=Paper=
{{Paper
|id=Vol-2154/paper1
|storemode=property
|title=New Concepts for Trust Propagation in Knowledge Processing Systems
|pdfUrl=https://ceur-ws.org/Vol-2154/paper1.pdf
|volume=Vol-2154
|authors=Markus Jager,Josef Kung
|dblpUrl=https://dblp.org/rec/conf/atal/0003K18
}}
==New Concepts for Trust Propagation in Knowledge Processing Systems==
<pdf width="1500px">https://ceur-ws.org/Vol-2154/paper1.pdf</pdf>
<pre>
         New Concepts for Trust Propagation
          in Knowledge Processing Systems

                         Markus Jäger and Josef Küng

         Institute for Application Oriented Knowledge Processing (FAW)
                Faculty of Engineering and Natural Sciences (TNF)
                 Johannes Kepler University Linz (JKU), Austria
                      {markus.jaeger, josef.kueng}@jku.at



      Abstract. Everybody has a sense of trusting people or institutions, but
      how is trust defined? It always depends on the specific field of research
      and application and is different most of the time, which makes it hard
      to answer this question in general at a computational level.
      Thinking on knowledge processing systems we have this question twice.
      How can we define and calculate trust values for the input data and,
      much more challenging, what is the trust value of the output? Meeting
      this challenge we first investigate appropriate ways of defining trust.
      Within this paper we consider three different existing trust models and
      a self developed one.
      Then we show ways, how knowledge processing systems can handle these
      trust values and propagate them through a network of processing steps
      in a way that the final results are representative. Therefore we show the
      propagation of trust with the three existing trust models and with a
      recently self developed approach, where also precision- and importance-
      values are considered. With these models, we can give insights to the
      topic of defining and propagating trust in knowledge processing systems.

      Keywords: Trust; Propagation; Knowledge Processing Systems; Trust
      Metrics; Trust Models; Precision; Fusion; Knowledge; Provenance;


1   Introduction
The main subject of our research is the topic of how knowledge processing sys-
tems can work with trust values. While going deeper into this research field,
several questions arise.
   In our work, we try to figure out, how trust can be defined and mea-
sured and in particular the question of how knowledge processing systems
can deal with trust? Furthermore we investigate the topic of how several
trust values can be combined (in general and in knowledge process-
ing) and how can trust values be propagated through several steps of
a knowledge processing system?
   We try to give answers to these questions by investigating possibilities of
trust measurement, combination and propagation and try to propose a sound
and all-encompassing way of handling these topics.
    The rest of this paper is structured as following: section 2 defines common
terminologies and shows related work in our field of research.
    As the term ”trust” has a significant high importance in our research, we
dedicated an extra section for defining trust in knowledge processing - section 3.
In this section, we investigate different models of measuring or determining trust.
We briefly introduce our recently developed and already published approach,
where trust- and precision values are handled, processed and propagated by
taking into account several importances in section 4.
    Section 5 examines the question, how trust can be propagated through knowl-
edge processing systems. Here we cover the different trust measuring models,
which were introduced in section 3 and 4 and how trust can be propagated in
these different models. Section 6 shows the application of the presented trust
propagation models in a scenario. We close this paper with section 7 by giving
a summary of our work and an outlook for further research.


2     Related Work
In this section we provide some insights into important terms which are relevant
to our work.Furthermore we discuss the fusion of sensor precision values.

2.1   Trust
The meaning of the term ”Trust” always depends on the specific environment
and field of research and application. In a recent publication about trust, we
state: ”The question of ’How can we trust anything/anybody?’ is discussed since
the beginning of mankind, but what does this topic mean in context to today’s
technology age and especially for the information technology?” [11].
    The three main types of applicable trust by Rousseau et al. [18] are (1)
trusting beliefs, (2) trusting intentions, and (3) trusting behaviours, where these
three types are connected to each other.
    Another point of view is the similarity of trusting people and trusting tech-
nology, especially information technology, where the main difference is within
the application of trust in the specific area [16].
    Also a very interesting publication about trust in information sources is from
Hertzum et al. [9]. They compare the concept of trust between people and virtual
agents, based on two empirical studies. Some relational aspects concerning trust
in the industrial marketing and management sector can be found in ”Concerning
trust and information” from Denize et al. [8].

