<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <title-group>
        <article-title>Contextual Binding and Deception Detection</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <string-name>Jim Q. Chen</string-name>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Ph.D.</string-name>
        </contrib>
        <contrib contrib-type="author">
          <string-name>National Defense University</string-name>
        </contrib>
        <contrib contrib-type="author">
          <string-name>U.S.A.</string-name>
        </contrib>
      </contrib-group>
      <abstract>
        <p>Deception is frequently used in cyber attacks. Detecting deception is always a challenge, as witnessed in attacks in social media and other online environments. Contexts can help to identify deception. Unfortunately, there is not much literature available in this aspect. This paper explores the unique properties of contextual binding. It examines roles that it plays. It also proposes a novel approach in detecting deception utilizing contextual binding in the cyber domain.</p>
      </abstract>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>Introduction</title>
      <p>
        A context is defined in Webster dictionary as “the
interrelated conditions in which something exists or occurs” or as
“the parts of a discourse that surround a word or passage
and can throw light on its meaning”.
(http://www.merriamwebster.com/dictionary/context) It is also defined in
        <xref ref-type="bibr" rid="ref2">Brézillon (1999)</xref>
        and
        <xref ref-type="bibr" rid="ref3">Brézillon (2002)</xref>
        as “a collection of
relevant conditions and surrounding influences that make a
situation unique and comprehensible”. In this sense, a
context helps to disambiguate meaning and find out the actual
referent. Hence, it is essential in data mining and big data
analytics. There are various approaches in context analysis.
        <xref ref-type="bibr" rid="ref4">Brézillon (2003)</xref>
        uses contextual graphs to address the
dynamic of context.
        <xref ref-type="bibr" rid="ref9">Gaifman (2008)</xref>
        uses syntactically
represented context operators in the analysis of contextuality.
        <xref ref-type="bibr" rid="ref10">Grossi, Dignum, and Meyer (2006</xref>
        ) propose a notion of
contextual terminology to “reason within contexts
(intracontextual reasoning)” and to “reason also about contexts
and their interplay (inter-contextual reasoning)”.
        <xref ref-type="bibr" rid="ref13">Rebuschi
and Lihoreau (2009)</xref>
        address “the connections between
knowledge and context” with the contextual epistemic
logic.
        <xref ref-type="bibr" rid="ref12">McCarthy (1993)</xref>
        discusses “formalizing contexts as
first class objects”. From Linguistic perspective, Key
(1989) mentions contextual operators, which, in his term,
“are lexical items or grammatical constructions whose
semantic value consists, at least in part, of instructions to find
in, or impute to, the context a certain kind of information
structure and to locate the information presented by the
Copyright held by the author. All rights reserved.
sentence within that information structure in a specified
way”. All these researches scrutinize contexts from varied
perspectives. They all show the significance of contexts in
nailing down meaning or interpretation. In the same spirit,
this paper proposes the Contextual Binding Conditions and
the Detection Condition utilizing contextual operators, on
the basis of linguistic samples. These conditions are then
applied to both language disambiguation and deception
detection in the cyber domain. The success of the
application confirms the validity of these conditions.
      </p>
    </sec>
    <sec id="sec-2">
      <title>Challenges</title>
      <p>As discussed previously, contexts are used in linguistic
analysis, in building intelligent agents, and in other
artificial intelligence fields. However, no formal methods that
employ contexts or contextual operators have ever been
used in detecting deception in the cyber domain. This
paper intends to explore this possibility.</p>
      <p>Before moving on, let us be aware of two types of
deception.</p>
      <p>
        <xref ref-type="bibr" rid="ref7">Caddell (2004</xref>
        ) defines two types of deception. They are
fabrication and manipulation. He states, “If false
information is created and presented as true, this is fabrication.”
“Manipulation, on the other hand, is the use of information
which is technically true, but is being presented out of
context in order to create a false implication.”
        <xref ref-type="bibr" rid="ref1">Almeshekah and
Spafford (2014)</xref>
        also state: “Deception always involves
two basic steps, hiding the real and showing the false”.
