applying data science to move beyond keywords for social analysis

20
pplying Data Science to ove Beyond Keywords or Social Analysis

Upload: datasift

Post on 15-Apr-2017

358 views

Category:

Technology


1 download

TRANSCRIPT

Page 1: Applying Data Science to Move Beyond Keywords for Social Analysis

Applying Data Science toMove Beyond Keywords for Social Analysis

Page 2: Applying Data Science to Move Beyond Keywords for Social Analysis

Richard CaudleDirector

Developer Relations

Claudio WeeraratneDirectorProduct

Management

Page 3: Applying Data Science to Move Beyond Keywords for Social Analysis

DATASIFT FORUM

Page 4: Applying Data Science to Move Beyond Keywords for Social Analysis

RUN ON THE BANKS?

Page 5: Applying Data Science to Move Beyond Keywords for Social Analysis

RUN ON THE BANKS?

Page 6: Applying Data Science to Move Beyond Keywords for Social Analysis

AMBIGUITY OF NATURAL LANGUAGE

Page 7: Applying Data Science to Move Beyond Keywords for Social Analysis

RUN ON THE BANKS?

Page 8: Applying Data Science to Move Beyond Keywords for Social Analysis

MOVING BEYOND KEYWORDS

bank

similarity x

with-

draw

similarity y

AND

interaction.content any "rbs,lloyds,hsbc,barclays"AND interaction.content any "withdraw,close,cashpoint,atm"

Page 9: Applying Data Science to Move Beyond Keywords for Social Analysis

CONCEPT MODELING

Page 10: Applying Data Science to Move Beyond Keywords for Social Analysis

KEYWORD RELATIONSHIPS

Page 11: Applying Data Science to Move Beyond Keywords for Social Analysis

CONCEPT MODEL

rbs

VECT

OR S

PACE

#rbs

runningbacks

#hsbc

Page 12: Applying Data Science to Move Beyond Keywords for Social Analysis

OUR APPROACH• Produce a vector space where words are grouped by

their context• Context of a word is given by surrounding words• Perform unsupervised machine learning to learn topics• word2vec is a well known implementation• gensim is a Python library that simplifies word2vec

usage• Resulting model is queryable for similarity (of word

vectors)• Language-agnostic solution

Page 13: Applying Data Science to Move Beyond Keywords for Social Analysis

LEARNING SIMILARITY

Learn to predict a word from surrounding words

"I'm heading to #rbs to close my account"

rbs

account

closerbs

account

close

hsbcbarclays

withdrawbalance

cash money

(1000's posts)

CONCEPT 'BANK'NE

URAL

NET

WOR

K

Page 14: Applying Data Science to Move Beyond Keywords for Social Analysis

LEARNING SIMILARITY

Page 15: Applying Data Science to Move Beyond Keywords for Social Analysis

DEMO

Page 16: Applying Data Science to Move Beyond Keywords for Social Analysis

IMPROVED FILTERING & CLASSIFICATION

interaction.content similar "bank,hsbc:0.7"AND interaction.content similar "withdraw:0.8"

interaction.content any "rbs,lloyds,hsbc,barclays"AND interaction.content any "withdraw,close,cashpoint,atm"

CONCISEINTUITIVE

MAINTAINABLE

UP-TO-DATEHIGHER COVERAGE

ACCURACY

Page 17: Applying Data Science to Move Beyond Keywords for Social Analysis

IMPROVING OUR PLATFORM

• Further validation of approach• Operationalization of model production

• Creation new models for different audiences• Automated updating of models

• Implementation of 'similarity' in CSDL

Page 18: Applying Data Science to Move Beyond Keywords for Social Analysis

Q&A

Page 19: Applying Data Science to Move Beyond Keywords for Social Analysis

LEARN MOREdatasift.com/forum

Page 20: Applying Data Science to Move Beyond Keywords for Social Analysis

THANK YOU