MULTI-TAPE TWO-LEVEL MORPHOLOGY: A Case Study in Semitic Non-linear Morphology

MULTI-TAPE TWO-LEVEL MORPHOLOGY: A Case Study in Semitic Non-linear Morphology George Anton Kiraz* COMPUTER LABORATO1;:Y, UNIVEIlSITY OF CAMP, RIDCI,; (St John's Colh!ge) E-mail. George. Kiraz~cl. cam. ac. uk April 22, 199d Abstract This I)aper presents an implemented multi-tal)e two- level model capable of describing Semitie non-linear morphology. The computational fl'arnework behind the ettrrcnt work is motivated by [Kay 1987]; the fimnal- ism presented here is an extension to the formalism re- ported by [Puhnan art(1 Hepl)le. 1993]. The objectives of the current work are: to stay as close as possible, in spirit, to standard two-level morl)hology, to stay close to the linguistic description of Semitic stems, and to present a model which can be used with ease by the Semitist. The. Imper illustrates that if finite-state transducers (FSTs) in a standard two-level morphology model are replaced with multi-tape attxiliary versions (AFSTs), one can account for Semitic root-andq)attern morphology using high level notation. 1 INTRODUCTION This paper aims at presenting a computational mor- phology model which can handle the non-linear phenomenon of Semitic morphology. The ap- proach presented here builds on two-level mori)hology [Koskennienfi 1983], extending it to achieve the desired objective. Tit('. contril)ution of this l)almr tnay ])e Slllll- marised as follows: With regards to the two-level model, we extend this model by allowing it to have multiI)le tapes on the lex- ical level and retaining the one tape on the surface level; hence, 'multi-tape two-level morphology'. Feasi- ble pairs in the standard two-level model become 'fea- sible tuple pairs' in our multi-tape model. With regards to the formalism, we have. chosen a twodevel formalism and extended it to be al)le to write multi-tape two-level grammars which involve non-linear operations. To achieve this, we made all lexieal expressions n-tuple regular expressions. In ad- dition, we introduced the notion of 'ellipsis', which in- *Supported by a Benefitctor Studentship from SI+ Jolm's Col- lege. q~llis research was done tllld(!r the SUlmrvision <ff I)r Steph(!n G. Pulman whom I thank fro' guidance, support and feedback. q'hanks to 13r ,Iohn Carroll for editm'ial comments, Arturo 'lh'u- jillo for useful 'chats' ;rod Tanya Bow(h!n for Prolog tips, dicates the (optional) omission from left-context lexical e×I)ressions of tui)les; this accounts for spr(~a(ling. Two-level implementations either work directly on rules or compile rules into FSTs. For the latte.r cats(:, we propose, an au×iliary finite-state transduce.r into which multi-tape two-level rules can be co)replied. Tit(.' ma- chine scans %Ulfle imirs ' instead of pairs of symbols. 'Fhe outline of the paper is as follows: Sect;ion 2 in- troduces the root-and-pattern nature of Semitic roof phology. Section 3 provides a review of the previous prol)osals iBr han(lling Semitie morphology. Section 4 t)resents our proposal, extending two-level morphology anti l)roposing a formalism which is adequate, for writ- ing non-linear grammars using high level notation. Sex> tion 5 al)i)lies our model on the Arabic verb. Section 6 I)resents an auxiliary automaton into which multi-tape two-level rules can/)e compiled. Finally, section 7 giw;s eonchtding remarks. 2 1~10 ()T- A ND- PATTI,;I)~N MORPItOL- …

Content truncated to 500 words