      </p>
      <p>
        How can deception, specifically fabrication and
manipulation, be detected in the cyber domain? This is one of the
challenges we are facing. Even though there is not much
literature available in this aspect, researches in the similar
fields may be looked at.
        <xref ref-type="bibr" rid="ref5">Burgoon, Blair, Qin, and
Nunamaker (2003</xref>
        ) use decision trees in detecting
deception within online chat messages. They utilize “16
linguistic features that can be automated to return assessments of
the likely truthful or deceptiveness of a piece of text”.
They find out that “deceivers do utilize language
differently than truth tellers”. However, it has to be pointed out the
16 linguistic features are pre-defined so that deception
employing features other than these 16 features may not be
detected.
      </p>
      <p>
        <xref ref-type="bibr" rid="ref15">Zhou, Shi, and Zhang (2008</xref>
        ) develop the Statistical
Language Models (SLMs), which consider “all of the
words in a text as potential features without relying on the
extraction of a predefined set of cues to deception”. Word
dependencies are learned to “capture semantic
relationships and dependency relationships among words so as to
approximate the meaning of sentences, which can benefit
deception detection”. This method is better than the
method used in
        <xref ref-type="bibr" rid="ref5">Burgoon, Blair, Qin, and Nunamaker (2003</xref>
        ).
However, it is relatively time-consuming, as “SLMs
consider all possible n-grams as features and implicitly
represent the importance of those features according to their
contribution to the quality of language modelling”. In
addition, this approach does not address the detection of
fabrication or manipulation as it is not designed for that
purpose.
      </p>
      <p>
        To address all these challenges and to figure out a
holistic and dynamic solution,
        <xref ref-type="bibr" rid="ref8">Chen and Duvall (2014)</xref>
        propose
the Operational-Level Cybersecurity Strategy Formation
Framework, which consists of a Contextual Analysis
Component among other components in making strategic
decisions. This paper further explores the inner-workings of the
Contextual Analysis Component.
      </p>
    </sec>
    <sec id="sec-3">
      <title>Proposal</title>
      <p>A novel approach is proposed here in this section, where
the contextual binding relationship is explored and then
used in detecting abnormal behavior. Within a contextual
binding relationship, a contextual binding operator plays a
crucial role. It helps to set up the baseline in a context, in
which the Detection Condition can be applied.</p>
    </sec>
    <sec id="sec-4">
      <title>The Contextual Binding Conditions</title>
      <p>A contextual binding operator is deterministic in
disambiguation. Let us take a look at a linguistic example
demonstrated in the English sentence in (1) below.
(1) John likes to buy a booki and read iti within three
days.</p>
      <p>Any English speaker knows that the following
interpretation is acceptable: “John likes to buy a book and read the
book that he buys within three days.” Any English speaker
also knows that the interpretation below is not acceptable:
“John likes to buy a book and read the map within three
days.”</p>
      <p>The pronoun “it” in this sentence refers back to the noun
phrase “the book”, which precedes the pronoun in the same
sentence. To a certain extent, pieces of information
provided previously can serve as contextual components for
entities in the same sentence or following sentences. In this
particular case, the agent of the action is the noun phrase
“John” in the subject position; the patient of the action is
the noun phrase “a book” in the object position; and the
predicate, i.e. the action, is “purchasing and reading x,
which is a book in this case”. Now, the noun phrase “the
book” plays the role of a contextual operator, which
becomes available for the interpretation of the pronoun “it”,
as it satisfies the condition of being a singular non-human
entity, just like the pronoun “it”, in this particular context.
Based on the observation of the contextual relationship, it
may be claimed that the pronoun “it” is contextually bound
by the contextual operator of “what”, namely, “a book” in
this particular case. In this contextual binding case, if the
pronoun “it” refers to another entity, such as “a map”,
rather than the entity “a book”, the interpretation is
immediately recognized as being abnormal or regarded as being
unacceptable. This clearly reveals a contextual binding
relationship, which can be defined as follows:</p>
      <p>The Basic Contextual Binding Condition:
Assume X is an entity, and CO is a contextual operator.
