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
ログから生まれる施策 / actions born from logs
Search
Hiroka Zaitsu
December 13, 2016
Technology
1
5.4k
ログから生まれる施策 / actions born from logs
WEB DIRECTORS SESSION Vol.2
http://peatix.com/event/219845
Hiroka Zaitsu
December 13, 2016
Tweet
Share
More Decks by Hiroka Zaitsu
See All by Hiroka Zaitsu
Vertex AI Matching Engine と CLIP を使って EC サービスの類似画像検索機能を作る / Development of similar image search function for EC services using Vertex AI Matching Engine and CLIP
zaimy
0
470
BigQuery の日本語データを Dataflow と Vertex AI でトピックモデリング / Topic modeling of Japanese data in BigQuery with Dataflow and Vertex AI
zaimy
1
4.7k
データサイエンティストの仕事紹介 / Data Scientist Job Introduction
zaimy
1
460
GMOペパボのサービスと研究開発を支えるデータ基盤の裏側 / Inside Story of Data Infrastructure Supporting GMO Pepabo's Services and R&D
zaimy
1
1.2k
正則化とロジスティック回帰/machine-learning-lecture-regularization-and-logistic-regression
zaimy
0
7.5k
ECサイトにおける閲覧履歴を用いた購買に繋がる行動の変化検出 / Change Detection in Behavior Followed by Possible Purchase Using Electronic Commerce Site Browsing History
zaimy
0
760
trinity で Cloud Composer に ワークフローを簡単デプロイ / Easy workflow deployment to Cloud Composer with trinity
zaimy
0
740
ハンドメイド作品を対象としたECサイトにおける大量生産品の検出 / Detection of Mass-produced Goods at EC Site to Trade Handmade Goods
zaimy
3
4.4k
キャリアキーノート2018 / Career Keynote 2018
zaimy
1
1.9k
Other Decks in Technology
See All in Technology
SWC Transformerから見るTypeScript関数記述ベストプラクティス
fujiyamaorange
1
180
kcp: Kubernetes APIs Are All You Need #techfeed_live / TechFeed Experts Night 28th
ytaka23
1
190
My road to OSEE Part1
yunolay
0
110
シンプルなHITL機械学習と様々なタスクにおけるHITL機械学習
naohachi89
0
320
「知的単純作業」を自動化する、地に足の着いた大規模言語モデル (LLM) の活用
nrryuya
8
8.3k
グイグイ系QAエンジニアでやっていくよ!
____rina____
0
770
動画配信サービスのフロントエンド実装に学ぶ設計原則
yud0uhu
1
130
Databricksの生成AI戦略
taka_aki
1
370
OPENLOGI Company Profile for engineer
hr01
1
2.2k
QAエンジニアが伝えたい品質保証の羅針盤 / Compass for Quality Assurance
mii3king
2
330
Cloudflare WorkersがPythonに対応したので試してみた
miura55
0
190
Kaggleで学ぶ系列データのための深層学習モデリング
yu4u
7
1.7k
Featured
See All Featured
Practical Orchestrator
shlominoach
183
9.8k
Statistics for Hackers
jakevdp
790
220k
Building Applications with DynamoDB
mza
88
5.7k
From Idea to $5000 a Month in 5 Months
shpigford
377
45k
Being A Developer After 40
akosma
67
580k
Clear Off the Table
cherdarchuk
86
310k
Navigating Team Friction
lara
179
13k
How to name files
jennybc
65
94k
What the flash - Photography Introduction
edds
64
11k
The MySQL Ecosystem @ GitHub 2015
samlambert
244
12k
Put a Button on it: Removing Barriers to Going Fast.
kastner
58
3.1k
Debugging Ruby Performance
tmm1
70
11k
Transcript
ϖύϘͷϩάz׆༻zج൫ʮ#JHGPPUʯ ࡒେՆ(.01FQBCP *OD 8&#%*3&$50344&44*0/7PM ϩά͔Βੜ·ΕΔࢪࡦ
σʔλαΠΤϯςΟετσΟϨΫλʔ ࡒେՆ![BJNZ NJOOFࣄۀ෦
ϋϯυϝΠυϚʔέοτNJOOF IUUQTNJOOFDPN
࣍ w8FCαʔϏεͷߦಈϩά wϖύϘͷϩάz׆༻zج൫ʮ#JHGPPUʯ wࢪࡦͷ׆༻ྫ
8FCαʔϏεͷߦಈϩά
8FCαʔϏεͷߦಈϩά wϢʔβʔ͕αʔϏεΛར༻ͨ͠ࡍͷཤྺ wʮ͍ͭʯʮ୭͕ʯʮԿΛʯߦͬͨͷ͔ w݁Ռ͚ͩͰͳͦ͘͜ʹࢸΔաఔɺ݁Ռ్தͰఘΊͯ͠·ͬͨϢʔ βʔͷߦಈ͔Δ
ߦಈϩάΛͬͯ ΑΓྑ͍αʔϏε!
