What are the other metrics that we can use in Sequential Pattern Mining, when using SPADE algorithm?

Question

I'm reading this awesome page with interest metrics used in Association Rules:
https://michael.hahsler.net/research/association_rules/measures.html

As I have sequential data, I decided to use the arulesSequences from R, which do Sequential Pattern Mining, and create the rules using the SPADE algorithm.
Here are some rules created:

head(as(rules, 'data.frame'))
           rule                              support         confidence         lift
1    <{A}> => <{B}>                        0.026485890       0.13160987      0.8112745
2    <{D}> => <{B}>                        0.009853382       0.03726893      0.2297345
3     <{C}> => <{B}>                       0.063455778       0.10779325      0.6644632
4   <{C},{A}> => <{B}>                     0.018524358       0.24607330      1.5168542
6    <{D}> => <{E}>                        0.015607757       0.14494876      3.1703792
7    <{A}> => <{F}>                        0.011587577       0.05757932      1.2593987

I'm thinking if makes sense to calculate some other metrics, like Chi Squared test (to test the null hypothesis that Lift = 0 for a specific rule), or calculate the Standardized Lift.
And if makes sense, there's some function in this package or in another to calculate these new metrics?
I have my doubts because the cspade() function only creates rules with support, confidence and lift.

Answer 1

This book is one of the most useful resources I've found for pattern mining. Chapter 5 (available as a sample chapter) talks about a few properties of interest measures, such as whether the measure is invariant to inversion, scaling, and null addition. When choosing an interest measure it's worth thinking about what conditions are most important.

I'm not overly familiar with R, but the interestMeasure package looks like what you want. Otherwise the networkx package in Python contains some additional interest measures, or implementing them yourself shouldn't be too hard.