methodology ontology

1
methodology - ontology

• who identifies who?
• how?
• how to structure input of research?

2
D2.1 Nabeth 2004

• ontology in computer sciences
• explicit specification of a conceptualisation
• ontology (what is)
• epistemology (what can we know)
• methodology (how can we produce knowledge)

3

• ontology in computer science is an instrument to
  clarify and share the use of terms (pragmatic
  approach)
• better not get into a discussion on real
  meaning or true identity
• better see the difference between 1st and 3rd
  person perspective or self and same

4
identity concept

• modeling
• (mix of 1st/3rd p. perspective)
• I, me and self (Mead)
• true identity, assigned identity, abstracted
  identity (Durand)
• identities and territories (contexts, Nabeth?)
• relational and dynamic concept of identity as
  nexus of different roles, evershifting

5
identification concept

• 3rd person perspective
• risks, mechanisms, protection against, management
• importance 1st person perspective
• (organisations, national state)

6
Inventory of terms and some categorisation

• definitions, illustrations and references,
  relations between terms
• beginnings of the construction of a semantic
  network
• lexical (syntactic, definitions that relate a
  term to other terms)

7
Canhoto Backhouse

• categorization theory and semiotics
• case-study
• EU-directive to combat money laundering
• objective the same in all MSs
• wide variation in submission levels

8
Suspicious Transaction Report

• STR to Financial Intelligence Unit
• trade off between false negatives and false
  positives, reporting institution is stimulated to
  over-report, law enforcement agents should
  minimize false positives
• over-reporting creates backlog

9

• how to reduce false negatives and false
  positives
• how to expand knowledge to refine the
  identification of suspicious financial
  transactions

10

• role of automatic monitoring
• role of intuition (practical wisdom, experience,
  refined judgement)
• traditional methodology too much oriented towards
  technological design and legal regulation, plea
  for incorporation of semiotics

11
semiotics I

• physical level
• records of actions and users
• empirical level
• aggregation of data at client level
• syntactical level
• automatic monitoring systems

12
semiotics II

• semantic level
• legal landscape, differentiation MSs
• pragmatic level
• cognitive prototype developed by professionals
  with significant experience
• social level
• formal/informal norms, cultural context

13
profiling/categorisation

• how to generate profiles that yield few false
  negatives and few false positives (balance
  between the two will depend per context)
• this is a matter of both privacy and security
• but privacy/security is also about not being
  profiled (anonymity, pseudonymity,
  unlinkability)

14
refinement of identification

• syntactical level
• develop intelligent automatic monitoring systems
• pragmatic level
• learning theory interpretation of automatically
  generated profiles, how to generate/recognise new
  patterns
• social level
• diversity of socio-cultural norms

15
semiotics

• recognition of intersubjective perspective in
  objectification
• beyond reification of ontologies
• beyond reductive interpretations of identity
• challenge how to further refine profiling
  technologies while protecting indeterminate
  identity (freedom to (re)define yourself)