2.2   Provenance
When we come into trust concerning trusting in data and trusting the sources
of data, the term ”Data Provenance” comes into account. It means the origin
and complete processing history of any kind of data. A quite good introduction
and overview can be found in [2] and [3].
    Several problems concerning data provenance are covered in [5].
    Recent research work on provenance can be found in the following literature:
”Trust Evaluation Scheme of Web Data Based on Provenance in Social Semantic
Web Environments” [19] and ”Transparently tracking provenance information in
distributed data systems” [4].
    ”Research of Data Resource Description Method oriented Provenance” [20]
and ”A semantic Foundation for Provenance Management” [17] provide more
theoretical and conceptual foundations for the usage of provenance.

2.3   Risk
Risk in general addresses the potential of losing something with a special personal
value. It is also seen as an intentional interaction with uncertainty, where the
outcome is hard to predict.
    Rousseau et al. [18] say that ”Risk is the perceived probability of loss, as in-
terpreted by a decision maker [...]. The path-dependent connection between trust
and risk taking arises from a reciprocal relationship: risk creates an opportunity
for trust, which leads to risk taking.”

2.4   Precision & Multi Data Sensor Fusion
The link on related work of fusion precision values in sensor networks can be
found in our recent publication ”Focussing on Precision- and Trust-Propagation
in Knowledge Processing Systems” [12]. The concluding findings are, that sensor
fusion is motivated to avoid problems which come from the use of single sensors
(e.g. sensor deprivation, limited spatial and temporal coverage, imprecision and
uncertainty). Fusion processes in the sensor domain are often categorized in three
levels: (1) raw data fusion (low level), (2) feature fusion (medium level), and (3)
decision fusion (high level).


3     Defining Trust in Knowledge Processing
The main question in our work is, how knowledge processing systems can handle
and work with trust. In this context, the first step is to find a definition of trust,
which is suitable for this scientific domain. In this section we investigate different
applicable models of measuring or determining trust. Therefore we describe three
existing ways: the binary trust model, the probabilistic trust model and the
opinion-space trust model.

3.1   Trust, Certainty and Precision in Knowledge Processing
To the best knowledge of the authors, there is no related work dealing with this
topic directly – neither for processing trust and certainty, nor for the aggregation
of trust, (un)certainty or precision. A good approach for measuring trust is given
in [7]. Recent research on modeling uncertainty is given in [15] and in [6].
   The propagation/fusion of (sensor) precision values has been evaluated in
recent publications, as stated in section 2.4. Some of the investigated propa-
gation/fusion methods of sensor precision values seem quite promising also for
the application on trust. Nevertheless we focus on models that cover only trust
values, as presented in the following sections.
   Another approach is the refactoring of a trust value from a given precision,
which will be covered in our future work.


3.2   Binary Trust Model

One of the easiest ways to represent trust values in an understandable and
applicable way is the usage of a binary trust model. In this model the possible
trust values can either be 0 or 1. Therefore the only differentiation is to fully
trust a subject (trust = 1) or not (trust = 0).
     In our opinion, the model is very hard to apply in a real world domain because
it is very hard to get a trust value of 1 anyway. The definition of possible states
in the binary trust model, can be seen in formula 1.

                                   T =0 ∨ 1                                    (1)
Formula 1: Boundaries for the scope of trust in the binary trust model.


3.3   Probabilistic Trust Model

A very application oriented and realistic way of representing trust is the usage of
a probabilistic trust model. In this model the possible trust values range from 0
to 1 and can, for example be seen as a type of percentage view. The value of not
trusting a subject (trust = 0) and fully trusting a subject (trust = 1 or 100%)
is like in the binary trust model, but here the grading can be more precise, as
there are (theoretically) infinite states of trust between 0 and 1 (or 100%). The
range of possible states of trust in the probabilistic trust model can be seen in
formula 2.

                                    0≤T ≤1                                     (2)
Formula 2: Boundaries for the scope of trust in the probabilistic trust model.


