Morphological analyzer

• US 5,323,316 A
• Filed: 02/01/1991
• Issued: 06/21/1994
• Est. Priority Date: 02/01/1991
• Status: Expired due to Term

• Current Assignee: ANPA Incorporated
• Original Assignee: Wang Laboratories Incorporated
• Inventors: Kadashevich, A. Julie, Harvey, Mary F., Clark, Cheryl
• Primary Examiner(s): Envall, Jr., Roy N.
• Assistant Examiner(s): BAI, ARI
• Application Number:
  US07/649,779
• Time in Patent Office: 1,236 Days
• Field of Search: 364/419, 364/419.01, 364/419.10, 364/419.11, 364/419.16
• US Class Current: 704/9
• CPC Class Codes: G06F 40/253 Grammatical analysis; Style critique

Abstract

A morphological analyzer for analyzing the morphology of an input word, the analyzer including a recognition engine which includes means for finding a stem within the input word; and means for identifying suffixes attached to the stem, wherein the finding means and suffix identifying means cooperate to conduct morphological analysis of the input word from the root to the affix and wherein the recognition engine performs inflectional and derivational analysis, the derivational analysis using more than two derivational suffixes.

Claims

1. A morphological analyzer for analyzing the morphology of an input word, the analyzer comprising a recognition engine including:

a memory for storing said input word;

a linguistic stem processor for reading said input port and for finding a linguistic stem within said input word; and

a suffix processor for identifying suffixes attached to said linguistic stem, wherein said linguistic stem finding processor and suffix identifying processor cooperate to conduct morphological analysis of the input word from the root to the affix and wherein said recognition engine performs inflectional and derivational analysis, said wherein each derivational analysis is performed recursively using more than two derivational suffixes within each said input word.

2. A morphological analyzer for analyzing the morphology of an input word, the analyzer comprising a recognition engine including:

a memory for storing said input word;

a stem processor for reading said input word and for finding a stem within said input word by starting near the beginning of the input word and proceeding through the input word one letter at a time to find said stem; and

a suffix processor for identifying suffixes attached to said stem, wherein said stem finding processor and suffix identifying processor cooperate to conduct morphological analysis of the input word from the root to the affix and wherein said recognition engine performs inflectional and derivational analysis, said derivational analysis using more than two derivational suffixes.

3. A morphological analyzer for analyzing the morphology of an input word, the analyzer comprising a recognition engine including:

a memory for storing said input word;

a stem processor for reading said input word and for finding a stem within said input word; and

a suffix processor for identifying suffixes attached to said stem by proceeding through the portion of the input word remaining after the stem one letter at a time to find said suffixes, wherein said stem finding processor and suffix identifying processor cooperate to conduct morphological analysis of the input word from the root to the affix and wherein said recognition engine performs inflectional and derivational analysis, said derivational analysis using more than two derivational suffixes.

4. A morphological analyzer for analyzing the morphology of an input word, the analyzer comprising a recognition engine including:

a memory for storing said input word;

a stem processor for reading said input word and for finding a stem within said input word; and

a suffix processor for identifying suffixes attached to said stem, wherein said stem finding processor and suffix identifying processor cooperate to conduct morphological analysis of the input word from the root to the affix and wherein said recognition engine performs inflectional and derivational analysis, said derivational analysis using more than two derivational suffixes;

wherein

said recognition engine in performing in stem and suffix recognition functions proposes a surface null after at least some of the characters of the input word.

5. A morphological analyzer for analyzing the morphology of an input word, the analyzer comprising a recognition engine including:

a memory for storing said input word;

a stem processor for reading said input word and for finding a stem within said input word, wherein the stem finding means starts its recognition task at the n^th letter of said input word, skipping the first n-1 letters of said input word, n being a variable which takes on different integer values depending upon the size of the input word; and

a suffix processor for identifying suffixes attached to said stem, wherein said stem finding means and suffix identifying means cooperate to conduct morphological analysis of the input word from the root to the affix and wherein said recognition engine performs inflectional and derivational analysis, said derivational analysis using more than two derivational suffixes.

