# Annotate same text with different label

**URL:** <https://support.prodi.gy/t/annotate-same-text-with-different-label/5418>\
**Category:** Uncategorized\
**Tags:** usage, ner, solved, streams\
**Created:** [March 4, 2022, 12:37pm UTC](https://support.prodi.gy/t/annotate-same-text-with-different-label/5418 "2022-03-04T12:37:35Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![matt.whitby](https://sea2.discourse-cdn.com/flex020/user_avatar/support.prodi.gy/matt.whitby/32/2837_2.png) [@matt.whitby](https://support.prodi.gy/u/matt.whitby)\
**Post date:** [March 4, 2022, 12:37pm UTC](https://support.prodi.gy/t/annotate-same-text-with-different-label/5418/1 "2022-03-04T12:37:36Z")

</div>

If I've annotated a given piece of text with a given label, e.g. PERSON, how to I go back and annotate the same text with a new label, e.g. ORGANISATION? It's quicker (for me) to annotate one label, then go back and go through with a second label.

---

<div class="post-metadata">

**Author:** ![ines](https://sea2.discourse-cdn.com/flex020/user_avatar/support.prodi.gy/ines/32/3_2.png) [@ines](https://support.prodi.gy/u/ines)\
**Post date:** [March 6, 2022, 2:32pm UTC](https://support.prodi.gy/t/annotate-same-text-with-different-label/5418/2 "2022-03-06T14:32:12Z")

</div>

Yes, that's definitely a good strategy! You can always load back an existing dataset as the input source using the [`dataset:` prefix](https://prodi.gy/docs/api-loaders#datasets), use your next label and save the results to a new dataset containing both labels.

For example, your first step would be:

```bash
prodigy ner.manual data_person your_examples.jsonl --label PERSON

```

And in the next step, you'd do:

```bash
prodigy ner.manual data_person_org dataset:data_person --label ORG

```

If you ever make a mistake or want to try out different combinations of labels, you can always go back and re-use your original dataset.