3.4   Opinion-Space Trust Model

A very well developed model for measuring trust values is the opinion-space
model by Audun Jøsang and S.J. Knapskog from 1998: ”A Metric for Trusted
Systems” [14].
    They introduce an evidence space and an opinion space which are two equiv-
alent models for representing human beliefs, which can be summarized as trust
in their model. We focus on the opinion-space trust model, which consists of
the values belief b, disbelief d, and uncertainty u. These three values represent
the trust, which is determined. The sum of the three values is always 1, so the
interpretation of trust has to be clarified for the current domain of application.
The opinion-space trust model can be seen in figure 1.




         Fig. 1: The opinion-space trust model / opinion triangle from [14].




                         b + d + u = 1, {b, d, u} ∈ [0, 1]3                    (3)
Formula 3: Boundaries for the scope of trust in the opinion-space model.


4     Specification of our Approach

In our recent research, we designed a convenient approach for propagating trust
and precision values through multi-step knowledge processing systems, where
also a factor importance was introduced and considered in the calculation. Our
approach was evaluated and published in several conferences before e.g. in [10,
12, 13] and tested on some artificial and real world scenarios.


4.1   Principle Idea

Following, we describe the idea of our approach. The main components are:

 – any Source (S), which provides data; there can be multiple sources.
 – any Data (D), which is provided by one source; for our model, every source
   usually provides one or more data (elements).
 – any Knowledge Processing System (KPS), which processes data from one or
   more sources; each KPS itself produces new data as output.

The main values in our approach are:

 – Trust value (T) of source (S), which defines how trustable the source is. The
   system has to be seen as a whole environment, hence the trust level for one
   source should always be the same.
 – Precision value (P) of data (D), which describes how precise, reliable, confi-
   dent or steady the provided data is.
 – Importance value (I) of one input data (D), decided by the current knowledge
   processing system (KPS) for the current step of computation.
      Our approach is sketched in figure 2.




                     Fig. 2: Graphical description of our approach.




4.2     Scopes & Calculation
In our approach, the scopes of possible values are fixed as following:
 – Trust T of source S, for each S, has to be greater than 0 and less or equal
   than 1, where each value of T for each S has to be the same (if used multiple
   times) - a higher value represents higher trust:
                                        0<T ≤1                                (4)
 – Precision P of data D, for each D, has to be greater than 0 and less or equal
   than 1, where each value of P for each D has to be the same (if used multiple
   times) - a higher value represents higher precision:
                                        0<P ≤1                                (5)
 – Importance I of data D, decided by the knowledge processing system:
    • 0.5 for values which are not very important
    • 1.0 for regular values, where no special impact on importance is given
    • 1.5 for very important values, concerning the current data processing

                                     I = 0.5 | 1 | 1.5                        (6)
5     Propagation of Trust in Knowledge Processing Systems

In this section we compare different methods of fusing and propagating trust
values in knowledge processing systems.


5.1   Propagating Trust in the Binary Trust Model

As simple the determination of trust in the binary trust model is, as simple is
the propagation of of these values. The only way of a reasonable propagation
of binary trust values, is to take the results of the logical conjunction over all
values. This means, if only one input value is 0, the output is 0 as well. This
can be explained, because it is not justifiable to trust an output of a processing
where one input value wasn’t trusted.
                                             n
                                             ^
                                  Tnew =         Ti                           (7)
                                           i=1

Formula 7: Calculating Tnew over all T1-n in the binary trust model.


5.2   Propagating Trust in the Probabilistic Trust Model

In the probabilistic trust model, several trust values can be fused by multiplying
them, presupposed the providing sources are independent.
                                           n
                                           Y
                                  Tnew =         Ti                           (8)
                                           i=1

Formula 8: Calculating Tnew over all T1-n in the probabilistic trust model.


5.3   Propagating Trust in the Opinion-Space Trust Model

In this model, a reasonable way of propagation is to consider all input values in
the opinion-space trust model. This means that the values of belief, disbelief and
uncertainty have to be propagated through the system in a way that the output
can again be identified with separate belief, disbelief and uncertainty values
again. We propose a very clear way of propagating the three values through a
knowledge processing system: using the arithmetic mean for all three separated
values to gain the new output values.
                                                 n
                                             1X
                        bnew |dnew |unew =         (bi |di |ui )              (9)
                                             n i=1

Formula 9: Calculating bdunew over all bdu1−n in the opinion-space trust model.
5.4   Propagating Trust in the Arithmetic Mean Trust Model
This model of propagation is based on our approach, which is presented in sec-
tion 4. It calculates the propagated trust value based on the input values from
the sources, weighted with different important values, which are decided by the
knowledge processing system itself individually for every step of processing.
                                          n
                                      1X
                             Tnew =         (Ti × Ii )                      (10)
                                      n i=1
Formula 10: Calculating Tnew over all T1-n related to I1-n in the arithmetic mean
trust model.