6. A morphological analyzer for analyzing the morphology of an input word, the analyzer comprising a recognition engine including:

a memory for storing said input word;

a stem processor for reading said input word and for finding a stem within said input word, the stem finding means employing a two level morphological model; and

a suffix processor for identifying the suffixes attached to said stem, the suffix identifying means also employing the two level morphological model wherein the two level morphological model is derived from a set of spelling change rules which translates a surface string into its lexical string, and is implemented by a single automation, wherein said stem finding processor and suffix identifying processor cooperate to conduct morphological analysis of the input word from the root to the affix and wherein said recognition engine performs inflectional and derivational analysis, said derivational analysis using more than two derivational suffixes and wherein the rules in said set of spelling change rules are of the form W . X/YZ, wherein W is a lexical character, X is a surface character, Y is a first pair string describing an environment preceding W . X and Z is a second pair string describing an environment following W . X.

7. A morphological analyzer for analyzing the morphology of an input word, the analyzer comprising:

a lexicon memory for storing a general lexicon, including a suffix lexicon memory for storing a suffix lexicon and a stem lexicon memory for storing a stem lexicon,

a memory for storing said input word; and

a recognition engine, including

a stem processor for reading said input word and for referring to said stem lexicon for finding a stem within said input word; and

a suffix processor for referring to said stem finding processor and suffix identifying processor cooperate to conduct morphological analysis of the input word from the root to the affix and wherein said recognition engine performs inflectional and derivational analysis, said derivational analysis using more than two derivational suffixes wherein said suffix lexicon contains REL suffixes wherein REL suffixes are those suffixes which when added to a root form a distantly related word that is also a root form.

8. The morphological analyzer of claim 7 wherein said suffix lexicon contains formative suffixes.

9. The morphological analyzer of claim 7 wherein said suffix lexicon contains more than one boundary character.

10. The morphological analyzer of claim 7 wherein said stem lexicon contains valid words and abstract words.

11. The morphological analyzer of claim 7 wherein said stem lexicon contains a plurality of irregular words and for each of at least some of said plurality of irregular words, said stem lexicon identifies a cross-reference to one or more corresponding base forms for that irregular word, wherein each of said base forms is a valid word.

12. The morphological analyzer of claim 7 wherein said stem lexicon identifies, for certain of the stems, a cross reference to irregular forms of that stem.

13. The morphological analyzer of claim 7 wherein said stem lexicon contains a plurality of abbreviations and, for each of said plurality of abbreviations, said stem lexicon identifies a cross-reference to a corresponding stem which gives an expanded form of that abbreviation.

14. The morphological analyzer of claim 7 wherein:

said stem lexicon identifies, for each stem within said stem lexicon, a corresponding continuation class tree, wherein the continuation class tree for a given stem represents the valid combinations of suffixes that can be attached to that stem; and

,

said suffix lexicon contains a plurality of suffixes and each of said plurality of suffixes is represented by a corresponding continuation class.

15. The morphological analyzer of claim 7 wherein said general lexicon further comprises a continuation class lexicon, wherein said continuation class lexicon contains a plurality of continuation class trees, each of said continuation class trees corresponding to a different stem found within said stem lexicon and representing the valid combinations of suffixes that can be attached to that stem.

16. A morphological analyzer for analyzing the morphology of an input word, the analyzer comprising:

a memory for storing said input word;

a lexicon memory for storing a general lexicon, including

a stem lexicon memory for storing a stem lexicon; and

a suffix memory for storing a single suffix lexicon containing a plurality of suffixes, each of said plurality of suffixes being represented by a corresponding continuation class, said suffix lexicon being shared by all continuation classes; and

a recognition engine including

a stem processor for reading said input word and for referring to said stem lexicon for finding a stem within said input word; and

a suffix processor for referring to said suffix lexicon for identifying suffixes attaches to said stem, wherein said stem finding processor and said suffix identifying processor cooperate to conduct morphological analysis of the input word from the root to the affix and wherein said recognition engine performs inflectional and derivational analysis, said derivational analysis using more than two derivational suffixes.

