Finding common word sequences

Previously we saw how to create a query that finds word sequences of length three ("3-tuples"). We used that query as a subquery in a traditional SQL query to find the most common 3-tuples in the text document. You will now perform a similar task to find the most common 5-tuples.

Dataframe text_df is available. It contains the first five chapters of the Sherlock Holmes text. It has columns: word, id, part, title. The id column is an integer such that a word that comes later in the document has a larger id than a word that comes before it. The part column separates the data into chapters. The dataframe text_df is also registered as temporary table called text. Our objective is to create a dataset where each row corresponds to a 5-tuple, having a count indicating how many times the tuple occurred in the dataset.

Create a query query that finds the 10 most common 5-tuples in the dataset.

Take Hint (-30 XP)

Exercise

Finding common word sequences

Instructions