1
TREC 06: Finding opinionated posts, either positive or negative, about a query 2006 TREC Blog corpus: 80K blogs, 300K post 50 test queries BlogVox opinion extraction system Document and sentence level scorers Combined scores using an SVM meta-learner Data cleaning: splogs and post identification Modeling Influence Opinions and Modeling Influence Opinions and Structure in Social Media Structure in Social Media Akshay Java Advisor: Tim Finin Modeling Influence Thesis Statement Influence is Topical Ongoing Research An accurate model of influence on the Blogosphere must analyze and combine many contributing factors, including topic, social structure, opinions, biases and time. We will develop, implement and experimentally evaluate such a model to demonstrate its improved accuracy over models based on any of these factors.’’ Temporal Trends Indicate Influence Popular Topics in Feeds That Matter “Topic Ontology” derived from 83K user feed subscriptions consisting of 500K feeds. Provides a readership-based metric for influence. Tech Slashdot Gizmodo Wired Politics Dems Reps Dailykos Talkingpoi nts Michellemalki n RightwingNews Epidemic Based Influence Models Linear Threshold Model Σ b wv ≥ θv w is the active neighbor of v, θv intrinsic threshold for a node Greedy Heuristic • Assign random θv • Compute approx influenced set • At each step, add the node that increases the marginal gain in the size of the influenced set Limitations Selected nodes may belong to different topics • Social structure not considered • Static View of the network Extended Model Finds influential nodes for a topic • Models opinions, bias and trust • Identifies communities and social impact • Tracks temporal evolution of a meme • Richer framework to model influence Opinions and Bias Influence Readers Bias towards MSM sources A generalized framework for influence in Social Media Predictive Models for topical trends and influence Link Polarity and Trust Improved sentiment analysis Generative Models for the Blogosphere Who started talking about the topic first? Who were the early adopters? Who were the influencers? Who was the source of the information? What are the future trends to watch out for?

TREC 06 : Finding opinionated posts, either positive or negative, about a query

  • Upload
    signa

  • View
    25

  • Download
    1

Embed Size (px)

DESCRIPTION

Modeling Influence Opinions and Structure in Social Media. Akshay Java Advisor: Tim Finin. Thesis Statement. Modeling Influence. - PowerPoint PPT Presentation

Citation preview

Page 1: TREC 06 :  Finding  opinionated  posts,  either positive or negative, about a query

TREC 06: Finding opinionated posts, either positive or negative,about a query2006 TREC Blog corpus:80K blogs, 300K post50 test queries

BlogVox opinion extraction systemDocument and sentence level scorersCombined scores using an SVM meta-learner

Data cleaning: splogs and post identification

Modeling Influence Opinions and Modeling Influence Opinions and

Structure in Social MediaStructure in Social Media Akshay Java Advisor: Tim Finin

Modeling InfluenceThesis Statement

Influence is Topical

Ongoing Research

An accurate model of influence on the Blogosphere must analyze and combine many contributing factors, including topic, social structure, opinions, biases and time. We will develop, implement and experimentally evaluate such a model to demonstrate its improved accuracy over models based on any of these factors.’’

Temporal Trends Indicate Influence

Popular Topics in Feeds That Matter

“Topic Ontology” derived from 83K user feed subscriptions consisting of 500K feeds. Provides a readership-based metric for influence.

Tech

Slashdot

Gizmodo

Wired

Politics

Dems Reps

Dailykos

Talkingpoints

Michellemalkin

RightwingNews

Epidemic Based Influence ModelsLinear Threshold Model Σ bwv ≥ θvw is the active neighbor of v, θv intrinsic threshold for a node

Greedy Heuristic• Assign random θv• Compute approx influenced set• At each step, add the node that increases the marginal gain in the size of the influenced set

Limitations• Selected nodes may belong to different topics• Social structure not considered • Static View of the network

Extended Model• Finds influential nodes for a topic• Models opinions, bias and trust • Identifies communities and social impact• Tracks temporal evolution of a meme• Richer framework to model influence

Opinions and Bias Influence Readers

Bias towards MSM sources

A generalized framework for influence in Social Media

Predictive Models for topical trends and influence

Link Polarity and Trust

Improved sentiment analysis

Generative Models for the Blogosphere

Who started talking about the topic first? Who were the early adopters?

Who were the influencers? Who was the source of the information?

What are the future trends to watch out for?