6     Scenarios
6.1   Artificial Scenario
The scenario is a complex environment and consists of several knowledge process-
ing systems, where multiple processing steps are taken and where some output
values are used as new input values. We introduce six sources (S1 to S6 ) with
different trust values (T1 to T6 ), each providing one or two data with different
precision values for multiple knowledge processing systems (KPSA to KPSE ),
which weight the different importances. All following calculated propagation
models consider the trust values. The usage of the precision- and importance
values is only considered in the arithmetic mean trust model of our approach.




                            Fig. 3: Artificial Scenario.
Propagation with the binary trust model Table 1 shows the initial trust
values in the scenario for propagation with the binary trust model.

               Table 1: Artificial Scenario, Binary Trust Model
S1 : T1 =1    S2 : T2 =1   S3 : T3 =1      S4 : T4 =1   S5 : T5 =1     S6 : T6 =0
                       TA = T1 ∧ T2 ∧ T3 = 1 ∧ 1 ∧ 1 = 1                     (11)
                       TB = T3 ∧ T4 ∧ T5 = 1 ∧ 1 ∧ 1 = 1                     (12)
                          TC = T5 ∧ T6 = 1 ∧ 0 = 0                           (13)
Applying formula 7 in the calc. 11-13 we get the results for TA , TB and TC .

                      TD = T1 ∧ TA ∧ TB = 1 ∧ 1 ∧ 1 = 1                      (14)
                         TE = TD ∧ TC = 1 ∧ 0 = 0                            (15)
By applying formula 7 in the calculations 14 and 15 from the propagation of
trust in the binary trust model, the result is Tnew =0. Here we notice again the
problem of the binary trust model, where only one untrusted input (T6 ) ruins
the trust in the overall system.


Propagation with the probabilistic trust model Table 2 shows the initial
trust values in the scenario for propagation with the binary trust model.

             Table 2: Artificial Scenario, Probability Trust Model

S1 : T1 =0.95 S2 : T2 =0.87 S3 : T3 =0.75 S4 : T4 =0.66 S5 : T5 =0.50 S6 : T6 =1.00
               TA = T1 × T2 × T3 = 0.95 × 0.87 × 0.75 = 0.6198               (16)
               TB = T3 × T4 × T5 = 0.75 × 0.66 × 0.50 = 0.2475               (17)
                     TC = T5 × T6 = 0.50 × 1.00 = 0.50                       (18)
Applying formula 8 in the calc. 16-18 we get the results for TA , TB and TC .

             TD = T1 × TA × TB = 0.95 × 0.619 × 0.247 = 0.1457               (19)
                 TE = TD × TC = 0.14575 × 0.50 = 0.07287                     (20)
By applying formula 8 in the calculations 19 and 20 from the propagation of
trust in the probabilistic trust model, the result is Tnew =0.07287 or 7.287%.
Again, we recognize the problem of inputs with ”lower” trust values (T4 and
T5 ) which lead to very low overall trusting outcome, especially in multi step
knowledge processing systems.


Propagation with the opinion-space trust model Table 3 shows the initial
values for belief, disbelief and uncertainty in the scenario for propagation with
the opinion-space trust model.
            Table 3: Artificial Scenario, Opinion-Space Trust Model
            S1 :   b=1                   d=0                u=0
            S2 :   b=0                   d=1                u=0
            S3 :   b=0                   d=0                u=1
            S4 :   b = 0.3               d = 0.4            u = 0.3
            S5 :   b = 0.5               d = 0.5            u=0
            S6 :   b = 0.8               d=0                u = 0.2

