If you look at some compression techniques (such as LZW Compression https://en.wikipedia.org/wiki/Lempel%E2%80%93Ziv%E2%80%93Welch ) they make use of the concept of hashing techniques to compress files. We are going to do something similar. Consider the phrase

To be or not to be, that is the question.

Even though it does not make much sense to do this as the phrase is short. We are going to create bins for each word, and record the position of the instances of the words in the proper bins. So write a program that first has a series of bins

to

be

or

not

etc.

Because the word to occurs in positions 1 and 5, the bin for the word to will contain 1 and 5

to: 1 5

be: 2 6

or: 3

Write an program that will take the phrase, create bins for each word, and then record the instance position of each word in the bin. It will then output each bin (word) and the locations of the word in the phrase. So if the input was;

I want what I want and I know what I want.

The output would be;

I: 1 4 7 10

want: 2 5 11

what: 3 9

and: 6

know: 8

I have created a JSFiddle Shell to get you started that is here http://jsfiddle.net/reaglin/rm96Luws/