JP6206874B2 - Case component extraction program - Google Patents

Description

  The present invention relates to a case component extraction program for extracting a case component from a character string described in an electronic data document such as a patent specification.

  Conventionally, the purpose is to check a patent specification in a short time.Especially, even when the number of claims is kept as small as possible, including the description of the intermediate concept, the scope of rights, Patent specification analysis display devices that can be checked in consideration of the patent rate have been proposed (see, for example, Patent Document 1). This patent specification analysis display device has an advantage that the directionality described in the specification itself can be easily checked including intermediate concepts.

  However, the disclosed technique of Patent Document 1 has a problem in that it is not possible to perform an analysis focusing on the breadth of the inventive concept defined in the claims, in other words, the breadth of the technical scope.

  For this reason, Patent Document 2 discloses a patent specification analysis display device capable of counting and displaying the degree of limitation of the invention defined in the claims of the patent specification. According to the technology disclosed in Patent Document 2, the concept of a case component is newly adopted as a minimum extraction unit in a numerical method for quantifying and quantifying the breadth of the technical scope of the patented invention. The number of case components (number of case components) is a quantitative index that has the highest correlation with the breadth of the technical scope of the patented invention. The number of case components is obtained by quantifying how much the constituent elements are limited in the description of the claims. The number of case components counts and counts only the elements necessary to realize the verb proposition among the nouns (including noun phrases) that depend on the verb defined in the claims. is there.

JP 2006-155151 A JP 2009-259154 A

  However, since the disclosed technique of Patent Document 2 is not an algorithm suitable for practical use in actually extracting the number of case components from the claims, the case component extraction flow becomes very complicated, There was a problem that efficiency would deteriorate.

  In addition to this, if there is a limit to the quantification and quantification of the patent breadth based on the number of case components, the noun concept hierarchy (higher concept or lower concept) is not reflected in the parameters. For example, a communication device and a mobile phone are in a relationship of a higher concept and a lower concept, and the definition of a communication device can cover the scope of rights more conceptually than the definition of a mobile phone. When defined as a mobile phone, the scope of rights can only cover mobile phones and may not be included in the scope of rights until a tablet terminal is used as a communications device. By doing so, there is a high possibility that they can all be included in the scope of rights. In other words, the breadth of the technical scope of the patented invention is determined based on two levels: a grammatical structural unit size relationship and a noun concept hierarchy (superordinate concept or subordinate concept). The former can be detected with high accuracy by using the number of case components as an index, but the concept hierarchy of the latter noun cannot be detected by the number of case components.

  For this reason, it is necessary to propose an evaluation index for the breadth of a patent that takes into account the concept hierarchy of such nouns in addition to the number of case components. In other words, the breadth of rights is measured based on the number of case components that depend only on a simple phrase structure, and there is a problem that the above-mentioned needs cannot be met.

  Therefore, the present invention has been devised in view of the above-described problems, and the object of the present invention is to actually determine the grammatical structural unit size relationship and the nouns of each from the claims. It is an object of the present invention to provide a case component extraction program capable of evaluating the breadth of a patent and a case component extraction program capable of extracting the number of case components with high accuracy in consideration of both concept hierarchies.

A case component extraction program according to claim 1 is a preprocessing step for extracting a character string from the description of a claim in an electronic data patent specification, and a morphological analysis for the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “ at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of A word comprehension step that summarizes as one word or “abbreviated”, “slightly”, or “sequential” + a noun as one word, and a noun phrase from the character string that includes the word that has undergone the word inclusion step Noun phrase extraction step to extract, and in the noun phrase extraction step, the case component as a noun phrase required to realize the action of the verb related to each component constituting the scope of claims from the extracted noun phrase And a noun included in each case component extracted in the case component extraction step with reference to a concept dictionary in which the concept of the noun is stored in a hierarchy from the upper concept to the lower concept The computer is caused to execute a concept hierarchy specifying step for specifying each concept hierarchy.

  The case component extraction program according to claim 2 is the invention according to claim 1, wherein the number of case components extracted in the case component extraction step and the noun concept in each case component specified in the concept hierarchy specifying step And a post-processing step for outputting or storing an evaluation value calculated based on the number of extracted case components and a noun conceptual hierarchy in each case component. To do.

  According to a third aspect of the present invention, in the second aspect of the invention, in the post-processing step, the post-processing step sets an evaluation value such that the broader the patent is, the higher the concept hierarchy is. As the concept hierarchy becomes lower, the evaluation value is calculated so that the breadth of the patent becomes narrower.

  The case component extraction program according to claim 4 is the invention according to any one of claims 1 to 3, wherein, in the concept hierarchy specifying step, the concept dictionary includes noun concepts hierarchized for each technical field. Reference is made to a concept dictionary that is stored and corresponds to the technical field of the patent specification to be extracted.

  The case component extraction program according to claim 5 is the invention according to claim 4, wherein the concept hierarchy specifying step uses a concept dictionary in which the concept hierarchy is normalized for each technical field.

  The case component extraction program according to claim 6 is the invention according to claim 2 or 3, wherein, in the post-processing step, each case component is weighted according to a concept hierarchy specified for each noun phrase included therein, The evaluation value is calculated based on the sum of the weighted case components.

  The case component extraction program according to claim 7 is the invention according to any one of claims 1 to 6, wherein in the concept hierarchy specifying step, the concept dictionary newly received via the network, or other software or A concept dictionary newly acquired from another database is used.

The case component extraction program according to claim 8 includes a preprocessing step of extracting a character string from the description of the scope of claims in the patent specification converted into electronic data, and a morphological analysis of the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “ at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of A word comprehension step that summarizes as one word or “abbreviated”, “slightly”, or “sequential” + a noun as one word, and a noun phrase from the character string that includes the word that has undergone the word inclusion step Noun phrase extraction step to extract, and in the noun phrase extraction step, the case component as a noun phrase required to realize the action of the verb related to each component constituting the scope of claims from the extracted noun phrase Each case extracted in the case component extraction step with reference to information on the appearance frequency of nouns by analyzing the description of the patent specification in advance with a conventional patent as a population. characterized in that to execute a frequency rank specifying step of specifying the rank of the current frequency output of nouns included in component respectively to the computer.

