This ISCX is a benchmark intrusion detection dataset with contains 7 days of synthetically recorded packet details replicating the real time network traffic by labelling the attacks. I would like to use a neural classifier to import this data and classify them for DDOS. The problems are very big, how can I resolve them?
@Abebe Abeshu Diro
Each dataset should have a documentation that explains the complete time of attacks and benign activities. Pls check this page and scrol down to find the detail of the attack and benign activities:
http://www.unb.ca/cic/datasets/ids-2017.html
@Aad ba
For any network traffic datasets, if you have issue for two flows with the same IPs, the best way for labelling them is checking the timestamp. If you know the attack time, then you can label the flows properly.
@Mohamed Debashi
Pls try the new datset CICIDS2017, that has the labelled .csv files also.
By applying Sampling technique to extract the some samples of the data and after taking sample apply the concept of training and testing.
All you need to do is to fill a request form and they will mail you the login credentials to download. But its a huge dataset.
I have worked on this data
do one thing u should do some prepossessing method before applying any classifier
otherwise do it in rapid miner tool
If u want this data set just, apply request form and wait
i waited around three months until they gave me pwd and username
@Theyazn : since you worked on the dataset, could you please tell me how can I identify the HTTP DOS or Brute force attack from the huge packet details.. What kind of pre processing methos would you suggest to detect Dos or brute force attacks
I have been told by UNB that the dataset is no longer available but maybe something will come out next year.
Hi, we already changed our website on last month and you can find all datasets now in the new address:
http://www.unb.ca/research/iscx/dataset/index.html
two new Datasets also will be up soon.
Good luck
Hi,
If you have our dataset, then you can use the data for your algorithms. If you have any issue pls contact us based on the contact address for each dataset in ISCXdataset webpage.
Good luck
Hi,
If you have our dataset, then you can use the data for your algorithms. If you have any issue pls contact us based on the contact address for each dataset in ISCXdataset webpage.
Good luck
This dataset looks interesting .Unfortunately it is confusing me.It is in pcap file.When you say about being labeled how do find it? Or should we make our own dataset (preprocessed) based on this wireshark capture ?Please help me to understand.
I don't know how this dataset is said to be labeled. Anyone who explains?
Hi Guys,
If you have any question please read the related documents and contact us directly. For example, in IDS2012, the author mentioned that each day is for which attack and also which day is benign.
BTW, the new dataset and ISCXFlowMeter which is a feature extraction application are ready:
http://unb.ca/cic/research/datasets/
Good luck,
@Arash Habibi Lashkari, thanks so much. You mean that if we know the date of the collection, indirectly it tells us attack/normal. But, how can we know how much on the data is attack/normal on the date both of them collected?
@Arash Habibi Lashkari, Hello. I understand that access to dataset is through mentioned website and/or by email request? I would like to use it for my research, so I've sent you a mail concerning this matter. Thank you.
I read about this dataset and explored even the traffic using Wireshark, These is no any labelling to this dataset, just saying that day for example has DoS attack. it is not labeled packet by packet for me. I could not see any labelling if excited. I wasted a lots of time trying to understand how it is labeled. I sent an email to get an answer how it is labeled but no answer. any one explain??
If any one knows how it is labeled please let me know I can help you with the rest, regards..
Hi Folks,
Please use our new IDS dataset that has the .csv files completely labelled. You can fine the .pcap files in the dataset also:
http://www.unb.ca/cic/datasets/ids-2017.html
@Abebe Abeshu Diro
Each dataset should have a documentation that explains the complete time of attacks and benign activities. Pls check this page and scrol down to find the detail of the attack and benign activities:
http://www.unb.ca/cic/datasets/ids-2017.html
@Aad ba
For any network traffic datasets, if you have issue for two flows with the same IPs, the best way for labelling them is checking the timestamp. If you know the attack time, then you can label the flows properly.
@Mohamed Debashi
Pls try the new datset CICIDS2017, that has the labelled .csv files also.
I would like to use your dataset to analyse my IDS algorithm, but I cannot find any download link on the website you provided...
Please post a link from where the dataset can be downloaded...
@Li Min @Emmanuel Ogu
Write an email to Mr. Habibi, we will provide you with a temp download link.
- Badges
- Science topic