=Paper=
{{Paper
|id=Vol-423/paper-10
|storemode=property
|title=Uncertainty Reasoning for the World Wide Web: Report on the URW3-XG Incubator Group
|pdfUrl=https://ceur-ws.org/Vol-423/paper10.pdf
|volume=Vol-423
|dblpUrl=https://dblp.org/rec/conf/semweb/LaskeyL08
}}
==Uncertainty Reasoning for the World Wide Web: Report on the URW3-XG Incubator Group==
https://ceur-ws.org/Vol-423/paper10.pdf
Uncertainty Reasoning for the World Wide Web:
Report on the URW3-XG Incubator Group
Kenneth J. Laskey# Kathryn Blackmond Laskey
MITRE Corporation, M/S H305 Department of Systems Engineering
7515 Colshire Drive and Operations Research
McLean, VA 22102-7508 USA George Mason University
klaskey@mitre.org 4400 University Drive
Fairfax, VA 22030-4444 USA
klaskey@gmu.edu
Abstract. The Uncertainty Reasoning for the World Wide Web Incubator
Group (URW3-XG) was chartered as a means to explore and better define the
challenges of reasoning with and representing uncertain information in the con-
text of the World Wide Web. The objectives of the URW3-XG were: (1) To
identify and describe situations on the scale of the World Wide Web for which
uncertainty reasoning would significantly increase the potential for extracting
useful information; and (2) To identify methodologies that can be applied to
these situations and the fundamentals of a standardized representation that
could serve as the basis for information exchange necessary for these method-
ologies to be effectively used. This paper describes the activities undertaken by
the URW3-XG, the recommendations produced by the group, and next steps re-
quired to carry forward the work begun by the group.
1 Introduction
The Uncertainty Reasoning for the World Wide Web Incubator Group (URW3-XG)
was proposed [1] during the 2006 Uncertainty Reasoning for the Semantic Web
workshop [2] as a means to explore and better define the challenges of reasoning with
and representing uncertain information in the context of the Semantic Web. In addi-
tion, it was intended to identify situations in which the combination of semantics and
uncertainty could further the Web Services vision of quickly and efficiently compos-
ing services and data resources to address the needs of users in an ever-changing
world.
The 2006 workshop included a Use Case Challenge [3] to generate an initial col-
lection of use cases and to gauge the interest of the workshop participants in continu-
# ·
The author's affiliation with The MITRE Corporation is provided for identification purposes
only, and is not intended to convey or imply MITRE's concurrence with, or support for, the
positions, opinions or viewpoints expressed by the author.
ing as a W3C XG or through some other collaboration venue. The Use Case Chal-
lenge generated a lively interchange of ideas, and the participants overwhelmingly
agreed to create the XG to continue the work.
2 W3C Incubator (XG) Process
As noted in [1], the World Wide Web Consortium (W3C) [4] created the Incubator
process [5] to provide a formal, yet flexible venue to better understand Web-related
challenges and their potential solutions. It encourages a public exploration of issues
and potential solutions before the solutions are mature enough for standardization. It
also provides a “head start” if the Incubator experimental group, the XG, is able to
adequately formulate principles and techniques that gain consensus in the wider
community.
The URW3-XG was in operation [6] from 5 March 2007 until its final report [7]
was published by the W3C on 22 April 2008. The group included 25 participants from
North and South America, Europe, and Australia. Participants came from a range of
time zones spanning 18 hours. The group conducted over 20 telecons, with an average
duration between 90 and 120 minutes. In addition, face-to-face meetings of subsets of
the XG were held at the 5th ISWC (Busan - Korea) and the SUM conference in Col-
lege Park, Maryland USA. The telecons were supported by the W3C resources (e.g.
telecon bridge, IRC, RSSAgent, etc). Meeting results and action items were cata-
logued in online Minutes [6].
The objectives of the URW3-XG were twofold:
• To identify and describe situations on the scale of the World Wide Web
for which uncertainty reasoning would significantly increase the potential
for extracting useful information; and,
• To identify methodologies that can be applied to these situations and the
fundamentals of a standardized representation that could serve as the basis
for information exchange necessary for these methodologies to be effec-
tively used.
3 Results of the URW3-XG Effort
The Final Report [7] was the major deliverable of the URW3-XG. It describes the
work done by the XG, identifies elements of uncertainty that need to be represented to
support reasoning under uncertainty for the World Wide Web, and provides an over-
view of the applicability to the World Wide Web of various uncertainty reasoning
techniques (in particular, probability theory, fuzzy logic, and belief functions) and the
information that needs to be represented for effective uncertainty reasoning to be
possible. The report concludes with a discussion on the benefits of standardization of
uncertainty representation to the World Wide Web and the Semantic Web and pro-
vides a series of recommendations for continued work. The report also includes a
Reference List of work relevant to the challenge of developing standardized represen-
tations for uncertainty and exploiting them in Web-based services and applications.
