Dimensions of argumentation in social media EKAW2012 2012 10 11
Upcoming SlideShare
Loading in...5
×
 

Like this? Share it with your network

Share

Dimensions of argumentation in social media EKAW2012 2012 10 11

on

  • 769 views

Paper presentation for EKAW 2012, The 18th International Conference on Knowledge Engineering and Knowledge Management. ...

Paper presentation for EKAW 2012, The 18th International Conference on Knowledge Engineering and Knowledge Management.

Position paper with Brian Davis and Adam Wyner: http://jodischneider.com/pubs/ekaw2012.pdf

Conference website: http://ekaw2012.ekaw.org/

Abstract: Mining social media for opinions is important to governments and businesses. Current approaches focus on sentiment and opinion detection. Yet, people also justify their views, giving arguments. Understanding arguments in social media would yield richer knowledge about the views of individuals and collectives. Extracting arguments from social media is difficult. Messages appear to lack indicators for argument, document structure, or inter-document relationships. In social media, lexical variety, alternative spellings, multiple languages, and alternative punctuation are common. Social media also encompasses numerous genres. These aspects can confound the extraction of well-formed knowledge bases of argument. We chart out the various aspects in order to isolate them for further analysis and processing.

Statistics

Views

Total Views
769
Views on SlideShare
755
Embed Views
14

Actions

Likes
3
Downloads
7
Comments
1

3 Embeds 14

http://learn.vccs.edu 8
https://si0.twimg.com 3
https://twitter.com 3

Accessibility

Categories

Upload Details

Uploaded via as Microsoft PowerPoint

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
  • Slide 19 should look like this:
    http://www.slideshare.net/jodischneider/envisioning-a-discussion-dashboard-for-collective-intelligence-of-web-conversations-cscw2012collectiveintelligence/9
    (Thanks Jack Park for reminding me about that! My local copy, bizarrely unlike the downloaded version, is still fine...)

