Utilizing Suffix Meanings in a Hybrid Approach to Classification

1
Utilizing Suffix Meanings in a Hybrid Approach to
Classification

• Aaron Loehrlein
• 2004 SLIS Doctoral Student Research Forum
• September 18, 2004

2
Objective

• explore methods that combine the advantages of
  automatic and manual classification
• identify methods for an automatic partial
  classification of concepts
• determine the extent to which terms organized by
  suffix meanings form the basis for a useful
  classification system

3
Previous research regarding suffixes

• Stemming in order to conflate terms (Harman,
  1991 Savoy, 1993)
• subjected, subject, subjective, subjects
• Identifying phrases within a sentence (Okada et
  al, 2001)
• natural language, language processing,
  natural language processing

4
EPA Topics website
(Source EPA Topics, http//www.epa.gov/epahome/to
pics.html)
5
EPA Classification
(Source EPA Topics, http//www.epa.gov/epahome/to
pics.html)
6
Identify every term in the lexical base

• Term No. of occurrences
• Funds 1
• Furans 1
• Garden 1
• Garment 1
• Gas 1
• Generation 6
• Generator 3
• Generators 1
• Genetic 1
• Geothermal 1
• Glass 1

• Term No. of
  occurrences
• Global 11
• Grants 18
• Great 2
• Green 1
• Greenhouse 1
• Ground 10
• Group 2
• Growth 2
• Guidance 4
• Guidelines 1
• Halons 1

723 terms total (source EPA Topics,
http//www.epa.gov/epahome/topics.html)
7
Identify the most popular ending strings in the
lexical base

• String No. of occurrences
• ve 13
• ive 12
• tive 12
• ative 7
• ctive 4

• String No. of occurrences
• al 45
• cal 8
• ical 8
• gical 4
• ogical 4
• logical 4

8
Identify every suffix in the English language

• Examples
• -podium
• -podous
• -pods
• -pogon
• -poieses
• -poiesis
• -poietic
• -polis
• -polises

• -poo
• -pora
• -pore
• -pores
• -praxes
• -praxia
• -praxias
• -praxis
• -praxises

1,571 suffixes total (Source Merriam-Webster
Unabridged, http//unabridged.merriam-webster.com/
)
9

• -al
• action or process ltbestowalgt
• agricultural animal bilateral
  disposal montreal natural
• of, relating to, or characterized by
  ltdirectionalgt
• agricultural animal bilateral
  disposal montreal natural
• pharmaceutical product ltbarbitalgt
• agricultural animal bilateral
  disposal montreal natural
• -ative
• of, relating to, or connected with ltquantitativegt
• alternative cooperative cumulative
  initiative
• tending to lttalkativegt
• alternative cooperative cumulative
  initiative
• -ial
• of, relating to, or characterized by ltmanorialgt
• confidential industrial remedial
  terrestrial

(Source Merriam-Webster Unabridged,
http//unabridged.merriam-webster.com/)
10
Unit of analysis suffix / meaning pair

• -al / actions, processes
• -al / characterized by
• -al / pharmaceutical products
• -ative / characterized by
• -ative / characterized by an action
• -ial / characterized by

11
Consult a larger lexical base in the same domain

• pesticide
• EPA topic classes
• (723 terms total)

• acaricide
• biopesticide
• decide
• fungicide
• herbicide
• insecticide
• miticide
• mosquitocide
• pesticide
• rodenticide
• termiticide
• EPA resource metadata terms
• (13,189 terms total)

12
Organize meanings

• -acetic acetic acid or vinegar
• -id binary chemical compounds, compounds
  regarded as binary
• -ene unsaturated carbon compounds
• Assigned to the superordinate chemicals,
  chemical compounds
• -acetic chemicals, chemical compounds / acetic
  acid or vinegar
• -id chemicals, chemical compounds / binary
  chemical compounds, compounds regarded as
  binary
• -ene chemicals, chemical compounds /
  unsaturated carbon compounds

13
Organize meanings (cont.)

• -escent reflecting or emitting light
• -biotic life, a specific mode of life
• -logical the study of
• Assigned to the superordinate characteristics
• -escent characteristics / reflecting or
  emitting light
• -biotic characteristics / life, a specific
  mode of life
• -logical characteristics / the study of