  The case component extraction program according to claim 9 is the invention according to claim 8, wherein the number of case components extracted in the case component extraction step and the frequency of nouns in each case component specified in the frequency rank specification step. A post-processing step that outputs or stores ranks, or outputs or stores evaluation values calculated based on the number of extracted case components and the frequency rank of nouns in each case component, To do.

  The case component extraction program according to claim 10 is the evaluation value according to claim 9, wherein in the post-processing step, the evaluation value is such that the breadth of the patent becomes wider as the frequency rank becomes higher. It is characterized in that the evaluation value is calculated such that the width of the patent becomes narrower as the rank becomes lower in frequency.

  The case component extraction program according to claim 11 is the invention according to any one of claims 8 to 10, wherein in the frequency rank specifying step, the frequency rank is stored for each technical field, and the patent specification to be extracted The frequency rank corresponding to the technical field of the book is referred to.

  The case component extraction program according to claim 12 is the invention according to any one of claims 8 to 11, wherein, in the frequency rank specifying step, the frequency rank is normalized for each technical field, or is analyzed. It is characterized by using one normalized according to the patent application year.

  The case component extraction program according to claim 13 is weighted according to the frequency rank specified for each noun phrase included in each case component in the post-processing step in the invention according to claim 9 or 10. The evaluation value is calculated based on the sum of the weighted case components.

  The case component extraction program according to claim 14 is the invention according to any one of claims 8 to 13, wherein in the frequency rank specifying step, the frequency rank information newly received via the network or other software Alternatively, frequency rank information newly acquired from another database is used.

The case component extraction program according to claim 15 includes a preprocessing step of extracting a character string from the description of the scope of claims in the patent specification converted into electronic data, and a morphological analysis of the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “ at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of A word comprehension step of summarizing words as one word, or summing any one of "abbreviation", "slightly", and "sequential" + nouns, and a noun phrase from the character string including the word that has undergone the word inclusion step A noun phrase extraction step for extracting a noun phrase, and in the noun phrase extraction step described above , a case as a noun phrase required for realizing the action of a verb related to each component constituting the scope of claims from the extracted noun phrase . A noun included in each case component extracted in the case component extraction step with reference to information regarding the appearance frequency of nouns by analyzing the description of the patent specification to be analyzed, and a case component extraction step for extracting components and wherein the output be executed and a frequency rank specifying step of specifying each rank of the current frequency to the computer.

  The case component extraction program according to claim 16 is the invention according to claim 15, wherein the number of case components extracted in the case component extraction step and the frequency of nouns in each case component specified in the frequency rank specification step. A post-processing step that outputs or stores ranks, or outputs or stores evaluation values calculated based on the number of extracted case components and the frequency rank of nouns in each case component, To do.

The case component extraction program according to claim 17 is the "characteristic" in "characterizing" in the character string in the case component extraction step in the invention according to any one of claims 1 to 16. the characterized in that to execute a computer to extract a noun phrase, excluding from the string as "at least rated component.

  An information transmission system according to claim 18 includes an evaluation value calculation means for calculating the evaluation value by the case component extraction program according to any one of claims 2 to 7 and 9 to 17, and an evaluation value calculation means. A server that stores at least one of the obtained evaluation value and a figure or data based on the evaluation value, and the evaluation value requested from the client side and a figure or data based on the server are read out from the server And transmitting means for transmitting to the client side via the network.

A case component extraction program according to claim 19 includes a preprocessing step of extracting a character string from the description of the scope of claims in a patent specification converted into electronic data, and a morphological analysis of the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “ at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of A word comprehension step of summarizing words as one word, or summing any one of "abbreviation", "slightly", and "sequential" + a noun, and a noun phrase from the character string including the word that has undergone the word inclusion step In the noun phrase extracting step and the noun phrase extracting step, when A and B are defined as noun phrases in the extracted noun phrases, A and B are alternatively defined Is characterized by excluding one of A and B from the case components to be extracted and causing the computer to execute a case component extraction step of extracting the remaining noun phrases as at least the case components.

The case component extraction program according to claim 20 includes a preprocessing step of extracting a character string from the description of the scope of claims in the patent specification converted into electronic data, and a morphological analysis of the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of A word comprehension step of summarizing words as one word, or summing any one of "abbreviation", "slightly", and "sequential" + nouns, and a noun phrase from the character string including the word that has undergone the word inclusion step A noun phrase extraction step for extracting a noun phrase and a case component as a noun phrase required to realize the action of a verb related to each component constituting the scope of claims from the noun phrase extracted in the noun phrase extraction step Is extracted from the case component to be extracted, the computer is executed to exclude a word composed of either “ka” or “ ya ” as a noun + parallel particle.

A case component extraction program according to claim 21 includes a preprocessing step of extracting a character string from the description of the scope of claims in a patent specification converted into electronic data, and a morphological analysis of the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “ at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of A word comprehension step of summarizing words as one word, or summing any one of "abbreviation", "slightly", and "sequential" + nouns, and a noun phrase from the character string including the word that has undergone the word inclusion step In extracting the noun phrase from the noun phrase extracted step and the noun phrase extracted in the noun phrase extracting step, "~", "~", and "to" are included in the extracted character string. , If the relationship between the subject defined in any of “to” and the verb that depends on the subject is included, and the wording corresponding to the subject has already been described above Excludes the word corresponding to the subject from the case component to be extracted, and if the word corresponding to the subject is not yet described in the upper row, the word corresponding to the subject is to be extracted. Case ingredients to be processed for inclusion in ingredients Characterized in that to execute the steps in computer output.

A case component extraction program according to claim 22 includes a preprocessing step of extracting a character string from the description of a claim in an electronic data patent specification, and a morphological analysis of the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “ at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of A word comprehension step of summarizing words as one word, or summing any one of "abbreviation", "slightly", and "sequential" + nouns, and a noun phrase from the character string including the word that has undergone the word inclusion step A noun phrase extraction step for extracting a noun phrase and a case component as a noun phrase required to realize the action of a verb related to each component constituting the scope of claims from the noun phrase extracted in the noun phrase extraction step In order to extract the verbs, the verbs are sequentially extracted, the identification step for identifying whether the extracted verbs are listed in the upper stage, and as a result of the identification in the identification step, the extracted verbs are already described in the upper stage If it has been identified, the noun phrase that depends on the identified verb, and the noun phrase that indicates the subject on which the verb already described above depends An identity determination step for determining identity between a noun phrase identified in the identifying step and the noun phrase identified in the identifying step and a noun phrase that depends on the extracted verb, and a noun phrase that depends on the extracted verb Among them, a case component determined to be the same in the identity determination step is caused to cause a computer to execute a case component extraction step having an exclusion step of excluding it from case components to be extracted.

