What is unstructured data? Data that is not organized into a rigid structure, such as text, video or photos, is considered as such. Until a few years ago, extracting information from these sources was very complex and, above all, time-consuming. However, with technological advances in recent years, whether in terms of algorithms and techniques, or computational power, this task has become much easier. Nowadays, classifying segments of a video, identifying/cataloging images or extracting information from texts is something that is relatively common. However, only universities, research centers or large companies are taking advantage of this technological revolution, because, often, small and medium-sized companies simply do not even know about such a possibility.

‍

To raise awareness of the potential of these technologies and methods, and mainly their impact on tourism and hospitality, with this post I will start a set of publications on Natural Language Processing (NLP), one of the areas with the greatest growth today .

In a simplistic way, it can be said that NLP is a sub-area of ​​artificial intelligence, which aims to allow computers to understand and process human languages . This understanding and processing are normally divided into a set of tasks that are sometimes applied together, namely:

• Sentiment analysis : allows you to analyze the polarity of sentiment in a text (negative or positive sentiment);
• Similarity analysis : allows you to compare the similarity between texts;
• Textual coherence : allows you to analyze and study the coherence of a text’s writing;
• Text to speech and speech to text conversion : allows you to convert voice recordings to text and vice versa;
• Terminology extraction : allows you to extract specific terms from an area based on texts from that same area;
• Text generation : allows you to automatically create texts;
• Entity identification : allows you to identify entities in a text (without prior knowledge of the names or type of entities);
• Topic identification : allows you to identify topics addressed by a set of texts;
• Entity connection : allows you to identify connections between entities based on a set of texts;
• Automatic translation : allows you to automatically translate texts;
• Automatic text summarization : allows you to summarize large texts in a few paragraphs or sentences;
• Among many others.

In this sequence, the next post will be dedicated to some examples of the potential of PLN in tourism and hospitality, starting with the analysis of sentiments and extraction of terminology applied to comments published by hotel customers on the various websites for this purpose. With these examples, we will be able to see that the information extracted from the texts of thousands of online comments is much richer and has the potential to be actionable in terms of management than the information extracted from the ratings of these comments.

‍