Natural language parser
View Patent ↗The present invention provides a method and a parser for syntactically analyzing an input string. The parser applies a plurality of rules which describe syntactic properties of the language of the input strings. The plurality of rules comprise two types of rules. A first type of rules comprises immediate dominance rules and linear precedence rules. A second type of rules being sequence rules. All rules of the plurality of rules are applied according to a predefined order to the input string. This new incremental parsing architecture has advantages with respect to grammar engineering and allows a more efficient parsing.
1. A method for parsing an input string that includes elements by applying a plurality of rules which describe syntactic properties of the language of the input string, the plurality of rules comprises two types of rules,
a first type of rules consisting of immediate dominance rules, which define dominance relations between constituents of the grammar rules, and linear precedence rules, which define an order of grammar constituents, and
a second type of rules defining dominance and precedence relations of grammar constituents in a fixed sequence order, wherein the rules of said plurality of rules are applied according to a predefined rule order, and said plurality of rules is distributed into layers that are applied in a predefined layer order, with each layer containing only the first type or the second type of rules.
2. The method of claim 1 wherein an output of a layer of said layers is an input of a subsequent layer, and said predefined layer order applies constraint phenomena in increasing complexity.
3. The method of claim 1 wherein a rule of said plurality of rules is further defined by at least one context feature.
4. The method of claim 1 wherein an immediate dominance rule further defines the first constituent and the last constituent of a rule.
5. The method of claim 1 wherein a rule of said second type may include a constituent definition representing any constituent.
6. The method of claim 5 wherein a rule of said second type may further use a definition of replacing the longest sequence of constituents between the preceding and subsequent constituent.
7. The method of claim 5 wherein the definition of any constituent may be further defined by a context feature.
8. The method of claim 1 wherein a rule of said second type may be further defined by a disjunction of constituents.
9. The method of claim 1 wherein a processed input string is represented by a sequence of sub-trees.
10. The method of claim 9 wherein each successful application of a rule results in a combination of sub-trees.
11. The method of claim 1 wherein the rule applied first combines the largest numbers of constituents.
12. The method of claim 1 wherein the order of constituent definitions in a rule of said second type is based on features of the constituents.
13. The method of claim 1 wherein a given constituent occurs in any number of times.
14. The method of claim 1 wherein the second type of rules include a symbol that represents any category for a set of the elements at a particular position, and the symbol is restricted by a feature-value constraint.
15. The method of claim 14 , further including an operator to replace a longest sequence between first and second phrases.
16. The method of claim 15 , further including a clause that includes a sub-category feature, the first phrase, the operator and the second phrase.
17. The method of claim 16 wherein a context definition of the elements is based on a sequence of the first and second phrases.
18. The method of claim 14 wherein the any category between mandatory categories is represented by the symbol.
19. The method of claim 1 , wherein each rule is only applied once during the parsing.
20. A parser for parsing an input string that includes elements, said parser comprising a grammar rule database storing a plurality of rules, which describe syntactic properties of the language of the input string, the plurality of rules comprises two types of rules,
a first type of rules consisting of immediate dominance rules, which define dominance relations between constituents of the grammar rules, and linear precedence rules, which define an order of grammar constituents, and
a second type of rules defining dominance and precedence relations of grammar constituents in a fixed sequence order, and said parser being adapted to apply the rules of said plurality of rules according to a predefined order, and said plurality of rules is distributed into layers that are applied in a predefined layer order with each layer containing only the first type or the second type of rules.
21. The parser of claim 20 , wherein each rule is only applied once during the parsing.
22. A computer program product for use in a computer system for parsing input strings, the computer program product comprising a computer readable medium having a computer readable program code thereon, the computer readable program code causing the computer system to parse an input string that includes elements by applying a plurality of rules which describes syntactic properties of the language of the input string, the plurality of rules comprises two types of rules,
a first type of rules consisting of immediate dominance rules, which define dominance relations between a constituents of the grammar rules, and linear precedence rules, which define an order of grammar constituents, and
a second type of rules defining dominance and precedence relations of grammar constituents in a fixed sequence order, wherein the rules of said plurality of rules are applied according to a predefined order, and said plurality of rules is distributed into layers that are applied in a predefined layer order with each layer containing only the first type or the second type of rules.
23. The computer program product of claim 22 , wherein each rule is only applied once during the parsing.