1 / 73

V.G.Vinod Vydiswaran Department of Computer Science University of Illinois at Urbana-Champaign

Can you believe what you read online?: Modeling and Predicting Trustworthiness of Online Textual Information. V.G.Vinod Vydiswaran Department of Computer Science University of Illinois at Urbana-Champaign December 5 th , 2011. Web content: structured and free-text.

ojackson
Télécharger la présentation

V.G.Vinod Vydiswaran Department of Computer Science University of Illinois at Urbana-Champaign

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Can you believe what you read online?: Modeling and Predicting Trustworthiness of Online Textual Information V.G.VinodVydiswaran Department of Computer Science University of Illinois at Urbana-Champaign December 5th, 2011

  2. Web content: structured and free-text Are all these pieces of information equally trustable?

  3. Even reputed sources can make mistakes Some sources / claims can be misleading on purpose.

  4. Blogs and forums give information too Sources may not be “reputed”, but information can still be trusted.

  5. RESULT : Duck’s quack echoes! Echoing quack of a duck Is trustworthiness always objective?

  6. Treating cancer Should trustworthiness be subjective?

  7. Is drinking alcohol good for the body? “A review article of the latest studies looking at red wine and cardiovascular health shows drinking two to three glasses of red wine daily is good for the heart.” Dr. Bauer Sumpio, M.D. , Prof., Yale School of Medicine Journal of American College of Surgeons,2005 “Women who regularly drink a small amount of alcohol — less than a drink a day — are more likely to develop breast cancer in their lifetime than those who don’t.” Dr. Wendy Chen, Asst. Prof. (Medicine), Harvard Medical School Journal of the American Medical Association (JAMA), 2011 Is drinking alcohol good for the body? Are these sources “reliable”? What other sources/documents say about this information?

  8. Every coin has two sides Should trustworthiness be user-dependent? People tend to be biased, and may be exposed to only one side of the story Confirmation bias Effects of filter bubble For intelligent choices, it is wiser to also know about the other side What is considered trustworthy may depend on the person’s viewpoint

  9. Milk is good for humans… or is it? Milk contains nine essential nutrients… Dairy products add significant amounts of cholesterol and saturated fat to the diet... The protein in milk is high quality, which means it contains all of the essential amino acids or 'building blocks' of protein. Milk proteins, milk sugar, and saturated fat in dairy products pose health risks for children and encourage the development of obesity, diabetes, and heart disease... It is long established that milk supports growth and bone development rbST [man-made bovine growth hormone] has no biological effects in humans. There is no way that bST [naturally-occurring bovine growth hormone] or rbST in milk induces early puberty. Given these evidence docs, users can make a decision Drinking of cow milk has been linked to iron-deficiency anemia in infants and children One outbreak of development of enlarged breasts in boys and premature development of breast buds in girls in Bahrain was traced to ingestion of milk from a cow given continuous estrogen treatment by its owner to ensure uninterrupted milk production.

  10. Actors in the trustworthiness story Source Claim Users Drinking alcohol is good for the body. Dr. Bauer Sumpio Dr. Wendy Chen Evidence “A review article of the latest studies looking at red wine and cardiovascular health shows drinking two to three glasses of red wine daily is good for the heart.” News Corpus Women who regularly drink a small amount of alcohol — less than a drink a day — are more likely to develop breast cancer in their lifetime than those who don’t. Data Medical sites Forums ClaimVerifier Blogs

  11. ClaimVerifier: Thesis contribution Free-text claims Source Claim Users Evidence Novel system for users to validate textual claims Incorporate textual evidence into trust models Data ClaimVerifier

  12. Challenges in building ClaimVerifier Source Claim How to assign truth values to textual claims? Users Are sources trustworthy? How to present evidence? How to address user bias? Evidence How to build trust models that make use of evidence? How to find relevant pieces of evidence ? What kind of data can be utilized? Data ClaimVerifier

  13. Outline of the talk 1 4 3 1 2 3 2 4 Measuring source trustworthiness Using forums as data source Content-based trust propagation models User biases and interface design

