# Siren NLP plugin configuration

**URL:** <https://community.siren.io/t/siren-nlp-plugin-configuration/379>\
**Category:** Siren Platform/Investigate Tech Support Community\
**Created:** [May 4, 2021, 10:36am UTC](https://community.siren.io/t/siren-nlp-plugin-configuration/379 "2021-05-04T10:36:46Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![habsburg](https://avatars.discourse-cdn.com/v4/letter/h/e79b87/32.png) [@habsburg](https://community.siren.io/u/habsburg)\
**Post date:** [May 4, 2021, 10:36am UTC](https://community.siren.io/t/siren-nlp-plugin-configuration/379/1 "2021-05-04T10:36:46Z")

</div>

Hi,

I’m trying to add two Siren NLP ingest processors (NER & Taxonomy) to a pipeline but am getting the error shown below. Any help will be much appreciated!

```
{

    "error": {

        "root_cause": [
            {
                "type": "exception",
                "reason": "java.lang.IllegalArgumentException: Cannot deserialize instance of `java.lang.String` out of START_ARRAY token\n at [Source: UNKNOWN; line: -1, column: -1] (through reference chain: uk.gov.nca.annot8.components.opennlp.processors.NER$Settings[\"nerType\"])",
                "processor_type": "siren-nlp"
            }
        ],

        "type": "exception",
        "reason": "java.lang.IllegalArgumentException: Cannot deserialize instance of `java.lang.String` out of START_ARRAY token\n at [Source: UNKNOWN; line: -1, column: -1] (through reference chain: uk.gov.nca.annot8.components.opennlp.processors.NER$Settings[\"nerType\"])",

        "processor_type": "siren-nlp",
        "caused_by": {
            "type": "illegal_argument_exception",
            "reason": "Cannot deserialize instance of `java.lang.String` out of START_ARRAY token\n at [Source: UNKNOWN; line: -1, column: -1] (through reference chain: uk.gov.nca.annot8.components.opennlp.processors.NER$Settings[\"nerType\"])",
            "caused_by": {
                "type": "mismatched_input_exception",
                "reason": "Cannot deserialize instance of `java.lang.String` out of START_ARRAY token\n at [Source: UNKNOWN; line: -1, column: -1] (through reference chain: uk.gov.nca.annot8.components.opennlp.processors.NER$Settings[\"nerType\"])"
            }
        }
    },
    "status": 500

```

}

**Siren-nlp Configuration**  
{

```
    "processors": [
        {
            "siren-nlp": {
                "fields": [
                    "title",
                    "snippet"
                ],
                "processors": [
                    {
                        "class": "Taxonomy",
                        "settings": {
                            "index": "cars",
                            "idField": "id",
                            "preferredTermField": "preferred_term",
                            "synonymFields": [
                                "synonyms"
                            ],
                            "parentsField": "parents",
                            "caseSensitive": false,
                            "exactWhitespace": false,
                            "plurals": true,
                            "type": "taxonomy-cars",
                            "additionalData": true
                        }
                    },
                    {
                        "class": "NER",
                        "settings": {
                            "nerType" : [
                                "Location", "Organization", "Person"
                            ]
                        }
                    }
                ]
            }
        }
    ]
}
```

---

<div class="post-metadata">

**Author:** ![Andrew\_Winter](https://yyz2.discourse-cdn.com/flex032/user_avatar/community.siren.io/andrew_winter/32/33_2.png) [@Andrew\_Winter](https://community.siren.io/u/Andrew_Winter)\
**Post date:** [May 4, 2021, 11:00am UTC](https://community.siren.io/t/siren-nlp-plugin-configuration/379/2 "2021-05-04T11:00:05Z")

</div>

For the NER processor the **nerType** must be a single string : Location, Organization or Person.  
To have multiple NER types in a pipeline, use multiple NER processors, each with a different nerType.

---

<div class="post-metadata">

**Author:** ![habsburg](https://avatars.discourse-cdn.com/v4/letter/h/e79b87/32.png) [@habsburg](https://community.siren.io/u/habsburg)\
**Post date:** [May 4, 2021, 2:34pm UTC](https://community.siren.io/t/siren-nlp-plugin-configuration/379/3 "2021-05-04T14:34:23Z")

</div>

It works like a charm now, thanks!