mine_fuzzy_model: Mines eventlog data using Fuzzy Algorithm
In PlaypowerLabs/fuzzymineR: Build Fuzzy Process Model from Event Data

Description Usage Arguments Details Value References See Also Examples

View source: R/mine_fuzzy_model.R

Takes eventlog as input, applies fuzzy mining algorithm and returns a list of metrics which specifies significance for different activities and their precedence relations.

1	mine_fuzzy_model(eventlog, verbose = FALSE)

`eventlog`	An eventlog object
`verbose`	logical. Should output messages from system command be printed in console?

mine_fuzzy_model uses fuzzy mining algorithm to mine eventlog data and generates a list of metrics. This list of metrics is then used for creating process models. Fuzzy mining algorithm is widely used to create process models because of its adaptiveness.

It generates a fixed number of metrics. There are total of 9 metrics generated by this algorithm. Out of this, 2 metrics are for the activities which will be nodes in our process models. Other 7 would represent precedence relation between them and would form edges in the process model. These precedence relation are of two types - significance and correlation. Out of 7, 2 would be significance relations and 5 would be correlation metrics.

Frequency Unary Significance : The more often an activity is observed in event log more the value of frequency significance of that activity.

Routing Unary Significance : The higher the number and significance of predecessors for a node (Activity) differs from the number and significance of its successors the more important that node is for routing in the process.

Frequency Binary Significance : The more often two activities are observed after one another, the more significant their precedence relation

Distance Binary Significance : This metric calculates the significance of two activity that occur in a trace but not after one another. More the distance, less is the weightage of that significance value. Hence it takes into account the significance of all the global relations.

Proximity Binary Correlation : This metric captures the activities that occur shortly after one another and measures there correlation. This metric is important for identifying the clusters of activites that occur in short time frame and hence can be seen as one logical activity.

Originator Binary Correlation : Correlation between two activities that is determined by the name/ID of the person who have triggered the event. Similar the names, the higher the correlation between two activities. Eg: Resource person names : (sales_x, sales_y) and (technical_x, technical_y)

Endpoint Binary Correlation : Correlation between two activities is determined by the name of the two events. More similar the names are more is the correlation between activites. Eg : (Check_x, Check_y)

Data Type Correlation : Correlation bewteen two activities is higher when sub-sequent activities share large amount of data types(attributes keys).

Data Value Correlation : Corrrelation between two activities is determined by the values of these shared data types or common attributes. More similar the value higher is the correlation.

For getting more insights about these metrics and what is their importance, please go to the reference website.

A list of metrics which contains significance of each activities and their precedence relations with each other for several parameters

Fuzzy Mining - Adaptive Process Simplification Based on Multi-Perspective Metrics

create_eventlog, viz_fuzzy_model

## Not run: 
#' library(fuzzymineR)
data("artificial_loan_process")
log <- create_eventlog(artificial_loan_process,
                       case_id = "case",
                       activity_id = "event",
                       timestamp = "completeTime")
metrics <- mine_fuzzy_model(log)

## End(Not run)