W.A. Woods. Experimental parsing system for transi- tion network grammar. In Natural Language Process ing, Algorithmic Press, 1971.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... most purely linguistic area, coordination has not received in depth treatment, if not recently (for example, see [Sag et al. 1984] and [Kaplan and Maxwell, 1988] Until the beginning of the 80 s almost no system (the most relevant exception being the SYSCONJ module, present in the LUNAR system [Woods, 1973]) has confronted coordination in a generalized manner. The 80 s have seen renewed computational interest in coordination that has brought new efforts (see [Kwasny and Sondheimer, 1981] Dahl and McCord, 1983] Font and Berwick, 1985] Lesmo and Torasso, 1985] Kosy, 1986] and [ProudJan and ....

William A. Woods. An Experimental Parsing System for Transition Network Grammars. In: R. Rustin ted). Natural Language Processing. Prentice-Hall, Englewood Clifl, New Jersey. 1973. 6 189

....but it appears that a proper and general treatment must recognize coordination as a metagrammatical construction, in the sense that metarule, general system operations, or second pass operations such as transformations, are needed for its formulation. Early attempts at such a general treatment [39, 4] were inefficient due to combinatorial explosion. A logic grammar rendition of coordination in terms of logic grammars [8] solved these inefficiencies through the addition of a semantic interpretation component that produced a logical form from the output of the parser and dealt with scoping ....

W.A. Woods. An experimental parsing system for transition network grammars. In Natural Language Processing, Rustin R. (ed.), pages 145--149, 1973.

....when the NP is already a complete NP. A grammar for a natural language can be represented in different ways. When represented simply as a set of rewrite rules, a grammar is suitable for a top down analysis as done by an augmented transition network (ATN) Wanner and Maratsos, 1978; Woods, 1970; Woods, 1973] However, top down processing forces a sentence processor to make unnecessary commitments along the way with a consequent problem of frequent and wasteful backtracking that does not happen in human sentence processing behavior. One can perform a purely bottom up processing with a grammar ....

W. A. Woods. An experimental parsing system for transition network grammars. In R. Rustin, editor, Natural language processing. Algorithmics Press, 1973.

.... For example, the backtracking models embodied the theory that only a single syntactic interpretation need be maintained at any given time, so long as the processor could keep track of its decisions, undo them when an erroneous decision was discovered, and then reinterpret the input (e.g. Woods, 1973). The lookahead parsers tried to sidestep the problems inherent in backtracking by postponing any decision until enough input had been processed to guarantee a correct decision, thereby avoiding erroneous decisions to some extent (e.g. Marcus, 1980) Another approach to avoiding erroneous ....

Woods, W. A. 1973. An experimental parsing system for transition network grammars. In Rustin, R. ed. Natural language processing. New York: Algorithmics Press.

....layer , and so on. Generally to derive the theorems of the next layer , at least one theorem produced at the previous stage must be used. This process terminates when no more new theorems can be generated. 2. 3 Coordination Early work on coordination proposed meta grammatical treatments (e.g. [12, 3]) in which the appearance of a coordinating word, or conjunction (e.g. and , or , but ) is treated as a demon. When a conjunction appears in a sentence of the form A X conj Y B a process is triggered in which backing up is done in the parse history in order to parse Y parallel to X, and B is ....

W. Woods. An experimental parsing system for transition network grammars. In R. Rustin, editor, Natural Language Processing, pages 145--149. Algorithmic Press, New York, 1973.

....now consider processing strategies, where syntactic structure of shared material is characterised more indirectly by the state of the parser. PROCESSING STRATEGIES There have been several attempts to treat coordination by adapting pre existing parsing strategies. For example, ATNs were adapted by Woods (1973), DCGs by Dahl and McCord (1983) and chart parsers by Haugeneder (1992) Woods and Dahl McCord s system are similar. Haugeneder s system has very limited coverage. In Wood s SYSCONJ system, the parser can back up to various points in the history of the parse, and parse the second conjunct ....

Woods, W. (1973). An Experimental Parsing System for Transition Network Grammars. In R. Rustin, Ed., Natural Language Processing, Algorithmics Press, New York.

....initial commitments without any information in support of the commitments and later backtracks until its blind commitments turn out to be the right ones. Such a top down parser is used in models of language processing based on the Augmented Transition Network (ATN) model (e.g. the LUNAR system by Woods, 1973). The five principles above dictate that a model of language processing must make commitments to particular interpretations, make those commitments incrementally in processing the input, make commitments only when there is some knowledge that justifies the commitments, make the commitments as soon ....

