I found that for an IP 95891 mapping of (app+Device)---> OS is one-many\nThe change occurred in a span of 3 days. Does that means the user is changing his OS everyday? or is it an indication of some fraudulent clicking?
", "messages": [ { "id": "291627", "postDate": "03/06/2018 14:11:44", "content": "I found that for an IP 95891 mapping of (app+Device)---> OS is one-many\nThe change occurred in a span of 3 days. Does that means the user is changing his OS everyday? or is it an indication of some fraudulent clicking?
", "rawMarkdown": "I found that for an IP 95891 mapping of (app+Device)---> OS is one-many\nThe change occurred in a span of 3 days. Does that means the user is changing his OS everyday? or is it an indication of some fraudulent clicking?", "votes": null }, { "id": "291651", "postDate": "03/06/2018 15:14:22", "content": "You are right, I think it is caused by fraudulent traffic, some ad channels are responsible for this.
", "rawMarkdown": "You are right, I think it is caused by fraudulent traffic, some ad channels are responsible for this.", "votes": null }, { "id": "291782", "postDate": "03/06/2018 20:51:08", "content": "What if it was a shared IP?
", "rawMarkdown": "What if it was a shared IP?", "votes": null }, { "id": "293141", "postDate": "03/09/2018 08:47:59", "content": "Just a thought encoded IP is at user level or not is not sure. It might be at network level also or may be dynamic IP by service provider
", "rawMarkdown": "Just a thought encoded IP is at user level or not is not sure. It might be at network level also or may be dynamic IP by service provider", "votes": null }, { "id": "293493", "postDate": "03/10/2018 01:18:14", "content": "@Nagaraju I'm learning to do data exploration, curious about how you find out about such cases? Thanks!
", "rawMarkdown": "Nagaraju I'm learning to do data exploration, curious about how you find out about such cases? Thanks!", "votes": null }, { "id": "293603", "postDate": "03/10/2018 08:04:35", "content": "Hi chisq\nJust based on basic click stream data knowledge , I assumed that multiple OS values to single IP is not possible unless it's a shared IP. So sorted the data descending on the basis of number of clicks per IP and found this anomaly case \nHope it helps :)
", "rawMarkdown": "Hi chisq\nJust based on basic click stream data knowledge , I assumed that multiple OS values to single IP is not possible unless it's a shared IP. So sorted the data descending on the basis of number of clicks per IP and found this anomaly case \nHope it helps :)", "votes": null }, { "id": "293799", "postDate": "03/10/2018 16:38:15", "content": "I see. So first group (#of clicks) by IP and then sort by the number is an interesting trick I can probably apply to other cases. Thanks for sharing!
", "rawMarkdown": "I see. So first group (#of clicks) by IP and then sort by the number is an interesting trick I can probably apply to other cases. Thanks for sharing!", "votes": null }, { "id": "301719", "postDate": "03/23/2018 06:44:27", "content": "You might also like to group #clicks by (ip,device,os) to narrow it down to fewer groups of users. Or some grouping of os (it's long-tailed).
", "rawMarkdown": "You might also like to group #clicks by (ip,device,os) to narrow it down to fewer groups of users. Or some grouping of os (it's long-tailed).", "votes": null }, { "id": "2080092", "postDate": "12/29/2022 20:54:54", "content": "what about pirvate ips pools
", "rawMarkdown": "what about pirvate ips pools", "votes": null } ], "comments": [ { "id": 2080092, "author_name": "noman223", "author_url": "", "post_date": "12/29/2022 20:54:54", "content": "what about pirvate ips pools
", "votes": null, "replies": [] }, { "id": 291651, "author_name": "aaronyin", "author_url": "", "post_date": "03/06/2018 15:14:22", "content": "You are right, I think it is caused by fraudulent traffic, some ad channels are responsible for this.
", "votes": null, "replies": [] }, { "id": 291782, "author_name": "muhammadalfiansyah", "author_url": "", "post_date": "03/06/2018 20:51:08", "content": "What if it was a shared IP?
", "votes": null, "replies": [] }, { "id": 293141, "author_name": "vishaljindal0", "author_url": "", "post_date": "03/09/2018 08:47:59", "content": "Just a thought encoded IP is at user level or not is not sure. It might be at network level also or may be dynamic IP by service provider
", "votes": null, "replies": [] }, { "id": 293493, "author_name": "chisq1564", "author_url": "", "post_date": "03/10/2018 01:18:14", "content": "@Nagaraju I'm learning to do data exploration, curious about how you find out about such cases? Thanks!
", "votes": null, "replies": [ { "id": 293603, "author_name": "rajuspartan", "author_url": "", "post_date": "03/10/2018 08:04:35", "content": "Hi chisq\nJust based on basic click stream data knowledge , I assumed that multiple OS values to single IP is not possible unless it's a shared IP. So sorted the data descending on the basis of number of clicks per IP and found this anomaly case \nHope it helps :)
", "votes": null, "replies": [] }, { "id": 293799, "author_name": "chisq1564", "author_url": "", "post_date": "03/10/2018 16:38:15", "content": "I see. So first group (#of clicks) by IP and then sort by the number is an interesting trick I can probably apply to other cases. Thanks for sharing!
", "votes": null, "replies": [] }, { "id": 301719, "author_name": "smcinerney", "author_url": "", "post_date": "03/23/2018 06:44:27", "content": "You might also like to group #clicks by (ip,device,os) to narrow it down to fewer groups of users. Or some grouping of os (it's long-tailed).
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "291627": "I found that for an IP 95891 mapping of (app+Device)---> OS is one-many\nThe change occurred in a span of 3 days. Does that means the user is changing his OS everyday? or is it an indication of some fraudulent clicking?", "291651": "You are right, I think it is caused by fraudulent traffic, some ad channels are responsible for this.", "291782": "What if it was a shared IP?", "293141": "Just a thought encoded IP is at user level or not is not sure. It might be at network level also or may be dynamic IP by service provider", "293493": "Nagaraju I'm learning to do data exploration, curious about how you find out about such cases? Thanks!", "293603": "Hi chisq\nJust based on basic click stream data knowledge , I assumed that multiple OS values to single IP is not possible unless it's a shared IP. So sorted the data descending on the basis of number of clicks per IP and found this anomaly case \nHope it helps :)", "293799": "I see. So first group (#of clicks) by IP and then sort by the number is an interesting trick I can probably apply to other cases. Thanks for sharing!", "301719": "You might also like to group #clicks by (ip,device,os) to narrow it down to fewer groups of users. Or some grouping of os (it's long-tailed).", "2080092": "what about pirvate ips pools" }, "source": "meta" }