(i)If X is directly related to CO in such a setting:
COi {……Xi……}
then Xi is contextually bound by COi.</p>
      <p>The entity Xi is directly related to COi iff COi provides a
context that the interpretation of Xi solely depends on.</p>
      <p>Now, the contextual relationship in (1) can be captured
in the following schematic configuration:</p>
      <p>CObook {……Xbook……}</p>
      <p>Applying the Basic Contextual Binding Condition, the
pronoun “it” is contextually bound by the contextual
operator “book”. This means that the pronoun “it” has to be
interpreted as “the book” if the Basic Contextual Binding
Condition is obeyed.</p>
      <p>Given COs = {agent, patient, activity, time, location,
environment, background, precedence, etc.}, the
Restrictive Contextual Binding Condition can also be defined.</p>
      <p>
        Before we define this condition, let us see how
        <xref ref-type="bibr" rid="ref12">McCarthy (1993)</xref>
        handles the time component. In discussing the
relations among contexts,
        <xref ref-type="bibr" rid="ref12">McCarthy (1993)</xref>
        examines the
specialize-time (t, c), which he refers to as “a context
related to c in which the time is specialized to have the value t”.
The axiom that he comes up with is as follows:
      </p>
      <p>C0: specialize-time (t, c1, c2) ˄ ist (p, c1) ⊃ ist (c2,
attime (t, p))
This axiom refers to two assertions. The first one is in
Context1, i.e. c1, the proposition p is true in the context of c1.
The second one is in Context2, i.e. c2, which is a subset of
the set c1, and which has the specialize-time t, the
proposition is true at time t. The second assertion is a subset of the
first assertion.</p>
      <p>From this perspective, the time component further
narrows down the interpretation of the entity with the time
aspect.</p>
      <p>Let us take a look at another linguistic example
demonstrated in the English sentence in (2) below.</p>
      <p>(2) Yesterday John bought a booki at the bookstore. He
enjoyed reading iti.</p>
      <p>Here, the pronoun “it” in the second sentence refers back
to the noun phrase “a book”, sitting in the object position
of the first sentence and serving as the patient of the action.
The temporal adverbial phrase “yesterday” refers to the
time component of the context. The locality adverbial
phrase “at the bookstore” refers to the locality component
of the context. Comparing the noun phrase “a book” in (1)
with the noun phrase “a book” in (2), one may notice that
the former refers to a general book while the latter refers to
a specific book, i.e. the book bought yesterday at the
bookstore. In this sense, the latter in (2) may be considered
as a subset of the former in (1).</p>
      <p>
        Following
        <xref ref-type="bibr" rid="ref12">McCarthy (1993)</xref>
        , the assertion within a
specialized time, location, environment, and/or background is
treated as a subset of a general assertion.
      </p>
      <p>This can be captured schematically as follows in
defining the Restrictive Contextual Binding Condition.</p>
      <p>The Restrictive Contextual Binding Condition:
Assume X is an entity, and CO is a contextual operator.
(ii) In a specialized time, location, environment,
background, if X[Y,Z] is directly related to CO[Time,
Locality] in such a setting:</p>
      <p>COi[Timej, Localityk] {……Xi[Yj, Zk]……}
then Xi[Yj, Zk] is contextually bound by COi[Timej,
Localityk].</p>
      <p>The entity Xi[Yj, Zk] is directly related to COi[Timej,
Localityk] iff (if and only if) COi[Timej, Localityk] provides
a context that the interpretation of Xi[Yj, Zk] solely depends
on.</p>
      <p>Obviously, COi[Timej, Localityk] is more restrictive than
COi. In this sense, the Restrictive Contextual Binding
Condition is a subset of the Basic Contextual Binding
Condition.</p>
      <p>Now, the contextual relationship in (2) can be captured
in the following schematic configuration:</p>
      <sec id="sec-4-1">
        <title>CObook[Timeyesterday, Localityatbookstore]</title>
        <p>{……Xbook[Yyesterday, Zatbookstore]……}</p>
        <p>Applying the Restrictive Contextual Binding Condition,
the pronoun “it” is contextually bound by the contextual
operator “CObook[Timeyesterday, Localityatbookstore]”. This
means that the pronoun “it” has to be interpreted as “the
book bought yesterday at the bookstore” if the Restrictive
Contextual Binding Condition is obeyed.</p>
        <p>The Detection Condition can be derived from the above
two conditions:
(iii) If X is in such a contextual binding configuration:
COi {……Xm……}
where Xm is supposed to be contextually bound by
COi but not so, then an abnormal circumstance is
detected.</p>
        <p>Likewise,
(iv) If X is in such a contextual binding configuration:
COi[Timej, Localityk] {……Xm[Yj, Zk]……}
where Xm[Yj, Zk] is supposed to be contextually
bound by COi[Timej, Localityk] but not so, then an
abnormal circumstance is detected.</p>
        <p>As shown above, a contextual operator helps to form the
contextual binding relationship and to resolve ambiguity. A
