# Failing at intent classification

**URL:** <https://forum.rasa.com/t/failing-at-intent-classification/13815>\
**Category:** Rasa Open Source\
**Created:** [August 1, 2019, 12:40pm UTC](https://forum.rasa.com/t/failing-at-intent-classification/13815 "2019-08-01T12:40:15Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![shubham](https://avatars.discourse-cdn.com/v4/letter/s/85e7bf/32.png) [@shubham](https://forum.rasa.com/u/shubham)\
**Post date:** [August 1, 2019, 12:40pm UTC](https://forum.rasa.com/t/failing-at-intent-classification/13815/1 "2019-08-01T12:40:15Z")

</div>

I am training my model which has following data but I am failing at intent classification. can anyone suggest why its happening and suggest better parameter for intent classifier to improve dection.

| intent | Examples of intents |
| --- | --- |
| budget\_details | 13 |
| goodbye | 24 |
| affirm | 29 |
| deny | 36 |
| greet | 42 |
| product\_submission | 2776 |
| request\_waiver | 2976 |
| customer\_details | 3011 |
| product\_details | 4913 |
| product\_information | 26516 |

@akelad @JiteshGaikwad @Juste @JulianGerhard

---

<div class="post-metadata">

**Author:** ![JulianGerhard](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/juliangerhard/32/3781_2.png) [@JulianGerhard](https://forum.rasa.com/u/JulianGerhard)\
**Post date:** [August 1, 2019, 1:20pm UTC](https://forum.rasa.com/t/failing-at-intent-classification/13815/2 "2019-08-01T13:20:27Z")

</div>

Hi @shubham

this is a bit vague to be honest. Could you please provide some verbose information?

1. Are you able to train the bot?
2. Are you able to start the bot?
3. Could you start the bot with `-vv` flag and post the output here?

Besides it seems a bit unbalanced where “a bit” means a lot. Is your classifier working but you are not satisfied with its performance? If so, please post your pipeline, such that we can analyze the problem properly!

Regards Julian

---

<div class="post-metadata">

**Author:** ![SamS](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/sams/32/3497_2.png) [@SamS](https://forum.rasa.com/u/SamS)\
**Post date:** [August 1, 2019, 3:07pm UTC](https://forum.rasa.com/t/failing-at-intent-classification/13815/3 "2019-08-01T15:07:29Z")

</div>

Hey @shubham, may I see the confusion matrix produced by `rasa test nlu`? Essentially, I want to see how exactly your intent classifier fails.

---

<div class="post-metadata">

**Author:** ![shubham](https://avatars.discourse-cdn.com/v4/letter/s/85e7bf/32.png) [@shubham](https://forum.rasa.com/u/shubham)\
**Post date:** [August 2, 2019, 6:09am UTC](https://forum.rasa.com/t/failing-at-intent-classification/13815/4 "2019-08-02T06:09:35Z")

</div>

> [@shubham](#):
>
> 4913

language: en

pipeline:

- name: “SpacyNLP” model: “en\_core\_web\_lg”

- name: “SpacyTokenizer”

- name: “SpacyFeaturizer”

- name: “RegexFeaturizer”

- name: DucklingHTTPExtractor url: [http://localhost:8000](http://localhost:8000) dimensions:

- name: “CRFEntityExtractor” features: [[“low”, “title”,“pos”, “pos2”], [“bias”, “low”, “prefix5”, “prefix2”, “suffix5”,“digit”, “suffix3”,“suffix2”,“upper”, “title” ,“pattern”], [“low”, “title”, “upper”, “pos”, “pos2”]] BILOU\_flag: true max\_iterations: 50 L1\_c: 0.1 L2\_c: 0.1

- name: “EntitySynonymMapper”

- name: CountVectorsFeaturizer

- name: “EmbeddingIntentClassifier”

policies:

- name: MemoizationPolicy
- name: KerasPolicy
- name: MappingPolicy
- name: FormPolicy

---

<div class="post-metadata">

**Author:** ![SamS](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/sams/32/3497_2.png) [@SamS](https://forum.rasa.com/u/SamS)\
**Post date:** [August 5, 2019, 7:25am UTC](https://forum.rasa.com/t/failing-at-intent-classification/13815/5 "2019-08-05T07:25:39Z")

</div>

Hey @shubham, thanks for posting the pipeline, I think it will come handy a bit later.

For now, can, please, you explain what is not working and ideally provide the kind of information (either error messages or the intent confusion matrix, etc.) that would help me or @JulianGerhard help you, as we explained earlier? Thanks.