Learning Hebrew Roots: Machine Learning with Linguistic Constraints

Learning Hebrew Roots: Machine Learning with Linguistic Constraints Ezra Daya Dept. of Computer Science University of Haifa 31905 Haifa Israel edaya@cs.haifa.ac.il Dan Roth Dept. of Computer Science University of Illinois Urbana, IL 61801 USA danr@cs.uiuc.edu Shuly Wintner Dept. of Computer Science University of Haifa 31905 Haifa Israel shuly@cs.haifa.ac.il Abstract The morphology of Semitic languages is unique in the sense that the major word-formation mechanism is an inherently non-concatenative process of inter- digitation, whereby two morphemes, a root and a pattern, are interwoven. Identifying the root of a given word in a Semitic language is an important task, in some cases a crucial part of morphological analysis. It is also a non-trivial task, which many humans find challenging. We present a machine learning approach to the problem of extracting roots of Hebrew words. Given the large number of po- tential roots (thousands), we address the problem as one of combining several classifiers, each predict- ing the value of one of the root’s consonants. We show that when these predictors are combined by enforcing some fairly simple linguistics constraints, high accuracy, which compares favorably with hu- man performance on this task, can be achieved. 1 Introduction The standard account of word-formation processes in Semitic languages describes words as combina- tions of two morphemes: a root and a pattern.1 The root consists of consonants only, by default three (although longer roots are known), called radicals. The pattern is a combination of vowels and, possi- bly, consonants too, with ‘slots’ into which the root consonants can be inserted. Words are created by interdigitating roots into patterns: the first radical is inserted into the first consonantal slot of the pattern, the second radical fills the second slot and the third fills the last slot. See Shimron (2003) for a survey. Identifying the root of a given word is an im- portant task. Although existing morphological an- alyzers for Hebrew only provide a lexeme (which is a combination of a root and a pattern), for other Semitic languages, notably Arabic, the root is an essential part of any morphological analysis sim- 1An additional morpheme, vocalization, is used to abstract the pattern further; for the present purposes, this distinction is irrelevant. ply because traditional dictionaries are organized by root, rather than by lexeme. Furthermore, roots are known to carry some meaning, albeit vague. We be- lieve that this information can be useful for compu- tational applications and are currently experiment- ing with the benefits of using root and pattern infor- mation for automating the construction of a Word- Net for Hebrew. We present a machine learning approach, aug- mented by limited linguistic knowledge, to the prob- lem of identifying the roots of Hebrew words. To the best of our knowledge, this is the first appli- cation of machine learning to this problem. While there exist programs which can extract the root of words in Arabic (Beesley, 1998a; Beesley, 1998b) and Hebrew (Choueka, 1990), they are all depen- dent on labor intensive construction of large-scale lexicons which are …

Content truncated to 500 words