    Possibly changing the size caused this problem?
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment
  • . The issue is whether it is the right product for the buyer, which is a matter not only of the pros and cons, but also of the explanations and counterarguments given. In our view, current approaches detect problems, but obscure the chains of reasoning about them.Image: http://www.nickmilton.com/2012/03/when-people-trust-crowds.html
  • Isn’t it funny that people tweet about thisDifference between cakes and biscuits? When stale, cakes go hard, biscuits go soft. Hence Jaffa Cakes are cakes. (Was official EU ruling).https://twitter.com/robeastaway/status/135838892694839296
  • http://www.amazon.co.uk/review/R34F9M7871K2W0/ref=cm_cr_rev_detmd_pl?ie=UTF8&asin=B004M8S152&cdForum=FxM7A99WYT0UJ9&cdMsgID=Mx35ZEPGH4SV7JK&cdMsgNo=1&cdPage=1&cdSort=oldest&cdThread=Tx1C124R97075W9&store=photo#Mx35ZEPGH4SV7JK
  • Maximal consistent sets
  • Isn’t it funny that people tweet about this
  • Our goal is to extract and reconstruct argumentation into formal representations which can be entered into a knowledge base. Drawing from existing approaches to subjectivity, topic identification, and knowledge extraction, we need to indicate disagreements and other relationships between opinions, along with justifications for opinions. This is currently done by hand. The goal really is to figure out how to automate the analysis. Issues include the informality of language in social media, the amount of implicit information, and various ‘meta’ information that contributes to the argument reconstruction, as we later discuss.
  • Our goal is to extract and reconstruct argumentation into formal representations which can be entered into a knowledge base. Drawing from existing approaches to subjectivity, topic identification, and knowledge extraction, we need to indicate disagreements and other relationships between opinions, along with justifications for opinions. This is currently done by hand. The goal really is to figure out how to automate the analysis. Issues include the informality of language in social media, the amount of implicit information, and various ‘meta’ information that contributes to the argument reconstruction, as we later discuss.
  • 4 Dimensions of ExpressionTo extract well-formed knowledge bases of argument, we must first chart out the various dimensions of social media, to point the way towards the aspects that argumentation reconstruction will need to consider, so that we later can isolate these aspects.Social media encompasses numerous genres, each with their own conversational styles, which affect what sort of rhetoric and arguments may be made. One key feature is the extent to which a medium is used for broadcasts (e.g. monologues) versus conversations (e.g. dialogues), and in each genre, a prototypical message or messages could be described, but these vary across genres due to social conventions and technical constraints. De Moor and Efimova compared rhetorical and argumentative aspects[4] of listservs and blogs, identifying features such as the likelihood that messages receive responses, and whether spaces are owned communities or by a single individual, and the timeline for replies [5]. Important message characteristics include the typical and allowable message length (e.g. space limitations on microblogs) and whether messages may be continually refined by a group (such as in StackOverflow).Metadata associated with a post (such as poster, timestamp, and subject line for listservs) and additional structure (such as pingbacks and links for blogs) can also be used for argumentation. For example, a user’s most recent post is generally taken to identify their current view, while relationships between messages can indicate a shared topic, and may be associated with agreement or disagreement.Users are different, and properties of users are factors that contribute not only to substance of the user’s comment, but as well to how they react to the comments of others. These include demographic information such as the user’s age, gender, location, education, and so on. In a specific domain, additional user expectations or constraints could also be added. Different users are persuaded by different kinds of information. Therefore, to solve peoples’ problems, based on knowledge bases, when dealing with inconsistency, understanding the purposes and goals that people have would be useful.Therefore, the goals of a particular dialogue also matter. These have been considered in argumentation theory: Walton & Krabbe have categorized dialogue types based on the initial situation, participant’s goal, and the goal of the dialogue [11]. The types they distinguish are inquiry, discovery, information seeking, deliberation, persuasion, negotiation and eristic. These are abstractions–any single conversation moves through various dialogue types. For example, a deliberation may be paused in order to delve into information seeking, then resumed once the needed information has been obtained.Higher level context would also be useful: different amounts of certainty are needed for different purposes. Some of that is inherent in a task: Reasoning about what kind of