Working Group 4 Creative Systems for Knowledge Management in Life Sciences
Jan 01, 2016
Working Group 4
Creative Systems for Knowledge Management in
Life Sciences
Purpose of this Talk
We are researching methods which we believe could provide non-standard solutions to complex problems
We need concrete problems to identify possible interactions between the working groups
Structure of Talk
Individual research directions
General techniques for creative reasoning
A case study
Computational Bioinformatics Laboratory, Imperial College London
Progol system– Learning of concepts in bioinformatics– Theory behind, and implementation of ILP– Applications:
• Predictive toxicology, secondary structure in proteins, learning metabolic pathways
HR system– Discovering in mathematics (and bioinformatics)– Theory behind, and implementation of ATF– Applications:
• Adding to databases: Integer sequences, TPTP library• Finding invariants, inventing CSP constraints, tutorials
Scientific Discovery via integration of techniques
Centre for Computational CreativityCity University, London
Formal frameworks for describing and reasoning about creative behaviour– Compare seach methods and outcomes– Define value etc and reason about properties of
definitions Pattern discovery and matching technogies for
multidimensional datasets– Discover/locate geometrically identical structural
regions, possibly with gaps in multi-D data– Example: 3D representations of atoms in space for
pharmacophore bonding models
University of A CorunhaHybrid Society (HS)
Development framework to validate and to allow the learning of various computational models of tasks which require creativity and a social behaviour
HS is based on machines and humans living together in a virtual and “egalitarian” society
Solves the problem of Value in a dynamic context.
Allows the comparison of different computer paradigms and systems.
Allow the collaboration between humans and computer systems
Allows the use of adaptive techniques such us Evolutionary Computation and Artificial Neural Networks
Creative Systems GroupUniversity of Coimbra Computational Models of Creativity
– Analogy– Evolution– Conceptual Blending
Models of Surprise Hybrid Societies for Creativity
Assessment
University of Edinburgh
Lakatos-style reasoning:- Experts interact to build a common theory
- Counterexamples used to modify conjectures; clarify concepts; improve proofs
- Ways of evaluating machine creativity
Universidad Complutense de Madrid
Ongoing research work:– Knowledge intensive CBR
• CBRArm: framework for CBR + ontologies
– Generating narrative and metaphorical texts, NLG architectures, CBR for text generation
– CBR for Knowledge Management • Java documentation, helpdesks
– Information Filtering + User Modeling– Computer games
Creative Reasoning
Reasoning in non-standard ways to produce:– “interesting”/valued/unexpected outputs– emergent complex behaviour
Reconceptualise existing knowledge structures to get new knowledge structures with added value– using in a different way than they were intended– lateral connections that weren’t there already
Heuristic reasoning – Including sound and unsound methods
Post hoc verification – value measurements for the domain are a pre-requisite
General Techniques
Conceptual blending Metaphorical/analogical reasoning Inductive inference Hypothesis repair Evolutionary methods
Conceptual Blending
Blend: Rutherford Atom
Input: Atom (1) Input: Solar System(2)
E = electron N = nucleus r = rotates around N much bigger than E
S
P E
N
N=S
r
E = P
o
P = planet S = sun o = orbits around S much bigger than P
Electron = Atomic Planet Nucleus = Atomic Sun Gravity-like force keeps the electrons in orbit about the nucleus
BUT:
Electrons have a statistical rather
than absolute position in space
r=o
similar
similar
Metaphorical Reasoning
“A poison secreted by certain animals”
Venom (1)
Poison (1)
isa
isa
isa
Substance
Entity
Object
“Anything that harms or destroys”
Poison (2)
Destructiveness
isa
isa
isa
Quality
Abstraction
Attribute
“An artist who is master of a particular style”
Insult (1)
Disrespect
isa
isa
isa
Communication
Abstraction
Relation
Inductive Inference
Predictive Induction– Know the positives/negatives of a concept– Search for a concept which fits categorisation
• Use examples as evidence for predictive accuracy• Cross validate results
Descriptive Induction– Search for rules which associate background
predicates, using data as empirical evidence– (Sometimes) use deduction to prove rules found
Hypothesis Repair
Using a counterexample to repair a faulty hypothesis by:– Generalising from counterexample to a
property then stating the exception in the hypothesis
– Generalising from the positives and then limiting the hypothesis to these
Evolutionary Methods
Exploration of complex search spaces– in non-uniform ways– Based on biologically inspired evolutionary
notions such as gene recombination, mutation, fitness functions
– Dynamically adaptive systems
Potential Applications
Levels of discovery– You know what you are looking for,
• But you don’t know what it looks like
– You don’t know what you are looking for• But you know you are looking for something
– You didn’t know you were even looking for anything
Levels of search– At the object level (millions/billions of data points)– At the semantics level (tens of thousands of terms)– At the meta-level (scores of techniques)
Possible (General) Application:Ontology Maintenance Ontologies standardise concepts
– And standardise relationships between them Many areas see the need for ontologies
– Including scientific domains such as life sciences Very important that the ontology represents
current scientific thinking Need to continually maintain ontology
– New nodes– New links
Need to continually interpret ontology– Large scale structures
Case Study – Gene Ontology
~14,000 terms from biology/genetics– Process, function, structure– Structured into hierarchies using isa/partof
Each term has genes associated– ~ 1.3 million genes (from, e.g., GenBank)
Aims to unify biology– Databases are in a bad state
• Different interpretations/notations/standards
Gene Ontology (Example)
Methods for Ontology Maintenance Mining rules between concepts using inductive
techniques (adds edges)– Project to use HR for this in progress– Project to use Progol to learn terminology
Conceptual blending– Invent new concepts (nodes)
Metaphorical reasoning– Look at structure to reorganise links
Hypothesis repair– Explain genes which are seemingly misclassified
Proactive and Reactive Applications Proactive
– Attempt to make discoveries in GO– Give value added when someone submits a
new term to the ontology Reactive
– A new gene is added which (using sequence alignments) is associated with “wrong” concept
– Creatively re-organise ontology to fix problem
The Bottom Line We have solutions but not problems
– With respect to Life Sciences Our application domains are disparate
– But our methods are general We’re already thinking about certain
tasks/problems in life sciences– Predictive toxicology– Protein structure prediction
And we’re inventing our own problems– Maintaining the Gene Ontology
But we really need to discuss what it is that standard techniques do not yet give you– And see what creative systems/techniques can do