LAPIN

Introduction

LAPIN is a software package to mine frequent patterns in a sequence database. Given a collection of customer sequences and a minimum support threshold, LAPIN finds all the frequent sub-sequences whose support is above the threshold.

Usage Agreement

   1. Downloading is for internal research usage only. Redistribution and commercial usage are not permitted.
   2. Use of the downloaded software is confined to performance test. For other interest, please contact the authors.
   3. If you use this software, please cite the following paper if possible. Thank you.

LAPIN: Effective Sequential Pattern Mining Algorithms by Last Position Induction for Dense Databases. Zhenglu Yang, Yitong Wang, and Masaru Kitsuregawa. Proc. of 12th International Conference on Database Systems for Advanced Applications (DASFAA'07), pp. 1020-1023, Bankok, Thailand, Apr. 2007.

Note

   1. Feedback and application description are always welcome.
   2. Contact My email address for bugs and questions about LAPIN.

Download

   LAPIN

Installation

LAPIN is implemented using C/C++ with STL support. Currently, LAPIN is distributed as executable code format for Windows platform.

Run

   lapin -i input_file -s support [-o output_file]
  
  • Usage example: lapin -i test.data -s 0.9 -o pattern.txt, where lapin is the executable file name, test.data is the input dataset, 0.9 is the relative support, pattern.txt is the output file.
    •   
  • The output can be turned off.
    •   
  • The minimum support should be in relative support format.


    • More detail about input and output file format can be found in the README file.


    Data sets

    Synthetic data

    data converter toolkit

  • We convert the data, which is created by the IBM data generator from Zaki's homepage: http://www.cs.rpi.edu/~zaki/software/
    • to LAPIN and PrefixSpan format.

    Real data

    For LAPIN         For PrefixSpan  


    PrefixSpan software can be downloaded from the author's homepage.