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
Sentiment Analysis
Search
Matt Kirk
October 21, 2013
Technology
0
160
Sentiment Analysis
Talk given in Portland October 21, 2013
Matt Kirk
October 21, 2013
Tweet
Share
More Decks by Matt Kirk
See All by Matt Kirk
Overcoming Big Data Apathy
hexgnu
2
330
Other Decks in Technology
See All in Technology
AI人生苦節10年で会得したAIがやること_人間がやること.pdf
shibuiwilliam
1
260
alecthomas/kong はいいぞ
fujiwara3
6
1.3k
【CEDEC2025】『Shadowverse: Worlds Beyond』二度目のDCG開発でゲームをリデザインする~遊びやすさと競技性の両立~
cygames
PRO
1
250
株式会社島津製作所_研究開発(集団協業と知的生産)の現場を支える、OSS知識基盤システムの導入
akahane92
1
1.3k
Findy Freelance 利用シーン別AI活用例
ness
0
180
多様なニーズに応える Movable Type ラインナップ 全紹介
masakah
0
110
解消したはずが…技術と人間のエラーが交錯する恐怖体験
lamaglama39
0
170
地域コミュニティへの「感謝」と「恩返し」 / 20250726jawsug-tochigi
kasacchiful
0
120
【CEDEC2025】LLMを活用したゲーム開発支援と、生成AIの利活用を進める組織的な取り組み
cygames
PRO
1
2.3k
経験がないことを言い訳にしない、 AI時代の他領域への染み出し方
parayama0625
0
290
少人数でも回る! DevinとPlaybookで支える運用改善
ishikawa_pro
5
2.2k
人に寄り添うAIエージェントとアーキテクチャ #BetAIDay
layerx
PRO
8
1.8k
Featured
See All Featured
Reflections from 52 weeks, 52 projects
jeffersonlam
351
21k
Thoughts on Productivity
jonyablonski
69
4.8k
What's in a price? How to price your products and services
michaelherold
246
12k
Agile that works and the tools we love
rasmusluckow
329
21k
A Tale of Four Properties
chriscoyier
160
23k
"I'm Feeling Lucky" - Building Great Search Experiences for Today's Users (#IAC19)
danielanewman
229
22k
Mobile First: as difficult as doing things right
swwweet
223
9.8k
Designing for humans not robots
tammielis
253
25k
Improving Core Web Vitals using Speculation Rules API
sergeychernyshev
18
1k
Unsuck your backbone
ammeep
671
58k
I Don’t Have Time: Getting Over the Fear to Launch Your Podcast
jcasabona
33
2.4k
GraphQLとの向き合い方2022年版
quramy
49
14k
Transcript
Sentiment Analysis Matthew Kirk - Modulus 7 Monday, October 21,
13
The Challenge Monday, October 21, 13
Big Data Languages Java Python R Julia Clojure Monday, October
21, 13
Ruby has tools too Monday, October 21, 13
Monday, October 21, 13
Ruby is not Complex Math Monday, October 21, 13
A Solution • Listen To Speeches • Watch Others •
Read Books and Materials Monday, October 21, 13
Sentiment Monday, October 21, 13
The Goal is to Classify Data Monday, October 21, 13
Monday, October 21, 13
Machine Learning Monday, October 21, 13
Step 1: Collect Data • https://github.com/jperla/sentiment-data • Finance • Harvard
Negative Words • Hu Liu Words (Positive / Negative) • Pang Lee Movie Reviews Monday, October 21, 13
Now What? Monday, October 21, 13
Chess Monday, October 21, 13
The Features of a Chess Piece • Color • Location
on Board • Type (Rook, Queen, Pawn...) • Material • Length • Width • Temperature Monday, October 21, 13
Linear Regression Monday, October 21, 13
Support Vector Machines Monday, October 21, 13
That’s It! Monday, October 21, 13
Oh right Sentiment Analysis Monday, October 21, 13
Features of Text Monday, October 21, 13
We will use words only Monday, October 21, 13
Monday, October 21, 13
Linear Regression has a Curse Monday, October 21, 13
Support Vector Machines • Using the example of linear support
vector machines (our chess example) we’d end up with a similar response to the clustering Monday, October 21, 13
But there’s a trick (The Kernel Trick) Monday, October 21,
13
Parameter C • Find through Cross Validation • Try all
C’s from 2^-15 to 2^15 accepting the lowest error. Monday, October 21, 13
Demonstration Play with the tool at sentiment-analyzer.heroku.com Monday, October 21,
13
Testing Monday, October 21, 13
Conclusion This is just the beginning Go learn more become
more adept at data analysis Monday, October 21, 13
QR CODE! Monday, October 21, 13
modulus7.com/svm-talk @mjkirk Monday, October 21, 13
http://en.wikipedia.org/wiki/File:Kernel_Machine.png http://rickmanelius.com/article/do-you-dread-emails Comedy Central http://en.wikipedia.org/wiki/File:Chess_board_opening_staunton.jpg Star Trek Photo Credits Monday,
October 21, 13