A major part of the work was development of a set of use cases illustrating condi-
tions under which uncertainty reasoning is important. Another major effort was the
development of an Uncertainty Ontology that was used to categorize uncertainty
found in the use cases. These products are described briefly in the following sections.
Section 4 then details the conclusions and recommendations from the report.
3.1 The Uncertainty Ontology
The Uncertainty Ontology is a simple ontology developed to demonstrate some ba-
sic functionality of exchanging uncertain information. It was used to classify the use
cases developed by the URW3-XG with the intent of obtaining a relatively complete
coverage of the functionalities related to uncertainty reasoning about information
available on the World Wide Web. The top level of the ontology is shown in Figure 1.
According to the ontology, uncertainty is associated with sentences that make asser-
tions about the world, and are asserted by agents (human or computer). The uncer-
tainty derivation may be objective (via a formal, repeatable process) or subjective
(judgment or guess). Uncertainty type includes ambiguity, empirical uncertainty,
randomness, vagueness, inconsistency and incompleteness. Uncertainty models in-
clude probability, fuzzy logic, belief functions, rough sets, and other mathematical
models for reasoning under uncertainty. Uncertainty nature includes aleatory (chance;
inherent in the phenomenon) or epistemic (belief; due to limited knowledge of the
agent).
Figure 1 Top level of URW3-XG Uncertainty Ontology
While this ontology served the purpose of focusing discussion of the use cases, al-
lowing use case developers to show examples of annotation of uncertainty, the ontol-
ogy was only meant to provide a starting point to be refined through an iterative proc-
ess. Further development of a more complete ontology for annotating uncertainty is
one of the XG’s recommendations.
3.2 The URW3-XG Uncertainty Use Cases
Building on the work started during the Use Case Challenge, the URW3-XG de-
veloped 16 use cases to identify how the representation of uncertainty would help to
address issues in Web reasoning that cannot be properly represented with current
deterministic approaches. The use cases were developed for the most part using a
common template. Occurrences of uncertainty in the use case descriptions were an-
notated with information from the Uncertainty Ontology. One use case, entitled Buy-
ing Speakers, is shown in the Appendix.
The analysis of the use cases indicated that a representation of uncertainty would
be required to represent both uncertainty inherent in the data and uncertainty related
to the processing of data and the delivery of processing results. This will be discussed
further in section 4.
4 Key Conclusions and Recommendations
In automated data processing, we often face situations where Boolean truth-values are
unknown, unknowable, or inapplicable. This is true for a wide variety of data and
information processing applications, and therefore it should be no surprise that the
methodologies considered by the XG are popular in contexts other than the Web. The
use cases considered by the XG concerned reasoning challenges specific to the Web,
such as discovery of Web Services, order processing via Web Services, and the like.
The XG’s work confirmed the hypothesis that a unified model for uncertainty annota-
tion of Web resources would provide value for deductive engines, and this could be
further facilitated by an ontology characterizing the types and sources of uncertainty.
The work with the Uncertainty Ontology suggested that a finer grained extension
may be useful. Such an extension could provide a means to visualize a possible evolu-
tion of an upper level Uncertainty Ontology. The conclusions go on to focus espe-
cially on finer classification of Machine Agents and uncertainty caused by lack of
knowledge of a machine agent.
With respect to the kinds of uncertainty observed in the use cases, it was noted that
uncertainty may be an inherent part of the data or may be related to the processing
that produces results. In the first case, the standardization should provide a single
syntactical system so that people can identify and process this information quickly.
For example, one may want to be able to communicate information that Study X
shows that people with property Y have an Z% increased likelihood of disease W.
The ability to communicate such information using a common interchange syntax
could be extremely useful in a number of web-based applications. Such characteriza-
tions of data uncertainty may require something like uncertain extensions to OWL
(i.e., probabilistic, fuzzy, belief function, random set, rough set, and hybrid uncertain
extensions to OWL).
The second kind of uncertainty involves reasoning on the part of the tools used to
access and share web information. For example, if a web service uses uncertainty
reasoning to find and rank hotel rooms, the need would be to represent meta-
information about the reasoning models and assumptions. This could facilitate the
development of trust models, or allow the identification of compatible web services to
increase the likelihood that the results are consistent with the user preferences. Here
the representation would include determining how to represent the meta-information
on processing and deciding how detailed the meta-information would need to be and
where it would reside.
The deliberations and conclusions of the URW3-XG led to the following recom-
mendations:
• A principled means for expressing uncertainty will increase the usefulness of
Web-based information and a standard way of representing that information
should be developed.
• Different use cases appear to lend themselves to different uncertainty formal-
isms, indicating the standard representation should provide a means to un-
ambiguously identify the formalism providing the context for assigning other
uncertainty characteristics and values.
