Searching and extracting information in a text sequence (scientific article, search engine query, discussion forum post) requires a named entity recognition (NER) process. However, the data available to carry out this process vary according to their nature and field of study. In this article, we focus on the performance of named entity recognition systems, on the one hand, and on their complexity and ability to process data from different origins, on the other. A comparative study between several state-of-the-art approaches, applied to different types of data (search engine queries and discussion forum posts) related to the tyre sector, is proposed in order to select the approach that best suits our use case. To do this, we will rely on the results of evaluation metrics of automatic learning models such as precision, recall, F-score.