                           bA = (1 + 0 + 0)/3 = 0.33                         (21)
                           dA = (0 + 1 + 0)/3 = 0.33                         (22)
                           uA = (0 + 0 + 1)/3 = 0.33                         (23)
                         bB = (0 + 0.3 + 0.5)/3 = 0.266                      (24)
                          dB = (0 + 0.4 + 0.5)/3 = 0.3                       (25)
                          uB = (1 + 0.3 + 0)/3 = 0.433                       (26)
                           bC = (0.5 + 0.8)/2 = 0.65                         (27)
                            dC = (0 + 0.5)/2 = 0.25                          (28)
                             uC = (0 + 0.2)/2 = 0.1                          (29)
Applying formula 9 in the calc. 21-29 we get the results for TA , TB and TC .

                       bD = (1 + 0.33 + 0.266)/3 = 0.533                     (30)
                        dD = (0 + 0.33 + 0.3)/3 = 0.211                      (31)
                       uD = (0 + 0.33 + 0.433)/3 = 0.255                     (32)
                        bE = (0.533 + 0.65)/2 = 0.59166                      (33)
                        dE = (0.211 + 0.25)/2 = 0.23055                      (34)
                         uE = (0.255 + 0.1)/2 = 0.177                        (35)
By applying formula 9 in the calculations 30-35 from the propagation of trust in
the opinion-space trust model, the result is bnew =0.59166, dnew =0.23055, and
dnew =0.177. A very mean and realistic result in particular, when we observe the
well distributed input values. In our opinion, the outcome is very representative.

Propagation with the arithmetic mean trust model Table 4 shows the
initial trust values in the scenario for propagation with the weighted arithmetic
mean trust model.

          Table 4: Artificial Scenario, Arithmetic Mean Trust Model
           S1 : T1 =1.0          D11 : P11 =0.9     KPSA : IA1 =0.5
           S2 : T2 =0.4          D2 : P2 =0.3       KPSA : IA2 =1.0
           S3 : T3 =0.8          D31 : P31 =0.8     KPSA : IA3 =1.5
                                 D32 : P32 =0.5     KPSB : IB1 =1.0
           S4 : T4 =0.2          D4 : P4 =0.2       KPSB : IB2 =1.5
           S5 : T5 =0.5          D51 : P51 =1.0     KPSB : IB3 =0.5
                                 D52 : P52 =0.7     KPSP : IP1 =1.0
           S6 : T6 =0.9          D6 : P6 =1.0       KPSC : IC2 =1.0

The calculation steps for this scenario can be found in [10]. The results are
Tnew =0.6625 and Pnew =0.6875, which are very promising and realistic.
6.2   Agricultural Scenario
Because the scenario from the last subsection is completely artificial, we applied
our approach in the course of a project in the agricultural domain funded by the
European Union, to have comparable computations.
    Therefore, we are now referring to the DPM (Disease Pressure Model, used
in the Project CLAFIS [1]) for calculating an accurate daily risk value. This
shows how certain a specific disease outbreak for a specific agricultural field can
be. The DPM uses input values from a FMIS (farm management information
system), which stores information such as this year’s and last year’s crop as
well as the used tillage method. The needed weather data comes from several
weather stations, which gathers information, such as temperature, relative hu-
midity, amount of rainfall, and wind speed is gathered. This was a very practical
scenario for the application of our approach, where several different trustable
sources where used as input. This application of our approach (the weighted
arithmetic mean trust propagation model) was evaluated by experts from the
agricultural domain and showed very good results. The detailed calculation and
evaluation can be found in [10].


7     Summary & Outlook
We addressed the question of how to determine trust values in general and how
knowledge processing systems can handle these values in particular. Addition-
ally, we investigated several models for the definition of trust and presented our
recently developed approach. Furthermore we showed ways to propagate trust
values from the investigated trust models through multi step knowledge process-
ing systems and also a new way of trust propagation from our approach. We gave
example calculations on a scenario with all propagation models and compared
the results.
    Our approach was evaluated in a real world scenario in the frame of a co-
operation project with colleagues from practice. An evaluation was done by
interviewing experts from this scientific and application domain and the results
were encouraging. Our aim is to develop a complete model for incorporating
trust, precision and importance values into knowledge processing systems. This
approach then could be applied to all other processing systems as well. Such a
system, which could be used in a broad variety of applications, definitively would
be very useful in practice.