  14. ClaimVerifier: Trustworthiness factors Source Claim Source Trustworthiness 1 Users User bias may affect perceived trustworthiness Evidence 4 Knowing why something is true is important Data 3 Information from blogs and forums ClaimVerifier 2

  15. 1 Identify trustworthy websites (sources) Case study: Medical websites Joint work with ParikshitSondhi and ChengXiangZhai (ECIR 2012)

  16. Variations in online medical information

  17. Problem Statement • For a (medical) website • What features indicate trustworthiness? • How can you automate extracting these features? • Can you learn to distinguish trustworthy websites from others?

  18. “cure back pain”: Top 10 results health2us.com Content Presentation Financial interest Transparency Complementarity Authorship Privacy

  19. Trustworthiness of medical websites HON code Principles • Authoritative • Complementarity • Privacy • Attribution • Justifiability • Transparency • Financial disclosure • Advertising policy Our model (automated) • Link-based features • Transparency • Privacy Policy • Advertising links • Page-based features • Commercial words • Content words • Presentation • Website-based features • Page Rank

  20. Research questions HON code principles link, page, site features Yes Learned SVM and used it to re-rank results • For a (medical) website • What features indicate trustworthiness? • How can you automate extracting these features? • Can you learn to distinguish trustworthy websites from others? • Bias results to prefer trustworthy websites?

  21. Use classifier to re-rank results +8.5% Relative

  22. 2 Understanding what is written (evidence) Case study: Scoring medical claims based on health forums (KDD 2011 Workshop on Data Mining for Medicine and Healthcare)

  23. Many medical support groups available

  24. Scoring claims via community knowledge Claim DB Claim DB Claim DB Claim DB Claim Essiac tea is an effective treatment for cancer. Chemotherapy is an effective treatment for cancer. Evidence & Support DB

  25. Problem statement Given • A corpus of community generated content • E.g. forum postings, mailing lists • A database of relations (“claims”) • E.g. [disease, treatments], [treatments, side-effects] can we • Rank and score the claims based on their support / verifiability, as demonstrated in the text corpus • Build scoring function to rate databases as a whole

  26. Key steps • Relation Retrieval • Query Formulation • Parse result snippet • Find sentiment expressed in snippet • Score snippets • Aggregate them to get claim score A Collect relevant evidence for claims B Analyze the evidence documents C Score and aggregate evidence Ranked treatment claims

  27. A Collect relevant evidence for claims What is a Relation? Relation Entity Entity protect type type PennState Sandusky ORG PER cured by Cancer Essiac tea Treatment Disease • Entities • Nouns • Objects of interest • Possibly typed • Relation words • Verbs • Binary, for now • Usually with roles (entities participating in the relation have specific roles)

  28. A Collect relevant evidence for claims Relation Retrieval Relation • Query Formulation • structured relation • possibly typed • Query Expansion • Relation: with synonyms, words with similar contexts • Entities: with acronyms, common synonyms • Query weighting Entity Entity cure treat help prevent reduce type type cured by Entity 1 Entity 2 Treatment Disease Impotence ED Erectile Dysfunction Infertility Chemotherapy Chemo

  29. Key steps revisited Relation • Relation Retrieval • Query Formulation • Parse result snippet • Find sentiment expressed in snippet • Score snippets • Aggregate them to get claim score A Collect relevant evidence for claims Entity Entity type type B Analyze the evidence documents C Score and aggregate evidence Ranked treatment claims

  30. C Score and aggregate evidence Scoring post snippets Number of snippets found relevant (popularity) Number of evidence contexts extracted Orientation of the posts (positive / negative) Percentage of opinion words Subjectivity of opinion words used (strong, weak) Length of posts Relevance of the post to the claim

  31. Treatment effectiveness based on forums Ignored Treatment claims Source Claim Users QUESTION: Which treatments are more effective than others for a disease? Evidence Forum posts describing effectiveness of the treatment Data Health forums and medical message boards ClaimVerifier

  32. Evaluation: Corpus statistics • Collection of nine health forums and discussion boards • Not restricted to any specific disease or treatment class

  33. Treatment claims considered

  34. Results: Ranking valid treatments • Datasets • Skewed: 5 random valid + all invalid treatments • Balanced: 5 random valid + 5 random invalid treatments • Finding: Our approach improves ranking of valid treatments, significant in Skewed dataset.