....left unspecified in this model. Many early models were sequential models based on the augmented transition network (ATN) formalism. These models employed an ATN parser to perform syntactic analysis and fed the output of the ATN to a semantic processor. Examples of ATN based systems such as LUNAR (Woods, 1970; 1973) and SHRDLU (Winograd, 1973) are well known landmarks in the early history of artificial intelligence and computational modeling of natural language processing. Though some models including SHRDLU interleaved syntactic and semantic processing, the architecture was still sequential with semantic ....

Woods, W. A. (1973). An experimental parsing system for transition network grammars. In Rustin, R., editor, Natural language processing. Algorithmics Press.

....but it appears that a proper and general treatment must recognize coordination as a metagrammatical construction, in the sense that metarule, general system operations, or second pass operations such as transformations, are needed for its formulation. Early attempts at such a general treatment [29, 4] were inefficient due to combinatorial explosion. A logic grammar rendition of coordination in terms of logic grammars [9] solved these inefficiencies through the addition of a semantic interpretation component that produced a logical form from the output of the parser and dealt with scoping ....

W.A. Woods. An experimental parsing system for transition network grammars. In Natural Language Processing, Rustin R. (ed.), pages 145--149, 1973.

....but it appears that a proper and general treatment must recognize coordination as a metagrammatical construction, in the sense that metarule, general system operations, or second pass operations such as transformations, are needed for its formulation. Early attempts at such a general treatment [23, 3] were inefficient due to combinatorial explosion. A logic grammar rendition of coordination in terms of logic grammars [5] solved these inefficiencies through the addition of a semantic interpretation component that produced a logical form from the output of the parser and dealt with scoping ....

W.A. Woods. An experimental parsing system for transition network grammars. In Natural Language Processing, Rustin R. (ed.), pages 145--149, 1973.

....the tutorial session, plan pedagogical strategies and tactics, and determine natural language utterances. The discourse management network has three successive levels: the tutoring approach, the strategy, and the tactics. These are organized into states of an augmented transition network (atn) [62], with additional meta rules which can affect the transitions [63] In a nutshell, an atn is a finite state machine in which nodes and arcs can store information in global registers for later use, and arcs are dynamically activated or deactivated based on these register values and other input. ....

W.A. Woods, "An Experimental Parsing System for Transition Network Grammars," in Natural Language Processing (ed. Rustin). New York: Algorithmics Press, 1972, pp. 113154.

....and only if there is some proof that it performs a transition between some suitable initial and final This research was supported by an SERC research fellowship. states. It is worth noting at this early stage that dynamic grammars are not lexicalised rehashes of Augmented Transition Networks (Woods, 1973). ATNs use a finite number of states combined with a recursion mechanism, and act essentially in the same way as a top down parser. They are not particularly suited to incremental interpretation. To get an idea of how logics (instead of the more usual algebras) can be used to specify dynamic ....

Woods, W. (1973) An Experimental Parsing System for Transition Network Grammars. In Rustin, R. (ed.), Natural Language Processing, Algorithmics Press, New York.

....difficult to encode in a purely lexicalist framework. As an example, it briefly describes an analysis of English non constituent coordination. The analysis has much in common with both grammar based accounts of coordination (e.g. Gazdar 1981, Steedman 1985) and procedural accounts such as SYSCONJ (Woods 1973). 2. DYNAMIC SPECIFICATIONS This section provides a formal notation for describing the dynamics of a process i.e. the states and the possible transitions between states. The notation is designed to enable modelling of the transitions between abstract characterisations of mental states during ....

....label, which is chosen from a finite set of symbols. For PDAs, each label is a pair consisting of an element from a finite set, and a list of symbols of arbitrary length (the stack) The symbols on the stack are also chosen from a finite set. In the closely related Recursive Transitions Networks (Woods 1973), each element on the stack is a syntactic category (e.g. s, np or vp) For TMs, each label is a pair of an element from a finite set, and a tape, unbounded to left and right (the tape can be replaced by two arbitrary length lists) Although TMs have the most complex state labels of the three ....

[Article contains additional citation context not shown here]

Woods, W.: 1973, `An Experimental Parsing System for Transition Network Grammars,' in Rustin, R. (ed.), Natural Language Processing, Algorithmics Press, New York.

No context found.

W.A. Woods. Experimental parsing system for transi- tion network grammar. In Natural Language Process ing, Algorithmic Press, 1971.

No context found.

Woods, W.A. An Experimental parsing System for Transition Network Grammars. In Rustm (editor), Natural Language Processing. Algorithmic Press, 1973.

No context found.

Woods, W.A. [1970] An experimental parsing system for transition network grammars. In R.Rustin (Ed.) Natural Language Processing. Prentice Hall: Englewood Cliffs, New Jersey.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC