Graham, S., Harrison, M., and Ruzzo, W. (1980). An improved context-free recognizer. ACM Transactions on Programming Languages and Systems, 2(3):415--462.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....well suitedtoEarleyparsing. In our description of Earley parsing up to now, we have not mentioned the use of automata. However, the idea to use efficient deterministic automata as the Earley s algorithm recognizes that B derives the empty string with a series of PREDICTOR and COMPLETER steps [16]. #. S. E E. a a. A . #. #. #. AAAA. a a a E E 1 23 4 5 6 7 FIGURE 4. LR(0) automaton for the grammar in Figure 2. Shading indicates the start state. #. S. E. AAAA. S. #. E. ....

....[18] and Earley like parsers [19, 20, 21] Conceptually, this has the effect of precomputing groups of Earley items which must appear together in an Earley set, thus reducing the amount of work the Earley algorithm must perform at parse time. We note that the parsing algorithm described in [16] is related to an Earley parser and employs precomputation as well. Figure 4 shows the LR(0) automaton for the grammar in Figure 2. The construction of LR(0) automata is treated in most compiler texts [7] and is not especially pertinent to this discussion. Each LR(0) automaton state consists ....

Graham, S. L., Harrison, M. A. and Ruzzo, W. L. (1980) An improved context-free recognizer. ACM Trans. Program. Languages Syst., 2, 415--462.

....was proposed by Ruzzo [Ruz80] This recognizer, however, as that presented in [GR88] works only for grammars in Chomsky Normal Form. This restriction makes hard to apply the algorithm in compiler design since grammars in Chomsky Normal Form do not reflect the semantic structure of languages (see [GHR80] for a discussion on disadvantage of grammars in Chomsky Normal Form) This problem has been recently solved by Bruschi and Pighizzini [BP92] that proposed the first parallel parser working with every context free grammar. In this paper we show how the algorithm presented in [BP92] can be ....

S. Graham, M. Harrison and W. Ruzzo,"An improved context--free recognizer," ACM Transactions 2(3) : 415--462, 1980.

....different parsing algorithms. While we used the basic examples from the Shieber et al. article, one can also implement all the different deduction schemes from Sikkel (1997) This also includes advanced algorithms such as left corner or head corner parsing, the refined Earley algorithm proposed by Graham et al. 1980), or (unification based) ID LP parsing as defined in Morawietz (1995) or any improved version of any of these. Furthermore, because of the logical semantics of CHR with their soundness and completeness, all correctness and soundness proofs for the algorithms can be directly applied to this ....

Graham, G., Harrison, M. G. and Ruzzo, W. L. (1980). An improved context--free recognizer, ACM Transactions on Programming Languages and Systems 2 (3).

....wrapping may mention these either as requirements or as products. We use a variant of the Cocke Kasami Younger parsing algorithm [3] as modified by V. Pratt for LINGOL [69] and extended by C. Landauer for METER [38] 55] to use Early algorithm type pre filtering of the WKB entries as rules (see [27] for a description and complexity analysis of a similar combination parser, 35] for a general description of this style of parser, called a chart parser) This 9 parser matches both goals and available information simultaneously (both bottom up and top down) which is both forward and backward ....

Susan L. Graham, Michael A. Harrison, Walter L. Ruzzo, "An Improved Context-Free Recognizer", ACM Transactions on Programming Languages and Systems, Volume 2, No. 3, pp. 415-462 (July 1980)

....from the choice of paths through the automata, grammatical structure also helps determine the highest weighted parse. 4 Cubic Time Parsing This section begins by reviewing the general idea of chart parsing, presenting a general method drawn from context free dotted rule methods such as [Graham et al. 1980, Earley 1970] Second, we will see why this method is inecient when applied to bilexical grammars in the obvious way. Finally, a more ecient (cubictime) algorithm is presented, which applies the general chart parsing method rather di erently. 4.1 Generalized Chart Parsing Method The input is a ....

Graham, S.L., Harrison, M.A. and Ruzzo, W.L. An Improved Context-Free Recognizer. ACM Transactions on Prog. Languages and Systems 2(3):415-463.

