> For the complete documentation index, see [llms.txt](https://docs.nested-knowledge.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.nested-knowledge.com/review-workflow/tagging/match-tags-to-your-exports.md). # Match Tags to your Exports Learn how to match tags to your export spreadsheets. **Can I upload a data extraction spreadsheet and have AI extract my studies into it?** Not currently. Nested Knowledge doesn't yet offer the ability to turn a spreadsheet's columns into tags. You build a tag hierarchy that matches your spreadsheet first, either manually or from a template. Once it exists, [**Adaptive Smart Tags**](/review-workflow/tagging/ai-tools-in-tagging/adaptive-smart-tags.md) can extract your studies into it, and your export comes out with the same columns. Your tag hierarchy is the structure of your export: each tag becomes a column or set of columns, and each study becomes a row. Matching the two before you start extracting avoids rebuilding the hierarchy and re-checking studies later.

Tags represented as column headers in a spreadsheet

### Recommended Approach ```mermaid flowchart LR A[Design your export table] --> B[Build it in the Tag Hierarchy] B --> C[Extract data in Tagging] C --> D[Check the live Export / Dashboard] D -.->|refine columns or rows| B ``` {% hint style="success" %} Save a finished hierarchy as an [organizational template](/account-management/organization-management/organization-settings/organizational-templates.md) for future reviews. Build a matching table in [Dashboard](/review-workflow/synthesis/set-up-synthesis-outputs/dashboard-editor.md) as soon as your tags exist. It updates as included studies are tagged. {% endhint %}

Can I use an AI chatbot to plan my tag hierarchy from my spreadsheet?

Yes. A general AI chatbot, such as Claude or ChatGPT, can read your data extraction spreadsheet and propose a hierarchy, tag types, and AST questions. You still build the hierarchy in Nested Knowledge yourself and review every question before extraction. Check your organization's policy before uploading client or confidential spreadsheets to an external tool. Copy the prompt below, fill in the bracketed parts, and attach your spreadsheet. > I'm building a tag hierarchy in Nested Knowledge (NK) so its AI extraction tool, Adaptive Smart Tags (ASTs), can extract study data into the attached data extraction spreadsheet. NK can't import a spreadsheet as a hierarchy, so I need a plan I can build by hand. > > First, read these NK documentation pages: > > * Configure Tagging (tag types and how they export): > * Match Tags to your Exports: > * Adaptive Smart Tags: > * Prompt Guidance (how to write AST questions): > > Then, for each of these sheets: \[list sheets, or "every sheet except ..."] > > 1. Propose a root tag. > 2. Map each column to one tag type: Text, Numeric, Text Options, Select with child tags, Table, or Index Table. Use an Index Table when several groups of columns share the same identifying columns, such as arm or intervention: name the index columns and group the rest into sub-tables of related columns. Use a Table for repeating rows of one kind with no shared identifying columns. > 3. Write the AST question for every tag, and a header that stands on its own for every table column. Follow Prompt Guidance: define the concept, state the units or summary statistic, and say what to return when data isn't reported. > 4. Show the hierarchy as an indented outline, in the order tags should appear. > > Ask me before assuming anything the spreadsheet doesn't make clear. Plan the hierarchy only; don't extract data.

### 1. Build Your Tag Hierarchy ### Match Columns To Tag Types Choose a tag type for every column in your spreadsheet. Each section below shows how the tag looks in your hierarchy and how it looks in your export. All tag types can be combined. For setup steps, see [**Configure Tagging**.](/review-workflow/tagging/configure-tagging.md) #### Text For an open-ended answer with one value per study, such as study objective. This is the default for every tag.

Study Objective is shown as an example text tag configured in the hierarchy and nested under a Study Characteristics tag. — Text Tag: Study Objective configured in the hierarchy

An export table is shown with columns: Title, Author, and Study Objective. — Text Tag: Study Objective displayed as a column in the Dashboard Export table

Configure a [Text Tag.](/review-workflow/tagging/configure-tagging.md#text-default) #### Numeric For a single number you want to sort or chart, such as total sample size. Use a separate tag per statistic for values reported in different ways, such as mean and median age.