medical treatment to seek for a long-term illness, based on PatientsLikeMe, requires more certainty than deciding what to buy based on product reviews.Informal language is very typically found in social media. Generic language processing issues, with misspellings and abbreviations, slang, language mixing emoticons, and unusual use of punctuation, must be resolved in order to enable text mining (and subsequently argumentation mining) on informal language. Indirect forms of speech, such as sarcasm, irony, and innuendo, are also common. A step-by-step approach, focusing first on what can be handled, is necessary.Another aspect of the informality is that much information is left implicit. Therefore, inferring from context is essential. Elliptical statements require us to infer common world knowledge, and connecting to existing knowledge bases will be needed.We apply sentiment techniques to provide candidates for argumentation mining and especially to identify textual markers of subjectivity and objectivity. The arguments that are made about or against purported facts have a different form from the arguments that are made about opinions. Arguments about objective statements provide the reasons for believing a purported fact or how certain it is. Subjective arguments might indicate, for instance, which users would benefit from a service or product (those similar to the poster). Another area where subjective arguments may appear is discussions of the trust and credibility about the people making the arguments.
  • Social media encompasses numerous genres, each with their own conversational styles, which affect what sort of rhetoric and arguments may be made. One key feature is the extent to which a medium is used for broadcasts (e.g. monologues) versus conversations (e.g. dialogues), and in each genre, a prototypical message or messages could be described, but these vary across genres due to social conventions and technical constraints. De Moor and Efimova compared rhetorical and argumentative aspects of listservs and blogs, identifying features such as the likelihood that messages receive responses, and whether spaces are owned communities or by a single individual, and the timeline for replies [5]. Important message characteristics include the typical and allowable message length (e.g. space limitations on microblogs) and whether messages may be continually refined by a group (such as in StackOverflow).
  • Metadata associated with a post (such as poster, timestamp, and subject line for listservs) and additional structure (such as pingbacks and links for blogs) can also be used for argumentation. For example, a user’s most recent post is generally taken to identify their current view, while relationships between messages can indicate a shared topic, and may be associated with agreement or disagreement.
  • Users are different, and properties of users are factors that contribute not only to substance of the user’s comment, but as well to how they react to the comments of others. These include demographic information such as the user’s age, gender, location, education, and so on. In a specific domain, additional user expectations or constraints could also be added.
  • Different users are persuaded by different kinds of information. Therefore, to solve peoples’ problems, based on knowledge bases, when dealing with inconsistency, understanding the purposes and goals that people have would be useful. Therefore, the goals of a particular dialogue also matter. These have been considered in argumentation theory: Walton & Krabbe have categorized dialogue types based on the initial situation, participant’s goal, and the goal of the dialogue [11]. The types they distinguish are inquiry, discovery, information seeking, deliberation, persuasion, negotiation and eristic. These are abstractions–any single conversation moves through various dialogue types. For example, a deliberation may be paused in order to delve into information seeking, then resumed once the needed information has been obtained. Higher level context would also be useful: different amounts of certainty are needed for different purposes. Some of that is inherent in a task: Reasoning about what kind of medical treatment to seek for a long-term illness, based on PatientsLikeMe, requires more certainty than deciding what to buy based on product reviews. Informal language is very typically found in social media. Generic language processing issues, with misspellings and abbreviations, slang, language mixing emoticons, and unusual use of punctuation, must be resolved in order to enable text mining (and subsequently argumentation mining) on informal language. Indirect forms of speech, such as sarcasm, irony, and innuendo, are also common. A step-by-step approach, focusing first on what can be handled, is necessary. Another aspect of the informality is that much information is left implicit. Therefore, inferring from context is essential. Elliptical statements require us to infer common world knowledge, and connecting to existing knowledge bases will be needed. We apply sentiment techniques to provide candidates for argumentation mining and especially to identify textual markers of subjectivity and objectivity. The arguments that are made about or against purported facts have a different form from the arguments that are made about opinions. Arguments about objective statements provide the reasons for believing a purported fact or how certain it is. Subjective arguments might indicate, for instance, which users would benefit from a service or product (those similar to the poster). Another area where subjective arguments may appear is discussions of the trust and credibility about the people making the arguments.