deception can be detected if the entity is supposed to be
contextually bound by a contextual operator but it is not so
in a configuration.</p>
        <p>Let us apply these conditions to the case in (1).</p>
        <p>In (1), there is such a configuration:</p>
      </sec>
      <sec id="sec-4-2">
        <title>COwhat {……Xwhat……}</title>
        <p>This can be rewritten as follows:</p>
      </sec>
      <sec id="sec-4-3">
        <title>CObook {……Xbook……}</title>
        <p>Here, the pronoun “it” possesses the property “Xbook”,
which is contextually bound by CObook. Hence, this
interpretation is valid and acceptable.</p>
        <p>However, if the pronoun “it” in (1) refers to another
entity, say “the map”, rather the entity “the book” that is
mentioned in the first sentence, this contextual binding
relationship immediately ceases to exist. Below is the
configuration:</p>
      </sec>
      <sec id="sec-4-4">
        <title>CObook {……Xmap……}</title>
        <p>Here, the pronoun “it”, which possesses the property
“Xmap”, is supposed to be contextually bound by the
contextual operator “COmap” but not be contextually bound by
the contextual operator “CObook”. However, the contextual
operator “COmap” is not available. Hence, such a
configuration triggers the Detection Condition. The interpretation is
thus regarded as being invalid and unacceptable.</p>
        <p>Let us look at another linguistic example demonstrated
in the English sentence in (3) below.</p>
        <p>(3) * John likes to buy a booki and read themi within
three days.</p>
        <p>Any speaker of English knows that this sentence is
awkward in the context where the pronoun “them” refers
back to the noun phrase “a book”, as there is a mismatch
between the third-person singular form and the
thirdperson plural form.</p>
        <p>The contextual relationship can be captured in the
following schematic configuration:</p>
      </sec>
      <sec id="sec-4-5">
        <title>CObook {……Xbooks……}</title>
        <p>Here, Xbooks is supposed to be contextually bound by
CObook but not so. Hence, an abnormal circumstance is
detected.</p>
        <p>Let us have a look at still another linguistic example
demonstrated in the English sentences in (4) below.
(4) * John likes to buy a cookbooki and cook iti
following the instruction.</p>
        <p>Any speaker of English knows that it is awkward to have
the pronoun “it” in this context to refer back to the noun
phrase “a cookbook”, because the noun phrase “a
cookbook” possesses the features: [+Object, -edible] while the
pronoun “it” possesses the features: [+Object, +edible] in
the sub-context of “cooking”. This mismatch in features
indicates that the noun phrase “a cookbook” and the
pronoun “it” refer to different entities. In other words, the
pronoun “it” is not contextually bound by the contextual
operator “a cookbook” in this particular case.</p>
        <p>The contextual relationship can be captured in the
following schematic configuration:</p>
      </sec>
      <sec id="sec-4-6">
        <title>COcookbook {……Xediblething……}</title>
        <p>As the pronoun “it” is not contextually bound by
COcookbook in its contextual domain, another abnormal
circumstance is detected.</p>
        <p>Assuming whatever is within the contextual operator is
normal, the variable is normal if and only if it is contextual
bound by its corresponding contextual operator. As shown
above, in order to be properly bound in its contextual
domain, the variable has to possess the same features or
properties as those of the contextual operator. Any
deviation triggers the Detection Condition.</p>
      </sec>
    </sec>
    <sec id="sec-5">
      <title>Deception Detection</title>
      <p>In this section, the Contextual Binding Conditions and the
Detection Condition are applied in the detection of
deception. Assume that what an application or an executable is
expected to do on the basis of its functional requirement is
included in the feature set of the contextual operator. As a
result, this sets up the baseline for the application or the
executable. The actual execution of the application or the
executable is a variable, which should be bound by the
contextual operator. If the actual execution involves more
features than or different features from what is contained in
the contextual operator, the deviation from being normal is
identified, the Detection Condition is triggered, and a
deception is detected.</p>
      <p>Let us examine fabrication first. A piece of malware is a
good example of fabrication. For instance, appended to the
executable “notepad.exe” is a piece of code that makes
possible for the executable to perform file transfer in
addition to its original functionality of text file editing. This
abnormal behavior can be easily detected with the help of
the Basic Contextual Binding Condition and the Detection
Condition.</p>
      <p>Assume what is expected for the original functionality of
the executable is contained inside a contextual operator as
a feature set. Assume the actual functionality of the
executable is contained within a variable as current features.