Sample Size is shown as an example numeric tag configured in the hierarchy and nested under a Study Characteristics tag. — Numeric Tag: Sample Size configured in the hierarchy

An export table is shown with columns: Title, Author, and Sample Size. — Numeric Tag: Sample Size displayed as a column in the Dashboard Export table

Configure a [Numeric Tag.](/review-workflow/tagging/configure-tagging.md#numeric) #### Text Options For one answer from a short, fixed list, such as study design. It exports as a single column, which is usually the closest match to an existing spreadsheet.

Study Design is shown as an example text options tag configured in the hierarchy and nested under a Study Characteristics tag. — Text Options Tag: Study Design configured in the hierarchy

An export table is shown with columns: Title, Author, and Study Design. — Text Options Tag: Study Design displayed as a column in the Dashboard Export table

Configure a [Text Options Tag.](/review-workflow/tagging/configure-tagging.md#text-options) #### Select With Child Tags For a column where a study can have more than one answer, or where each answer needs its own evidence or follow-up questions, such as types of safety outcomes reported. Turn on **Multiple Responses** when more than one answer can apply.

Various efficacy outcomes are shown as an example select tag configured in the hierarchy. — Select Tag: Safety Outcomes and child tags configured in the hierarchy

An export table is shown with columns: Title, Author, and Safety Outcomes. — Select Tag: Safety Outcomes displayed as a column with child tags and annotations in the Dashboard Export table

Configure a [Select Tag.](/review-workflow/tagging/configure-tagging.md#choose-a-question-type) #### Table For a repeating group of columns where you can't predict how many rows a study needs, such as adverse events, outcomes, or timepoints. Rows are added as data is extracted, with no fixed limit.

Adverse events is shown as an example table tag configured in the hierarchy. — Table Tag: Adverse Events configured in the hierarchy

An export table is shown with columns: Title, Author, and Adverse Events. — Table Tag: Adverse Events displayed with configured columns in the Dashboard Export table

Configure a [Table Tag.](/review-workflow/tagging/configure-tagging.md#table) #### Index Table For several groups of columns that share the same identifying columns, such as baseline demographics and comorbidities reported by arm and dose. The shared columns form the **index**, and each group becomes a **sub-table** against it, so the arm is entered once. Set up every index row before extracting; rows can't be added mid-extraction.

Safety is shown as an example index table tag configured in the hierarchy. — Index Table Tag: Safety configured in the hierarchy

Configure an [Index Table Tag.](/review-workflow/tagging/configure-tagging.md#index-table) {% hint style="info" %} **Unsure whether to split your data?** Each tag generates its own [Smart Insight](/review-workflow/synthesis/set-up-synthesis-outputs/smart-insights.md), and separate tables can always be combined in your export. {% endhint %} ### 2. Pair With Adaptive Smart Tags The tag type decides the shape of your export. The question on each tag decides what [**Adaptive Smart Tags**](/review-workflow/tagging/ai-tools-in-tagging/adaptive-smart-tags.md) extracts into it, and how well. Writing those questions is up to you: "What was the total sample size enrolled?" gives the AI an extraction task, while "Sample size" doesn't. See [**Prompt Guidance**](/ai-transparency/ai-in-nested-knowledge/prompt-guidance.md) before running extraction. ### 3. Check Your Live Export In Dashboard Build your spreadsheet as a table card in [**Dashboard**](/review-workflow/synthesis/set-up-synthesis-outputs/dashboard-editor.md) before you rely on extraction. The card updates as studies are tagged, so you can compare it against your spreadsheet after each Adaptive Smart Tags run. When a column comes out wrong, you fix the tag, re-run, and check the card again, instead of finding the problem in your final export. ### Build Your Export Table {% stepper %} {% step %} #### Open Dashboard Editor In the Synthesis menu, select **Dashboard Editor**, then select the **+** icon to add a Dashboard. Name it after your spreadsheet so it's easy to find. {% endstep %} {% step %} #### Add A Table card Select **Table**. In **Table Builder**, choose the bibliographic fields, tags, and studies that match your spreadsheet's columns. See [Table Builder](/review-workflow/exports/custom-tables.md) for the full steps. {% endstep %} {% step %} #### Save The Card Select **Save**. Turn on **Wrap Text** to read individual study rows more easily. Add one card per sheet in your spreadsheet if it has several. Columns will only be populated if data has been extracted either by a reviewer or AI. See [Apply Tags.](/review-workflow/tagging/apply-tags.md) {% endstep %} {% endstepper %}

