Device and method for hiding information and device and method for extracting information
First Claim
1. A device for hiding information in a text having at least one sentence, comprisingmeans for linguistically analyzing the text to produce linguistic text components, the linguistic text components being components of the sentence and the sentence, in addition to at least one additional linguistic text component, having exactly one predicate as a linguistic text component;
- means for determining a plurality of formulation alternatives for the text, wherein the means for determining is operative to determine the plurality of formulation alternativesby varying the order of the linguistic text components itself,by ascertaining synonyms for the linguistic text components and varying the order of the synonyms for the linguistic text components, orby ascertaining synonyms for at least one linguistic text component and varying the order of a synonym for the at least one linguistic text component and/for another linguistic text component of the sentence,wherein the means for determining is adapted to generate the formulation alternatives such that every formulation alternative is grammatically correct for the text and has essentially the same meaning as the text, and that every order of linguistic text components and every synonym ascertained have specific partial information allocated thereto,wherein the partial information allocated to the simultaneously possible formulation alternatives for a sentence result in an overall probability of 1.0 such that arbitrary information can be hidden in the text,wherein the means for determining is operative to use, for the allocation of partial information to the simultaneously possible formulation alternatives, an existing probability distribution or a probability distribution generated in accordance with specific rules known to a device for extracting hidden information from a text;
means for selecting a formulation alternative from the plurality of formulation alternatives in such a manner that the partial information that is allocated to the selected formulation alternative corresponds at least to a part of the information to be hidden; and
means for outputting a selected formulation alternative forming a modified text, with said information to be hidden being hidden in said modified text,wherein the means for linguistically analyzing includes a parser, wherein the parser is operative to recursively break down the sentence into the text components using a hierarchical tree structure, wherein the parser is further operative to determine the predicate and remaining elements of the sentence and to recursively search a linguistic head within the remaining elements of the sentence until the text components of the sentence are obtained,wherein the parser includes a dictionary/grammar stage having an entry for the predicate or the linguistic head,wherein the means for determining is operative to use the entry for the predicate and the linguistic head to exclude parsing alternatives resulting in grammatically incorrect formulation alternatives, andwherein the means for determining is operative to leave a sentence unchanged when valid formulation alternatives can not be found.
0 Assignments
0 Petitions
Accused Products
Abstract
A device for hiding information in a text comprises a mechanicanism for providing the text, means for linguistically analyzing the text to produce text components, for determining a plurality of formulation alternatives for the text by varying the order of the text components and, optionally, in addition by using synonyms for text components, determining every formulation alternative is grammatically correct for the text and has essentially the same meaning as the text. Certain partial information is allocated to every sequence and/or to every synonym or to every paraphrase.
69 Citations
30 Claims
-
1. A device for hiding information in a text having at least one sentence, comprising
means for linguistically analyzing the text to produce linguistic text components, the linguistic text components being components of the sentence and the sentence, in addition to at least one additional linguistic text component, having exactly one predicate as a linguistic text component; -
means for determining a plurality of formulation alternatives for the text, wherein the means for determining is operative to determine the plurality of formulation alternatives by varying the order of the linguistic text components itself, by ascertaining synonyms for the linguistic text components and varying the order of the synonyms for the linguistic text components, or by ascertaining synonyms for at least one linguistic text component and varying the order of a synonym for the at least one linguistic text component and/for another linguistic text component of the sentence, wherein the means for determining is adapted to generate the formulation alternatives such that every formulation alternative is grammatically correct for the text and has essentially the same meaning as the text, and that every order of linguistic text components and every synonym ascertained have specific partial information allocated thereto, wherein the partial information allocated to the simultaneously possible formulation alternatives for a sentence result in an overall probability of 1.0 such that arbitrary information can be hidden in the text, wherein the means for determining is operative to use, for the allocation of partial information to the simultaneously possible formulation alternatives, an existing probability distribution or a probability distribution generated in accordance with specific rules known to a device for extracting hidden information from a text; means for selecting a formulation alternative from the plurality of formulation alternatives in such a manner that the partial information that is allocated to the selected formulation alternative corresponds at least to a part of the information to be hidden; and means for outputting a selected formulation alternative forming a modified text, with said information to be hidden being hidden in said modified text, wherein the means for linguistically analyzing includes a parser, wherein the parser is operative to recursively break down the sentence into the text components using a hierarchical tree structure, wherein the parser is further operative to determine the predicate and remaining elements of the sentence and to recursively search a linguistic head within the remaining elements of the sentence until the text components of the sentence are obtained, wherein the parser includes a dictionary/grammar stage having an entry for the predicate or the linguistic head, wherein the means for determining is operative to use the entry for the predicate and the linguistic head to exclude parsing alternatives resulting in grammatically incorrect formulation alternatives, and wherein the means for determining is operative to leave a sentence unchanged when valid formulation alternatives can not be found. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 25, 26, 27, 28, 29)
-
-
19. A method for hiding information in a text having at least one sentence, comprising:
-
