Author or Argumentation: Exploring the Effect of Prior Beliefs And - PowerPoint PPT Presentation

Author or Argumentation: Exploring the Effect of Prior Beliefs And Personality Traits On Persuasiveness Nikolay Kolyada

� 2 Outline Background / Motivation • Research Questions • Dataset • Modelling Authors • Modelling Interests • Modelling Beliefs • Modelling Personality traits • Evaluation • Conclusion •

� 3 Persuasion Motivation

� 4 Advertising, Marketing Motivation https://itpulse.com.ng/wp-content/uploads/2017/07/social-media.jpg

� 5 Online forums Motivation https://www.google.com/url?sa=i&source=images&cd=&ved=2ahUKEwji-rfqypbjAhVBUlAKHTNMCq0QjRx6BAgBEAU&url=https%3A%2F%2Fwww.hiveage.com%2Fblog%2Fforum-marketing-small- businesses%2F&psig=AOvVaw3AMSY07yfEy62TGvi8x1D9&ust=1562168894717718

� 6 Research Questions How to represent person's interests, beliefs, and personality traits? ○ How to model the author similarity? ○ How to demonstrate the impact of the similarity of people's ○ interests, beliefs, and personality traits on their persuasiveness, if any? Motivation

� 7 Reddit Constructing Dataset

� 8 Reddit SUBREDDITS 887k USERS 43m COMMENTS 3.7b Constructing Dataset

� 9 ChangeMyView

� 10 ChangeMyView SUBMISSIONS USERS 174k DISCUSSIONS AVERAGE 60k COMMENTS PER SUBMISSION Constructing Dataset

� 11 Submission Original Post Title: Cars should be equipped with both angry and apologetic horns Reply #1 Reply #2 … … A standard and widely used method of conveying thanks and You jump a big step here, because apology already exists: blinking you don't explain why the traditional your hazards a few times. light wave/hand under mirror gesture isn't effective. I see it pretty much … every time … Ah, really? Didn't know that! ∆ Constructing Dataset

� 12 Submission Constructing Dataset

� 13 Pairs Dataset

� 14 Pairs dataset Constructing Dataset

� 15 Pairs dataset Constructing Dataset

� 16 Pairs dataset OP - subreddit_id - author - created ORIGINAL POST ORIGINAL POST - title - text REPLY - author - created - text PERSUASIVE PERSUASIVE NON-PERSUASIVE - delta COMMENT PERSUASIVE COMMENT COMMENT COMMENT DELTA True / False Constructing Dataset

� 17 Pairs dataset TIMESPAN 01.2013 - 09.2017 8025 PERSUASIVE PAIRS 8025 NON-PERSUASIVE PAIRS Constructing Dataset

� 18 Argumentation features ORIGINAL POST ORIGINAL POST ● Linguistic features BOW, articles, pronouns, links, positive/negative words, question marks, quotations PERSUASIVE PERSUASIVE NON-PERSUASIVE COMMENT PERSUASIVE COMMENT COMMENT ● Interplay features COMMENT Constructing Dataset

� 19 Argumentation features ● Linguistic features ● Interplay features Constructing Dataset

� 20 Modelling Author Interests

� 21 Modelling Interests Modelling Author Interests

� 22 Modelling Interests News and Discussion Hobbies Entertainment Science Politics Modelling Author Interests

� 23 Modelling Interests News and Discussion Hobbies Entertainment Science Politics 12 91 POST FREQUENCIES TF-IDF ... PCA 32 0 Modelling Author Interests

� 24 Modelling Author Beliefs

� 25 Modelling Author Beliefs Modelling Author Beliefs

� 26 Context Sentiment category tagret post text context sentiment ... Besides, Hillary Clinton News And Politics –> Hillary Clinton also has some weaknesses. -0.362 Republican ... Modelling Author Beliefs

� 27 Modelling Author Beliefs Modelling Author Beliefs

� 28 Modelling Author Beliefs News and Hobbies Entertainment Science Discussion Politics Modelling Author Beliefs

� 29 Modelling Author Beliefs News and Discussion Hobbies Entertainment Science Politics Donald Trump -0.6 Hillary Clinton -0.2 MEDIAN SENTIMENT SCORE ... WITHIN TOPICS BREXIT 0.6 Second Amendment 0.2 Modelling Author Beliefs

� 30 Modelling Personality Traits

� 31 Big Five Openness Neuroticism Conscientiousness PERSONALITY Agreeableness Extraversion OCEAN, CANOE Modelling Personality Traits