A table card titled Study Characteristics in Dashboard Editor. — Living Export Table in Dashboard

An exported spreadsheet titled Study Characteristics in Excel. — Downloaded Export Table in Excel

### 4. Export Not Quite Right? Troubleshoot Extraction. Run [**Adaptive Smart Tags**](/review-workflow/tagging/ai-tools-in-tagging/adaptive-smart-tags.md), then check a handful of studies in your table card against their source text before reviewing the rest. Patterns usually show up immediately. | What you see in the card | Likely cause | Fix | | ---------------------------------------------------------------------- | ---------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Data lands in the wrong column, or several values share one cell | The tag type or table columns don't match the spreadsheet | Change the tag's content type or columns in [**Configure Tagging**](/review-workflow/tagging/configure-tagging.md) | | Values come back in mixed formats, such as some means and some medians | The question doesn't state the output you want | State the units or summary statistic in the question. Split a **Numeric** tag into one tag per statistic. See [Prompt Guidance](/ai-transparency/ai-in-nested-knowledge/prompt-guidance.md) | | Unrelated content appears in a column | The question doesn't say what to leave out | Add what not to extract to the question | | Cells are empty where the study reports the data | The question is too narrow, or a required annotation is making the AI conservative | Broaden the question, or review the annotation setting. See [Choose Annotations Deliberately](/ai-transparency/ai-in-nested-knowledge/prompt-guidance.md#choose-annotations-deliberately) | After editing a tag, re-run **Adaptive Smart Tags**. A re-run covers new studies and tags whose question or configuration has meaningfully changed (e.g. rewording a question, changing the title or content type). Unchanged tags on studies already extracted are left as they are. See [Will re-running Adaptive Smart Tags overwrite data I've already reviewed?](/review-workflow/tagging/ai-tools-in-tagging/adaptive-smart-tags.md#will-re-running-adaptive-smart-tags-overwrite-data-ive-already-reviewed) Download the card as a .csv at any point to share or check the export outside Nested Knowledge. ### FAQs

Can I import data I've already extracted into a spreadsheet?

Yes, once the matching tags exist. [**Import Applied Tags**](/review-workflow/tagging/import-applied-tags.md) brings completed answers into studies already in your nest. It doesn't create the hierarchy.

Does my tag name have to match my spreadsheet's column name exactly?

No. Matching them makes it easier to cross-check your work against the original sheet, but it isn't required.

How do I know my export will look how I want?

Build your export table in [**Dashboard**](/review-workflow/synthesis/set-up-synthesis-outputs/dashboard-editor.md) after setting up your hierarchy. It updates as data is extracted, so you can check it against your spreadsheet throughout. [See above.](#check-the-live-export-in-dashboard)

What if a column doesn't map cleanly to any tag type?

Use a Table Tag. Its cells accept any free text, so a reviewer can enter whatever value the column needs.

Do I need to build the whole hierarchy before I start tagging?

No. A new tag isn't applied retroactively, so adding one partway through only affects studies tagged from then on. Building the full hierarchy first means fewer changes mid-review.

### Next Steps


Configure Tagging	Build and refine the tag hierarchy for your review.	Configure Tagging
Import Tag Hierarchy Templates	Start from an existing hierarchy and adapt it.	Import Tag Hierarchy Templates
Import Applied Tags	Import data captured in an earlier review.	Import Applied Tags
Apply Tags	Capture data from your included studies.	Apply Tags
Dashboard Editor	Build a living table that updates as data is extracted.	Dashboard Editor
Prompt Guidance	Write clear questions for Adaptive Smart Tags.	Prompt Guidance

--- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.nested-knowledge.com/review-workflow/tagging/match-tags-to-your-exports.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.