The variable, by definition, should be contextually bound
by the contextual operator. Schematically, this relationship
is represented below:</p>
      <sec id="sec-5-1">
        <title>COTextEditing {……XTextEditing……}</title>
        <p>This represents a normal situation, in which an executable
is doing what it is expected to do.</p>
        <p>When an extra functionality is added into this
executable, the contextual relationship gets changed, as shown
below:</p>
      </sec>
      <sec id="sec-5-2">
        <title>COTextEditing {……XTextEditing+FileTransfer……}</title>
        <p>Here, one of the actual functionalities of the executable,
i.e. “FileTransfer”, is not contextually bound by the
contextual operator. Thus, an unacceptable behavior is
immediately detected at the application level, even before it is
executed and at the time when a request for extra resource
utilization is made.</p>
        <p>This also applies to other pieces of malware, which
always make requests for additional resource utilization. If
this contextual analysis component is implemented within
the kernel of an operating system, anytime when a request
for resource utilization is received, if it is not contextually
bound by a contextual operator, the request is denied
immediately, an investigation is launched, and this activity is
logged.</p>
        <p>Let us check manipulation now. Stegonography is a
good example of manipulation. For instance, one may hide
a text file inside a graphic file. After this operation, the file
size of the modified graphic file may remain the same as
the file size of the original graphic file. At the first glance,
nothing seems to have happened. However, using a digital
forensic tool, one would see the systematic change of
hexadecimal code even though the change for each byte is
minor, say the change from “0x52” to “0x51” in one byte
and the change from “0x73” to “0x72” in another byte that
is 3 bytes after the previously changed byte. This becomes
obvious when one compares the code for the original
graphic file with the code for the modified graphic file. In
addition, the original file timestamp pattern, consisting of
the date created time, the date accessed time, the date
modified time, and the date last saved time, is changed.
Evidently, the Restrictive Contextual Binding Condition is
violated. Hence, the abnormal behavior in this type of
cases can also be detected.</p>
        <p>Assume both the expected code pattern and the expected
timestamp pattern are contained within the feature set of
the contextual operator. Assume the actual code pattern
and the actual timestamp pattern are contained as current
features in the variable. Also assume that the timestamps
are used to further restrict the actual code pattern, as
illustrated in the Restrictive Contextual Binding Condition. By
definition, the variable should be contextually bound by
the contextual operator. Schematically, this relationship is
represented below:</p>
      </sec>
      <sec id="sec-5-3">
        <title>COPattern1[TimePattern2]{……XPattern1[YPattern2]……}</title>
        <p>This represents a normal situation, in which an expected
pattern is obtained.</p>
        <p>When a graphic file is altered to accommodate a hidden
text file, the actual code pattern gets changed. Now, the
contextual relationship also gets changed, as shown below:</p>
      </sec>
      <sec id="sec-5-4">
        <title>COPattern1[TimePattern2]{……XPattern6[YPattern7]……}</title>
        <p>Here, the actual representation of the file is not
contextually bound by the contextual operator, because the actual
code pattern represented by “XPattern6” is different from the
expected pattern “XPattern1” contained in the contextual
operator and the actual timestamp pattern represented by
“XPattern7” is different from the expected pattern “XPattern2”
contained in the contextual operator. Hence, the
unacceptable behavior is detected at the code level.</p>
        <p>As shown above, the Contextual Binding Conditions and
the Detection Condition can successfully detect deception
such as fabrication and manipulation.</p>
      </sec>
    </sec>
    <sec id="sec-6">
      <title>Conclusion</title>
      <p>Detecting deception in cyberspace is a challenge. Based on
the analysis of the unique property of contextual operators
in a natural language, this paper proposes a contextual
binding mechanism that can be used to disambiguate
interpretation and identify invalid and unacceptable
interpretation in a natural language. The same mechanism can also
be used to detect deception in the cyber domain,
specifically fabrication and manipulation. This mechanism can not
only aid the decision-making in cyber conflicts or cyber
competitions but also lay the foundation for employing
contextual operators in an artificial intelligence system.</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          <string-name>
            <surname>Almeshekah</surname>
            ,
            <given-names>M.</given-names>
          </string-name>
          and
          <string-name>
            <surname>Spafford</surname>
            ,
            <given-names>E.</given-names>
          </string-name>
          <year>2014</year>
          .