The case component extraction program according to claim 23 includes a preprocessing step for extracting a character string from the description of the scope of claims in the patent specification converted into electronic data, and a morphological analysis for the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “ at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of Conclusion As one word, or "substantially", "somewhat", and the word entrapping step to combine either + noun "sequential" One word claims from the string containing the word through the word entrapping step A noun phrase extracting step for extracting a noun phrase as a noun phrase required to realize the action of a verb related to each component constituting the range of the noun phrase, and a case component from the noun phrase extracted in the noun phrase extracting step When extracting any of the words, either “any” or “any” extracted as a noun is excluded from the case components to be extracted, and there are multiple words that are directly related to “any” or “any” In this case, the computer is caused to execute a case component extraction step of excluding the case component to be extracted from the preceding noun phrase.

The case component extraction program according to claim 24 includes a preprocessing step of extracting a character string from the description of the scope of claims in a patent specification converted into electronic data, and a morphological analysis of the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “ at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of Conclusion As one word, or "substantially", "somewhat", and the word entrapping step to combine either + noun "sequential" One word claims from the string containing the word through the word entrapping step A noun phrase extracting step for extracting a noun phrase as a noun phrase required to realize the action of a verb related to each component constituting the range of the noun phrase, and a case component from the noun phrase extracted in the noun phrase extracting step In extracting the noun phrases, the nouns including the nouns + “target / free / possible / impossible / impossible / you” + (“na”, “ni”, “no”, “to”) A case component is extracted from a case component to be extracted from a phrase or a phrase in which a hiragana is added to these noun phrases.

A case component extraction program according to claim 25 includes a preprocessing step of extracting a character string from the description of the scope of claims in a patent specification converted into electronic data, and a morphological analysis of the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “ at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of A word comprehension step of summarizing words as one word, or summing any one of "abbreviation", "slightly", and "sequential" + nouns, and a noun phrase from the character string including the word that has undergone the word inclusion step A noun phrase extraction step for extracting a noun phrase and a noun phrase extracted from the noun phrase extracted in the noun phrase extraction step, a word composed of “noun or noun phrase” + “A” + “noun or noun phrase” In the case where A = “up to”, “to”, “to”, “in”, “ya”, or “no”, the individual “nouns” separated by “A” are extracted. A case component extraction step of separately extracting “noun phrases” as case components is executed by a computer.

A case component extraction program according to claim 26 includes a preprocessing step for extracting a character string from the description of the scope of claims in the patent specification converted into electronic data, and a morphological analysis on the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “ at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of A word comprehension step of summarizing words as one word, or summing any one of "abbreviation", "slightly", and "sequential" + nouns, and a noun phrase from the character string including the word that has undergone the word inclusion step In extracting the noun phrase from the noun phrase extracted in the noun phrase extraction step in the noun phrase extraction step and the noun phrase extraction step, the “no”, “ It is characterized in that the computer is caused to include a pronoun consisting of any of “it”, “that”, “which”, “here”, “there”, “that”, “where” in the case component to be extracted. To do.

The case component extraction program according to claim 27 includes a preprocessing step of extracting a character string from the description of the scope of claims in the patent specification converted into electronic data, and a morphological analysis of the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “ at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of A word comprehension step of summarizing words as one word, or summing any one of "abbreviation", "slightly", and "sequential" + nouns, and a noun phrase from the character string including the word that has undergone the word inclusion step In the noun phrase extraction step for extracting the noun phrase, and in the noun phrase extraction step, the noun form of the verb, in which the verb is connected to “ko”, is extracted from the noun phrase extracted in the noun phrase extraction step. A case component extraction step of extracting as a case component is executed by a computer.

The case component extraction program according to claim 28 includes a preprocessing step for extracting a character string from the description of the scope of claims in the patent specification converted into electronic data, and a morphological analysis on the character string processed in the preprocessing step. And after analyzing these into words by parsing, the noun + noun / noun + “no” + noun / noun + “no” + noun / noun + “target / free / possible” / noun + “To”, noun + “by”, “ at least” + quantity expression, noun + “do, can, let, be done” as one word, or “above”, “above”, Either “the”, “the”, or the conjunction is combined with the following word as one word, or the noun + “no” + “adjective” + noun / noun + “to be related to Any of A word comprehension step of summarizing words as one word, or summing any one of "abbreviation", "slightly", and "sequential" + nouns, and a noun phrase from the character string including the word that has undergone the word inclusion step In the noun phrase extraction step and the noun phrase extraction step, which are required to realize the action of the verb related to each component constituting the scope of claims from the noun phrase extracted in the noun phrase extraction step. A case component extraction step for extracting case components as a noun phrase, and in the word inclusion step, causes the computer to execute “abbreviation”, “slightly”, “sequential” + noun as one word It is characterized by that.

  A case component extraction program according to a twenty-ninth aspect of the present invention is the invention according to any one of the nineteenth to twenty-eighth aspects, wherein in the case component extraction step, the “characteristic” in “characteristic of” is included in the character string. The computer is caused to extract at least a noun phrase that is removed from the character string as a case component.

The case component extraction program according to claim 30 is the invention according to any one of claims 19 to 29, wherein in the case component extraction step, the number of extracted case components is counted and output or stored. , An evaluation value based on the number of case components, or the number of case components (hereinafter referred to as CN) as noun phrases required to realize the action of a verb related to each component constituting the claims . Is output or stored as an evaluation value based on the following calculation formula.
Evaluation value = {α-log (CN + β)} × γ
Where α = 3-8
β = 15-25
γ = 15-40

An information transmission system according to a thirty-first aspect is required to realize the action of a verb related to each component constituting the scope of the claims by the case component extraction program according to any one of the nineteenth to twenty-ninth aspects. Case component extraction means for extracting case components as noun phrases ;
A server that stores at least one of the number of case components obtained by the case component extraction means or an evaluation value based on the number of case components;
An information transmission system comprising: a transmission unit that reads out the number of case components or an evaluation value of a patent requested from the client side from the server and transmits it to the client side via a network.