  35. Measuring site “trustworthiness” Trustworthiness should decrease Database score Ratio of degradation

  36. Over all six disease test sets • As noise added to the claim database, the overall score reduces. • Exception: Arthritis, because it starts off with a negative score

  37. Conclusion: Scoring claims using forums It is feasible to score trustworthiness claims using signal from million of patient posts We scored treatment posts based on subjectivity of opinionwords, and extended the notion to score databases It is possible to reliably leverage this general idea of validating knowledge through crowd-sourcing

  38. Modeling trust over content and source 3 Source Claim How to assign truth values to textual claims? 1 Users Are sources trustworthy? Evidence How to build trust models that make use of evidence? 2 How to find relevant pieces of evidence ? What kind of data can be utilized? Data ClaimVerifier

  39. 3 Content-Driven Trust Propagation Framework Case study: News Trustworthiness (KDD 2011)

  40. Problem: How to verify claims? Trust Sources Evidence Claims Claim 1 “A duck’s quack doesn’t echo.” “PennState protected Sandusky.” Passages that give evidence for the claim Web Sources News coverage on the issue of “Immigration” is biased. News media (or reporters) News stories

  41. Typical fact-finding is over structured data Sources Claims Assume structured claims and accurate IE modules Claim 1 Claim 2 . . . Claim n

  42. Traditional 2-layer Fact-Finder model s1 c1 s2 Trustworthiness of source c2 s3 c3 s4 Veracity of claims c4 s5 No Context • Hub-Authority style • Prior research • TruthFinder: Yin, Han, & Yu, 2007 • Pasternack & Roth, 2010 • Many more …

  43. Incorporating Text in Trust Models Sources Evidence Claims Claim 1 Claim 2 . . . Claim n 1. Textual evidence 2. Supports adding IE accuracy, relevance,similarity between text Free-text claims Special case:structured data

  44. Evidence-based 3-layer model e1 s1 e2 c1 e3 s2 c2 e4 s3 e5 c3 e6 s4 e7 c4 s5 e8 e9 e10 • Defining the model • Defining the three parameters • Initialization • Framework layers • Scores for layers • Trust propagation over the framework • Handling influence between layers

  45. Understanding model parameters • Scores computed • : Claim veracity • : Evidence trust • : Source trust • Influence factors • : evidence similarity • : Relevance • : Source-Evidence influence • Initializing • Uniform distribution for • Retrieval score for

  46. Computing Trust scores • Veracity of a claim depends on • the evidence documents for the claim and their sources. • Trustworthiness of a source is based on the claims it supports. • Confidence in an evidence document depends onsource trustworthiness and confidence in other similar documents. Trust scores computed iteratively

  47. Computing Trust scores Trustworthiness of source of evidence ej Sum over all other pieces of evidence for claim c(ei) Similarity of evidence ei to ej Relevance of evidence ej to claim ci Trust scores computed iteratively Adding influence factors

  48. Possible extensions • Structured claims • No context Mutually exclusive claims, constraints[Pasternack & Roth, 2011] Structure on sources, groups [Pasternack & Roth, 2011] Source copying [Dong, Srivastava, et al., 2009]

  49. Application: Trustworthiness in News Sources Evidence Claims ? Claim 1 Biased news coverage on a particular topic or genre? News media (or reporters) News stories Which news stories can you trust? Whom can you trust? How true is a claim?

  50. News trustworthiness Data collected from NewsTrust (Politics category) Articles have been scored by volunteers on journalistic standards Scores on [1,5] scale Some genres inherently more trustworthy than others

More Related