          <article-title>Planning and Integrating Deception into Security Defenses</article-title>
          .
          <source>The New Security Paradigm Workshop (NSPW</source>
          <year>2014</year>
          ), Retrieved from http://www.meshekah.com/wpcontent/uploads/2014/10/planning_and
          <article-title>_integrating_deception_into_c omputer_security_defenses_Almeshekah_Spafford</article-title>
          .pdf.
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          <string-name>
            <surname>Brézillon</surname>
            ,
            <given-names>P.</given-names>
          </string-name>
          <year>1999</year>
          .
          <article-title>Context in problem solving: A survey</article-title>
          .
          <source>The Knowledge Engineering Review</source>
          ,
          <volume>14</volume>
          (
          <issue>1</issue>
          ), pp.
          <fpage>1</fpage>
          -
          <lpage>34</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          <string-name>
            <surname>Brézillon</surname>
            ,
            <given-names>P.</given-names>
          </string-name>
          <year>2002</year>
          .
          <article-title>Modeling and using context: Past, present, and future</article-title>
          .
          <source>Rapport de Recherche du LIP6</source>
          <year>2002</year>
          /010, Université Paris 6, France.
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          <string-name>
            <surname>Brézillon</surname>
            ,
            <given-names>P.</given-names>
          </string-name>
          <year>2003</year>
          .
          <article-title>Context Dynamic and Explanation in Contextual Graphs</article-title>
          .
          <source>Context 2003. Lecture Notes in Artificial Intelligence (LNAI) 2680</source>
          .
          <string-name>
            <surname>Blackburn</surname>
            ,
            <given-names>P.</given-names>
          </string-name>
          et al. (Eds). pp.
          <fpage>94</fpage>
          -
          <lpage>106</lpage>
          . Berlin: Springer-Verlag.
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          <string-name>
            <surname>Burgoon</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Blair</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Qin</surname>
            ,
            <given-names>T.</given-names>
          </string-name>
          , and
          <string-name>
            <surname>Nunamaker</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          <year>2003</year>
          .
          <article-title>Detecting Deception through Linguistic Analysis</article-title>
          .
          <source>Proceedings of First NSF/NIJ Symposium on Intelligence and Security Informatics.</source>
        </mixed-citation>
      </ref>
      <ref id="ref6">
        <mixed-citation>
          pp.
          <fpage>91</fpage>
          -
          <lpage>101</lpage>
          , Berlin: Springer-Verlag.
        </mixed-citation>
      </ref>
      <ref id="ref7">
        <mixed-citation>
          <string-name>
            <surname>Caddel</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          <year>2004</year>
          . Deception 101 - Primer on Deception. Strategic Studies Institute, U.S. Army War College.
        </mixed-citation>
      </ref>
      <ref id="ref8">
        <mixed-citation>
          <string-name>
            <surname>Chen</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          &amp;
          <string-name>
            <surname>Duvall</surname>
            ,
            <given-names>G.</given-names>
          </string-name>
          <year>2014</year>
          .