  In the present invention having the above-described configuration, it is possible to extract the case component as the degree of limitation of the invention defined in the claims of the patent specification with high accuracy. In addition, it is possible to evaluate the breadth of patents by taking into account both the size relationship of structural units in the grammar and the concept hierarchy of each noun.

It is a figure for demonstrating a case component. It is another figure for demonstrating a case component. It is a flowchart for performing the case component extraction program to which this invention is applied. It is a figure for demonstrating a concept hierarchy. It is a figure which shows the example which allocates a numerical value about a concept hierarchy. It is another flowchart for performing the case component extraction program to which this invention is applied.

  Hereinafter, as a form for carrying out the present invention, a case component extraction program for extracting a case component from a character string described in an electronic data document including a patent specification and the like will be described in detail with reference to the drawings. Explained.

  In the following example, a case where a case component is extracted from the description of the claims in the patent specification will be described as a document. However, the present invention is not limited to this, and the present invention is applied to any other document. Of course, you may do.

Case ingredient concept

  Case components have been proposed as a minimum extraction unit in a quantification method for quantifying and quantifying the breadth of the technical scope of the patented invention.

  The claims are described in a so-called requirement enumeration system, in which each component A, B, and C is defined by enumerating as "XX apparatus (method) including ~ A, ~ B, and C" " As a premise, each component A, B, and C depends on a verb phrase. That is, when each of the constituent elements A, B, and C is a subject, a verb phrase depends on each of them, and a sentence composed of the subject and a predicate can be established. When the verbs in these verb phrases form a sentence as a predicate, the combination of noun phrases is basically determined in order to realize the movement, state, and relationship expressed by the verb. If a verb is tentatively called the case rule of a verb, the verb requires a combination of noun phrases required to form a sentence according to the lexical meaning of the verb. Noun phrases required as sentence components can be called case components in supplementing verbs.

  For example, it is assumed that the component “signal generating means” is defined by the following B-1).

  B-1) “Signal generation means for generating a drive signal in response to a user request, and so on”

  At this time, a verb phrase “generates a drive signal in response to a user request” depends on the signal generation means. In order to realize and complete the action of the verb “Generate” in this verb phrase, the role of requesting the noun phrases “user request (according to)” and “driving signal ()” is dominant. These noun phrases are case components. And the noun phrases such as “user's request (according to)” and “drive signal ()” have similar relational meanings such as “operation start condition” and “target” for the verb “generate” respectively. Is the case. This verb “generate” should be realized by the lexical meaning of the noun phrases that realize this case component, and the case formed as the relative meaning of those noun phrases. It becomes the core part of the proposition. In the above example, the operation start condition for realizing the proposition by the verb “generate” must be “in response to the user's request”, and “drive signal” must be generated as the target. Therefore, these two are conditional numbers for the realization of the verb proposition.

  As shown in FIG. 1, the case where the number of conditions that must be cleared in order to realize a proposition using such a verb is schematically shown. Since the number of conditions for starting the action by the verb is reduced, the possibility that the proposition can be realized is improved. On the other hand, FIG. 2 schematically shows a case where the number of conditions that must be cleared for the realization of a proposition by a verb is increased. In such a case, the possibility that the proposition can be realized decreases. .

  As the condition number increases, the number of steps for determining whether or not the condition is satisfied before the operation is actually started increases. Even considering that there is a slight difference in the probability of satisfying each condition, the verb phrase depends on as the number of judgment steps for this condition increases, in other words, as the case component that defines the condition increases. This means that the possibility of falling under component A is reduced, and the technical scope is narrowed by the reduced possibility. Conversely, if the number of case components is small, this means that the verb phrase is likely to fall under the component A that it depends on, indicating that the technical scope will expand accordingly. .

  In this way, the number of case components (condition number) governs the possibility of action start by a verb, and thus the possibility of realizing a proposition, and this affects the breadth of the technical scope. By counting the number of case components included in this verb phrase with the minimum extraction unit of the conversion method as the case component, it is considered that numericalization according to the width of the technical range can be realized.

  In the example of B-1) described above, the two case components “depending on the user's request” and “drive signal” that depend on the verb “generate” to realize the proposition of the component “signal generation means” Therefore, the number of case components is 2. At this time, if there is no case component “in response to the user's request” that defines the operation start condition of the verb “generate”, the number of case components is 1, regardless of whether or not the user has requested. The “drive signal” can be “generated”, and the possibility of realizing the proposition of “signal generation means” is increased. This suggests that the component “signal generation means” can improve the possibility of verifying the identity of the technical elements of the suspected infringing product. This is considered to correspond to the ease of capturing the patented invention with respect to the product, and thus the technical scope of the patented invention. For this reason, it is considered that by expressing the proposition feasibility of the constituent element through the number of case components, it is possible to achieve more optimal quantification and quantification according to the width of the technical range.

  The claims are described in a so-called requirement enumeration system, in which each component A, B, and C is defined by enumerating as "XX apparatus (method) including ~ A, ~ B, and C" " The number of case components is determined for each of the components A, B, and C, and the sum is defined as the number of case components of the OO device as the patented invention. If the number of case components of component A is 1, the number of case components of component B is 3, and the number of case components of component C is 2, the case of the XX device as a patented invention constituted by these components is assumed. The number of components is 6, which is the sum of the components.

  When actually determining the number of case components of each component, in the above-mentioned example of B-1), the case components “according to the user's request” and “drive signal” that depend on the verb “generate” are set. Although the case component is extracted, the actual extraction operation of the case component is performed by using morphemes as marks such as “accordingly” and “to”.

  Table 1 shows examples of morphemes that are referred to as landmarks when extracting case components.

  The claims may be defined not only in the orthodox form as in B-1) but also as shown in C-1), for example.

  C-1) "Reflecting means for totally reflecting the light beam, the spot diameter of which is controlled by the first lens, by the reflecting plate,"

  For example, the component “reflecting means” constituting the scope of the claims depends on the verb “total reflection”. As a stipulation that defines the action start condition of the verb “totally reflect”, “light flux ()” and “reflector (by)” exist, and first count the number of case components 2. Can do. However, when focusing on one of the case components “light flux”, it can be seen that the verb phrase “the spot diameter is controlled by the first lens” is further dependent. Rather than defining all luminous fluxes as “totally reflecting” objects, it is technically more restrictive to define conditional luminous fluxes as “totally reflecting” objects. Feasibility of proposition decreases. For this reason, it is necessary to correct this decrease in proposition feasibility as the number of case components.

