Skip to content

About

Workflow event logs to ML-based predictive timelines: deterministic reconstruction, sequence features (waiting/time gaps), and early risk prediction for prioritization.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

temporal sequence Event Log

Scope of the project

  • Problem: Many cases are missing deadlines or took too long to complete, which leads to inferior experience.
  • Objective: Predict early which cases are likely to miss the dealine or take unusally long.
  • Decision: Escalate or reassign case under limited capacity.
  • Impact: Reduce no of cases that misses time-target, faster resolution, and better workload balance.

Exact constraint values live in configs/constraints.yaml.


Dataset

A timestamped event log (audit trail / change log) of case updates. Each case (case_id) appears multiple times — each row is one recorded update at update_timestamp, showing the case’s current status, assignment, and counters up to that moment.

Entity definitions

  • Case: one support request / issue (unique case_id)
  • Event/Record: one logged update to that case (a snapshot at update_timestamp)

Target labels (for prediction)

  • On-time resolution (met_deadline): whether the case completed within an expected time window.

  • (future scope) time-to-finish: duration from opened_timestamp to closed_timestamp (or resolved_timestamp)

License

  • Code: MIT (see LICENSE)
  • Dataset: CC BY 4.0 — see CITATION.md for attribution and source.

About

Workflow event logs to ML-based predictive timelines: deterministic reconstruction, sequence features (waiting/time gaps), and early risk prediction for prioritization.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages