Information Extraction from Biomedical Text

1
Information Extraction from Biomedical Text

• Jerry R. Hobbs
• Artificial Intelligence Center
• SRI International

2
Introduction

• Information Extraction
• Extract entities, relations, events
• Capture structured information
• Domain specific
• Focus only relevant parts
• Mainly on economic and military interest?
• Biomedical domain

3
Cascaded Finite-State Transducers

• Separate Processing into several stages
• FASTUS (Finite-State Automaton Text Understanding
  System)
• Earlier Stages
• Smaller linguistic objects
• Domain independent
• Later Stages
• Domain dependent patterns

4
Cascaded Finite-State Transducers

• Complex Words
• Basic Phrases
• Complex phrases
• Domain Patterns
• Merging Structures

5
Example

• gamma-Glutamyl kinase, the 1st enzyme of
  the proline biosynthetic pathway, was puried to a
  homogeneity from an Escherichia coli strain
  resistant to the proline analog
  3,4-dehydroproline. The enzyme had a native
  molecular weight of 236,000 and was apparently
  comprised of six identical 40,000-dalton subunits.

6
Target Database

• Reaction Object
• Attributes ID
• Pathway
• Enzyme
• ..
• Enzyme Object
• Attribute ID
• Name
• Molecular-Weight
• Subunit-Component
• Subunit-Number

7
Complex Words

• Recognizes
• multiword fixed phrases
• proper names
• Rich in the biological domain
• Use lexicon or ML and Statistic methods

• gamma-Glutamyl kinase,
• the 1st enzyme of the proline biosynthetic
• pathway, was purified to a homogeneity
• from an Escherichia coli strain resistant to
• the proline analog 3,4-dehydroproline.
• The enzyme had a native molecular
• weight of 236,000 and was apparently
• comprised of six identical
• 40,000-dalton subunits.

• gamma-Glutamyl kinase,
• the 1st enzyme of the proline biosynthetic
• pathway, was purified to a homogeneity
• from an Escherichia coli strain resistant to
• the proline analog 3,4-dehydroproline.
• The enzyme had a native molecular
• weight of 236,000 and was apparently
• comprised of six identical
• 40,000-dalton subunits.

8
Basic Phrases

• Segment a sentence into noun groups, verb groups,
  and particles
• Use Sager 1981 grammar

9
Complex Phrases
10
Complex Phrases

• Structures of basic and complex phrases, entities
  and events

11
Clause-Level Domain Patterns

• The enzyme had a native molecular weight of
  236,000 and was apparently comprised of six
  identical 40,000-dalton subunits.

12
Clause-Level Domain Patterns

• The enzyme had a native molecular weight of
  236,000 and was apparently comprised of six
  identical 40,000-dalton subunits.

13
Merging Structures

• First 4 levels processes within single sentence
• This level collect and combine information for
  on entity or relationship
• Three Criteria
• The internal structure of noun groups
• The nearness along some metric
• Consistency and compatibility of the 2 structures

14
(No Transcript)
15
Compile Time Transformations

• Subject-Verb-Object pattern ? linguistic patterns
  (passive, relative clauses, etc)

16
Types of Specialized Domains

• noun-driven approach
• The type of an entity is highly predictive of its
  role in event
• Loose S-V-O patterns
• verb-driven approach
• The role of the entities in events cannot be
  predicted from their type
• Tight S-V-O patterns

17
Limitation of IE Technology

• MUC (1990)
• Name recognition 95 recall and precision
• Event recognition 60 recall and precision
• Possible reasons
• Process of merging
• Only works with explicit information
• Common cases are covered, how about those rare
  cases?