Word annotation system

• US 4,864,501 A
• Filed: 10/07/1987
• Issued: 09/05/1989
• Est. Priority Date: 10/07/1987
• Status: Expired due to Term

• Current Assignee: Vantage Technology Holdings LLC
• Original Assignee: Houghton Mifflin Harcourt Publishing Company (Houghton Mifflin Harcourt Co. (Private Equity))
• Inventors: Carus, Alwin B., Kucera, Henry
• Primary Examiner(s): Fleming, Michael R.
• Application Number:
  US07/106,224
• Time in Patent Office: 699 Days
• Field of Search: 364/200, 364/900, 364/419, 382/57, 382/36, 382/37, 382/38
• US Class Current: 704/8
• CPC Class Codes: G06F 40/253 Grammatical analysis; Style critique

Abstract

A system for annotating digitally encoded text includes a dictionary of base forms. For each base form, a first set of tags represents possible grammatical and syntactic properties of the word, and may encode inflectional paradigms of the base form, or feature agreement behavior and special processing. If a text word is not found in the dictionary, an inflectional analyzer looks up one or more base forms derived from the word, and if found, and annotates them with their dictionary tags. A morphological analyzer assigns tags to words not retrieved in the dictionary. The morphological analyzer recognizes words formed by prefixation and suffixation, as well as proper nouns, ordinals, idiomatic expressions, and certain classes of character strings. The tagged words of a sentence are then processed to parse the sentence.

Claims

1. Apparatus for annotating digitally encoded natural language text words, such apparatus comprising

a dictionary database including a plurality of encoded word base forms, wherein a base form is stored together with a first set of data encoding the possible uses or features of words corresponding to the base form, and with a second set of data encoding the synthesis of inflections of the base form,

look-up means for identifying a base form of a text word, such look-up means including

(i) means for detecting a characteristic inflectional ending occurring in the text word to produce a candidate base form, and

(ii) means for determining whether the candidate base form is a word base form in the dictionary data base, and in that event assigning data stored with the base form to the text word, and

means for assigning a dummy base form and a set of data codes to a word for which the look-up means retrieves no base form from the dictionary database.

2. Apparatus according to claim 1, further comprising

grammatical processing means, operative on digitally encoded text words and on the first and second sets of data codes thereof for determining a parse of a sentence.

3. Apparatus according to claim 1, further comprising

matching means, for detecting when a word is approximately identical to an inflection of a base form but differs therefrom, and

error means, for displaying an error message associated with the difference.

4. Apparatus for annotating digitally encoded natural language text words, such apparatus comprising

a dictionary database including a plurality of word records, each record including a set of tags indicative of properties of a word

look up means for looking up a text word in the dictionary database and retrieving its set of tags when the text word is identified in the dictionary database, and

morphological analyzer means, operative on a text word which is not identified in the dictionary database, for determining a set of tags and dummy base form by inspection of the morphology or context of such word.

5. Apparatus according to claim 4, further comprising

grammatical processing means, operative on digitally encoded text words and on the tags for determining a parse of a sentence of the text.

6. Apparatus according to claim 4, wherein the morphological analyzer means comprises

means for identifying a text word which is derived from a word having a dictionary base form by a process of suffixation.

7. Apparatus according to claim 4, wherein the morphological analyzer means comprises

means for identifying an idiomatic expression.

8. Apparatus according to claim 4, wherein the morphological analyzer means comprises

means for identifying a word consisting of a permissible literal-numeric combination.

9. Apparatus according to claim 4, wherein the morphological analyzer means comprises

means for recognizing a characteristic inflectional portion of a text word and for assigning thereto a provisional base form and provisional tags.

10. Apparatus according to claim 9, wherein the morphological analyzer means comprises

means for creating a word record including said provisional base form and provisional tags.

11. Apparatus according to claim 4, wherein the morphological analyzer means comprises

means for identifying a text word which is derived from a word having a dictionary base form by a process of prefixation.

12. Apparatus according to claim 4, wherein the morphological analyzer means includes

a stored table of word endings wherein the class of words having such ending is characterized by corresponding tags, each such ending being stored with its corresponding tags, and

means for looking up endings of a text word in said stored table and retrieving the tags corresponding to a said ending found in the table.

13. Apparatus according to claim 12, wherein the morphological analyzer means further includes means for identifying certain classes of words by inspection of character types prior to looking up endings of unidentified words in said table.

14. Apparatus according to claim 13, wherein said means for identifying by inspection of character types includes means for identifying at least one of abbreviations, proper nouns, cardinals, ordinals and alphanumeric names.