment Bar at the top of the page is a powerful tool for narrowing the author data displayed in the table. All filtering occurs in real time — the Stats Bar and the Authors Table update immediately as you apply filters.

Use the query field to search using boolean operators:

Important: Boolean operators must be typed in UPPERCASE to be recognized.

The Segment Overlay Search bar appears below the Detailed Stats Bar and allows you to apply a second boolean query on top of your active segment without modifying it. Enter a query and click Apply to layer the search on the current segment. The overlay query is independent of the Segment Bar and can be cleared separately.

• Date Range — Limit the dataset to a specific time window using presets or a custom range
• Platform — Filter by one or more social media platforms
• Risk Signals — Filter for authors flagged as Bot-like, Toxic, or Anomalous
• Cohorts — Filter by one or more author cohort classifications
• Additional Filters — Depending on your project configuration, additional filters such as engagement ranges or entity filters may be available

You can save your current filter combination as a Segment for quick reuse in future sessions. See the Segment Bar User Guide for full details.

• Start with the Stats Bar — Review the high-level author, post, and engagement counts before diving into individual rows to confirm your filters are scoped correctly
• Use Risk Signal columns together — Add Bot-like, Toxic, and Anomalous columns simultaneously to quickly triage accounts of interest during threat investigations
• Save effective filter combinations — When you find a filter set that surfaces relevant authors, save it as a Segment for fast retrieval in future sessions

• Add both location columns for geographic analysis — Profile Location and Detected Location together give you a more complete picture than either alone
• Focus your view — Remove columns you do not need for a given workflow to reduce visual noise and keep the table scannable

• Filter first — Use the Segment Bar to narrow your dataset before analyzing individual author rows; this keeps the table performant and results focused
• Sort by risk signal columns — Sorting by Bot-like or Anomalous flags quickly surfaces the highest-priority accounts for review

• Check your active Segment Bar filters — they may be too restrictive or applied to a field with no matching data
• Verify that your date range includes the expected time window
• Confirm the correct project is selected in the project selector

• The Bot-like flag is only set for accounts that have been evaluated and flagged by Blackbird AI's detection models — not all authors will have this signal present

• Cohort classifications are assigned based on project-specific configurations — an empty Cohorts field means the author has not been assigned to any cohort within your current project

• Profile Location is sourced directly from the author's social media profile and may be blank if the author has not provided a location

• Estimated Location requires Actor Intelligence:

  • Actor Intelligence currently supports X. Additional platforms will be supported in the future.
  • A minimum number of posts is required to estimate location.

The Authors page gives analysts a focused, filter-aware view of every account active in their project dataset:

• Detailed Stats for instant, filter-responsive metrics including author count, engagement, top cohorts, and bot-like author totals
• Authors Table for a fully sortable, customizable grid with columns spanning account details, engagement, cohort membership, and risk signals
• Segment Bar for powerful boolean search and multi-dimensional filtering across the full author dataset