Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Sign up for free
Menu
Search
Features
All features
Private URLs
Password Protection
Custom URLS
Scheduled publishing
Remove Branding
Restrict embedding
Deck Collections
Notes
Features
All features
Private URLs
Password Protection
Custom URLS
Scheduled publishing
Remove Branding
Restrict embedding
Deck Collections
Notes
Explore
Featured decks
Featured speakers
Programming
Technology
Storyboards
Explore
Featured decks
Featured speakers
Programming
Technology
Storyboards
Pricing
Search
Sign in
Sign up for free
Streamingdatenstrukturen zum Analysieren von Nu...
Search
Torsten Bøgh Köster
November 03, 2015
Technology
930
1
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
Streamingdatenstrukturen zum Analysieren von Nutzeraktionen in Echtzeit
Torsten Bøgh Köster
November 03, 2015
More Decks by Torsten Bøgh Köster
See All by Torsten Bøgh Köster
Wenn die Vibes nicht reichen: Observability für GenAI Anwendungen
tboeghk
0
7
LLMs im Griff: Observability, Tracing und Security
tboeghk
0
94
LLMs im Griff: Observability, Tracing und Security
tboeghk
1
69
Oder mache ich es lieber selbst? Wie sich Kosten und Geopolitik auf Cloud-Betrieb auswirken
tboeghk
0
120
Taking an abandoned Solr search from zero to GenAI hero
tboeghk
0
88
Oder mache ich es lieber selbst? Wie sich Kosten und Geopolitik auf Cloud-Betrieb auswirken
tboeghk
0
66
🔪 How we cut our AWS costs in half
tboeghk
0
470
Shared Nothing Logging Infrastructure
tboeghk
0
140
Beyond Cloud: A road trip into AWS and back to bare metal
tboeghk
1
130
Other Decks in Technology
See All in Technology
Getting Started with Google Antigravity
glaforge
0
320
WasmをRustにする
sansantech
PRO
0
290
USB PDを使ってスマホを労わりたい
raspython3
0
180
Active Record Connection Adaptersにおける長年の課題に再び取り組む
yahonda
0
220
データエンジニアの次の戦場は "プロセスエンジニアリング"
sansantech
PRO
1
120
ソフトウェア ファクトリー - 再現性のあるソフトウェアの工場づくり / Software factory
iktakahiro
7
5.1k
hackers_champloo_2026_tidb
makocchi
0
170
PHPカンファレンス関西の招待状 / Invitation to PHP Conference Kansai
ykanoh
0
210
1つのリクエストを追って理解する - llm-dのP/D分離
akhino
1
170
今さら理解するウェブフォントver.2026
takanorip
3
910
二歩目のフロントエンド・ウェブアクセシビリティ
camiha
2
670
宣言型部分更新は何を解決しどこへ向かうのか
shuta13
1
480
Featured
See All Featured
RailsConf 2023
tenderlove
30
1.6k
How to optimise 3,500 product descriptions for ecommerce in one day using ChatGPT
katarinadahlin
PRO
3
3.9k
Redefining SEO in the New Era of Traffic Generation
szymonslowik
1
480
Impact Scores and Hybrid Strategies: The future of link building
tamaranovitovic
0
460
A Modern Web Designer's Workflow
chriscoyier
699
190k
Are puppies a ranking factor?
jonoalderson
2
4.1k
Why Our Code Smells
bkeepers
PRO
340
58k
The Art of Programming - Codeland 2020
erikaheidi
57
14k
The Invisible Side of Design
smashingmag
301
52k
Chrome DevTools: State of the Union 2024 - Debugging React & Beyond
addyosmani
10
1.4k
Designing Powerful Visuals for Engaging Learning
tmiket
1
610
How STYLIGHT went responsive
nonsquared
100
6.3k
Transcript
Streamingdatenstrukturen zum Analysieren von Nutzeraktionen in Echtzeit WJAX 2015 Torsten
Bøgh Köster (Shopping24) 3. November 2015
Agenda
@tboeghk CTO shopping24 internet group Search Technology Meetup Hamburg Search,
build, delivery, code quality, road bike
None
Open Source Power. Delivered.
search system @ shopping24
Anwendungsfall 1
Produkte gezielt löschen
Bloomfilter
Funktionsweise Bloomfilter
Lokale Bloomfilter
Anwendungsfall 2
Relevante Produkte je Suchanfrage
Benutzeraktionen einfangen
Benutzeraktionen verarbeiten
None
You cannot scale into real time!
Stream Mining
Logstash FTW!
Popularitätswerte als Rankingfaktor
Mit Mandanten exponentielle Datenpunkte
None
The Count-Min-Sketch: A Bloomfilter on Steroids
Wie geht das?
None
Relevanz von Datenpunkten im zeitlichen Verlauf
Exponential Decay
Punisher.java
Anwendungsfall 3
Populäre Suchen in der Autocompletion boosten
Heavy Hitters a.k.a. TopK
Und so geht’s
Und sonst so?
BitSet / SparseFixedBitSet: Non-probabilistic existence test
HyperLogLog: Estimating cardinality
Data Sampling: Reduce large data sets using statistics. Use for:
expensive computations
Data Sampling: Existence computation reduces large data sets to constant
~700pcs
Packed Ints: Reduce heap size for large integer arrays
Packed Ints: Further heap reduction with an offset
None
@see ‣T-Digest (Ted Dunning): https:// www.mapr.com/blog/better-anomaly- detection-t-digest-whiteboard- walkthrough ‣Realtime personalization
(Mikio Braun): http:// blog.mikiobraun.de/2014/05/bbuzz- realtime-personalization- recommendation-stream-mining.html ‣Algorithms and data structures that power Lucene and Solr (Adrien Grand): http:// berlinbuzzwords.de/session/ algorithms-and-data-structures- power-lucene-and-elasticsearch ‣HypeLogLog in Reds: http:// redis.io/commands#hyperloglog ‣Sketching & Scaling Series: http://blog.kiip.me/engineering/ sketching-scaling-part-1-what- the-is-sketching/
Questions? @tboeghk developer.s24.com
[email protected]