....can be used with almost no modification. The standard polynomial time context free parsing algorithms that are most frequently used are essentially variations of the Earley top down parsing algorithms (Earley, 1986; Stolcke, 1995; Nederhof Satta, 1997) or of the CYK bottomup parsing algorithm (Graham et al. 1980; Nederhof, 1994) with more or less improvement for the coupling with a speech recognizer (Thomason, 1986; Fred Leito, 1993; Jurafksy et al. 1995) 5 which in that case would be the number of frames, one per 10 ms. 6 perplexity is value quite well inversely correlated with recognition ....

GRAHAM S. L., HARRISON M. A. & RUZZO W. L. (1980). An improved context-free recognizer. ACM Transactions on Programming Languages and Systems, 2(3), 415--462.

....and for which any parse in the original grammar would be reconstructible from a parse in the CNF grammar. In short, we can, without loss of generality, assume that the SCFGs we are dealing with are in CNF. In fact, our algorithm generalizes straightforwardly to the more general Canonical Two Form (Graham et al. 1980) format, and in the case of bigrams (n = 2) it can even be modified to work directly for 2 Gamma Gamma Gamma Gamma Omega Omega Omega Omega Omega Omega Gamma J J J JJ Omega Omega Omega Omega Omega Omega Gamma J J J J J . Y Z X w ....

Graham, Susan L., Michael A. Harrison, & Walter L. Ruzzo. 1980. An improved contextfree recognizer. ACM Transactions on Programming Languages and Systems 2.415--462.

....and part of speech (just assume that all numbers have the same stem) The P lex parameters therefore have to be estimated from training data like other parameters. 6 Chapter 3 Parsing LoPar is an implementation of the well known left corner parsing algorithm for context free grammars (see e.g. [Graham et al. 1980]) The chart is organised as a two dimensional array of lists. Each list is a linked list of edges. Each edge contains information about its start position, the grammar rule and the dot position (the position after the last daughter constituent which is covered by the edge) The first dimension ....

Graham, S., Harrison, M., and Ruzzo, W. (1980). An improved context-free recognizer. ACM Transactions on Programming Languages and Systems, 2(3):415--462.

....grammar are found. ffl The parser uses subsumption checking to reduce the size of the chart. Essentially, an edge is not added to the chart if it is less general than a preexisting edge, and preexisting edges are removed from the chart if the new edge is more general. ffl The parser is on line (Graham, Harrison, and Russo, 1980), essentially meaning that all edges that end at position i are constructed before any that end at position i 1. This feature is particularly desirable if the final architecture of the speech understanding system couples Gemini tightly with the speech recognizer, since it guarantees for any ....

Graham, S., Harrison, M., and Ruzzo, W. (1980). "An Improved Context-Free Recognizer ", ACM Transactions on Programming Languages and Systems, Vol. 2, No. 3, pp. 415462.

....The search algorithm can then tabulate each sub derivation and reuse it in building any derivation that shares that sub derivation. 3 Such tabular algorithms are widely used in parsing and language modeling with appropriate kinds of grammars (Younger, 1967; Kay, 1986; Earley, 1970; Lang, 1974; Graham, Harrison, et al. 1980; Tomita, 1987) because they support exhaustive search algorithms with polynomial space and time with respect to sentence length. Furthermore, tabular algorithms can be readily extended to dynamic programming algorithms to search for optimal derivations with respect to appropriate evaluation ....

Graham, S. L., Harrison, M. A., and Ruzzo, W. L. (1980). An improved context-free recognizer. ACM Transactions on Programming Languages and Systems, 2(3):415--462.

....accruing from the choice of paths through the automata, grammatical structure also helps determine the highest weighted parse. 4 Cubic Time Parsing This section begins by reviewing the general idea of chart parsing, presenting a general method drawn from context free dotted rule methods such as [Graham et al. 1980, Earley 1970] Second, we will see why this method is inefficient when applied to bilexical grammars in the obvious way. Finally, a more efficient (cubictime) algorithm is presented, which applies the general chart parsing method rather differently. 4.1 Generalized Chart Parsing Method The ....

Graham, S.L., Harrison, M.A. and Ruzzo, W.L. An Improved Context-Free Recognizer. ACM Transactions on Prog. Languages and Systems 2(3):415-463.

....the field of computational linguistics. The standard methods for CFG parsing are the CKY algorithm (Kasami, 1965; Younger, 1967) and Earley s algorithm (Earley, 1970) both of which have a worst case running time of O(gN 3 ) for a CFG (in Chomsky normal form) of size g and a string of length N . Graham et al. 1980) give a variant of Earley s algorithm which runs in time O(gN 3 = log N ) Valiant s parsing method is the asymptotically fastest known (Valiant, 1975) It uses Boolean matrix multiplication (BMM) to speed up the dynamic programming in the CKY algorithm: its worst case running time is O(gM (N ....

