Representing, Proving and Sharing Trustworthiness of Web Resources Using Veracity
1. Representing, Proving and Sharing Trustworthiness of Web Resources Using Veracity GrégoireBurel, Amparo E. Cano, Matthew Rowe and Alfonso Sosa, OAK Group - University Of Sheffield EKAW2010, Lisbon, Portugal 12 October 2010
3. Trustworthiness and WWW What is truewhat is false? (Can I rely on this information ?) 13+ Billion triples 500+ Million Facebook Users 36+ Billion pages
4. Trustworthiness and WWW Questions of Trust: (Policies, provenance and content) Who said this ? Really ? Is this person trustworthy ? Is this information trustworthy ? Is the information not compromised? Is it true ? Why ? Still ? (…)
9. So what is trustworthiness ? (1) A proposition is trustworthy if the carried information is reliable, commonly accepted as true given pre-existing trusted knowledge and stable over time.
10. So what is trustworthiness ? (2) Many components and factors: (Depending on the context…) Willingness, confidence, risk and control (Mayer and Davis, 1995) Content Trust vs. Entity Trust and 19 factors of trust (Gil and Artz, 2007) Scope, Content Services and People (Golbeck, 2006) (…) Social Trust (= Entity Trust) vs. Rational Trust (≈ Content Trust)
11. So what is trustworthiness ? (2) Many components and factors: (Depending on the context…) Willingness, confidence, risk and control (Mayer and Davis, 1995) Content Trust vs. Entity Trust and 19 factors of trust (Gil and Artz, 2007) Scope, Content Services and People (Golbeck, 2006) (…) Social Trust (= Entity Trust) vs. Rational Trust (≈ Content Trust)
12. Veracity, shared Model of Rational Trustworthiness… A tripartiteTrustworthiness model and an Ontology based on Social Trustworthiness and Rational Trustworthiness.
13. Veracity, shared Model of Rational Trustworthiness… A tripartiteTrustworthiness model and an Ontology based on Social Trustworthiness and Rational Trustworthiness. Social Trustworthiness: The validity of an information can only be verified by trusting the person that wrote the statement or confirms it.
14.
15.
16. The Veracity ontology How much [0…1] How much? [0..1] WOT FOAF True or false ? http://purl.org/net/veracity/ns
17. Veracity in Real-Life Did Einstein work at the University of Berlin and was a Physicist ? Alice Alice needs to know if Einstein was really a Physicist and worked at the university of Berlin. Unfortunately she doesn’t know anything about Einstein: Can she trust the author of the proposition ? (Social Trust) Does the author know about Einstein ? (Rational Trust, Knowledge Justification) Is there any other existing information meaning the same ? (Rational Trust, Knowledge Reference) Does anybody else know about it ?
19. Veracity in Real-Life Can I trust this author (Stephen Hawking) ? Alice Social You don’t know who he is. You should probably not trust him. Veracity
20. Veracity in Real-Life Can I trust this author (Stephen Hawking) ? Alice Social You don’t know who he is. You should probably not trust him. Veracity Doesthis author know about Einstein? Alice
21. Veracity in Real-Life Can I trust this author (Stephen Hawkins) ? Alice Social You don’t know who he is. You should probably not trust him. Veracity Doesthis author know about Einstein? Alice Rational, Knowledge Justification Yes, He is known to be a famous Physicist. Veracity
23. Veracity in Real-Life Is there any other resources supporting this statement? Alice Rational, Knowledge Reference Yes, there is other trustworthy resources asserting similar propositions. Veracity
24. Veracity in Real-Life Is there any other resources supporting this statement? Alice Rational, Knowledge Reference Yes, there is other trustworthy resources asserting similar propositions. Veracity Is there anybody else confirming that statement? Alice
25. Veracity in Real-Life Is there any other resources supporting this statement? Alice Rational, Knowledge Reference Yes, there is other trustworthy resources asserting similar propositions. Veracity Is there anybody else confirming that statement? Alice Yes, some people confirmed the current proposition. However, you should check if they are trustworthy.
34. Conclusions Veracityis an abstractmodel and ontology . Veracity introduces the concept of social trustworthiness and rational trustworthiness Veracity supports social and rational trustworthiness. Veracity support recursive trust assertion. Veracity can be applied with different trust metrics. Veracity is decentralised. Veracity does not rely on trust assumptions. Veracity ensure the integrity of trust statements (Digital Signature).
