hierarchical aspect and sentiment model, context-dependent conceptualisation

51
Hiearchical Aspect-Sentiment Model & Context-Dependent Conceptualization Alice Oh [email protected] http://uilab.kaist.ac.kr/ April 11, 2013

Upload: alice-oh

Post on 05-Dec-2014

735 views

Category:

Technology


2 download

DESCRIPTION

 

TRANSCRIPT

Page 1: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Hiearchical Aspect-Sentiment Model & Context-Dependent Conceptualization

Alice Oh [email protected] http://uilab.kaist.ac.kr/ April 11, 2013

Page 2: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Overview

¤ Hierarchical Aspect-Sentiment Model (AAAI 2013) ¤  Suin Kim, et al.

¤ Collaboration with Microsoft Research Asia

¤ Context-Dependent Conceptualization (IJCAI 2013) ¤ Dongwoo Kim, Haixun Wang, Alice Oh

¤ Collaboration with Microsoft Research Asia

2

Page 3: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Users & Information Lab @ KAIST

3

Page 4: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Hiearchical Aspect-Sentiment Model (AAAI-13) Suin Kim, Jianwen Zhang, Zheng Chen, Alice Oh, and Shixia Liu

4

Page 5: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Hierarchical aspect-sentiment model ¤ Goal: To discover a hierarchy of aspects and associated

sentiments from a corpus of online reviews

¤ Assumptions ¤  Each sentence expresses a single aspect and a single sentiment ¤  An aspect (e.g., “battery life”) consists of neutral, positive, and

negative words

¤ Model: A hierarchical aspect-sentiment joint model using the recursive Chinese restaurant processes (rCRP)

¤ Results ¤  A reasonable hierarchy of aspects discovered without supervision ¤  Sentiment classification accuracy comparable other recent

sentiment-aspect joint models

5

Page 6: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Aspect-sentiment hierarchy

6

Goals •  To discover and organize the aspects and associated sentiments into a hierarchy •  To determine the aspect in each sentence •  To determine the sentiment of each sentence

Page 7: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Comparison to other models 7

General Specific

Positive

Neutral

Negative

ASUM  &  JST

Multigrain  Topic  Model

General

Specific

Positive Negative

Reverse  JST

Hierarchical  Aspect-­Sentiment  Model

General Specific

Positive

Neutral

Negative

ASUM  &  JST

Multigrain  Topic  Model

General

Specific

Positive Negative

Reverse  JST

Hierarchical  Aspect-­Sentiment  Model

General Specific

Positive

Neutral

Negative

ASUM  &  JST

Multigrain  Topic  Model

General

Specific

Positive Negative

Reverse  JST

Hierarchical  Aspect-­Sentiment  Model

General Specific

Positive

Neutral

Negative

ASUM  &  JST

Multigrain  Topic  Model

General

Specific

Positive Negative

Reverse  JST

Hierarchical  Aspect-­Sentiment  Model

General Specific

Positive

Neutral

Negative

ASUM  &  JST

Multigrain  Topic  Model

General

Specific

Positive Negative

Reverse  JST

Hierarchical  Aspect-­Sentiment  Model

Page 8: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

HASM

8

Page 9: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Aspect-sentiment hierarchy

9

•  Aspects tend to be general near the root and specific toward the leaves •  Each aspect node consists of positive and negative polarity •  Each sentence in a review is generated from a single aspect and sentiment •  Each word in a sentence is either neutral or subjective

Page 10: Hierarchical aspect and sentiment model, Context-dependent conceptualisation
Page 11: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

“The screen is clear and the picture quality is outstanding.”

Page 12: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

“The screen is clear and the picture quality is outstanding.”

Page 13: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

the screen is and the picture

clear quality outstanding

Page 14: Hierarchical aspect and sentiment model, Context-dependent conceptualisation
Page 15: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

“A short battery life undermines portability.”

Page 16: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

“A short battery life undermines portability.”

Page 17: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

A battery life portability

short undermines

Page 18: Hierarchical aspect and sentiment model, Context-dependent conceptualisation
Page 19: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

HASM: Experiments & Results

¤ Data: Amazon reviews on laptops (10,014) and DSLRs (20,862)

¤ Aspect-sentiment hierarchies

¤ Quantitative evaluation ¤ Topic specialization

¤ Hierarchical affinity

¤ Aspect-sentiment consistency

¤ Fine-grained sentiment classification

¤ User scenario

19

Page 20: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

20

Page 21: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Topic specialization

Evaluates the general-to-specific nature of the hierarchy by comparing the average distance of the aspect nodes from the root at each tree depth

Page 22: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Hierarchical affinity

Measures whether a parent-child pair shows smaller distance compared to a non-parent-child pair, one at level L and another at level L+1

Page 23: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Aspect-sentiment consistency

Measures how in-node topics are statistically coherent by comparing •  average intra-node topic distance •  average inter-node topic distance ttt

ttt

ttt

ttt ttt

Page 24: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Sentiment classification accuracy