Graham, Susan L., Michael A. Harrison, and Walter L. Ruzzo. 1980. An improved context-free recognizer. ACM Transactions on Programming Languages and Systems, 2(3):415--462.

....analyses (or partial analyses) for further sentences. A similar bootstrapping approach is found in [Simmons 1990] The precise grammatical theory we use to hand analyze sentences should not be crucial, so long as it is applied consistently and is not unduly large. Parsing Algorithms Following [Graham et al. 1980], we adopt the following notation. An arbitrary context free grammar is given by G = V; Sigma; P; S) where V is the vocabulary of all symbols, Sigma is the set of terminal symbols, P is the set of rewrite rules, and S is the start symbol. For an input sentence w = a 1 a 2 : an , let w i;j ....

....the substring a i 1 : a j and w i = w 0;i denote the prefix of length i. We use Greek letters (ff; fi; to denote symbol strings in V . Tabular dynamic programming algorithms are the methods of choice for ordinary context free recognition [Cocke and Schwartz 1970, Earley 1970, Graham et al. 1980] Each entry t i;j in a table or chart, t, holds a set of symbols or rules that match w i;j . A symbol A matches w i;j if A ) w i;j . Some of these methods use dotted rules to represent progress in matching the input. For all A fffi in P , A ff Delta fi is a dotted rule of G. The dotted ....

[Article contains additional citation context not shown here]

Graham, S.L., Harrison, M.A. and Ruzzo, W.L. "An Improved Context-Free Recognizer, " ACM Transactions on Programming Languages and Systems 2 (3). 1980, pp. 415-463.

....far, some of which are due specifically to the probabilistic aspects of parsing. This section summarizes the necessary modifications to process null productions correctly, using the previous description as a baseline. Our treatment of null productions follows the (non probabilistic) formulation of Graham et al. 1980), rather than the original one in Earley (1970) 4.7.1 Computing ffl expansion probabilities The main problem with null productions is that they allow multiple prediction completion cycles inbetween scanning steps (since null productions do not have to be matched against one or more input ....

....idea in the probabilistic formulation of Earley s algorithm is the collapsing of recursive predictions and unit completion chains, replacing both with lookups in precomputed matrices. This idea arises in our formulation out of the need to compute probability sums given as infinite series. Graham et al. 1980) use a non probabilistic version of the same technique to create a highly optimized Earley like parser for general CFGs that implements prediction and completion by operations on Boolean matrices. 19 The matrix inversion method for dealing with left recursive prediction is borrowed from the LRI ....

Graham, Susan L., Michael A. Harrison, & Walter L. Ruzzo. 1980. An improved contextfree recognizer. ACM Transactions on Programming Languages and Systems 2.415--462.

....same state. CHAPTER 6. EFFICIENT PARSING WITH STOCHASTIC CONTEXT FREE GRAMMARS 141 section summarizes the necessary modifications to process null productions correctly, using the previous description as a baseline. Our treatment of null productions follows the (non probabilistic) formulation of Graham et al. 1980), rather than the original one in Earley (1970) 6.4.7.1 Computing ffl expansion probabilities The main problem with null productions is that they allow multiple prediction completion cycles in between the scanning steps (since null productions do not have to be matched against one or more input ....

....WITH STOCHASTIC CONTEXT FREE GRAMMARS 158 6.6 Implementation Issues This section briefly discusses some of the experience gained from implementing the probabilistic Earley parser. Implementation is mainly straightforward and many of the standard techniques for context free grammars can be used (Graham et al. 1980). However, some aspects are unique due to the addition of probabilities. 6.6.1 Prediction Due to the collapsing of transitive predictions, this step can be implemented in a very efficient and straightforward manner. As explained in Section 6.4.5, one has to perform a single pass over the current ....

[Article contains additional citation context not shown here]

GRAHAM, SUSAN L., MICHAEL A. HARRISON, & WALTER L. RUZZO. 1980. An improved context-free recognizer. ACM Transactions on Programming Languages and Systems 2.415--462.

