[SPARK-7212] Frequent pattern mining for sequential item sets

• Type: New Feature
• Status: Resolved
• Priority: Major
• Resolution: Won't Fix
• Component/s: MLlib
• Target Version/s: 1.5.0
• Assignee: Feynman Liang
• Reporter: Joseph K. Bradley
• Shepherd: Xiangrui Meng
• Created: 28/Apr/15 21:38
• Updated: 06/Jul/15 23:12
• Resolved: 06/Jul/15 23:12

Description

Currently, FPGrowth handles unordered item sets. It would be great to be able to handle sequences of items, in which the order matters. This JIRA is for discussing modifications to FPGrowth and/or new algorithms for handling sequences.

Issue Links

• is superceded by: SPARK-6487 Add sequential pattern mining algorithm PrefixSpan to Spark MLlib
• relates to: SPARK-7211 Improvements for FPGrowth
• links to: [Github] Pull Request #6997 (feynmanliang)
• links to: [Github] Pull Request #7240 (mengxr)