How to discover syntagmatic and paradigmatic relations with MeTA


#1

Hi,

I’m reading the book “Text Data Management and Analysis: A Practical Introduction to Information Retrieval and Text Mining” and trying to figure out some concepts.
In chapter 13 “Word Association Mining” book tells us about syntagmatic and paradigmatic relations. Then it follows BM25 algorithm, entropy and mutual information.

I’ve already looked in the forum, MeTA and metapy tutorials. Can you give an example in MeTA which applies the syntagmatic and paradigmatic relations?

Such as exercise 13.5 in the book: Use MeTA to implement one or both of the word association mining methods. Use the default unigram tokenization chain to read over a corpus and create feature vectors for each term ID. Then, given a query term, return the most similar terms.

Thank you.