Analyzes Clickstreams Based on Markov Chains
A set of tools to read, analyze and write lists of click sequences
on websites (i.e., clickstream). A click can be represented by a number,
character or string. Clickstreams can be modeled as zero- (only computes
occurrence probabilities), first- or higher-order Markov chains.
- Incorrect absorbing probabilities when predicting the next states of a given pattern and markov chain object when one of the next states is an absorbing state (thanks to Sebastian Lechner for reporting the bug).
- Compatibility problem with arules 1.4.0
- Citation file was using deprecated arguments.
- Incorrect start probabilities.
- Incorrect prediction of cluster assignment when clustering clickstreams with an order > 0.
- Compatibility problem with igraph 1.0.0
- Possibility to estimate markov chains with linear programming (default).
- Speed of several functions (thanks to Dennis Bohle firstname.lastname@example.org for contributing some ideas to improve the performance of the package)
- Method for converting a Clickstreams-object to a data frame consisting of session-specific state frequencies.
- Method for converting a Clickstreams-object to an arules::transactions-object.
- NAs when reading large clickstream data.
- Deleted print-method for S4 classes.
- Added the possibility to define the number of digits in a Markov-chain-plot.
- Bug in function clusterClickstreams with orders higher than the smallest clickstream
- Improved summary of clickstream clusters.
- Wrong line breaks in readClickstreams.
- Wrong direction in MarkovChain plots.
- Wrong edge titles in MarkovChain plots.
- Function for clustering lists of clickstreams.
- Functions for printing and summarizing clickstream clusters.
- Function for writing a clickstream list to file.
- Possibility to use absorbing probabilities for predicting next states.