  In such a case, in the verb phrase “spot diameter is controlled by the first lens” that depends on “light flux”, the two case components “first lens ( )) And “spot diameter ()” are present, so add 2 case components. As a result, the total number of case components for C-1) is 4.

  In addition, description C-1) of this claim may be defined as C-2) below depending on the creator of the specification.

  C-2) “Flux control means for controlling the spot diameter of the light flux by the first lens, and reflection means for totally reflecting the light flux by the reflecting plate, and so on”.

  The technical range formed by the “light flux controlling means” and the “reflecting means” in C-2) is substantially the same as C-1). In C-2), the content of the verb phrase that depends on the “light flux” in C-1) is redefined as a proposition to be realized by the “light flux control means”. In fact, even in C-1), in order to produce the defined "light flux", some means or member will be used, but it is only deliberately defined as "light flux control means". . Similarly, when the number of case components is counted for C-2), the number of case components is 2 for the light flux control means and the number of case components is 2 for the reflection means. The result is the same as for C-1). Become.

  As described above, since the scope of the technical range is counted through the number of case components required to realize the proposition of the constituent element, the description method and the number of words differ from each other substantially. C-1) and C-2) consisting of the same technical scope can be expressed with the same number of case components. If the minimum extraction unit is a component, C-1) is 1 and C-2) is 2. If the minimum extraction unit is a word, C-1) is 7, and C-2) is 9. In view of this, it is considered that the number of case components can be a minimum extraction unit of a more suitable numerical method.

  The configurations of C-1) and C-2) may be defined by the following C-3).

C-3) “Flux control means for controlling the spot diameter of the light flux by the first lens, and reflection means for totally reflecting the light flux whose spot diameter is controlled by the light flux control means by the reflector, and so on”

  The wavy line portion of C-3) corresponds to the portion of “the luminous flux” in C-2). However, the content of the verb phrase described in this wavy line is a proposition that has already been realized by the straight line part that depends on the "light flux control means". This is merely a confirmation to prevent doubt about interpretation of the “light beam” as the object that the “means” totally reflects. For this reason, when the number of case components is calculated, if the description of the wavy line portion is also counted in the same manner, the content of the underlined portion is double counted, resulting in a deterioration in measurement accuracy.

  For this reason, it is possible to eliminate the shift in the number of case components due to the difference in the description order and description method of the claims by taking care not to count the description of the wavy line already executed as a proposition. Actually, it is necessary to determine whether or not the proposition realized by the description of the straight line portion and the description of the wavy line portion is the same.

  The numerical method using the minimum extraction unit as a case component is particularly useful when determining the identity of propositions. Case components, unlike words, have the property of being governed by verbs, so that the case components are the same means that in addition to the content of the case component, the verb that governs it must also be the same. Yes. If the case component and the verb that governs it are the same, then the proposition realized by this is the same. For this reason, it is possible to identify whether or not the propositions are the same through the determination of the identity of the case components, and thus it is possible to easily identify whether or not the proposition is a double count. Moreover, the difference in the description method between the specification creators does not affect the obtained numerical value.

  In the above-described example, the case where the description of the claims is described in the component requirement enumeration type is described as an example. However, even if it is described in another format, it can be counted in the same manner. For example, in the case of being described in the Jepson type, the description of “writing”, which is a known part, is also counted by the same method. At this time, in the case where the counting method described in “note writing” is not known, counting may be performed after rewriting from the Jepson type to the component requirement enumeration type. Incidentally, the rewriting method from the Jepson type to the component requirement enumeration type is conventionally known.

  Further, even when the scope of claims is described in a flow-through type, it can be counted by the same method. For example, if C-2) is written in a flow-through style, it will be as shown in C-4) below.

  C-4) "The spot diameter of the light beam is controlled by the first lens, and this light beam is totally reflected by the reflector.

  Compared to C-2), C-4) described in this flow-through type is different in that the words “light flux controlling means” and “reflecting means” are omitted, but the rest is the same. . “Flux control means” and “reflecting means” are subject words and are not counted, but only count noun phrases that depend on this subject. There is no difference in the number of components.

Case component extraction process

  The case component is extracted based on the following process. What is hereinafter referred to as “count” means that all are extracted as case components.

  FIG. 3 shows a flowchart of a case component extraction program to which the present invention is applied. In this flowchart, first, in pre-processing step S11, a character string is extracted from the description of the claims in the patent specification converted into electronic data. Then, predetermined preprocessing is performed on the extracted character string.

  In this preprocessing, first, if there is a part enclosed in parentheses in the character string, it is deleted. In other words, if there is a description in the character string “with the light beam control means for controlling the spot diameter (φ) of the light beam by the first lens”, for the convenience of subsequent case component extraction, Delete “(φ)” in parentheses. Incidentally, when there is no part enclosed in parentheses in the character string, the process is omitted.

  Next, the wording corresponding to the name of the invention is extracted. In this process, the word following “characterized by” in the character string is regarded as the name of the invention. For example, when the ending of the character string ends with “information filtering device characterized by”, a process is performed in which “information filtering device” following “characterized by” is regarded as the name of the invention.

  In other cases, a process is performed in which each word from the end of the character string is confirmed in the reverse direction until a word other than a punctuation mark or noun appears as the name of the invention. For example, in the case where the ending ends with “vehicle having“ ˜ ”, words other than the noun“ having ”appear when words are confirmed one by one from the end of the character string. A process (vehicle) in which a word other than the noun appears is regarded as the name of the invention. In addition, in the case of “medium on which is recorded”, words other than the noun “recorded” appear when words are confirmed one by one from the end of the character string. A portion (medium) where a word other than the noun appears is regarded as the name of the invention.

  At this time, in this pre-processing step S11, it is possible to search the character string for the same word as the recognized invention name and replace it with a predetermined word (for example, “invention”). As a result, when the name of the invention has a certain number of characters, it is possible to process so as to be surely determined as one noun in the syntax analysis. Note that the present invention is limited to the above-described method as long as a character string is extracted from the description of the claims in the patent specification converted into electronic data, and at least the name of the invention is extracted from the character string. However, the present invention may be executed based on any other method. Note that this preprocessing step S11 is not an essential step and may be omitted.