Acknowledgment
The research leading to these results, has received funding partly from the Eu-
ropean Union Seventh Framework Programme (FP7/2007-2013) under grant
agreement no.604659 in the project CLAFIS and partly from the federal county
of Upper Austria.
References
 1. CLAFIS: Crop, livestock and forests integrated system for intelligent automation,
    2013-2016. EU Seventh Framework Programme NMP.2013.3.0-2.
 2. Peter Buneman and Susan B Davidson. Data provenance–the foundation of data
    quality, 2010.
 3. Peter Buneman, Sanjeev Khanna, and Wang-Chiew Tan. Data provenance: Some
    basic issues. In Sanjiv Kapoor and Sanjiva Prasad, editors, FST TCS 2000, volume
    1974 of Lecture Notes in Computer Science. Springer Berlin Heidelberg, 2000.
 4. P.C. Castro, M. Pistoia, and J. Ponzo. Transparently tracking provenance infor-
    mation in distributed data systems, August 7 2014. US Patent App. 13/761,916.
 5. Wang chiew Tan. Research problems in data provenance. IEEE Data Engineering
    Bulletin, 27:45–52, 2004.
 6. Gianpaolo Cugola, Alessandro Margara, Matteo Matteucci, and Giordano Tambur-
    relli. Introducing uncertainty in complex event processing: model, implementation,
    and validation. Computing, 97(2):103–144, 2015.
 7. Chenyun Dai, Dan Lin, Elisa Bertino, and Murat Kantarcioglu. An approach to
    evaluate data trustworthiness based on data provenance. In Proceedings of the 5th
    VLDB Workshop on Secure Data Management, SDM ’08. Springer-Verlag, 2008.
 8. Sara Denize and Louise Young. Concerning trust and information. Industrial
    Marketing Management, 2007.
 9. Morten Hertzum, Hans H.K Andersen, Verner Andersen, and Camilla B Hansen.
    Trust in information sources: seeking information from people, documents, and
    virtual agents. Interacting with Computers, 14(5):575 – 599, 2002.
10. Markus Jäger and Josef Küng. Introducing the factor importance to trust of
    sources and certainty of data in knowledge processing systems - a new approach
    for incorporation and processing. In Proceedings of the 50th HICSS, 2017.
11. Markus Jäger, Stefan Nadschläger, and Nhan Trong Phan. Towards the trustwor-
    thiness of data, information, knowledge and KPS in smart homes, IDIMT 2015.
12. Markus Jäger, Jussi Nikander, Stefan Nadschläger, Van Quoc Phuong Huynh, and
    Josef Küng. Focussing on precision- and trust-propagation in knowledge processing
    systems. 4th Future Data and Security Engineering. Springer, 2017.
13. Markus Jäger, Trong Nhan Phan, Christian Huber, and Josef Küng. Incorporating
    trust, certainty and importance of information into knowledge processing systems
    - an approach. 3rd Future Data and Security Engineering. Springer, 2016.
14. Audun Jøsang and S.J. Knapskog. A metric for trusted systems. 1998.
15. Alexander Karlsson, Björn Hammarfelt, H. Joe Steinhauer, Göran Falkman, Nas-
    rine Olson, Gustaf Nelhans, and Jan Nolin. Modeling uncertainty in bibliometrics
    and information retrieval: an information fusion approach. Scientometrics, 2015.
16. D. Harrison McKnight. Trust in Information Technology. The Blackwell Encyclo-
    pedia of Management: Operations management. Blackwell Pub., 2005.
17. Sudha Ram and Jun Liu. A semantic foundation for provenance management.
    Journal on Data Semantics, 1(1):11–17, 2012.
18. Denise Rousseau, Sim Sitkin, Ronald Burt, and Colin Camerer. Not so different
    after all: A cross-discipline view of trust. Academy of Management Review, 1998.
19. Sangwon Yoon, Kitae Choi, Jaeyeol Park, Jongtae Lim, Kyoungsoo Bok, and Jae-
    soo Yoo. Trust evaluation scheme of web data based on provenance in social
    semantic web environments. Journal of KIISE, 43(1):106–118, 2016.
20. Yan-peng Zhao, Chao-fan Dai, and Xiao-yu Zhang. Research of data resource
    description method oriented provenance. In Proceedings of the 22nd Int. Conf. on
    Industrial Engineering and Engineering Management 2015. Springer, 2016.

</pre>