•  Sentiment classification using short (<100 characters) reviews

•  Small set contains positive reviews of 5 stars, negative reviews of 1 star

•  Large set contains positive reviews of 4~5 stars, negative reviews of 1~2 stars

Page 25: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

User scenario

Visualization of hierarchical aspect-sentiments for a user who is looking for a camera with good picture quality under low lights, a good LCD screen, and high-end lenses

Page 26: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Context-dependent Conceptualization (IJCAI 2013) Dongwoo Kim, Haixun Wang, Alice Oh

26

Page 27: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Semantic relatedness

Apple reveals new iPad

Microsoft introduces Surface

Surface vs iPad

Samsung’s new android tablets

iPhone 5, the best smart phone ever

By Topic Modeling

iPad Apple

Microsoft iPhone

Software Samsung

SmartPhone Android

Software Company iOS

Mobile Phones

Page 28: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Contextual relatedness

Apple reveals new iPad

Fruit Company

Food Fresh fruit Fruit tree

Brand Crop

Flavor Item

Manufacturer

Device Platform Technology Mobile device Tablet Portable device Tablet computer Gadget Apple product Output device

Page 29: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Conceptualization given semantic context

Apple reveals new iPad

Fruit Company

Food Fresh fruit Fruit tree

Brand Crop

Flavor Item

Manufacturer

Device Platform Technology Mobile device Tablet Portable device Tablet computer Gadget Apple product Output device

iPad

A

pple

M

icro

soft

iP

hone

So

ftw

are

Sam

sung

Sm

artP

hone

A

ndro

id

Soft

war

e C

ompa

ny

iOS

Mob

ile P

hone

s Semantic Context of Sentence

Concept of Apple Concept of iPad

Page 30: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Conceptualization given semantic context

Apple reveals new iPad

Fruit Company

Food Fresh fruit Fruit tree

Brand Crop

Flavor Item

Manufacturer

Device Platform Technology Mobile device Tablet Portable device Tablet computer Gadget Apple product Output device

iPad

A

pple

M

icro

soft

iP

hone

So

ftw

are

Sam

sung

Sm

artP

hone

A

ndro

id

Soft

war

e C

ompa

ny

iOS

Mob

ile P

hone

s Semantic Context of Sentence

Concept of Apple Concept of iPad

Reinforcing concepts Based on context

Fruit Company

Food Fresh fruit Fruit tree

Brand Crop

Flavor Item

Manufacturer

Page 31: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Context-dependent conceptualization

company 0.104 client 0.078 tree 0.069

corporation 0.050 computer 0.047

software company 0.041 oems 0.025 laptop 0.020

personal computer 0.019 host 0.019

Concept of Apple

Apple and iPad

fruit 0.039 food 0.035

company 0.026 brand 0.024 flavor 0.021 crop 0.020 juice 0.018

fresh fruit 0.017 plant 0.017 snack 0.015

Apple and Orchard

company 0.063 brand 0.041 client 0.038

corporation 0.033 tree 0.028

business 0.028 computer 0.027

crop 0.027 software company 0.022 computer company 0.021

Page 32: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Context-dependent conceptualization

Concept of Jordan

Jordan and Basketball

Jordan and Iraq

country 0.172 state 0.107 place 0.088

arab state 0.070 arab country 0.067

muslim country 0.052 arab nation 0.045

middle eastern country 0.042 islamic country 0.040

regime 0.023

place 0.284 player 0.240 team 0.177

nation 0.106 host country 0.041

professional athlete 0.021 great player 0.020 role model 0.020

shoe 0.018 offensive 0.016

country 0.172 state 0.107 place 0.088

arab state 0.070 arab country 0.067

muslim country 0.052 arab nation 0.045

middle eastern country 0.042 islamic country 0.040

regime 0.023

Page 33: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Experiments and Results

¤ Frame elements

¤ Word similarity in context

¤ Query-ad clickthrough

Page 34: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Experiments and Results

¤ Frame elements ¤ Background: Semantic role labeling depends heavily on

annotated data such as FrameNet

¤ Problem: Building FrameNet requires expertise, and while FrameNet contains 170k annotated sentences, it lacks coverage

¤ Approach: Expand FrameNet using CDC

1.  Conceptualize the frame elements given a sentence as the context

2.  Find other instances given the most probable concepts

¤ Experiment: Compare likelihood of frame elements in unseen sentences in FrameNet

Page 35: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Frame elements

Given sentence :

in  the  I   cook   them   oven  

1.  What is the frame of this sentence ? 1)  abusing 2) closure 3) apply_heat

Page 36: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Frame elements

in  the  I   cook   them   oven  

Given sentence :

1.  What is the frame of this sentence ? 1)  abusing 2) closure 3) apply_heat

2.  What is the frame element of the word ‘oven’ 1) cooker 2) food 3) heat_source

Page 37: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Frame elements

inthe  I   cook   them   oven  

FE: Cooker FE: Food

FE: Heat source

Frame: Apply_Heat

Lexical Unit (Target)

Final Goal :

FE (Frame Element)

Page 38: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Frame elements: conceptualization for expansion