  • Isn’t it funny that people tweet about this
  • Isn’t it funny that people tweet about thisDifference between cakes and biscuits? When stale, cakes go hard, biscuits go soft. Hence Jaffa Cakes are cakes. (Was official EU ruling).
  • Delete the article]...hasn't played since 2008. His 66-73 record is far from stellar and, in my opinion, does not merit an article.
  • It was good to see that EKAW is developing a serious focus on proving that the technologies developed in this field can scale up to actual use. This year, an in-use paper category was introduced and I think this might develop into an important platform research needed to scale up some of the technologies to enterprise scale. If the number of contributors will rise, and the right reviewing criteria are developed, this could be a great contribution http://www.vangrondelle.com/2010/10/ekaw-2010/
  • Abstract argumentation frameworks have been well-developed to support reasoning with inconsistent information starting with [6] and much subsequent research ([1], [2], [3]). An abstract argument framework, as introduced by Dung, [6] is a pair AF = hA; attacki, where A is a set of arguments and attack a binary relation on A. A variety of semantics are available to evaluate the arguments. For example, where AF = hfA1; A2; A3; A6; A7g; fatt(A6; A1); att(A1; A6); att(A7; A2)gi, then the preferred extensions are: fA3, A6, A7g and fA2, A3, A7g. However, Dung’s arguments are entirely abstract and the attack relation is stipulated. In other words, it is unclear why one argument attacks another argument, as there is no content to the arguments. In order to instantiate arguments we need argumentation schemes, which are presumptive patterns of reasoning [10]. An instantiated argumentation scheme, such as Position To Know, has a textual form such as: 1. Ms. Peters is in a position to know whether Mr. Jones was at the party. 2. Ms. Peters asserts that Mr. Jones was at the party. 3. Therefore, presumptively, Mr. Jones was at the party. This has a formal representation in a typed logical language with functions from argument objects to predicates. The language formally represents the propositions required of the scheme as well as aspects of defeasible reasoning [12].
  • Abstract argumentation frameworks have been well-developed to support reasoning with inconsistent information starting with [6] and much subsequent research ([1], [2], [3]). An abstract argument framework, as introduced by Dung, [6] is a pair AF = hA; attacki, where A is a set of arguments and attack a binary relation on A. A variety of semantics are available to evaluate the arguments. For example, where AF = hfA1; A2; A3; A6; A7g; fatt(A6; A1); att(A1; A6); att(A7; A2)gi, then the preferred extensions are: fA3, A6, A7g and fA2, A3, A7g. However, Dung’s arguments are entirely abstract and the attack relation is stipulated. In other words, it is unclear why one argument attacks another argument, as there is no content to the arguments. In order to instantiate arguments we need argumentation schemes, which are presumptive patterns of reasoning [10]. An instantiated argumentation scheme, such as Position To Know, has a textual form such as: 1. Ms. Peters is in a position to know whether Mr. Jones was at the party. 2. Ms. Peters asserts that Mr. Jones was at the party. 3. Therefore, presumptively, Mr. Jones was at the party. This has a formal representation in a typed logical language with functions from argument objects to predicates. The language formally represents the propositions required of the scheme as well as aspects of defeasible reasoning [12].
  • . The issue is whether it is the right product for the buyer, which is a matter not only of the pros and cons, but also of the explanations and counterarguments given. In our view, current approaches detect problems, but obscure the chains of reasoning about them.
  • . The issue is whether it is the right product for the buyer, which is a matter not only of the pros and cons, but also of the explanations and counterarguments given. In our view, current approaches detect problems, but obscure the chains of reasoning about them.
  • . The issue is whether it is the right product for the buyer, which is a matter not only of the pros and cons, but also of the explanations and counterarguments given. In our view, current approaches detect problems, but obscure the chains of reasoning about them.
  • . The issue is whether it is the right product for the buyer, which is a matter not only of the pros and cons, but also of the explanations and counterarguments given. In our view, current approaches detect problems, but obscure the chains of reasoning about them.

