Now I am trying to include apriori algorithm in my website -shopping cart - and I choose php array and mysql to do that . I think my code gonna be complicated and not as fast as I wish especially in real time , Is there any suggestion or open source tool could help ? Note : The ways I tried are connecting to weka and R but they aren't effective

Am not sure what you are looking for. Can you explain with an example?

why not record the sales on different table and sort the best combination or group of products that sell the most. Don't make it too complicated. All you need to do is define your own restrictions. e.g number of item sold, which pair of products are the top selling. But IMHO, it is more of a speculation though. We are betting on the odds that customers will be buying the same items based on the Algo's approximation.

The thing is nobody can calculate the degree of the influence of the environment. For example, we cannot prove that in the month of December ice cream and sugar cone combination will have the same sales as during the months of June, July, August and early September. Relying on this combination can cause an over-froozed ice cream and soggy sugar cones. Besides, no one can really predict what will be the trend a week from today.

This is how the data mining company makes money, and so as the buy on red and sell on green infomercials.

What is certain though is that, people will buy paper and mechanical pencil with eraser or paper ,pencils without eraser and eraser. But less likely to buy just an eraser and very seldom even-never buy a combination of paper and eraser. Did you see the logic behind this? So, based on this example we have discovered two weak products. What are they? Can you tell me the Algo's probable approximation? What would be the Algo's top product combination if any?