  After preprocessing step S11 is completed, the process proceeds to word inclusion step S12. In this word inclusion step S12, the character string processed in the preprocessing step is subjected to morphological analysis and syntax analysis to be decomposed into words, and then processing is performed with a predetermined character string as one word. That is, in this word inclusion step S12, a process of grouping words is performed so that the extraction process in the subsequent stage can be easily performed.

In this word inclusion step S12, the character string processed in the preprocessing step S11 is decomposed into words, and then noun + noun, noun + “no” + noun, noun + “no” + noun, noun + “ For extraction of case components such as "manual / free / possible", noun + "to", noun + "by", "at least" + quantity expression, noun + "to / can / do" Words that should be one word (inclusive word) are grouped as one word. In addition, “the above”, “above”, “the”, “the”, a conjunction (“this”, “that”, “that”, “which”) and the following word and one word (inclusive word) As a summary. In addition, a process of collecting nouns + “no” + “adjectives” + nouns and “nouns +“ becomes related to, related to, and related to ”as one word (inclusive word) is performed. Further, in this word inclusion step S12, a process of combining “abbreviated”, “slightly”, “sequential” + nouns into one word (inclusive word) may be performed.
In general, when morphological analysis is performed by natural language processing, words are decomposed into a minimum unit. However, in case component extraction, inclusive words such as nouns + nouns should be counted as one word. Therefore, based on the predetermined rule as described above, it is necessary to perform a process for collecting the comprehensive words as one word among the minimum unit words decomposed by the morphological analysis. With this word inclusion step, it is possible to extract the case components with high accuracy even using the automatic decomposition result of the words by morphological analysis.

  In performing the above-described processing, in the word inclusion step S12, so-called morpheme analysis technology and syntax analysis technology are appropriately used. A morpheme is “the smallest unit that has meaning in natural language”. Morphological analysis is the decomposition of sentences into morpheme units. For example, “Taro ate lunch” can be broken down into the morphemes shown in Table 2 below.

  As for the actual state of morphological analysis in practical use, in natural language processing, if the unit of “morpheme” in its original meaning is decomposed, the granularity becomes finer than that useful for analysis. For this reason, the resolution is often limited to “coarse” units. For example, although it is decomposed into 8 parts of speech in the above example, in actual analysis, it may be decomposed into 3 units as shown in Table 3 below.

  The output of general morpheme analysis / syntax analysis is the granularity information of each “clause” unit and the detailed structure in the block of each “clause” unit (for example, “Taro (noun)-is a (particle)”. ) In two layers (or can be configured later).

  By the way, this "lunch rice" can be considered to be decomposed into at least two morphemes of "daytime" and "rice". In order to perform the extraction, it is preferable not to decompose more than “lunch”.

  The extent to which these morphemes are put together can be freely adjusted using morphological analysis technology. However, in the present invention, these morphemes are expressed as “Taro is” or “Lunch” as described above. , Process so that nouns or noun phrases are grouped as a unit.

  In this word inclusion step S12, further syntax analysis may be performed. The result of the morphological analysis described above is a one-dimensional word string having only context. Parsing this word string into a two-dimensional “syntax tree” is syntactic analysis.

  In a syntax tree, each word typically has at most one “parent” and zero or more “children”. The “parent” of a word is the destination of the word “related”. A “child” is a word related to the word. For example, “Hanako bought a red bag” has “Red” as a child and “Bought” as a parent. In other words, “red” is related to “鞄”, so “red” is a child for “鞄”. In addition, “buy” is related to “buy”, so “buy” is “parent” for “mochi”. In addition, “Bought” has “Maki” and “Hanako” as children. As a result, this syntax tree can be expressed as shown in Table 4.

  After finishing the word inclusion step S12 described above, the process proceeds to a noun phrase extraction step S13. In this noun phrase extraction step S13, morphological analysis and syntax analysis as described above may be used as necessary, and noun phrases are extracted from the character string. By the way, this noun phrase has a concept including only a noun. After completing the extraction of all nouns or noun phrases from the character string, the process proceeds to case component extraction step S14.

In this case component extraction step S14, a flag is set for the noun or noun phrase extracted in the noun phrase extraction step S13, which is not included in the case component. And the noun or noun phrase except the noun or noun phrase for which the flag is set is extracted as the case component. Finally, the total of the extracted case components is counted and displayed as the number of case components or stored.
Incidentally, in the present invention, any method may be used as long as a process for excluding the case component to be extracted is performed instead of setting a flag. That is, it is not essential to use a flag, and any method may be used as long as it can be selected from what is included in the case component to be extracted and what is excluded.

  Hereinafter, the target for which the flag is set in the case component extraction step S14 will be described.

  First, flags are set for all of the same words as the name of the invention extracted in the preprocessing step S11. Thereby, in this case component extraction step S14, the wording corresponding to the name of the invention can be excluded from the case components to be extracted. It should be noted that it is not essential to flag all of the same words as the name of the present invention, and a rule that does not set a flag may be used.

  In the case component extraction step S14, when a wording identical to the name of the invention is replaced with a predetermined wording, a flag is set for the replaced predetermined wording.

  In case component extraction step S14, when A and B are nouns or noun phrases, respectively, if A and B are alternatively defined, a flag is set for one of A and B. You may make it stand. Here, the case where A and B are alternatively defined is a case where “A or B”, “A or B”, or the like is described, for example. In such a case, if A is a noun, flag A. If A is not a noun and B is a noun, a flag is set for B. Thereby, in this case component extraction step S14, any one of the nouns or noun phrases that are alternatively described can be excluded from the case components to be extracted.

  Further, in the case component extraction step S14, a flag may be set for a word composed of a noun and a parallel particle (either “ka” or “ya”). For example, in the case where “by rope or wire” is described, the combination of the noun and the parallel particle becomes “rope”. Flag this. Since the description of “by rope or wire” merely lists examples of tools necessary for connection, such as ropes and wires, one of the nouns or noun phrases should be extracted. It is supposed to be excluded from the ingredients.

  Among the noun phrases excluding the same noun phrase as the name of the invention, by setting this flag, the processing operation to be excluded from the case components to be extracted is the “case component number extraction basic” of the following (1) to (4). It may be based on a “rule”.

