![Emotion Mining in Social Media Data · works in the area of sentiment mining from text. 2.1 Sentiment and Emotion Mining Authors Bruyne et al. [6] use ensemble of classi ers for the](https://static.fdocuments.in/doc/165x107/60f991ed9fa68110a069aa6b/emotion-mining-in-social-media-data-works-in-the-area-of-sentiment-mining-from-text.jpg)
From Sentiment to Emotion Analysis in Social Networks
46
1 From Sentiment to Emotion Analysis in Social Networks Jie Tang Department of Computer Science and Technology Tsinghua University, China
description
From Sentiment to Emotion Analysis in Social Networks. Jie Tang Department of Computer Science and Technology Tsinghua University, China. Social Networks. > 1000 million users The 3 rd largest “Country” in the world More visitors than Google. > 800 million users. - PowerPoint PPT Presentation
Transcript of From Sentiment to Emotion Analysis in Social Networks
1
From Sentiment to Emotion Analysis in Social Networks
Jie Tang
Department of Computer Science and TechnologyTsinghua University, China
2
Social Networks• >1000 million users• The 3rd largest “Country” in the world• More visitors than Google
• More than 6 billion images
• 2009, 2 billion tweets per quarter• 2010, 4 billion tweets per quarter• 2011, tweets per quarter
• >800 million users
• Pinterest, with a traffic higher than Twitter and Google
25 billion
• 2013, users, 40% yearly increase560 million
3
A Trillion Dollar Opportunity
Social networks already become a bridge to connect our daily physical life and the virtual web space
On2Off [1]
[1] Online to Offline is trillion dollar businesshttp://techcrunch.com/2010/08/07/why-online2offline-commerce-is-a-trillion-dollar-opportunity/
4
Core Research in Social Network
BIG Social Data
Social Theories Algorithmic Foundations
BA m
odel
Social
influence
Action
Social Network Analysis
Theory
Prediction Search Information Diffusion AdvertiseApplication
Macro Meso Micro
ER
model
Com
munity
Group
behavior
Dunbar
Social tie
5
Today, let us start with sentiment analysis…
6
“Love Obama”
I love Obama
Obama is great!
Obama is fantastic
I hate Obama, the worst president ever
He cannot be the next president!
No Obama in 2012!
Positive Negative
7
Homophily and Influence• Homophily—“birds of a feather flock together”
– A user in the social network tends to be similar to their connected neighbors.
• Originated from different mechanisms– Influence
• Indicates people tend to follow the behaviors of their friends– Selection
• Indicates people tend to create relationships with other people who are already similar to them
– Confounding variables• Other unknown variables exist, which may cause friends to behave
similarly with one another.
8
Twitter Data
• Twitter– 1,414,340 users and 480,435,500 tweets– 274,644,047 t-follow edges and 58,387,964 @ edges– Manually labeled thousands of users
[1] Chenhao Tan, Lillian Lee, Jie Tang, Long Jiang, Ming Zhou, and Ping Li. User-level sentiment analysis incorporating social networks. In KDD’11, pages 1397–1405, 2011.
An example: “#lakers b**tch!” =?√ X
9
Influence
Shared sentiment conditioned on type of connection.
10
Selection
Connectedness conditioned on labels
11
Question: what drives users’ sentiments?
12
Sentiment vs. Emotion
Charles Darwin: – Emotion serves as a purpose
for humans in aiding their survival during the evolution.[1]
Emotion is the driving force of user’s sentiments…
Emotion stimulates the mind 3000 times quicker than rational thought!
[1] Charles Darwin. The Expression of Emotions in Man and Animals. John Murray, 1872.
13
“Happy” System
Location SMS & Calling
EmotionActivities
Can we predict users’ emotion?
14
Observations (cont.)
Location correlation(Red-happy)
Activity correlation
Karaoke
?
?
?
?
?
GYM
DormThe Old Summer Palace
Classroom
15
Observations
(a) Social correlation (a) Implicit groups by emotions
(c) Calling (SMS) correlation
16
Observations (cont.)
Temporal correlation
17
Methodologies
18
MoodCast: Dynamic Continuous Factor Graph Model
Jennifer
Happy
Happy
location
Neutral
Neutral
call
sms
Mike
Allen
MikeAllen
Jennifer today
Jennifer yesterday
?
Jennifer tomorrow
MoodCast
Predict
Attributes f(.)
Temporal correlation h(.)
Social correlation g(.)
Our solution
1. We directly define continuous feature function;
2. Use Metropolis-Hasting algorithm to learn the factor graph model.
[1] Jie Tang, Yuan Zhang, Jimeng Sun, Jinghai Rao, Wenjing Yu, Yiran Chen, and ACM Fong. Quantitative Study of Individual Emotional States in Social Networks. IEEE TAC, 2012, Volume 3, Issue 2, Pages 132-144.
19
Problem Formulation
Gt =(V, Et, Xt, Yt)
Attributes: - Location: Lab - Activity: Working
Emotion: Sad
Learning Task:
Time t
Time t-1, t-2…
20
Dynamic Continuous Factor Graph Model
Time t’
Time t
: Binary function
21
Learning with Factor Graphs
Temporal
Social
Attribute
y3
y4y5
y2y1
y'3
22
MH-based Learning algorithm
Random Sampling
Update
[1] Yuan Zhang, Jie Tang, Jimeng Sun, Yiran Chen, and Jinghai Rao. MoodCast: Emotion Prediction via Dynamic Continuous Factor Graph Model. In ICDM’10. pp. 1193-1198.