� 32 Modelling Personality Traits Modelling Personality Traits

� 33 Modelling Personality Traits BIG neuroticism extraversion openness agreeableness conscientiousness FIVE awful, other, folk, wonderful, completed, though, drinks, humans, together, adventure, lazy, restaurant, poet, visiting, stupid, worse, dancing, art, morning, boring, depressing, cats, universe, spring, adventures, irony, grandfather, narrative, walked, desperate, road, countless, giveaway, beautiful, enjoying, terrible, shots, century, staying, saying, stressful, computer, sexual, felt, utter, horrible glorious films cost extreme Modelling Personality Traits

� 34 Modelling Personality Traits BIG neuroticism extraversion openness agreeablenes conscientiousnes FIVE s s awful, other, folk, wonderful, completed, though, drinks, humans, together, adventure, lazy, restaurant, poet, visiting, stupid, worse, dancing, art, morning, boring, depressing, cats, universe, spring, adventures, irony, grandfather, narrative, walked, desperate, road, countless, giveaway, beautiful, enjoying, terrible, shots, century, staying, saying, stressful, computer, sexual, felt, utter, horrible glorious films cost extreme 12 91 ... 32 0 BIG FIVE 12 91 ... 32 0 TRAITS PERCENTILES 12 91 ... 32 0 WORD CATEGORIES Modelling Personality Traits

� 35 Author Similarity

� 36 Author Similarity POLITICS RELIGION HOBBIES ORIGINAL POST 12 91 ... 32 0 12 91 ... 32 0 12 91 ... 32 0 PERSUASIVE PERSUASIVE COMMENT PERSUASIVE POLITICS RELIGION HOBBIES COMMENT COMMENT 12 91 ... 32 0 12 91 ... 32 0 12 91 ... 32 0 Author Similarity

� 37 Author Similarity POLITICS 12 91 POLITICS ... 32 -0.6 0 -0.2 ... 0.6 0.2 POLITICS 12 91 ... 32 0 Author Similarity

� 38 Author Similarity POLITICS 12 91 POLITICS ... 32 -0.6 0 TOTAL -0.2 OVER ... 7 000 FEATURES 0.6 0.2 POLITICS 12 91 cosine( "haudpe", "TheWildThing" ) ... 32 0 Author Similarity

� 39 Features overview Feature groups: • interests • beliefs • personality traits • argumentation features

� 40 How to demonstrate the impact of the similarity of people's interests, ○ beliefs, and personality traits on their persuasiveness, if any?

� 41 Predicting the Persuasive Argument Opinion Malleability Prediction

� 42 Predicting the Persuasive Argument Interests ORIGINAL POST 12 91 ... 32 0 Beliefs PERSUASIVE 12 91 ... 32 0 OR Traits NON- PERSUASIVE PERSUASIVE COMMENT PERSUASIVE 12 91 ... 32 0 PERSUASIVE COMMENT COMMENT Argumentation 12 91 ... 32 0 Interests ORIGINAL POST 12 91 ... 32 0 Beliefs PERSUASIVE 12 91 ... 32 0 OR NON- Traits NON- PERSUASIVE PERSUASIVE 12 91 ... 32 0 COMMENT Argumentation 12 91 ... 32 0 Evaluation

� 43 Predicting the Persuasive Argument Evaluation

� 44 How easy is to persuade OP?

� 45 Opinion Malleability Evaluation

� 46 Opinion Malleability personality traits 12 12 91 91 RESISTANT ORIGINAL POST ... ... OR 32 32 MALLEABLE 0 0 BOW Evaluation

� 47 Opinion Malleability POSTS 8 121 MALLEABLE 3 186 UNIQUE AUTHORS 4 641 Evaluation

� 48 Opinion Malleability Evaluation

� 49 Opinion Malleability Evaluation

� 50 Future work • structure of arguments, quality of arguments • correlation between author traits and argumentation used • model the experience: deltas earned • different discussion platforms

� 51 Summary • Research questions How to represent person's interests, beliefs, and personality traits? ○ How to model the author similarity? ○ How to demonstrate the impact of the similarity of people's interests, beliefs, and ○ personality traits on their persuasiveness, if any? • Pairs dataset • Author features: Interests, Beliefs, Personality Traits • Experiments: Prediction of persuasive argument, Opinion malleability • Explore the impact of author similarity on persuasion