....kind of structure they produce to represent all parses of the analyzed sentence is an essential characteristic of these algorithms. Some of the published algorithms produce only a chart as described by Kay in [14] which only associates nonterminal categories to segments of the analyzed sentence [11, 39, 13, 3, 9], and which thus still requires non trivial processing to extract parse trees [26] The worst size complexity of such a 1 Ambiguity may also have a semantical origin. chart is only a square function of the size of the input 2 . However, practical parsing algorithms will often produce a more ....

Graham, S.L.; Harrison, M.A.; and Ruzzo W.L. 1980 An Improved Context-Free Recognizer. ACM Transactions on Programming Languages and Systems 2(3): 415-462.

....With appropriate care, it is also possible 2 Theoretically faster algorithms [29, 7] can achieve O(n 2:496 ) but with an unacceptable constant factor. Note also that we do not require the grammar to be in Chomsky Normal Form. 3 And unlike tabular algorithms such as Cocke YoungerKasami s [13, 15, 30, 11]. to read this output grammar as a shared forest (see appendix A) However its meaningful interpretation as a shared forest is dependent on the parsing schema (cf. 12, 16] used in constructing the PDT that produces it as output. Good definition and understanding of shared forests is essential ....

Graham, S.L.; Harrison, M.A.; and Ruzzo W.L. 1980 An Improved Context-Free Recognizer. ACM Transactions on Programming Languages and Systems 2(3): 415-462.

....information is extracted from the grammar in order to predict in a top down manner which rules could be applied at any given point during a parse, and eliminate from consideration others which could not possibly form part of a complete analysis at that point, given what has already been analysed. Graham, Harrison Ruzzo (1980) present a predictive recogniser that is derived from Earley s in fact they state that the CKY and Earley methods are almost identical [to it] but claim their recogniser to be conceptually simpler. Graham et al. describe versions of the recogniser which theoretically run in less than O(n ....

Graham, S., M. Harrison & W. Ruzzo (1980) "An improved context-free recognizer." ACM Transactions on Programming Languages and Systems, 2(3), 415--462.

....presented here is a modification of Earley s algorithm (Earley 1970) along the lines of Shieber (1984) and Seiffert (1991) The paper by Earley describes a parser and a recognizer. Since there exist algorithms for extracting all parse trees from the parse lists with some encoding (for example Graham, Harrison and Ruzzo (1980)) we can abstract away from the distinction and can view the recognizer in some sense as a parser. Strictly speaking, we just present a recognizer, but this distinction is somewhat irrelevant with unification based systems. Since the concept of a context free grammar is augmented to a TU ID LP ....

Graham, G., Harrison, M. G. and Ruzzo, W. L. (1980). An Improved Context-- Free Recognizer, ACM Transactions on Programming Languages and Systems 2 (3), pp. 415--462.

....turned off, its syntactic accuracy rate drops from 99 to 66 , and it runs far more slowly. The KITSS Algorithm The KITSS parsing algorithm (given as Algorithm 1 in Appendix A) is a variant of tabular or chart parsing methods for context free languages [Cocke and Schwartz 1970, Earley 1970, Graham et al. 1980] It scans the sentence from left to right, assembling possible partial interpretations of the sentence; but it continually discards interpretations that are statistically unlikely. The grammar rules and statistics are generated automatically by training on a bracketed corpus. The grammar is ....

Graham, S.L., Harrison, M.A. and Ruzzo, W.L. 1980. An Improved Context-Free Recognizer. ACM Transactions on Programming Languages and Systems 2(3):415-463.

No context found.

Graham, S., Harrison, M., and Ruzzo, W. (1980). An improved context-free recognizer. ACM Transactions on Programming Languages and Systems, 2(3):415--462.

No context found.

Graham, S.L., M.A. Harrison and W.L. Ruzzo 1980. An improved context-free recognizer, ACM trans. actio o Progr. Languages and Systems 2:415-

No context found.

GRAHAM,SUSAN L., MICHAEL A. HARRISON,&WALTER L. RUZZO. 1980. An improved context-free recognizer. ACM Transactions on Programming Languages and Systems 2.415--462.

No context found.

Susan L. Graham, Michael A. Harrison, and Walter L. Ruzzo. 1980. An improved contextfree recognizer. ACM Transactions on Programming Languages and Systems, 2(3):415-- 462, July.

No context found.

S. Graham, M. Harrison, and W. Ruzzo. "An Improved Context-Free Recognizer." ACM TOPLAS 2,3 (July 1980), 415--462.

First 50 documents

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