Frame Element : Heat_Source

… egg and chips was sizzling over camp-fires. … the pig sizzled on the flames , spitting fat … a large black kettle was  sizzling  on the hob. Droplets of coffee  sizzled  on the hotplate. … kitchen the meat  sizzled  in the oven and a big pan of potatoes … …   sizzled, now and then, upon the diminutive stove

☞ Conceptualize labeled frame elements with context

Labeled elements

Page 39: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Frame elements: conceptualization for expansion

Concept of Heat_Source FE

Extended Heat_Source FE with Probase :

Page 40: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Frame elements: experiment

Per-word heldout log-likelihood of the predicted frame elements using five-fold validation. The naïve approach is conceptualization using Probase without context (Song, IJCAI 2012).

Page 41: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Experiments and Results ¤  Frame elements

¤ Word similarity in context ¤  Background: Recent work in word similarity prediction uses

annotated data of words in sentential context ¤  Problem: Existing methods for word similarity are specifically

tailored for word similarity only. Naïve conceptualization does not consider sentential context.

¤  Approach 1.  Given two words and their sentential contexts, conceptualize

the words 2.  Estimate the similarity using cosine similarity of the concept

vectors ¤  Experiment: Compare the correlation between CDC-based

similarity and human judgment

Page 42: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Word similarity in context

¤ … Native Chinese cuisine makes frequent use of Asian leafy vegetables like bok choy and kai-lan and puts a greater emphasis on fresh meat …

¤ … American Chinese food is usually less pungent than authentic cuisine …

¤ Human evaluation = 9.2 (0~10 scale)

Page 43: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Word similarity in context

¤  ... This system would be implemented into the national response plan for bioweapons attacks in the Netherlands . Researchers at Ben Gurion University in Israel are developing a different device called the BioPen , essentially a “Lab-in-a-Pen” …

¤ … originally written in 1969 and performed extensively at the time by an Israeli military performing group , has become one of the anthems of the Israeli peace camp . During the Arab uprising known as the First Intifada , Israeli singer Si Heyman sang “Yorim VeBokhim” …

¤ Human evaluation = 8.1 (0~10 scale)

Page 44: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Word similarity in context: Results

Note: State-of-the-art word similarity method yields correlation of 0.66 (Huang ACL 2012)

Page 45: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Experiments and Results ¤  Frame elements

¤ Word similarity in context

¤ Query-ad clickthrough ¤  Background: Matching ads with user queries is an important but

difficult task. Clickthrough rate for sponsored links is generally very low.

¤  Problem: Ad bids and user queries are short sequences of keywords that do not benefit from full NLP techniques. But simple keyword expansion methods are inaccurate.

¤  Approach: Use CDC for both ad bids and queries and match them using cosine similarity of the concept vectors.

¤  Experiment: Using search results of Bing, compare the correlation of query-ad concept similarity and CTR.

Page 46: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Sponsored link bid keywords

Bid keywords for sponsored links= { Rockport, Shoes }

User Query = { Rockport men shoes }

Show sponsored links when bid keywords and query match!

Page 47: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Query-ad clickthrough

Ad-bids Query CTR rockport shoes rockport men boots 0.0201 rockport shoes florsheim shoes 0.0022 rockport shoes men dockers shoes 0.0000 replica watches breitling copy watches 0.0833 replica watches replica 0.0833 replica watches tiffany replica bracelet 0.0064

free email e mail 0.0454 free email windows mail 0.0294 free email set up free email account 0.0232

Page 48: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Equal weighting phrase conceptualization

company 0.366 brand 0.255 town 0.183 shoe 0.071

shoe company 0.058 neighboring town 0.054

popular name brand 0.010 top brand 3.49E-08

popular brand 3.01E-08 top name 2.38E-08

Bid keywords for sponsored links= { }

accessory 0.092 clothes 0.051

equipment 0.049 essential 0.045 garment 0.045

shoe 0.042 fashion accessory 0.034

touch 0.033 textile 0.029 surface 0.029

CDC

How to combine two CDC results?

Rockport,

CDC

Shoes

Page 49: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

URL title and Query Conceptualization

User Query = { Bayesian Topic Model }

Title of this page { Latent Dirichlet allocation – Wikipedia, the free encyclopedia }

Retrieve web pages based on concept similarities between URL-title and query

Page 50: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

IDF Weighting Phrase Conceptualization

Title of Web page { Latent Dirichlet allocation – Wikipedia, the free encyclopedia }

User Query = { Bayesian Topic Model }

Are these important concepts for retrieval?

How to combine CDC results of query and title?

Page 51: Hierarchical aspect and sentiment model, Context-dependent conceptualisation

Correlation between CTR and avg. similarity

CDC achieves higher correlations between average similarity and CTR

Model Correlation

CDC-IDF-100 CDC-IDF-200 CDC-IDF-300

0.818 0.827 0.838

CDC-EQ-100 CDC-EQ-200 CDC-EQ-300

0.932 0.952 0.955

Keyword IJCAI 11

0.259 0.243