KADLEC, Vladimír and Pavel SMRŽ. How Many Dots Are Really Needed for Head-Driven Chart Parsing? Lecture Notes in Artificial Intelligence. Berlin: Springer, 2006, 3831/2006, No 1, p. 483-492, 9 pp. ISSN 0302-9743.
Other formats:   BibTeX LaTeX RIS
Basic information
Original name How Many Dots Are Really Needed for Head-Driven Chart Parsing?
Name in Czech Kolik skutečně potřebujeme teček pro syntaktickou analýzu řízenou hlavou pravidla?
Authors KADLEC, Vladimír (203 Czech Republic, guarantor) and Pavel SMRŽ (203 Czech Republic).
Edition Lecture Notes in Artificial Intelligence, Berlin, Springer, 2006, 0302-9743.
Other information
Original language English
Type of outcome Article in a journal
Field of Study 10201 Computer sciences, information science, bioinformatics
Country of publisher Czech Republic
Confidentiality degree is not subject to a state or trade secret
Impact factor Impact factor: 0.302 in 2005
RIV identification code RIV/00216224:14330/06:00015577
Organization unit Faculty of Informatics
UT WoS 000235805500046
Keywords in English nlp; CFG; parsing
Tags CFG, NLP, parsing
Changed by Changed by: RNDr. Vladimír Kadlec, Ph.D., učo 3541. Changed: 3/2/2006 15:21.
Abstract
This paper presents an improved form of head-driven chart parser that is appropriate for large context-free grammars. The basic method --- HDddm (Head-Driven dependent dot move) --- is introduced first. Both variants that improve the basic approach are based on the same idea --- to reduce the number of chart edges by modifying the form of items (dotted rules). The first one ``unifies'' the items that share the analyzed part of the relevant rule (thus, only one dot is needed to mark the position before and after the covered part). The second method applies the inverse strategy, it ``eliminates'' the parts that have not been covered yet (no dot needed). All the discussed alternatives are described in the form of parsing schemata. We also shortly mention a tricky technique (employing a special trie-like data structure developed originally for Scrabble) that enables to minimize the extra information needed in the algorithms. We demonstrate the advantages of the described methods by the significant decreases in the number of edges for charts. The results are given for the standard set of testing grammars (and respective inputs) as well as for a large and highly ambiguous Czech grammar.
Abstract (in Czech)
V článku je prezentován vylepšený algoritmus syntaktické analýzy pomocí bezkontextových gramatik řízený hlavou pravidla. Nejdříve je popsána základní metoda -- HDddm. Obě další varianty, které vylepšují rychlost této základní metody jsou založeny na podobné myšlence -- redukovat množství generovaných hran tabulkového analyzátoru pomocí modifikace položek (pravidel s tečkami). První metoda sjednocuje položky, které mají stejnou již analyzovanou část pravidla. Druhá varianta naopak eliminuje tu část položky, která již byla analyzována. Jako datová struktura pro algoritmus je navržena struktura původně určená pro implementaci hry Scrabble. Výhody popsaných technik jsou demonstrovány na standardních testovacích datech pro angličtinu a také na vysoce nejednoznažné bezkontextové gramatice češtiny.
Links
GA201/05/2781, research and development projectName: Překlad českých vět do konstrukcí transparentní intenzionální logiky
Investor: Czech Science Foundation, Translation of Czech Sentences to Transparent Intensional Logic Constructions
PrintDisplayed: 26/7/2024 16:37