Dimensions of argumentation in social media EKAW2012 2012 10 11 Presentation Transcript

  • 1. Digital Enterprise Research Institute www.deri.ie Dimensions of Argumentation in Social Media Jodi Schneider1, Brian Davis1, and Adam Wyner2 1Digital Enterprise Research Institute, NUI Galway 2Department of Computer Science, University of Liverpool EKAW, Galway, Ireland Thursday 11th October 2012 Copyright 2011 Digital Enterprise Research Institute. All rights reserved. Enabling Networked Knowledge 1
  • 2. Want to reuse social media dataDigital Enterprise Research Institute www.deri.ie  Growing amount of social media data  Want to reuse data  Need to analyse & structure it Enabling Networked Knowledge
  • 3. Key PointDigital Enterprise Research Institute www.deri.ie The overall popularity of an opinion is not as important as the reasons supporting it Image: http://www.nickmilton.com/2012/03/when-people-trust-crowds.html Enabling Networked Knowledge
  • 4. ArgumentsDigital Enterprise Research Institute www.deri.ie Claim: Jaffa Cakes are cakes Justification: official EU ruling; cakes go hard when stale https://twitter.com/robeastaway/status/135838892694839296 Enabling Networked Knowledge 4
  • 5. ArgumentsDigital Enterprise Research Institute www.deri.ie  Claim: If Canon does not make the repair, you should take them to small claims court.  Justification: it’s easy, cheap, and can be done online; it’s within your rights Enabling Networked Knowledge 5
  • 6. Knowledge bases of social mediaDigital Enterprise Research Institute www.deri.ie  Growing amount of social media data  Want to reuse data  Need to analyse & structure it Envision: Knowledge bases of social media  Query  Make a representation we can reason with – even in the face of inconsistency. Enabling Networked Knowledge
  • 7. Oops!Digital Enterprise Research Institute www.deri.ie  People disagree  Inconsistency is bad for knowledge bases!  How do we reconcile & resolve the inconsistencies? Enabling Networked Knowledge
  • 8. Argumentation to the Rescue!Digital Enterprise Research Institute www.deri.ie  People disagree  Inconsistency is bad for knowledge bases!  How do we reconcile & resolve the inconsistencies?  Argumentation!  Claims  Justifications  Contrary Claims & Disagreements… Enabling Networked Knowledge
  • 9. Formal RepresentationsDigital Enterprise Research Institute www.deri.ie  Argumentation Frameworks  Variety of Semantics – can be used to choose best options Enabling Networked Knowledge
  • 10. Calculate best optionsDigital Enterprise Research Institute (non-contradictory opinions) www.deri.ie Wyner, van Engers, & Bahreini. From Policy-making Statements to First-order Logic. EGOV 2010 Enabling Networked Knowledge 10
  • 11. Formal RepresentationsDigital Enterprise Research Institute www.deri.ie  Argumentation Schemes  Appeal to Expert Opinion  Appeal to Popular Opinion  From Analogy  Slippery Slope  ....  Indicate Relevant “Critical Questions” for a discussion  Patterns for Information Extraction (SWAIE 2012) Enabling Networked Knowledge
  • 12. John Danaher, http://philosophicaldisquisitions.blogspot.ie/2010/03/argumentation-schemes-part-1.html
  • 13. GoalDigital Enterprise Research Institute www.deri.ie  Extract arguments from text.  Reconstruct argumentation into formal representations.  Make a representation that we can reason with even in the face of inconsistency. Enabling Networked Knowledge
  • 14. How to automate?Digital Enterprise Research Institute www.deri.ie  What aspects of social media are relevant to extracting and representing argumentation? Enabling Networked Knowledge
  • 15. Dimensions of (Argumentative)Digital Enterprise Research Institute Expression www.deri.ie  Genre  Metadata  Properties of users  Goals of a particular dialogue  Context  Informal language  Implicit info  Sentiment techniques  Subjectivity and objectivity Enabling Networked Knowledge
  • 16. Genre 16
  • 17. Metadata 17
  • 18. Properties of UsersDigital Enterprise Research Institute www.deri.ie Enabling Networked Knowledge 18
  • 19. GoalsDigital Enterprise Research Institute www.deri.ie ! " % ( )*+, ( #$ & -. " #$ 1 #/ % / )0" . 2#3/ 4" #. 15 % , 6)7 #$ 7 % )% )! " % ( #$ 8 #$ & Find and Verify -. 9 "3+ Need to Have Proof Prove (Disprove) Hypothesis Evidence Need to Find an Find and Defend a ! " : ( 3+ 64% Choose Best Hypothesis for Testing Explanation of Facts Suitable Hypothesis Acquire or Give -. 8 3; #/ % < ( =" & % . 0( . Need Information Exchange Information Information Dilemma or Practical Coordinate Goals and ! ( $>( 3 % " #/ . Decide Best Available Course of Action Choice Actions 2( 36 #6" . % Conflict of Opinions Persuade Other Party Resolve or Clarify Issue Get What You Most ? ( & / #/ % % . Conflict of Interests Reasonable Settlement Both Can Live With Want Verbally Hit Out at @ 6/ 4 3" Personal Conflict Reveal Deeper Basis of Conflict Opponent Enabling Networked Knowledge 19