23
Still Challenges
• Q1: Are there any other social factor that may affect the prediction results?
• Q2: How to scale up the model to large networks?
24
Q1: Conformity Influence
I love Obama
Obama is great!
Obama is fantastic
Positive Negative
2. Individual
3. Group conformity
1. Peer influence
[1] Jie Tang, Sen Wu, and Jimeng Sun. Confluence: Conformity Influence in Large Social Networks. In KDD’13, 2013.
25
Conformity Factors• Individual conformity
• Peer conformity
• Group conformity
All actions by user v
A specific action performed by user v at time t
26
Q2: Distributed Learning
SlaveCompute local gradient via random sampling
MasterGlobal update
Graph Partition by MetisMaster-Slave Computing
Inevitable loss of correlation factors!
27
Random Factor GraphsMaster: Optimize with
Gradient DescentSlave: Distributedly
compute Gradient via LBP
Master-Slave Computing
Gradients
Parameters
K=1K=2Continue to increase K
28
Model Inference• Calculate marginal probability in each subgraph
• Aggregate the marginal probability and normalize
29
Theoretical Analysis• Θ*: Optional parameter of the complete graph• Θ: Optional parameter of the subgraphs• Ps,j: True marginal distributions on the complete graph• G*
s,j: True marginal distributions on subgraphs• Let Es,j = log G*
s,j – log Ps,j， we have:
30
Experiments
31
Results for Sentiment Analysis• Twitter
– 1,414,340 users and 480,435,500 tweets– 274,644,047 t-follow edges and 58,387,964 @ edges
• Baseline– SVM Vote
• Measures– Accuracy and Macro F1
32
Performance
33
Results of Different Learning Algorithms
34
35
• Data Set
• Baseline– SVM– SVM with network features– Naïve Bayes– Naïve Bayes with network features
• Evaluation Measure:Precision, Recall, F1-Measure
#Users Avg. Links #Labels Other
MSN 30 3.2 9,869 >36,000hr
LiveJournal 469,707 49.6 2,665,166
Results for Emotion Analysis
36
Performance Result
37
Factor Contributions
• All factors are important for predicting user emotionsMobile
38
Online Applications: Emotion Analysis on Flickr
39
40
Framework: Images -Aesthetic Effects -Emotions
Model: Factor Graphs for images in Social Networks[1] Jia Jia, Sen Wu, Xiaohui Wang, Peiyun Hu, Lianhong Cai, and Jie Tang. Can We Understand van Gogh’s Mood? Learning to Infer Affects from Images in Social Networks. In ACM Multimedia. pp. 857-860.[2 ] Grand Challenge 2nd Prize Award
41
App1: Emotion Distribution on Flickr
Before Thanksgiving 2011 VS During Thanksgiving holiday
Happy, Cheerful, and Peaceful 100,000 Images from Flickr
42
App2: Modify Images with Emotional Words
Happy
Natural
Clear
Original Image
Summer?
Autumn?
Winter?
More than 180 different effects
43
Summary
44
Summary• Social networks bring revolutionary changes to
the Web and unprecedented opportunities for us
• Emotion stimulates minds 3000 times faster than rational thoughts!
• Embedding social theories into sentiment/emotion analysis can benefit many applications
45
Related Publications• Jie Tang, Jimeng Sun, Chi Wang, and Zi Yang. Social Influence Analysis in Large-scale
Networks. In KDD'09, pages 807-816, 2009.• Chenhao Tan, Lillian Lee, Jie Tang, Long Jiang, Ming Zhou, and Ping Li. User-level sentiment
analysis incorporating social networks. In KDD’11, pages 1397–1405, 2011.• Jie Tang, Sen Wu, and Jimeng Sun. Confluence: Conformity Influence in Large Social
Networks. In KDD’13.• Jie Tang, Yuan Zhang, Jimeng Sun, Jinghai Rao, Wenjing Yu, Yiran Chen, and ACM Fong.
Quantitative Study of Individual Emotional States in Social Networks. IEEE Transactions on Affective Computing (TAC), 2012, Volume 3, Issue 2, Pages 132-144. (Selected as the Spotlight Paper)
• Yuan Zhang, Jie Tang, Jimeng Sun, Yiran Chen, and Jinghai Rao. MoodCast: Emotion Prediction via Dynamic Continuous Factor Graph Model. In ICDM’10. pp. 1193-1198.
• Jia Jia, Sen Wu, Xiaohui Wang, Peiyun Hu, Lianhong Cai, and Jie Tang. Can We Understand van Gogh’s Mood? Learning to Infer Affects from Images in Social Networks. In ACM MM, pages 857-860, 2012.
• Xiaohui Wang, Jia Jia, Peiyun Hu, Sen Wu, Lianhong Cai, and Jie Tang. Understanding the Emotional Impact of Images. (Grand Challenge) In ACM MM. pp. 1369-1370. (Grand Challenge 2nd Prize Award)
46
Thanks you！Collaborators: Lillian Lee, Chenhao Tan (Cornell)
Ming Zhou, Long Jiang (Microsoft), Yuan Zhang (MIT)Jimeng Sun (IBM), Jinghai Rao (Nokia)
Sen Wu, Jia Jia, Xiaohui Wang, Yiran Chen, Wenjing Yu (THU)
Jie Tang, KEG, Tsinghua U, http://keg.cs.tsinghua.edu.cn/jietangDownload data & Codes, http://arnetminer.org/download