(1) Subject rule In the case component extraction step, if the extracted character string contains a relationship between the subject and the verb as a result of syntactic analysis, the sentence corresponding to the subject is If already described, a flag is set for the word corresponding to the subject. If the word corresponding to the subject has not yet been described in the upper stage, processing is performed so as not to flag the word corresponding to the subject. The following 1) to 3) are specific processing flows.

  1) First, the subject is first extracted through syntax analysis. "●●● is", "●●● is", "●●● is", "●●● is also", etc. are all regarded as the subject and extracted. (At this time, there is no relation to ●●● whether or not “previous”, “related” or “above” is attached)

  2) Sort and confirm whether the extracted subject (●●●) is listed before that. As a result, if there is an object that matches 100% with the extracted subject (●●●) or includes a part thereof, the process proceeds to 3). Other than that, since the extracted subject (●●●) has appeared for the first time, the extracted subject (●●●) is specified as a case component.

  3) When shifting to 3) in 2) above, the extracted subject (●●●) is flagged so that it is not counted as a case component.

  In other words, the extracted character string includes the relationship between the subject defined by any one of “to-ga”, “to-ha”, “to-to”, and “to-mo” and the verb depending on this subject. First, it is determined whether or not there is. As a result, when the extracted character string includes a relationship between the subject and the verb, it is confirmed whether or not the wording corresponding to the subject has already been described in the upper stage. If the word corresponding to the subject has already been described in the upper row, the word corresponding to the subject is excluded from the case components to be extracted, and the word corresponding to the subject is still written in the upper row. If not, since it has newly appeared, the word corresponding to the subject is included in the case component to be extracted.

  As a result, in the description of the claims having many subjects, it is possible to eliminate the unreasonable fact that the case components increase as the number of subjects increases.

(2) Double-defined rule If the invention-specific matters already defined in the previous stage or the processing operations defined in the previous stage are counted again in the subsequent stage, double counting will occur. For this reason, it is not counted in the subsequent stage.

(2-1) “Registration by the registration unit” → If the registration unit has already completed the action of performing registration , the registration unit does not register but the registration unit registers .

Another example) “ Index entry that includes the key ” appears at the first time. → Include is a verb, and each case component is counted separately. Next, “ index entry including the key ” has appeared for the second time → Since it received the previous one, it is not counted separately, but it is a single component.

(2-2) “Imaging means for creating A and A created by the imaging means ”

As for “ A created by the imaging means” , when the number of case components is to be counted purely, the part “by the imaging means” has a morpheme “by”, The “imaging means” can be said to be the number of case components that can be counted independently. However, the proposition of creating A by the imaging means has already been realized in the previous stage. For this reason, if the “imaging means” is counted again here, the proposition of creating A is realized again, so that the proposition is realized twice. Since it deviates from the purpose of the number of case components as a parameter in accordance with the feasibility of the proposition, here, the word “created by the imaging means ” is combined with “A” to obtain the number of case components To do.

(2-3) “ Number of printed sheets calculating means for calculating the number of printed sheets and the number of printed sheets calculated by the number of printed sheets calculating means”

As for “the number of printed sheets calculated by the above-mentioned number of printed sheets calculating means” , when the number of case components is to be counted purely, the part “by the number of printed sheet calculating means” has a morpheme “by”. Therefore, it can be said that “the number of printed sheets calculation means” is the number of case components that can be counted independently. However, the proposition of calculating the number of printed sheets by the printed number calculating means has already been realized in the previous stage. For this reason, if “the number of printed sheets calculation means” is counted again here, the proposition of calculating the number of printed sheets is realized again, so that the proposition is realized twice. Become. Since it deviates from the purpose of the number of case components as a parameter according to the feasibility of the proposition, here, the term “calculated by the above-mentioned number-of-print-number calculating means” is combined with “number of printed sheets” to obtain one case component It is a number.

  For those not counted as case components in (2-1) to (2-3), a process of setting a flag is performed.

  Further, this double definition rule may be based on the following processing operation. First, the subject on which the verb is involved is identified, and then the case component that satisfies the operation condition of the verb is extracted.

For example, in the case of the following sentence, “subject to which the verb depends”, “verb”, and “case component” are specified as follows.
“Rotation angle signal detecting means for detecting a rotation angle signal output from the output shaft of the internal combustion engine”

  Make a list like this one by one. It is not essential to create a list, and a pair of “subject subject to verb”, “verb”, and “case component” may be stored.

  It is assumed that, for example, the phrase “based on the rotation angle signal detected by the rotation angle signal detection means” is defined in the subsequent stage.

  The following logic is used to determine whether or not the “rotation angle signal detecting means” that depends on the verb “detected” is counted as a case component.

  The pair of “the rotation angle signal detecting means” and “detected (performed)” coincides with the previously stored pair. For this reason, this is not counted because it is a proposition already executed.

Since “~ rotation angle signal to be detected” does not depend on the verb “detected” (reversely modified by the verb), the case component is related to the verb “detected” in the first place. It is not necessary to determine whether or not, and it is counted as a case component.

  In this way, a pair of “subject to which the verb is dependent”, “verb”, and “case component” are stored, and for the “verb”, these “subject to which the verb is dependent”, “case” If a component that depends on the component comes out later, it is not counted as a double definition.

Example 1 Claim 1 of p2000-196550
“In a vehicle equipped with an internal combustion engine as a drive source, vehicle speed detection means for detecting the vehicle speed of the vehicle, braking operation detection means for detecting the state of an operation input to the braking device of the vehicle operated by the driver, Exhaust gas recirculation means for recirculating part of the exhaust gas from the exhaust path into the intake pipe, and when the vehicle is found to be stopped based on the vehicle speed detected by the vehicle speed detection means, And a control means for reducing the amount of exhaust gas recirculated by the exhaust gas recirculation means when the braking operation detection means detects the release of the operation input in the stopped state. Control device for internal combustion engine for vehicle. "
In this case, the following list can be created.

  For example, “based on the vehicle speed detected by the vehicle speed detecting means” of 6 is already executed in 2 because “by the vehicle speed detecting means” is related to the verb “detected”. It is not counted as a case component.

  No. 8 “recirculate exhaust gas into the intake pipe by the exhaust gas recirculation means” has already been realized in 5 because of the verb relationship “by the exhaust gas recirculation means” and “in the intake pipe” “recirculate”. The “exhaust gas” that is not counted is counted as a first-order component because it has come out for the first time because of the verb “recirculate”.

