Search

An extension of the problem of reconstruction of words from a given multiset of its subwords supposedly generated by unit shifts of a window of fixed length along such words. In the extension, feasible solutions must satisfy additional constraints. The case is considered when these constraints are specified by forbidden words. A solution to the problem is obtained as a result of searching for Euler paths or Euler cycles in a de Bruijn multidigraph with the additional operation of reduction of its edges and subsequent application of special algebraic operations of multiplication of adjacency matrices as in the first part of this article. [ABSTRACT FROM AUTHOR]

Recent data from compound word processing suggests that compounds are recognized via their constituent lexemes (Juhasz, Starr, Inhoff, & Placke, 2003). The present lexical decision experiment manipulated orthogonally the frequency of the constituents of compound words in two languages: Basque and Spanish. Basque and Spanish diverge widely in their morphological properties and in the number of existing compound words. Furthermore, the head lexeme (i.e., the most meaningful lexeme related to the whole-word meaning) in Spanish tends to be the second lexeme, whereas in Basque the percentage is more distributed. Results showed a facilitative effect of the frequency of the second lexeme, in both Basque and Spanish compounds. Thus, both Basque and Spanish readers decompose compounds into their constituents for lexical access, and this decomposition is carried out in a language-independent and blind-to-semantics manner. We examine the implications of these results for models of lexical access. [ABSTRACT FROM AUTHOR]

The article refers to instructions that were given to authors by the editor in chief of the periodical. Publishers of the journal prefer the submission of manuscripts and figures in electronic form in addition to a hard-copy printout. Authors are instructed to label diskette properly, giving exact details on the names of the files, the operating system and software used. It informs that publishers prefer articles submitted in word processing packages such as MS Word, WordPerfect, etc. running under operating systems MS DOS, Windows and Apple Macintosh, or in the file format LaTeX.

The article focuses on validation studies, which represent an endangered species. Personnel Psychology stopped publishing the Validity Information Exchange in the mid 60's. Major research journals stopped accepting manuscripts describing orthodox validation studies because these studies did not seem to advance the field sufficiently to warrant the space they would take in the journal. Another cause for concern is the nature of work. It has changed substantially and will change still further in the next decade. Machine operator jobs have become machine-tending jobs. Secretarial positions have become word processing and data processing positions. The service industry has exploded; creating job titles that did not exist thirty years ago. Still another change is the nature of applicant populations. They are growing more diverse and this trend will continue through at least the next two decades since the aging work population is being replaced with a younger cadre reflecting greater proportions of women, blacks, Hispanics, and other traditionally underrepresented groups.

A correction to the article "Towards automatic bounding box annotations from weakly labeled images" that was published in the previous issue is presented.