׆༻·Ͱͷஈ֊ wऩूߦಈϩά͕ग़ྗ͞ΕɺऔΓ·ͱΊΒΕ͍ͯΔঢ়ଶ wੳऔΓ·ͱΊͨߦಈϩάΛࢹ֮ԽɺੳͰ͖Δঢ়ଶ w׆༻ੳͨ͠ߦಈϩάΛͱʹܧଓతͳαʔϏεվળ͕ߦ͍͑ͯΔঢ়ଶ
#JHGPPU IUUQTJDPOTDPN
#JHGPPU wϖύϘͷϩάz׆༻zج൫ w෯͍δϟϯϧͷ8FCαʔϏε wϋϯυϝΠυϚʔέοτNJOOF wωοτγϣοϓ࡞Χϥʔϛʔγϣοϓ wϩϦϙοϓʂϨϯλϧαʔόʔ wϩάऩू͔Β׆༻·Ͱͷ֤ஈ֊ʹ͓͍ͯ൚༻తʹར༻Ͱ͖Δશࣾج൫
#JHGPPU rack-bigfoot Service Request Activity log Services DB Attribute Big
Cube Cube https://speakerdeck.com/monochromegane/pepabo-log-infrastructure-bigfoot Bandit algorithm/ Recommendation Re-targeting Feedback Name identification BI/Visualize
ऩू
໊دͤ wαʔϏεͷΞΧϯτͱΫϥΠΞϯτʢɾϒϥβʣΛඥ͚ͮ wैདྷΫϥΠΞϯτ͝ͱͷܭଌ wະϩάΠϯঢ়ଶͷΞΧϯτͦͷޙϩάΠϯͨ͠λΠϛϯάͰաڈʹ Ḫͬͯඥ͚ͮΒΕΔ wαʔϏεΛލ͍ͩඥ͚ͮՄೳ
ੳ
#JH$VCFͱ$VCF w#JHGPPU্ͷશͯͷϩάΛ#JH$VCFʹू wੳͷΓޱʢσΟϝϯγϣϯͱϝδϟʔʣ͕ܾ·ͬͨΒ$VCFʹूܭ wྫʣ࣌ؒ͝ͱจֹۚ wߴͳࢀর͕Մೳ Activity Big Cube Cube HiveQL
SQL Dashboard Ad-hoc query Analyst Managers, Product owners, Promotion groups
ࢹ֮Խ w5BCMFBV wIUUQXXXUBCMFBVDPN wΞυϗοΫͳੳෳࡶͳσʔλ݁߹ʹ w3FEBTI wIUUQTSFEBTIJP w୭͕ࢀরͰ͖ΔμογϡϘʔυʹ
ύεੳ wίϯόʔδϣϯʹؔ͢Δੳͷ͏ͪɺܦ࿏ʢύεʣʹओ؟Λஔ͍ͨͷ wύεͷ࣌ؒ͞ྨܕ wΞτϦϏϡʔγϣϯϞσϧϖʔδͷग़ݱҐஔʹΑΓॏΈΛม͑Δ referrer landing last cv ??? Point
Analytics referrer landing last Path Analytics cv
ࢪࡦͷ׆༻
ϢʔβʔͷϦςϯγϣϯ wߦಈϩά͔ΒϢʔβʔΛநग़ wΧʔτʹ౸ୡ͕ͨ͠ങΘͳ͔ͬͨϢʔβʔʢ͍ΘΏΔΧʔτམͪʣ wظؒʹಉ͡࡞ΛԿݟ͍ͯΔϢʔβʔ wಛఆͷ݅ʹ߹க͢Εϓογϡ௨ϝʔϧͰϦςϯγϣϯ wϨγʔτϝʔϧͱಉͷ։෧ HiveQL Re-targeting
ࠂ࿈ܞ wߦಈϩά͔ΒϢʔβʔΛηάϝϯτ wϢʔβʔʹؔ࿈ੑͷߴ͍ࠂΛදࣔ wطʹϦʔνͨ͠ϢʔβʔʹࠂΛදࣔ͠ͳ͍ʢσϦλʔήςΟϯάʣ
όϯσΟοτΞϧΰϦζϜ w࡞ݕࡧը໘͔ΒͷརӹΛ࠷େԽ͍ͨ͠ w$53ͷҟͳΔύλʔϯͷ࡞ݕࡧΞϧΰϦζϜ wച্ʹର͢Δൺॏ͕ଟ͍ҝʹγεςϜมߋͷϦεΫ͕ߴ͍ w࠷ྑͷύλʔϯΛ͍ͳ͕ΒΑΓྑ͍ύλʔϯΛಈతʹ୳͢
όϯσΟοτΞϧΰϦζϜ w&QTJMPO(SFFEZ"MHPSJUIN w֬ Џ Ͱͦͷ࣌ͷ࠷ظ͕ߴ͍ύλʔϯΛ༻ʢ׆༻ʣ w֬ЏͰϥϯμϜʹબͨ͠ύλʔϯΛ༻ʢ୳ࡧʣ Activity Epsilon-Greedy algorithm User
1-ε: exploitation ε/pattern: exploration Click / Not click Import
όϯσΟοτΞϧΰϦζϜ wτϨʔυΦϑͷղܾ w׆༻͔Γ͍ͯ͠ΔͱݱࡏͷظΑΓྑ͍Λݟ͚ͭΒΕͳ͍ w୳ࡧ͔Γ͍ͯ͠Δͱظ͕ߴ͍͕ར༻͞Εͳ͍
Ϩίϝϯσʔγϣϯ wڠௐϑΟϧλϦϯάʹΑΔʮ͋ͳͨʹ͓͢͢Ίͷ࡞Ոʯ !NPOPDISPNFHBOFʮNJOOFNFFUT)JWFNBMMʯ IUUQTTQFBLFSEFDLDPNNPOPDISPNFHBOFQFQBCPNJOOFNBUSJYGBDUPSJ[BUJPOJOIJWFNBMM
ϩάz׆༻zج൫
ϩάz׆༻zج൫ wᶃਓ͕ؒੳ݁ՌΛݟͯߦ͏੩తͳʮ׆༻ʯ wࢪࡦͷਫ਼্ wᶄࣗಈԽ͞ΕͨϑΟʔυόοΫʹΑΔಈతͳʮ׆༻ʯ wͳΊΒ͔ͳγεςϜ