  • 20. Goals of #election2012 tweets?Digital Enterprise Research Institute www.deri.ie Enabling Networked Knowledge 20
  • 21. Related PapersDigital Enterprise Research Institute www.deri.ie  Schneider and Wyner. “Identifying Consumers Arguments in Text”, SWAIE 2012 at EKAW 2012.  Wyner, Schneider, Atkinson, and Bench-Capon. “Semi-Automated Argumentative Analysis of Online Product Reviews.” In 4th International Conference on Computational Models of Argument (COMMA 2012).  Wyner and Schneider. Arguing from a point of view, Agreement Technologies 2012. 21 Enabling Networked Knowledge October 9, 2012
  • 22. AcknowledgmentsDigital Enterprise Research Institute www.deri.ie  Project Funding  Science Foundation Ireland Grant No. SFI/08/CE/I1380 (Líon-2)  Short-term scientific mission (STSM 1868) from the COST Action ICO801 on Agreement Technologies  FP7-ICT-2009-4 Programme, IMPACT Project, Grant Agreement Number 247228 Enabling Networked Knowledge 22
  • 23. Thanks!Digital Enterprise Research Institute www.deri.ie  Questions?  Contacts:  Jodi Schneider jodi.schneider@deri.org @jschneider  Brian Davis brian.davis@deri.org  Adam Wyner adam@wyner.info Enabling Networked Knowledge
  • 24. Digital Enterprise Research Institute www.deri.ie Examples of Arguments in Social Media Copyright 2011 Digital Enterprise Research Institute. All rights reserved. Enabling Networked Knowledge 24
  • 25. 25
  • 26. ArgumentsDigital Enterprise Research Institute www.deri.ie Claim: Jaffa Cakes are cakes Justification: official EU ruling; cakes go hard when stale https://twitter.com/robeastaway/status/135838892694839296 Enabling Networked Knowledge 26
  • 27. 27 http://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/Heath_Totten
  • 28. ArgumentsDigital Enterprise Research Institute www.deri.ie Claim: does not merit an article Justification: hasn’t played since 2008, mediocre record Enabling Networked Knowledge 28
  • 29. 29
  • 30. http://www.amazon.co.uk/review/R34F9M7871K2W0/ref=cm_cr_rev_detmd_pl?ie=UTF8&asin=B004M8S152&cdForum=FxM7A99WYT0UJ9&cdMsgID=Mx35ZEPGH4SV7JK&cdMsgNo=1&cdPage=1&cdSort=oldest&cdThread=Tx1C124R97075W9&store=photo#Mx35ZEPGH4SV7JK 30
  • 31. ArgumentsDigital Enterprise Research Institute www.deri.ie  Claim: If Canon does not make the repair, you should take them to small claims court.  Justification: it’s easy, cheap, and can be done online; it’s within your rights Enabling Networked Knowledge 31
  • 32. http://www.vangrondelle.com/2010/10/ekaw-2010/
  • 33. ArgumentsDigital Enterprise Research Institute www.deri.ie  Claim: The in-use paper category is important.  Justification: Provides a platform for research scaling up to enterprise scale. Enabling Networked Knowledge 33
  • 34. Digital Enterprise Research Institute www.deri.ie etc Copyright 2011 Digital Enterprise Research Institute. All rights reserved. Enabling Networked Knowledge 34
  • 35. Using ArgumentationDigital Enterprise Research Institute www.deri.ie  Mark & help resolve inconsistencies  Provide justifications & explanations Enabling Networked Knowledge 35
  • 36. Formalism: AbstractDigital Enterprise Research Institute argumentation frameworks www.deri.ie AF = hA; attacki, where A is a set of arguments and attack a binary relation on A. A variety of semantics are available to evaluate the arguments. Enabling Networked Knowledge
  • 37. Formalism:Digital Enterprise Research Institute argumentation schemes www.deri.ie  Presumptive patterns of reasoning  Position To Know  1. Ms. Peters is in a position to know whether Mr. Jones was at the party.  2. Ms. Peters asserts that Mr. Jones was at the party.  3. Therefore, presumptively, Mr. Jones was at the party. Enabling Networked Knowledge
  • 38. Key PointDigital Enterprise Research Institute www.deri.ie The overall popularity of an opinion is not as important as the reasons supporting it Image: http://www.nickmilton.com/2012/03/when-people-trust-crowds.html Enabling Networked Knowledge
  • 39. Key PointDigital Enterprise Research Institute www.deri.ie overwhelming numbers of people may not matter Photograph DAVID GILES/PA NEWS WIRE/AP PHOTO - February 7, 2011 New Yorker via http://www.vincentskeltis.com/blog/2011/2/7/crowd-crush.html Enabling Networked Knowledge
  • 40. Key PointDigital Enterprise Research Institute www.deri.ie as much as a particular reason against it http://www.mirror.co.uk/3am/celebrity-news/crowd-crush-mars-jls-no1-431245 Enabling Networked Knowledge
  • 41. Digital Enterprise Research Institute www.deri.ie Enabling Networked Knowledge