Step By Step Guide To Natural Language Processing (NLP) In Trading – Part II

Read the first part in this series for an overview of NLP in trading.

Preprocess the data

There are different problems associated with these two data sets. Unstructured data like Twitter feeds consist of many non-textual data, such as hashtags and mentions. These need to be removed before measuring the text’s sentiment. 

For structured data, the size of the text can easily cloud its essence. To solve this, you need to break the text down into individual sentences or apply techniques such as tf-idf to estimate the importance of words.

Convert the text to a sentiment score

Converting the text data to a numerical score is a challenging task. For unstructured text, you can use pre-existing packages such as VADER to estimate the sentiment of the news. If the text is a blog or an article then you can try breaking it down for VADER to make sense of it.

For structured text, you don’t have any pre-existing libraries that can help you convert the text to a positive or a negative score. So, you will have to create a library of your own.

When building such a library of relevant structured data, care should be taken to consider texts from similar sources and the corresponding market reactions to this text data.

For example, if the Fed releases a statement saying that “the inflation expectations are firmly anchored” and changes it to “the inflation expectations are stable”, then libraries like VADER won’t be able to tell the difference apart, but the market will react significantly.

To understand the score of the sentiment of such text, you need to develop a word-to vector model or a decision tree model using the tf-idf array.

Visit QuantInsti website to learn how to generate and backtest the trading model.