17. A morphological analyzer for analyzing the morphology of an input word, the analyzer comprising:

a memory for storing said input;

a lexicon memory for storing a general lexicon, including a suffix lexicon memory for storing a suffix lexicon and a stem lexicon memory for storing a stem lexicon; and

a recognition engine, including

a stem processor for reading said input word and for referring to said stem lexicon for finding a stem within said input word; and

a suffix processor for referring to said suffix lexicon for identifying suffixes attached to said stem, wherein said stem finding processor and suffix identifying processor cooperate to conduct morphological analysis of the input word from the root to the affix and wherein said recognition engine performs inflectional and derivational analysis, said derivational analysis using more than two derivational suffixes wherein for each of at least some of said suffixes, said suffix lexicon identifies a corresponding suffix rating indicating the closeness in meaning between a first word and a second word formed by adding that suffix to said first word.

18. The morphological analyzer of claim 17 wherein for each of at least some of said suffixes, said suffix lexicon identifies a corresponding suffix rating indicating whether that suffix is inflectional, derivational, or formative.

19. A morphological analyzer for analyzing the morphology of an input word, the analyzer comprising a recognition engine including:

a memory for storing said input word;

a stem processor for reading said input word and for finding a stem within said input word; and

a suffix processor for identifying suffixes attached to said stem, wherein said stem finding processor and suffix identifying processor cooperate to conduct morphological analysis of the input word from the root to the affix and wherein said recognition engine performs inflectional and derivational analysis, said derivational analysis using more than two derivational suffixes;

wherein said recognition engine returns a base history of said input word, said base history containing morphological information about said input word.

20. The morphological analyzer of claim 19 wherein said base history includes a stem record for each stem found within said input word and, for each stem record, one or more continuation class records identifying the suffixes added to that stem to produce said input word, wherein each of said continuation class records identifies a part of speech associated with the corresponding suffix.

21. The morphological analyzer of claim 20 further comprising a generation engine for generating words that are lexically related to said input word by using information returned through the base history supplied for the input word by said recognition engine.

22. The morphological analyzer of claim 21 wherein said generation engine generates said lexically related words by proposing lexical nulls at certain points during the generation of said lexically related words.

23. The morphological analyzer of claim 21 further comprising a lexicon, said lexicon containing a listing of stems and, for each stem, an associated continuation class tree, and wherein each continuation class tree controls which suffixes said generation engine adds to the stem with which it is associated.

24. The morphological analyzer of claim 21 further comprising a table identifying equivalent suffixes, wherein equivalent suffixes are those which form similar words when added to a base word, and wherein said generation engine uses said table to identify which suffixes are added to a given stem to produce lexically related words.

25. A method of analyzing the morphology of an input word, comprising the steps of:

storing the input word in a memory;

by operation of a stem processor, first reading the input word and finding a linguistic stem within said input word; and

then, by operation of a suffix processor, identifying all suffixes attached to said linguistic stem, wherein said suffix identifying step involves inflectional and derivational analysis, wherein each said derivational analysis is performed recursively using more than two derivational suffixes within each said input word.

26. A method of analyzing the morphology of an input word, comprising the steps of:

storing the input word in a memory;

by operation of a stem processor, reading the input word and finding a stem within said input word;

by operation of a suffix processor, identifying all suffixes attached to said stem, wherein said suffix identifying step involves inflectional and derivational analysis, said derivational analysis using more than two derivational suffixes; and

returning a morphological history of said input word.

27. The method of claim 26 further comprising the steps of storing said morphological history in a morphological history memory and, by operation of a word generation processor, generating words that are lexically related to said input word by adding suffixes to said input word, the addition of suffixes being based upon information supplied by said morphological history.