• Different uncertainty formalisms assign values to properties specifically re-
lated to the underlying meaning and processing of these values, and the rep-
resentation should support defining different standard properties for each
formalism without requiring changes to the representation itself.
• Sample representations for the most useful formalisms should be developed
both as exemplars and for their immediate use, with the ability to expand be-
yond the initial exemplars as circumstances might indicate to be prudent.
• Given that uncertainty can be present anywhere, the representation should
support associating uncertainty with any property or value expressible across
the Web.
An open question that remains when considering a standard uncertainty representa-
tion is whether existing languages (e.g. OWL, RDFS, RIF) are sufficiently expressive
to support the necessary annotations. If so, the development of such annotations might
merely require work on a more complete uncertainty ontology and possibly rules;
otherwise, the expressiveness of existing languages might need to be extended. As an
example of the latter, it might be advisable to develop a probabilistic extension to
OWL or a Fuzzy-OWL format or profiles associated with the type of uncertainty to be
represented. Further work is required to investigate the adequacy of the existing lan-
guages against the compiled use cases.
The means to associate the uncertainty representation with its subject was also be-
yond the scope of the URW3-XG. The conclusions noted that a mechanism similar to
that specified under Semantic Annotations for WSDL and XML Schema (SAWSDL)
[8].
5 Considerations for Next Steps
The work of the URW3-XG provided an important beginning for characterizing
the range of uncertainty that affects reasoning on the scale of the World Wide Web,
and the issues to be considered in designing a standard representation of that uncer-
tainty. However, the work to date likely falls short of what would be needed to char-
ter an effort to develop that representation. Additional work needed includes the fol-
lowing:
• The conclusions note the value of the Uncertainty Ontology developed
thus far, but it also notes the value of further work to extend the ontology;
• A representation is needed for uncertainty models but it was beyond the
scope of the current effort to decide whether extensions to existing Se-
mantic Web languages (e.g. OWL, RDFS, RIF) will be sufficient or
whether new representation standards will be needed;
• As SAWSDL provides a mechanism to associate semantics with certain
Web resources, it might also provide a useful model for associating a
standard representation of uncertainty information, but the feasibility of
such use has not been adequately considered.
The question to be answered is what future venue should be pursued to tackle these
issues and others that may become evident. There are several nonexclusive possibili-
ties, among which are
• Continue with the URSW workshop series, using it as a forum to discuss
advances in theory and practice;
• Approach other communities, such as those dealing with health care and
life sciences, and form a wider collaboration to both continue the research
aspects and to provide concrete problems against which to develop solu-
tions;
• Develop a charter for and establish a new XG to work the items recom-
mended by the URW3-XG;
• Investigate funding opportunities to formalize a dedicated effort to pursue
the issues and develop implementable solutions and tools in a reasonable
time frame.
This paper provides a summary of work to date. As the discussions of the attendees
at the 2nd URSW workshop provided the basis for the URW3-XG work, so the 4th
URSW workshop provides the opportunity to discuss these and possibly other options
and assess the consensus of the community for its next steps.
References
[1] Laskey, K. J.; Laskey, K. B.; and Costa, P. C. G. (2006) A Proposal for a W3C XG on
Uncertainty Reasoning for the World Wide Web. Proceedings of the second workshop on
Uncertainty Reasoning for the Semantic Web (URSW 2006), held at the Fifth Interna-
tional Semantic Web Conference (ISWC 2006), 5-9 November 2006, Athens, Georgia,
USA. Available at
http://c4i.gmu.edu/ursw/2006/files/papers/URSW06_P5_LaskeyCostaLaskey.pdf.
[2] Second workshop on Uncertainty Reasoning for the Semantic Web (URSW 2006), held
at the Fifth International Semantic Web Conference (ISWC 2006), 5 November 2006,
Athens, Georgia, USA. http://c4i.gmu.edu/ursw/2006/
[3] Laskey, K. J. (2006) Use Case Challenge. Proceedings of the second workshop on Uncer-
tainty Reasoning for the Semantic Web (URSW 2006), held at the Fifth International
Semantic Web Conference (ISWC 2006), 5-9 November 2006, Athens, Georgia, USA.
Available at http://c4i.gmu.edu/ursw/2006/files/talks/URSW06_UseCaseChallenge.pdf
[4] World Wide Web Consortium (W3C), http://www.w3.org/
[5] W3C Incubator Activity > About XGs, http://www.w3.org/2005/Incubator/about.html
[6] Uncertainty Reasoning for the World Wide Web Incubator Group (URW3-XG),
http://www.w3.org/2005/Incubator/urw3/
[7] URW3-XG Final Report, 31 March 2008. Available at
http://www.w3.org/2005/Incubator/urw3/XGR-urw3-20080331/
[8] Semantic Annotations for WSDL and XML Schema. W3C Recommendation, 28 August
2007. Available at http://www.w3.org/2002/ws/sawsdl/spec/.
[9] Agarwal, S.; and Lamparter, S. (2005) sMART - A Semantic Matchmaking Portal for
Electronic Markets. Proceedings of the 7th International IEEE Conference on E-
Commerce Technology. Munich, Germany, 2005.
