Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Jan Stępień - Tracking those who Track
Search
MunichDataGeeks
July 02, 2013
Technology
1
200
Jan Stępień - Tracking those who Track
Talk by Jan Stępień at the firsta Munich DataGeeks Meetup
Data: 02.07.2013
MunichDataGeeks
July 02, 2013
Tweet
Share
More Decks by MunichDataGeeks
See All by MunichDataGeeks
Florian Haselbeck- Advancing Synthetic Protein Design with Large Language Models
munichdata
0
52
Tobias Ladner- Formal Verification of Neural Networks in Safety-Critical Environments
munichdata
0
69
Uladzislau Sazanovich - JetBrains AI: Deep Dive
munichdata
0
57
Jan Hauffa - A Case Study on Retrieval-Augmented Generation for Document Q&A: Experiences and Future Perspectives
munichdata
0
80
Thomas Schmidt - Revolutionizing SQL Data Model Testing: Introducing SQL-Mock by DeepL
munichdata
0
40
Maximilian Duesberg - The Data is Clear - But Humans are not
munichdata
0
89
Dr.Christoph Mittendorf-Beyond Bard and Transformers: Unconventional ML Use Cases
munichdata
0
130
Heidi Seibold - Are (data) scientists bad at science?
munichdata
0
120
Roland Rodde- Vegetation management for powerlines with remote sensing data
munichdata
0
130
Other Decks in Technology
See All in Technology
マルチモーダル / AI Agent / LLMOps 3つの技術トレンドで理解するLLMの今後の展望
hirosatogamo
37
12k
AWS Lambdaと歩んだ“サーバーレス”と今後 #lambda_10years
yoshidashingo
1
180
心が動くエンジニアリング ── 私が夢中になる理由
16bitidol
0
100
いざ、BSC討伐の旅
nikinusu
2
780
ノーコードデータ分析ツールで体験する時系列データ分析超入門
negi111111
0
420
FlutterアプリにおけるSLI/SLOを用いたユーザー体験の可視化と計測基盤構築
ostk0069
0
100
iOSチームとAndroidチームでブランチ運用が違ったので整理してます
sansantech
PRO
0
150
BLADE: An Attempt to Automate Penetration Testing Using Autonomous AI Agents
bbrbbq
0
320
テストコード品質を高めるためにMutation Testingライブラリ・Strykerを実戦導入してみた話
ysknsid25
7
2.7k
B2B SaaSから見た最近のC#/.NETの進化
sansantech
PRO
0
890
これまでの計測・開発・デプロイ方法全部見せます! / Findy ISUCON 2024-11-14
tohutohu
3
370
アジャイルでの品質の進化 Agile in Motion vol.1/20241118 Hiroyuki Sato
shift_evolve
0
170
Featured
See All Featured
A Tale of Four Properties
chriscoyier
156
23k
YesSQL, Process and Tooling at Scale
rocio
169
14k
Cheating the UX When There Is Nothing More to Optimize - PixelPioneers
stephaniewalter
280
13k
Design and Strategy: How to Deal with People Who Don’t "Get" Design
morganepeng
126
18k
Rebuilding a faster, lazier Slack
samanthasiow
79
8.7k
How to Create Impact in a Changing Tech Landscape [PerfNow 2023]
tammyeverts
47
2.1k
The Invisible Side of Design
smashingmag
298
50k
Responsive Adventures: Dirty Tricks From The Dark Corners of Front-End
smashingmag
250
21k
Fashionably flexible responsive web design (full day workshop)
malarkey
405
65k
Learning to Love Humans: Emotional Interface Design
aarron
273
40k
"I'm Feeling Lucky" - Building Great Search Experiences for Today's Users (#IAC19)
danielanewman
226
22k
CSS Pre-Processors: Stylus, Less & Sass
bermonpainter
356
29k
Transcript
Tracking those who track us Jan Stępień
My name is Jan Stępień and I come from Warsaw
Data analysis is not just big data
Data analysis is fun
It all started with ads tracking “like” buttons other irrelevant
things
1. Use an adblock plugin 2. Block all network communication
to unwelcome domains
My machine website.com ads.website.com
My machine website.com ads.website.com
Let’s capture all those requests!
03.2012 – 06.2013 106 414 requests 322 distinct days approx.
330 requests per day
SQLite3 + Incanter + R + Weka
http_if_none_match http_referer http_accept_encoding http_accept http_cookie http_connection http_host http_user_agent http_version path_info
http_accept_charset http_accept_language http_cache_control http_if_modified_since request_method request_path request_uri query_string remote_host remote_addr script_name server_name server_port server_protocol http_dnt timestamp
timestamp
03 04 05 06 07 08 09 10 11 12
01 02 03 04 05 06 15k 10k 5k 0
00 01 02 03 04 05 06 07 08 09
10 11 12 13 14 15 16 17 18 19 20 21 22 23 100 0 200 300 400 500
8k 6k 4k 2k 0 Mo Tu We Th Fr
Sa Su
http_host
www.google-analytics.com 36197 static.adzerk.net 13983 edge.quantserve.com 11659 www.facebook.com 9641 ad.doubleclick.net 3822
pagead2.googlesyndication.com 3764 s.youtube.com 2173 b.scorecardresearch.com 1974 pubads.g.doubleclick.net 1465 googleads.g.doubleclick.net 1231
48.9% of requests sent to domains owned by Google
http_referer
22902 distinct referrers 4692 distinct domains
Let’s try to combine this dataset with something else
Weather influence?
ogimet.com Humidity, min/max/avg temperature, cloud coverage, visibility, rain/snow, wind speed/direction,
etc.
No correlations!
Tags at stackoverflow.com
http://stackoverflow.com/questions/123/title
data.stackexchange.com
Thanks, wordle.net!
Can be my WWW traffic grouped into clusters?
1. Group requests into 15 minute intervals 2. Count domains
per interval
5008 intervals Each described by over 4500 values
1. Select request from popular domains 2. Group requests into
15 minute intervals 3. Count domains per interval
5008 intervals Each described by 95 values Only 2% of
cells with non-zero values
Principal Component Analysis 95 domains → 16 descriptors
X-means K-means based clustering algorithm
cluster 0 1268 cluster 1 702 cluster 2 651 cluster
3 2387 What is the meaning behind these clusters?
3 stackoverflow.com
2 reddit.com redditmedia.com bbc.co.uk
1 linkedin.com dictionary.reference.com meetup.com
0 rubyonrails.pl developer.android.com tex.stackexchange.com amazon.com youtube.com
How accurate is this clustering? Let’s build a classifier on
the original data
0 1 2 3 ← classified as 1188 29 11
40 cluster 0 47 654 1 0 cluster 1 10 1 622 18 cluster 2 50 0 18 2319 cluster 3 cluster 0: rubyonrails.pl developer.android.com amazon.com youtube.com cluster 1: linkedin.com dictionary.reference.com meetup.com cluster 2: reddit.com redditmedia.com bbc.co.uk cluster 3: stackoverflow.com
Let’s wrap up
Data analysis is not just big data
Data analysis is fun
Thank you very much The picture of Warsaw is ©
Dennis Jarvis 2009