35. Questions ? VeracityOntology http://purl.org/net/veracity/ns GrégoireBurel, Amparo E. Cano, Matthew Rowe and Alfonso Sosa, {g.burel,e.cano,m.rowe,a.sosa}@dcs.shef.ac.uk OAK Group - University Of Sheffield
36. Questions ? VeracityOntology http://purl.org/net/veracity/ns GrégoireBurel, Amparo E. Cano, Matthew Rowe and Alfonso Sosa, {g.burel,e.cano,m.rowe,a.sosa}@dcs.shef.ac.uk OAK Group - University Of Sheffield
Hinweis der Redaktion
There is more an more people contributing content on the Web. In all this big cloud of information, it is very easy to find contradictory data. How can we decide which one is true and which one is false. Assesing trust on the web is fundamentetal A lot of information on the web All this huge amount of data introduces new challenges related in On what to trust an whom to trust.With the size of the web , is very easy to find contradictory information, where is hard to decide which one is true or false
What we could say though, is that trustworthiness is not equivalent to trust, and that trustworthiness is not provenance,
What we could say though, is that trustworthiness is not equivalent to trust, and that trustworthiness is not provenance,
First of all, let’s discuss a little bit about our understanding of trustworthiness. The definition of trustworthiness is not absolute but highly contextual, it encapsulates social and personal concepts such as reputability, popularity, reliability and likelihood
Some of this concepts are ambiguous..Assesing trust on the web is fundamentetal A lot of information on the web All this huge amount of data introduces new challenges related in On what to trust an whom to trust.With the size of the web , is very easy to find contradictory information, where is hard to decide which one is true or falseFirst of all, let’s discuss a little bit about our understanding of trustworthiness. The definition of trustworthiness is not absolute but highly contextual, it encapsulates social and personal concepts such as reputability, popularity, reliability and likelihood
First of all, let’s discuss a little bit about our understanding of trustworthiness. The definition of trustworthiness is not absolute but highly contextual, it encapsulates social and personal concepts such as reputability, popularity, reliability and likelihood
With this requirements in mind, we have introduced a lightweight, decentralised model that enables the assertion of explanatory trust. This model highlights differences among, social trustworthiness, rational trustworthiness, and the sharing of trustworthiness. We have a tripartite model.WithBob says “This information is true !”2.a) Bob says “This information is true” and “You can believe me because I have a proof that I know about it”2.b) Bob says “There is another information that tends to affirms the same (that this information is true)”
The term rational trust highlight the contrast between pure social trust and knowledge based trust
Is a layered model that allows recursive trust assertions on a piece of information. The knowledge justification sits on top of the social trustworthiness.With this requirements in mind, we have introduced a lightweight, decentralised model that enables the assertion of explanatory trust. This model highlights differences among, social trustworthiness, rational trustworthiness, and the sharing of trustworthiness. We have a tripartite model.WithBob says “This information is true !”2.a) Bob says “This information is true” and “You can believe me because I have a proof that I know about it”2.b) Bob says “There is another information that tends to affirms the same (that this information is true)”
Following this model of trustworthiness we have created the veracity ontology which includes the concepts of
Option for Alice to evaluate if this proposition is true are:In the first case, Alice just needs to compare if her beliefs match the propo- sition. This case is unlikely since Alice does not actually know anything about the scientist yetThe second case implies that the editor of the information is reliable in general or in the terms of Alice. For example, the editor of the piece may be a well-known historian or a person she trusts personallyThe third case would require Alice too look for information about the author of the proposition that confirms that he knows directly or not about Einstein. In the fourth case, Alice needs to look for external information or a reference that confirms the considered propositionFinally, in the last case,Alice would need to find somebody that knows directly about the topic. This means that she may ask somebody to confirm the proposition written in the Web page. Alice needs to perform a lot of manual tasks in order confirm or not the trustworthiness of information that may not be performed accurately.Moreover, if we consider that Alice could be a software agent, these tasks become impossible as all the techniques rely on subjective judgments in order to work properly.
Some of the requirements highlighted by this use case includes: Identify a proposition Describe the trustworthiness of a proposition Identify an agent Provide agent credentials Provide supporting information Security of assertions Reliable assertions No predefined trust assumptions