# NER and TextClassification

**URL:** <https://support.prodi.gy/t/ner-and-textclassification/281>\
**Category:** Uncategorized\
**Tags:** usage, textcat\
**Created:** [February 3, 2018, 6:03am UTC](https://support.prodi.gy/t/ner-and-textclassification/281 "2018-02-03T06:03:25Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![madhujahagirdar](https://avatars.discourse-cdn.com/v4/letter/m/e99b99/32.png) [@madhujahagirdar](https://support.prodi.gy/u/madhujahagirdar)\
**Post date:** [February 3, 2018, 6:03am UTC](https://support.prodi.gy/t/ner-and-textclassification/281/1 "2018-02-03T06:03:25Z")

</div>

Is it possible to pass custom trained NER model to textcat.train other than the word2vec model ?

---

<div class="post-metadata">

**Author:** ![honnibal](https://sea2.discourse-cdn.com/flex020/user_avatar/support.prodi.gy/honnibal/32/35_2.png) [@honnibal](https://support.prodi.gy/u/honnibal)\
**Post date:** [February 5, 2018, 5:05am UTC](https://support.prodi.gy/t/ner-and-textclassification/281/2 "2018-02-05T05:05:28Z")

</div>

`texcat.train` doesn’t take into account the named entity annotations, so your results shouldn’t be different either way.

If you really want to have features for the named entities, you could merge them into single tokens before running the text classifier. However, I suspect this won’t have a big impact on accuracy — I think the normal text classifier will give you very similar results.