          <article-title>On Operational-Level Cybersecurity Strategy Formation</article-title>
          .
          <source>Journal of Information Warfare</source>
          ,
          <volume>13</volume>
          (
          <issue>3</issue>
          ), pp.
          <fpage>79</fpage>
          -
          <lpage>87</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref9">
        <mixed-citation>
          <string-name>
            <surname>Gaifman</surname>
            ,
            <given-names>H.</given-names>
          </string-name>
          <year>2008</year>
          .
          <article-title>Contextual Logic with Modalities for Time and Space</article-title>
          .
          <source>Review of Symbolic Logic</source>
          ,
          <volume>1</volume>
          (
          <issue>4</issue>
          ), pp.
          <fpage>433</fpage>
          -
          <lpage>458</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref10">
        <mixed-citation>
          <string-name>
            <surname>Grossi</surname>
            ,
            <given-names>D.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Dignum</surname>
            ,
            <given-names>F.</given-names>
          </string-name>
          , and Meyer,
          <string-name>
            <surname>J.</surname>
          </string-name>
          <year>2006</year>
          .
          <article-title>Contextual Terminologies</article-title>
          .
          <source>Computer Logic in Multi-Agent Systems (CLIMA) VI, Lecture Notes in Artificial Intelligence (LNAI) 3900</source>
          .
          <string-name>
            <surname>Toni</surname>
            ,
            <given-names>F</given-names>
          </string-name>
          &amp; Torroni,
          <string-name>
            <surname>P</surname>
          </string-name>
          . (Eds). pp.
          <fpage>284</fpage>
          -
          <lpage>302</lpage>
          . Berlin: Springer-Verlag.
        </mixed-citation>
      </ref>
      <ref id="ref11">
        <mixed-citation>
          <string-name>
            <surname>Kay</surname>
            ,
            <given-names>P.</given-names>
          </string-name>
          <year>1989</year>
          . Contextual Operators: Respective, Respectively, and
          <string-name>
            <given-names>Vice</given-names>
            <surname>Versa</surname>
          </string-name>
          .
          <source>Proceedings of the Fifteenth Annual Meeting of the Berkeley Linguistics Society</source>
          . pp.
          <fpage>181</fpage>
          -
          <lpage>192</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref12">
        <mixed-citation>
          <string-name>
            <surname>McCarthy</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          <year>1993</year>
          . Notes on Formalizing Context.
          <source>Proceedings of the 13th International Joint Conference on Artificial Intelligence -</source>
          Volume
          <volume>1</volume>
          . pp.
          <fpage>555</fpage>
          -
          <lpage>560</lpage>
          . San Francisco, CA: Morgan Kaufmann Publishers Inc.
        </mixed-citation>
      </ref>
      <ref id="ref13">
        <mixed-citation>
          <string-name>
            <surname>Rebuschi</surname>
            ,
            <given-names>M.</given-names>
          </string-name>
          &amp;
          <string-name>
            <surname>Lihoreau</surname>
            ,
            <given-names>F.</given-names>
          </string-name>
          <year>2009</year>
          . Contextual Epistemic Logic.
        </mixed-citation>
      </ref>
      <ref id="ref14">
        <mixed-citation>
          <string-name>
            <surname>Retrieved</surname>
          </string-name>
          from http://www.academia.edu/8052225/Contextual_Epistemic_Logic.
        </mixed-citation>
      </ref>
      <ref id="ref15">
        <mixed-citation>
          <string-name>
            <surname>Zhou</surname>
            ,
            <given-names>L.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Shi</surname>
            ,
            <given-names>Y.</given-names>
          </string-name>
          , and
          <string-name>
            <surname>Zhang</surname>
            ,
            <given-names>D.</given-names>
          </string-name>
          <year>2008</year>
          .
          <article-title>A Statistical Language Modeling Approach to Online Deception Detection</article-title>
          .
          <source>IEEE Transactions on Knowledge and Data Engineering</source>
          ,
          <volume>20</volume>
          (
          <issue>8</issue>
          ). pp.
          <fpage>1077</fpage>
          -
          <lpage>1081</lpage>
          .
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>