hello. i want to download "the CM-SPADE algorithm (Fournier-Viger et al, 2014, powerpoint)" in here but this link is notfound please help me. thanks.



Edited 1 time(s). Last edit at 05/22/2014 05:48PM by webmasterphilfv.

Hi,

Thanks for letting me know. The link for the article is now fixed.

The link for the powerpoint has also been fixed

Best,



Edited 3 time(s). Last edit at 05/22/2014 05:51PM by webmasterphilfv.

your welcome my best feirend. I kind of want to repay you. you help me in priblems. so i say tanks.

oh. i have another question. what is the " Join count " in output of algorithms?

SPADE and CM-SPADE performs joins of ID-LISTS to calculate the support of a pattern.

For example, to calculate the support of a pattern <a,b>, they will perform the join of the idlist of a and the idlist of b.

The join count is the total number of join operations that is performed. For CM-SPADE it is usually smaller than SPADE because it uses a strategy (described in the paper) to avoid some join operations, which makes this algorithm faster than SPADE.

hello. tanks for reply. my friend, i have another question. we have two final work in MA: 1) Seminar
2) Thesis
i do my Seminar about the algorithms of sequence pattern. i want continue this field for my Thesis in social network. on the other word sequence pattern mining usage in social network! is this good field ? is this update field ? what title you suggest me? what title i search in web and sits for give papers? please quid me.

Hello,

You could certainly apply seq. pattern mining on some social network. But what you need to think is what is your end goal? Seq. Pattern mining is a technique. But you need to have some purpose to apply it.

A good thesis topic should not be "sequential pat mining in social network". It should be a thesis title that explains what is your goal.

For example, I will give you a random example. A thesis topic topic could be "Predicting friendships relationships on social networks using sequential patterns". This topic may not be good. But I just give an example a title that mention what is the goal of your research.

To find a topic, it takes time. I suggest to read about social network mining on Google or Google Shcholar to see what has been done.

For social network mining, another issue to consider is that you will need to have some data. There exists some public datasets but they may not have time or sequential data. Or you may want to collect your own data. In any case, the availability of the data is something to think about when choosing the topic.

Best,

hello. oh , your answer was wery gooooooooooooooooood. tanks. i think about any of two items that you say : true topic and data. i want work about "Trend Analysis on social network using sequence pattern" is it good?
are you have any site for dataset download?

Any topic can be good if you can do something new or something better than what has already been done. Now, to know what has already been done, you will need to read some articles on this topic. Since I did not do that, I don't know either.

There exists many websites offering datasets. You can search "social networks datasets" on Google to find some. Otherwise, another way is to contact the author of a paper by e-mail to ask them for their data, or to collect your own data.

webmasterphilfv tanks alot sir

Dear Sir,
Please give me few examples to understand the i-extension and s-extension.
Perform a depth-first search by joining items to each pattern by i-extension and s-extension concepts.

malsoru
let a and b two items.
(ab) is i-extension (a and b are occur in same time)
ab or ba are s-extension (a and b occur in different time).
its easy to understand.