Natural Language Processing – NLP

Whenever you dictate something into your iPhone or Android phone device then your dictation is quickly converted to text, because natural language processing algorithm is in action, if so What is Natural Language Processing – NLP? Why should we use it? What are the tools that we can use about NLP? What are some field used examples?.. etc.

Let’s start the first question; by Wikipedia definition; Natural language processing (NLP) is a subfield of computer science, information engineering, and artificial intelligence concerned with the interactions between computers and human (natural) languages, in particular how to program computers to process and analyze large amounts of natural language data. ^[1]https://en.wikipedia.org/wiki/Natural_language_processing

Based on this definition, we can say that the focus of Natural Language Processing is teaching machines to understand what is said in the spoken and written in the text, thus texts and sounds are understood and interpreted by machines. 

So why should we use this, when you read or see the NLP application areas, you decide why it is important and should be applied. Some examples of NLP applications areas that you use every day are:

All of these usage areas are the use of technology in order to improve their product or service, and some are actual software providers that make this technology accessible to businesses. If you are curious about the subject’s history, you can look at the natural language processing at Wikipedia.  

I used yelp dataset ^[2]https://www.kaggle.com/kchow23/yelp100000reviews in the sample Kaggle project.  The Yelp dataset is a subset of your reviews for use in personal, educational, and academic purposes. As JSON files, you can use yelp databases to learn NLP.  I conducted the training and testing of the yelp data by using the Scikit-learn libraries. 

When performing exploratory data analysis, I see that there is a correlation between text length and the star rating and in the same way text count. As shown in the graph;

I attempted to classify Yelp Reviews into 1 star or 5-star categories based on the text content in the reviews.  So just by looking at the length of the text, we make sure that the system has an idea that the comments are one or five stars. Term Frequency, Inverse Document Frequency, and CountVectorizer are the models that I used to preprocess the texts to classify before fitting the Multinomial Naive Bayes classifier algorithm model on the texts.

Confusion matrix and classification report results are amazing. According to the Multinomial Naive Bayes classifier model, average accuracy prediction is 90%. This gives a good insight into the machines as well as us.

You can view all the analyzes and codes of this NLP project from here.^[3]https://www.kaggle.com/resulcaliskan/nlp-project-004?scriptVersionId=16946608

NLP has started to cover our lives and continues to develop in the AI world. The idea is that there is much more to be done in this area on the whole market. By developing and implementing NLP models in the business world, we can increase our marketing content and gain considerable speed in growth.

Be healthy and happy, see you next article  🙂