CSE217 INTRODUCTION TO DATA SCIENCE LECTURE 3: SENTIMENT ANALYSIS - - PowerPoint PPT Presentation

CSE217 INTRODUCTION TO DATA SCIENCE

Spring 2019 Marion Neumann

LECTURE 3: SENTIMENT ANALYSIS

2

SENTIMENT ANALYSIS

…discover people’s opinions, emotions, feelings about a subject, topic, product, or service from text

Step 1: Get the data Step 2: Process text into features Step 3: Infer sentiment

SENTIMENT ANALYSIS

Recap: Data Science Workflow

3

collect & understand data clean & format data data problem use data to create solution scientific, social, or business problem sentiment analysis improve movie recommender

• r

gauging brand perception scrape web/twitter working with text data

• rule-based predictor
• machine learning

classifier

?

SENTIMENT ANALYSIS WORKFLOW

4 Feature Extraction

excluded bad

& Negation Handling

bad ping pong exclude rio 2016 Stemming bad ping pong excluded rio 2016

à rule-based prediction à machine learning classifier

RULE-BASED APPROACH

à Lab 3

5

DSFS p25 Control Flow

TEXT DATA

• Data representation à strings
• four kinds of string data

1) categorical data 2) free strings (that can be semantically mapped to categories) 3) structured string data 4) free-form text data

6

à What makes text different?

TEXT DATA

…is Big Data!

7

MACHINE LEARNING APPROACH

• Classification

8

great small location friends …

FEATURES FOR TEXT DATA

• bag of words

à does word occur in document yes/no à binary feature

• word counts

à how often does word occur? à count feature

• more advanced: n-grams, TF-IDF

9

Same great flavor and friendly service as in the S 18th street

• location. This location is not as small but it's hard to talk to friends.

Thankfully there is great outdoor seating to escape the noise.

FEATURE REPRESENTATION

• bag-of-words and word counts are vectors

10

PDSH p38 Arrays

features

• freview

counts

• r binary review

great

III

resin

D

• f's'EukeJvocasueary

1

horrible

D

i

170.000

words

easel

Tpositive

KEITEL

dictionary

D extremely

sparse features

many zeros since most word donot appear

in review

WHAT IS A CLASSIFIER?

• Rule-based
• list of positive and negative words results in fixed score

(+1, -1, or 0) for each word

• Classifier
• no fixed lists of positive/negative words
• each word gets a weight parameter ! assigned
• classifier = parameterized model of the

relationship between input and output/label

• e.g. label = sign(w(x + +) using a linear relationship
• classifier learns the weights from labeled training data

11 w(x is referred to as

• dot product,
• inner product, or
• scalar product

CLASSIFIER

• output (sentiment) is a binary class

12

Is this new review positive or negative?

• r

EVALUATION

• Which approach (rule-based or machine learning)

performs better?

• Measures:
• error rate (or misclassification rate) =

# #$%%&'(%%$)$*+ ,*%, -.$/,% # ,*%, -.$/,%

• average accuracy (= 1 − 23343 3562)

13

à How can we measure this?

14

• DSFS
• Ch4: Linear Algebra à Vectors (p49-53)
• Ch9: Getting Data (p105-108, p114-120)
• Ch20: Natural Language Processing (p239-244)

SUMMARY & READING

• Sentiment Analysis automatically identifies,

extracts, and analyzes emotions in text data.

• Text data needs to be preprocessed to get features

that can be used for prediction and learning.

• Linear classification is used to predict binary or

categorical targets.

Do not use the implementations introduced in this chapter à use NumPy Arrays! PDSH p38