Organize automatically into categories all types of content

Text Classification assigns one or more categories to a text to facilitate its management, allowing to filter, sort, or group texts. Search engines, newspapers, or e-commerce portals categorize their content or their products to facilitate the search and navigation. Understand your clients and what they say about your products by categorizing the conversations on social networks or contact centers.

MeaningCloud's Text Classification API

This API automatically categorizes your texts in a hierarchical classification or taxonomy. MeaningCloud can classify any kind of text, from web pages to social media content, of any length and in several languages. The API features various predefined standard classification models, but you can also define your own classification schemes or models.

A standard classification, for example IPTC (International Press Telecommunications Council), is used by the media to assign pieces of news to different sections: politics, sports or economy. This classification includes more than 1300 categories organized hierarchically in 3 levels. Therefore, news can be associated with broad but detailed categories such as 'politics - government - privatization' or 'sports - basketball - NBA'.

It also enables to define your own classification models, which can be as simple as a binary classification (ham or spam) or as complex as a taxonomy with multiple hierarchical levels. The classification models of MeaningCloud's Text Classification API combine a statistical model and/or classification rules. You can train the statistical model using example texts for each category and optionally refine the classification through specific rules.

Advantages of automatizing content classification. Applications

Organizing and describing content consistently is a complex task that requires the previous definition of a taxonomy, the criteria and also assign specialized human resources. Automatic classification opens a new range of possibilities which include both total automatization and support tools that reduce time and improve the quality of manual tagging processes. Through the use of automatic methods, it is possible to refine the classification and the categories over time, obtain more consistent results, faster and at lower cost.


Automatically analyze pieces of news or websites and assign thematic categories using standard models like IPTC.

Document categorization

Classify and manage automatically documents such as medical records, claims or financial reports according to your workflow or standardized taxonomies (for instance ICD-10 in medicine).

Content search and recommendation

Tag you contents or your products using categories as a way to improve browsing or to identify related content in your website.

Voice of the Client Analysis

Analyze every type of channels to measure the perception the clients have of your company or products. Use standard models to classify social interactions according to corporate reputation or to customer satisfaction.

Highlights of our Text Classification API

Machine learning and rules

Combines the application of machine learning with the versatility of rules defined by experts.

Multi-tag classification

Assign multiple categories to each document, ordered by a confidence measure.

Easy to use

We provide trained standard classification models: IPTC thematic classification, EuroVOC thesaurus or Corporate Reputation.

Multiple languages

MeaningCloud enables to classify documents in several languages: English, Spanish, French, Italian, Portuguese, and Catalan, and can be easily extended to other languages.

Customize your categories

Define your own classification models using rules, training texts, or both. Classify from 2 to hundreds categories using a hierarchical organization.

Combine it with Topics Extraction

The classification is appropriate for broad categories and they require to be defined previously. If you need to identify key words or ad-hoc categories, you can combine it with the Topics Extraction API.

Our clients

Unidad Editorial