当前位置:网站首页>Introduction of process mining based on event log - Chaudhuri

Introduction of process mining based on event log - Chaudhuri

2021-05-04 12:36:30 Jiedao jdon

This paper briefly introduces the data-driven process discovery technology , It's called process mining .

What is process mining ?

For any process improvement project , Generally, it is very important to know the current state through the original process model . Yes person Interview , Review existing documents , And use modeling tools to create process models . These process models are mainly based on the views of stakeholders on the process . For mainly manual processes , It could be a good way , Maybe the only way .

however , With IT The increase of system participation in process execution , A lot of information will be recorded in various systems or event logs . The fact that these traces appear , These are evidence of real events that have happened 、 Just as it is .

Process mining creates process models based on data from traces or event logs , This means that the current state operation is not based on perception , It's based on hard evidence of what actually happened .

Use the event log maintained by the information system , You can use specialized process mining software to visualize the actual process and all its possible paths , Including exceptions and upgrades . The generated process allows detailed analysis of the current state .

Process mining is about visualizing processes ;  It helps to analyze , But process mining tools don't provide solutions or processes .

There are many possible process mining use cases . Three situations are described below .

scene 1: Process discovery from the event log

Extract event logs and prepare for mining - in other words , Clean up and reformat data for process mining tools . The process generated by process mining tools is more reliable than that created by interviewing process stakeholders .

The event log reveals the real process , Not how its actors view it . You can find unusual situations ( It's likely to be overlooked in the interview ), And highlight the overall process performance . This process Discovery Scheme provides data for future process analysis and improvement , And provide the basis for further analysis , For example, process compliance or process path prediction based on historical data .

Imagine a customer service center . Customers register complaints ;  then , The system records the details as “ Case study ”, Customer service agents work to solve cases, write notes , Involve different teams as needed , Finally, close the case . The customer complained that , Every other day , Different agents will call them for more details , But then there was no reaction .

Process mining can read the log of each case , Then visualize the process as needed . This will find out “ real ” technological process , And show what's actually done ( And the unfinished work ). This information can be used for further analysis .

scene 2: Process consistency check

Focus on process consistency checking , The event log is checked according to the ideal process . The ideal process could be a solution 1( Process discovery ) Improved and optimized version of the process results . It can also be an independent design process . Process mining tools convert event log files into process models , Then examine this realistic process model against the ideal scenario . Diagnose the mismatch between the process model from the event log and the ideal process , Highlight and visualize invisibility - This allows further analysis and optimization of these process components .

for example , Every company needs to use a fixed process to maintain accounts to track financial transactions . Auditors use process mining methods to highlight deviations from accounting standards , To use a process driven approach . Once the real process non-compliance is highlighted , Deviation is further analysis and improvement The theme of .

scene 3: Process enhancement

Focus on process enhancement , Analyze the event log to enhance or improve the process . It starts with process discovery , Then there's process evaluation and optimization - then , Analyze various performance parameters of the process to improve Process performance . The ideal process model can also be used in this enhanced process , To identify and highlight mismatches .

for example , In order processing , You can precisely highlight the processing steps that represent bottlenecks . This is very similar to process simulation , But using real trade data that simulation rarely uses . If we find that “ Order review ” It takes the most time , We will make further analysis . Process changes will be based on the reality represented by the event logs in the system .

​​​​​​​

版权声明
本文为[Jiedao jdon]所创,转载请带上原文链接,感谢
https://chowdera.com/2021/05/20210504123223988E.html

随机推荐