linguistically analyzing the text to produce linguistic text components, the linguistic text components being components of the sentence and the sentence, in addition to at least one additional component, having exactly one predicate as a linguistic text component; determining a plurality of formulation alternatives for the text by varying the order of the linguistic text components itself, by ascertaining synonyms for the linguistic text components and varying the order of the synonyms for the linguistic text components, or by ascertaining synonyms for at least one linguistic text component and varying the order of a synonym for the at least one linguistic text component and of another linguistic text component of the sentence, with every formulation alternative being grammatically correct for the text and having essentially the same meaning as the text, with every order of linguistic text components and every synonym ascertained having specific partial information allocated thereto, wherein the partial information allocated to the simultaneously possible formulation alternatives for a sentence result in an overall probability of 1.0 such that arbitrary information can be hidden in the text, wherein, for the allocation of partial information to the simultaneously possible formulation alternatives, an existing probability distribution or a probability distribution generated in accordance with specific rules known to a device for extracting hidden information from a text is used; selecting a formulation alternative from the plurality of formulation alternatives in such a manner that the partial information that is allocated to the selected formulation alternative corresponds to the information to be hidden; and outputting the formulation alternatives that form a modified text, with said information to be hidden being contained in said modified text, wherein the step of linguistically analyzing includes a step of parsing by recursively breaking down the sentence into the linguistic text components using a hierarchical tree structure, by determining the predicate and remaining elements of the sentence and by recursively searching a linguistic head within the remaining elements of the sentence until the text components of the sentence are obtained, wherein the step of parsing furthermore using a dictionary/grammar stage having an entry for the predicate and the linguistic head, wherein the step of determining comprises using the entry for the predicate and the linguistic head to exclude parsing alternatives resulting in grammatically incorrect formulation alternatives, and wherein the step of determining leaves a sentence unchanged when valid formulation alternatives can not be found.
-
-
20. A device for extracting information hidden in a modified text, the modified text having at least one sentence, comprising:
-
means for linguistically analyzing the modified text to produce linguistic text components of the modified text, the linguistic text components being components of the sentence and the sentence, in addition to at least one additional component, having exactly one predicate as a linguistic text component; wherein the means for linguistically analyzing includes a parser, wherein the parser is operative to recursively break down the sentence into the text components using a hierarchical tree structure, wherein the parser is further operative to determine the predicate and remaining elements of the sentence and to recursively search a linguistic head within the remaining elements of the sentence until the text components of the sentence are obtained, wherein the parser includes a dictionary/grammar stage having an entry for the predicate and the linguistic head, wherein the means for determining is operative to use the entry for the predicate and the linguistic head to exclude parsing alternatives resulting in grammatically incorrect formulation alternatives, and means for producing partial information allocated to the sequence of the linguistic text components and, in case the modified text has information hidden in linguistic meanings of the linguistic text components, to the linguistic meanings of the linguistic text components, said means for producing partial information producing the same partial information which, in hiding the information to produce the modified text, was allocated to the sequence of linguistic text components and optionally to the linguistic meanings of the linguistic text components, wherein the partial information allocated to the simultaneously possible formulation alternatives for a sentence result in an overall probability of 1.0 such that arbitrary information can be hidden in the text, wherein the means for producing is operative to use, for the allocation of partial information to the simultaneously possible formulation alternatives, an existing probability distribution or a probability distribution generated in accordance with specific rules known to a device for extracting hidden information from a text, wherein the means for producing is operative to not produce partial information from a sentence when valid formulation alternatives can not be found by the means for linguistically analyzing, means for combining the partial information that was produced for the modified text by said means for producing partial information, in order to obtain the information hidden in the modified text; and means for outputting the hidden information. - View Dependent Claims (21, 22, 23, 24)
-
-
30. A method for extracting information hidden in a modified text, the modified text having at least one sentence, comprising:
-
linguistically analyzing the modified text to produce linguistic text components of the modified text, the linguistic text components being components of the sentence and the sentence, in addition to at least one additional component, having exactly one linguistic head as component; wherein the step of linguistically analyzing comprises parsing by recursively breaking down the sentence into the text components using a hierarchical tree structure, and by determining the predicate and remaining elements of the sentence and by recursively searching a linguistic head within the remaining elements of the sentence until the text components of the sentence are obtained, wherein the step of parsing furthermore comprises using a dictionary/grammar stage having an entry for the predicate and the linguistic head, wherein the means for determining is operative to use the entry for the predicate and the linguistic head to exclude parsing alternatives resulting in grammatically incorrect formulation alternatives; producing partial information allocated to the sequence of the linguistic text components and to the linguistic meanings of the linguistic text components, with the same partial information being produced which, in hiding the information to produce the modified text, was allocated to the sequence of linguistic text components and, in case the modified text has information hidden in linguistic meanings of the linguistic text components, was allocated to the linguistic meanings of the linguistic text components; wherein the partial information allocated to the simultaneously possible formulation alternatives for a sentence result in an overall probability of 1.0 such that arbitrary information can be hidden in the text, wherein the step of producing comprises using, for the allocation of partial information to the simultaneously possible formulation alternatives, an existing probability distribution or a probability distribution generated in accordance with specific rules known to a device for extracting hidden information from a text, wherein the step of producing comprises to not produce partial information from a sentence when valid formulation alternatives can not be found in the step of linguistically analyzing, combining the partial information that was produced for the modified text by said step of producing partial information, in order to obtain the information hidden in the modified text; and outputting the hidden information.
-
Specification