Text summarization API

One of the most popular tasks in the field of text analysis is summarization. Sometimes it is referred to as the gisting or automatic reviewing.

Supported languages: English, French, German, Portuguese, Italian, Spanish, Japanese, Chinese, Greek.

If we take any article, we can easily see that it is based on an amazingly simple core idea. The entire meaning of any text in the natural language is contained in a few sentences. While the rest of the text only clarifies and explains the key idea. Automatic summarization allows analyzing the text and finding one or several sentences that contain the core idea of the text. Such algorithms are used on many news websites. When you deal with the main page or the list of articles, you need to display a short announcement for each text to make it easier for the user to decide whether to read the article. The text summarization API allows any website or mobile app to do this totally automatically.

Let us take a closer look at the features of the automatic text summarization API.

Identification of the most important sentences

The system automatically determines the main topic of the text and the key theses. After that, the sentences that contain the most important statements are chosen from the entire text. When sending a request to the API, you can specify the number of the most important sentences you want to receive. If you want to create a text title, specify the required number of sentences equal to 1.

Extracting text keywords

Keyword extraction is a mandatory step in the process of text summarization. The system does not search for the most common words in the text. In fact, it determines the most important words, building a hierarchy of connections and meanings between them. The list of keywords forms the basis for determining the most important sentences.

Extracting the most important part

During the process of text summarization, the sentences found are likely to be from different parts of the text. It is obvious that the most important theses can be scattered around the article. One at the beginning, and another near the end. While extracting the most important, the system finds several sentences that are connected to each other in order to fully convey the meaning of the text. In the conditions of the real world, this approach is slightly less common than conventional text summarization. At the same time, many people consider this opportunity important. If you have no idea about the best option for you, we recommend that you try and choose the one that is most suitable for your content.