-2

My data mining problem is a p=next web page prediction using the exixting web data. For that I have a set of frequent sequences which are obtained using cspade algorithm in R. Now I am not sure how to mine set of associations so that I can predict the next page? can someone help?

Frequent sequencies look like this: 


     items     support 
       1 [{1}]  0.1640261 
       2 [{10}] 0.05112657 
       3 [{11}] 0.05818949 
       4 [{12}] 0.11333700 
       5 [{13}] 0.07773954 
       6 [{14}] 0.12036354 
       7 [{15}] 0.02950037 
       8 [{17}] 0.01111922 
       9 [{2}]  0.17708912 
      10 [{3}]  0.12320245 
      11 [{4}]  0.12297109 
      12 [{5}]  0.02524403 
      13 [{6}]  0.21933426 
      14 [{7}]  0.08134223 
      15 [{8}]  0.09659857 
      16 [{9}]  0.09111978 
      17 [{6}, 
         {9}]  0.01086563 
      18 [{9},   
         {9}]  0.04410508 
      19 [{9},   
         {9},   
         {9}]  0.02321639 
      20 [{9},   
         {9},   
         {9},   
          {9}]  0.01316606 
      21 [{8},   
          {8}]  0.06783368 
      22 [{8},   
          {8},   
          {8}]  0.05253996 
      23 [{8},   
          {8},   
          {8},   
          {8}]  0.04431926 
      24 [{8},   
          {8},   
          {8},   
          {8},   
          {8}]  0.03771097 
      25[{8},   
          {8},   
          {8},   
          {8},   
          {8},   
          {8}]  0.02928619 
      26 [ {8},   
          {8},   
          {8},   
          {8},   
          {8},   
          {8},   
          {8} ]  0.02185351
user1840131
  • 11
  • 1
  • 6
  • 1
    If you need help choosing a statistical method to analyze your data, consider posting to [stats.se] or [datascience.se] instead. Stack Overflow is for specific programming tasks. If you knew exactly what you wanted to do with your data, but didn't know *how* to do it in R, then make that more clear in your question. – MrFlick Apr 21 '15 at 22:30
  • thank you – user1840131 Apr 21 '15 at 23:07

1 Answers1

0

There is not a lot in this data, unfortunately.

Forget the 1 itemsets.

Choose the most confident rule you have for prediction. Match n-1 items, predict the last.

Effeftively there are three predictions in your data:

  • 6 visits 9 next
  • 8 stays at 8
  • 9 stays a 9
Has QUIT--Anony-Mousse
  • 76,138
  • 12
  • 138
  • 194