14
Organize meanings (cont.)

• Hierarchy of meanings Sample suffixes
• abstractions
• taxonomic divisions -an
• actions, processes -ing
• cause to be / come to be -en
• a killing -cide
• characteristics -ive
• capable, fit for, worthy of -ably
• the study of -logical
• entities
• chemicals, chemical compounds
• unsaturated carbon compounds -ene
• people -man
• things
• drawings, writings, records -gram
• states, qualities, conditions -ship

15
Determine specificity of meaning

• entities
• chemicals, chemical compounds
• Carbon
• Chlorofluorocarbons
• Hydrochlorofluorocarbons
• binary chemical compounds, compounds regarded as
  binary
• Bromide
• Chloride
• Cyanide
• Cyanides
• Monoxide
• Oxides
• Radionuclides
• Ride
• chemical elements, chemical radicals, ions having
  a positive charge
• Cadmium
• Uranium

• chemical radicals
• Biphenyls
• Butyl
• Methyl
• unsaturated carbon compounds
• Benzene
• Ethylbenzene
• Scene
• Styrene
• Toluene
• unsaturated hydrocarbons, bivalent radicals
• Dichloroethylene
• Perchloroethylene
• Tetrachloroethylene
• Trichloroethylene

16
Determine specificity of meaning

• entities
• chemicals, chemical compounds
• Benzene
• Biphenyls
• Bromide
• Butyl
• Cadmium
• Carbon
• Chloride
• Chlorofluorocarbons
• Cyanide
• Cyanides
• Dichloroethylene
• Ethylbenzene
• Hydrochlorofluorocarbons

• Methyl
• Monoxide
• Oxides
• Perchloroethylene
• Radionuclides
• Ride
• Scene
• Styrene
• Tetrachloroethylene
• Toluene
• Trichloroethylene
• Uranium

17
Determine specificity of meaning (cont.)

• entities
• people, things
• performing an action
• Contaminants
• Coordinators
• Disinfectants
• Generator
• Generators
• Grants
• Indicators
• Pollutant
• Pollutants
• Propellants
• Refrigerant
• Refrigerants

• things
• fossils, minerals, and rocks
• Vermiculite
• vehicles
• Automobile
• Automobiles
• Mobile

18
Determine specificity of meaning (cont.)

• entities
• people, things
• performing an action
• Contaminants
• Coordinators
• Disinfectants
• Generator
• Generators
• Grants
• Indicators
• Pollutant
• Pollutants
• Propellants
• Refrigerant
• Refrigerants

• things
• Automobile
• Automobiles
• Mobile
• Vermiculite

19
Final table format
20
Output of the final heuristic

• characteristics
• Agricultural
• Animal
• Antimicrobial
• Appeals
• Aquatic
• characteristics -- an action
• Administrative
• Alternative
• Alternatives
• Bioaccumulative
• characteristics -- capable of, fit for, worthy of
• Achievable
• Available
• Feasibility
• characteristics -- full of, abounding in, having
• Hazardous
• Nonhazardous

• characteristics -- resembling
• Chloroform
• Coliform
• characteristics -- the study of
• Biological
• Ecological
• Radiological
• Toxicological
• characteristics -- used to indicate possession
• Children's
• Seniors'
• entities
• Cryptosporidium
• entities -- people, things
• entities -- people, things -- producing or being
  produced by
• Carcinogens
• Nitrogen
• Pathogens

21
Manual validation and revision

• Still in progress
• Preliminary results
• Potentially more specific meanings for certain
  suffixes
• -ing might be most likely to apply to physical
  actions (mining, biking, swimming, etc.)
• -ance words seem to be particularly abstract
  (assistance, maintenance, etc.)

22
References

• Harman, D. (1991). How effective is suffixing?
  Journal of the American Society for Information
  Science, 42(1), 7-15.
• EPA Topics, http//www.epa.gov/epahome/topics.html
  (13 Apr. 2004).
• Savoy, J. (1993). Stemming of French word based
  on grammatical categories. Journal of the
  American Society for Information Science, 44(1),
  1-9.
• Okada, M., Ando, K., Lee, S.S., Hayashi, Y.,
  Aoe, J. (2001). An efficient substring search
  method by using delayed keyword extraction.
  Information Processing Management, 37, 741-761.
• Webster's Third New International Dictionary,
  Unabridged. Merriam-Webster, 2002.
  http//unabridged.merriam-webster.com (13 Apr.
  2004).