Appendix – Buying Speakers Use Case
1 - Purpose/Goals
Customer needs to make a decision on (1) whether to go to a store today or wait
until tomorrow to buy speakers, (2) which speakers to buy and (3) at which store.
Customer is interested in two speaker features: wattage and price. Customer has a
valuation formula that combines the likelihood of availability of speakers on a par-
ticular day in a particular store, as well as the two features. The features of wattage
and price are fuzzy. Optionally, Customer gets the formulas from CustomerService, a
Web based service that collects information about products, stores, statistics, evalua-
tions.
2 - Assumptions/Preconditions
• Customer either relies on the definitions provided by CustomerService or
is knowledgeable in both probability and fuzzy sets.
• Stores provide information to CustomerService. CustomerService keeps
information on both probabilistic models and fuzzy models.
• Customer has the capability of either obtaining or defining a combination
function for combining probabilistic information with fuzzy.
3 - Required Resources
• Data collected by CustomerService on the availability of items, which in
turn depends on restocking and rate of selling.
• Ontology of uncertainty that covers both probability and fuzziness.
4 - Successful End
Customer gets necessary information about the availability and types of speakers
from stores. This information is sufficient for customer to compute the required met-
ric.
5 - Failed End
Customer does not get necessary information and thus needs to go to multiple
stores, wasting in this way a lot of time.
6 - Main Scenario
1. Customer formulates query about availability of speakers in the stores within
some radius.
2. Customer sends the query to the CustomerService.
3. CustomerService replies with information about the availability of speakers.
CustomerService cannot say for sure whether a given type of speaker will be
available in a store tomorrow or not. It all depends on delivery and rate of sell.
Thus CustomerService provides the customer only with probabilistic informa-
tion.
4. Since part of the query involves requests that cannot be answered in crisp terms
(vagueness), CustomerService annotates its replies with fuzzy numbers.
5. CustomerService uses the uncertainty annotated information to compute a met-
ric.
6. Customer uses the resulting values of the metric for particular stores and for
particular types of speaker to decide whether to buy speakers, what type and
which store.
7. Additional background information or references: This use case was inspired
by Agarwal and Lamparter [9].
8. General Issues and Relevance to Uncertainty:
1. There is known probability distribution on the availability of particular
speaker type in particular stores on a particular day in the future. Say there
are two stores (not too close to each other) and the probability that speak-
ers of type X will be available in stores A and B tomorrow are Pr(X,
A)=0.4 and Pr(X, B)=0.6. The probabilities for all types of speakers are
represented in the same way.
• The uncertainty annotation process (UncAnn) was used.
• The agent issues a query (a sentence): Sentence. It is a complex sen-
tence consisting of three basic sentences. One related to the availabil-
ity, one to the wattage and one to the price of speakers.
• Each of these sub-sentences will have uncertainty Uncertainty associ-
ated with it.
• The uncertainty type related to the availability of particular speaker
type in the stores is of type UncAnn - UncertaintyType: Empirical.
• The uncertainty nature is UncAnn - UncertaintyNature: Aleatory.
• The uncertainty model is UncAnn - UncertaintyModel: Probability.
2. The customer has (or obtains from CustomerService) definitions of fea-
tures of wattage and price in terms of fuzzy membership functions. For
wattage, Customer has three such functions: weak, medium and strong.
These are of "trapezoid shaped" membership functions. Similarly, for price
Customer has three such membership functions: cheap, reasonable and ex-
pensive.
• The uncertainty type related to the features of wattage and price is of
type UncAnn - UncertaintyType: Vagueness.
• The uncertainty nature is UncAnn - UncertaintyNature: Epistemic.
• The uncertainty model is UncAnn - UncertaintyModel: FuzzySets.
3. The valuation has three possible outcomes, all are expressed as fuzzy
membership functions: bad, fair, good and super.
4. Customer knows the probabilistic information, since the probabilities are
provided by CustomerService. CustomerService uses the Uncertainty On-
tology for this purpose.
5. Customer has (or selects) fuzzy definitions of the features of wattage and
price. Again, the six membership functions that define these features are
annotated with the Uncertainty Ontology.
6. Customer has (or uses one suggested by CustomerService) a combination
function that computes the decision, d, based upon those types of input.
This function can be modified by each customer, however the stores need
to give input to CustomerService - the probabilities and the (crisp) values
of wattage and price for their products. The features are fuzzified by the
customer's client software. Customer uses the Uncertainty Ontology to an-
notate the fuzziness of particular preferences.