Describes a method for automatically segmenting words into their stems and affixes using certain statistical properties of a corpus to indicate where words should be divided. (Author/PF)

(Text is in French.)

A SERIES OF SUBROUTINES (READRULE, ANNOTATE, CHOOSE, STORERULE, ADD-ONE AND SUB-ONE, DOMINATE, SUBTREE, TERMINAL, AND EQUAL) WRITTEN UP IN THE "COMIT" PROGRAMING LANGUAGE TO PROVIDE A CONVENIENT WAY FOR LINGUISTS TO TEST ANY SET OF GRAMMAR RULES AND TO GENERATE SENTENCES, ARE GIVEN WITH EXPLANATIONS FOR EACH. THIS IS FOLLOWED BY COMPLETE…

The greatest fallacy of word counts, the author maintains, lies in the fact that advocates of frequency lists stress the high percentage without telling the whole truth. It has become common to pretend that a frequency list of 3,000 words covers 95 percent of the language, that it enables a person to speak and understand a foreign language by…

JOSTRAN is a JOSS dialect that expedites the construction of FORTRAN programs. JOSS is an interactive, on-line computer system. JOSS language programs are list-processed; i.e., each statement is interpreted at execution time. FORTRAN is the principal language for programing digital computers to perform numerical calculations. The JOSS language…

This memorandum describes a set of subroutines for the IBM 7040/44 computers for reading textual material with complex formats and coding conventions--questionnaires, library catalog cards, etc.--from any external medium into the high-speed store of the machine. Different kinds of information in the input are recognized by explicit markers,…

Describes a model for a front end query processor to deal with incomplete information in a database. The effectiveness of the processor in restricting the number of objects to be processed in a query and in aiding the interpretation of a query is discussed. (26 references) (Author/CLB)

Describes an algorithm for the calculation of term discrimination values that may be used when the interdocument similarity measure used is the cosine coefficient and when the document representations have been weighted using one particular term weighting scheme. (7 references) (Author/CLB)

The implications of some major linguistic and semiotic concepts for the machine manipulation of natural language text are examined. It is proposed that the study of meaning should contribute to the development of sophisticated algorithms for machine indexing. (Author)

Discusses an investigation of certain problems concerning the structural design of lexicons used in computational approaches to natural language understanding. Emphasizes three aspects of design: retrieval of relevant portions of lexicals items, storage requirements, and representation of meaning in the lexicon. (Available from ALLC, Dr. Rex Last,…

Discusses a corpus-based study that focuses on word epenthesis.Provides quantitative evidence that vowel epenthesis applies categorically in some environments and variably in others. Probabilistic analysis demonstrates that the variable pattern is constrained by a complex interplay of linguistic factors. Interprets such grammatical conditioning as…

Looks at the value of language corpora for designing teaching materials and argues that they are an essential tool for teachers wishing to keep up to date with language as it is actually used in real life. Provides a number of useful and practical teaching examples for designing teaching tasks and for raising students' awareness to the differences…

Discusses the use of mechanical validation of formal, combinatorial theories in phonetics, and uses a computational tool in analyzing the implementation of the Charm and Government theory of phonology. (GLR)

Examines linguistic issues involved in designing a computer concordance program for Spanish that automatically retrieves in context all forms of the verb from the infinitive. Results indicate that, although concordances remain popular with researchers in the humanities, their lack of language-specific morphological information is a detraction.…

Reports on a project showing that relational data models, including unnormalized models that allow the nesting of relations, cannot fully capture the structural properties of lexical information. A feature-based model that allows for a full representation of some nesting and defines a factoring mechanism is described and demonstrated. (38…