Skip to content

Document classification

Classification and taxonomies

Document classification determines what a text is about in terms of categories of a taxonomy.

Available taxonomies are:

Taxonomy English Spanish French German Italian
iptc
geotax
emotional-traits
behavioral-traits

In the Natural Language API terminology, taxonomy "x" is both a specific set of categories and the name of the API resources capable of classifying a text according to that set.

Taxonomies' resources have paths like this:

categorize/taxonomy name/language code

Boxed parts are placeholders, so for example:

https://nlapi.expert.ai/v2/categorize/iptc/en

is the URL of the iptc resource capable of performing the IPTC Media Topics classification of English texts.
These resources must be requested with the POST method, submitting the text to classify.

In the reference section of this manual you will find all the information you need to perform document classification using the API's RESTful interface, specifically:

Note

Even if you use the API through a client that hides the REST interface, whether it is made by you or is one of those provided by expert.ai, knowing the output format helps you understand and navigate the results.

Here is an example of performing IPTC Media Topics classification of a short English text:

This example is based on the Python client you can find on GitHub.

The client gets user credentials from two environment variables:

EAI_USERNAME
EAI_PASSWORD

Set those variables with your account credentials before running the sample program below.

The program prints the list of categories.

from expertai.nlapi.cloud.client import ExpertAiClient
client = ExpertAiClient()

text = "Michael Jordan was one of the best basketball players of all time. Scoring was Jordan's stand-out skill, but he still holds a defensive NBA record, with eight steals in a half."
taxonomy = 'iptc'
language= 'en'

output = client.classification(body={"document": {"text": text}}, params={'taxonomy': taxonomy, 'language': language})

print("Tab separated list of categories:")

for category in output.categories:
    print(category.id_, category.hierarchy, sep="\t")

This example is based on the Java client you can find on GitHub.

The client gets user credentials from two environment variables:

EAI_USERNAME
EAI_PASSWORD

Set those variables with your account credentials before running the sample program below.

The program prints the JSON response.

import ai.expert.nlapi.security.Authentication;
import ai.expert.nlapi.security.Authenticator;
import ai.expert.nlapi.security.BasicAuthenticator;
import ai.expert.nlapi.security.DefaultCredentialsProvider;
import ai.expert.nlapi.v2.API;
import ai.expert.nlapi.v2.cloud.Categorizer;
import ai.expert.nlapi.v2.cloud.CategorizerConfig;
import ai.expert.nlapi.v2.message.CategorizeResponse;
import ai.expert.nlapi.v2.model.CategorizeDocument;

public class Main {

    public static Authentication createAuthentication() throws Exception {
        DefaultCredentialsProvider credentialsProvider = new DefaultCredentialsProvider();
        Authenticator authenticator = new BasicAuthenticator(credentialsProvider);
        return new Authentication(authenticator);
    }

    public static Categorizer createCategorizer() throws Exception {
        return new Categorizer(CategorizerConfig.builder()
                                                .withVersion(API.Versions.V2)
                                                .withTaxonomy("iptc")
                                                .withLanguage(API.Languages.en)
                                                .withAuthentication(createAuthentication())
                                                .build());
    }

    public static void main(String[] args) {
        try {
            String text = "Michael Jordan was one of the best basketball players of all time. Scoring was Jordan's stand-out skill, but he still holds a defensive NBA record, with eight steals in a half.";

            Categorizer categorizer = createCategorizer();

            CategorizeResponse categorization = categorizer.categorize(text);


            // Output JSON representation

            System.out.println("JSON representation:");
            categorization.prettyPrint();


            // Tab separated list of categories.

            System.out.println("Tab separated list of categories:");
            CategorizeDocument data = categorization.getData();

            data.getCategories().stream().forEach(c -> System.out.println(c.getId() + "\t" + c.getHierarchy()));
        }
        catch(Exception ex) {
            ex.printStackTrace();
        }
    }
}

The following curl command posts a document to the document classification resource of the API's REST interface.
Run the command from a shell after replacing token with the actual authorization token.

curl -X POST https://nlapi.expert.ai/v2/categorize/iptc/en \
    -H 'Authorization: Bearer token' \
    -H 'Content-Type: application/json; charset=utf-8' \
    -d '{
  "document": {
    "text": "Michael Jordan was one of the best basketball players of all time. Scoring was Jordan'\''s stand-out skill, but he still holds a defensive NBA record, with eight steals in a half."
  }
}'

The server returns a JSON object.

The following curl command posts a document to the document classification resource of the API's REST interface.
Open a command prompt in the folder where you installed curl and run the command after replacing token with the actual authorization token.

curl -X POST https://nlapi.expert.ai/v2/categorize/iptc/en  -H "Authorization: Bearer token" -H "Content-Type: application/json; charset=utf-8" -d "{\"document\": {\"text\": \"Michael Jordan was one of the best basketball players of all time. Scoring was Jordan's stand-out skill, but he still holds a defensive NBA record, with eight steals in a half.\"}}"

The server returns a JSON object.

The following articles describe the capabilities of the available taxonomies.

Self-documentation resources

The API provides self-documentation resource to programmatically discover available taxonomies and their features. Learn more about this resource in the dedicated article.