  Thus, checking whether or not it has already been executed is not limited to focusing only on the “verb” and “subject subject”, but on the “verb” and “case component” pair. Also pay attention.

  9 “Release of the operation input is detected by the braking operation detection means”, “by the braking operation detection means” and “detected” satisfy the condition of 4 and are not counted. Canceling operation input appears for the first time. (The release of the operation input) is counted as a case component.

  The point is that when referencing the list during this extraction process, it should be taken into account that active and passive changes can be made to the verbs described in the list.

Example 2 Claim 1 of P2000-258811
“In a printer control apparatus for controlling a plurality of types of printers each having a different resolution, an image selection means for selecting an image to be printed from a plurality of images, the number of pixels of the image selected by the image selection means, and the above Based on the resolution, printer determination means for determining a printer to be used from among the plurality of types of printers, and image data representing the image selected by the image selection means to the printer determined by the printer determination means A printer control device having transmission means for transmitting the

  In this case, the following list can be created.

  The “number of pixels of the image selected by the image selection means” of 4 is already executed because there is a pair of “the image selection means” for the verb “select”.

  Similarly, “to the printer determined by the printer determination unit” of 6 is already executed because there is a pair “by the printer determination unit” for the verb “determined”.

  In addition, “the image selected by the image selection unit” of 7 is already executed because there is a pair “image selection unit” for the verb “select”.

  What has already been executed is not counted as a case component.

  That is, in this double definition rule, after all, based on the following algorithm, it is determined whether to include or exclude from case components to be extracted.

動詞をベースにして、これに係り受けする格成分と、その動詞が係り受けする主体を特定する。次に、上段に同一の動詞（能動態と受動態の違いはあるにしても）があるか否かをチェックする。その結果、上段に同一の動詞が無い場合は、特に二重定義のチェックは行わない。これに対して、上段に同一の動詞がある場合には、以下の工程２へと移行する。 (2-4) Step 1

Based on the verb, the case component that depends on this and the subject that the verb depends on are specified. Next, it is checked whether or not there is the same verb (even if there is a difference between active and passive) in the upper row. As a result, if there is no identical verb in the upper row, no double definition check is performed. On the other hand, when there is the same verb in the upper stage, the process proceeds to the following step 2.

  (2-5) Step 2

  For each verb, the identity of the subject and case component to which the verb depends is determined. The above is between the “subject to which the verb depends” and “case component” related to the same verb (number 5) that appears first, and the case component related to the same verb (number 8) that appears later. It is an example which determines identity.

  (2-6) Step 3

  If there is the same thing between the “subject subject to the verb” and the “case component” for the same verb that appears first, and the case component for the same verb that appears later, The case component of the same verb that comes from is not considered a case component (because the proposition has already been executed in the upper row). Those that are not actually regarded as case components are identified by setting a flag.

  The above processing operation is similarly performed for all other verbs.

  In other words, the algorithm picks up verbs sequentially. Check if the picked-up verb has already appeared before. If it appears, the identity is determined between the “subject subject to the verb” and the “case component” relating to the same verb appearing first and the case component relating to the same verb appearing later. As a result of the determination, if the same thing appears, the case component of the same verb that comes out later is not regarded as the case component.

  At this time, it may be realized through the following steps.

  An identification step of sequentially extracting verbs and identifying whether the extracted verbs are already defined in the upper stage. As a result of identification in the identification process, when it is identified that the extracted verb has already been defined in the upper stage, the noun or noun phrase that depends on the verb already defined in the upper stage, and the upper stage A specifying step of specifying a noun or a noun phrase indicating a subject on which a verb already defined in (2) depends. An identity determination step of determining identity between the noun or noun phrase specified in the specifying step and the noun or noun phrase that depends on the extracted verb. Of the nouns or noun phrases that depend on the extracted verb, those that are determined to be the same in the identity determination step are excluded from the case components to be extracted by setting a flag.

  In the identification step, words composed of a plurality of nouns or noun phrases connected by morphemes may be sequentially extracted to identify whether the extracted words are already defined in the upper stage. Then, in the exclusion processing step, when the extracted wording is identified as a result of the identification in the identification step, the extracted wording is excluded from the case components to be extracted. A flag may be set to

For example, when “ when connected to the multi-control unit ...” Is displayed, the underlined portion is extracted as a case component for the first time. Further, in the same claim, if the same wording appears after the second time, " When connected to the multi-control unit ...", after the second time, these are collectively 1 to avoid double counting. The processing operation is performed so as not to extract the case components as case components.

  That is, “noun or noun phrase” + “morpheme” + “noun or noun phrase” is also checked for double definition.

(3) Case component exclusion of method invention (3-1) Product invention pattern Exactly the same invention is defined by a writing method and a requirement enumeration method as follows.
・ Writing method " ~ B is arranged in ~ A ,
C consisting of this B are connected,
Device characterized in that D is attached to C above "
= 7 case components

・ Requirement enumeration method
~ B arranged in A above ;
Connected to said B, a C consisting of,
A device characterized by comprising D attached to C above "
Number of 4th grade components (8th grade components with components)

  When the constituent elements are not counted as in the requirement enumeration method, the number of underlined case components is the number of conditions necessary for realizing the proposition to be realized by each constituent element. In the writing system, the number of case components underlined is the number of conditions necessary for the proposition to be realized with the title (device) of the invention.

  Even in the requirement enumeration method, the proposition realization target of the number of case components to be obtained is not a component but ultimately a device. That is, in the requirement enumeration method, the device includes A, B, C, and D. Then, it can be considered that each of A, B, C, and D realizes each proposition. When it is defined that “a device includes A, B, C, and D”, the conditions necessary for realizing the proposition that A, B, C, and D are “provided” by the subject “device” are as follows. Therefore, it becomes a case component.

  For this reason, even if it is a requirement enumeration method and each component is counted as a case component, it is logically consistent. In addition, the disparity due to the difference between the requirement enumeration method and the writing method can be eliminated.

(3-2) Pattern / requirement enumeration system of method invention
B step of adding P to O produced by the above A step ;
C process of heating Q obtained in the B process with R ;
And a D step of extracting heated S in the C step.
Underlined part: 8 Number of steps: 4 Total 12
・ Writing system “ P is added to O manufactured by the A process ,
Q obtained in the above step B is heated with R ,
A method characterized by extracting S heated in the above-mentioned step C "
Underlined part: 8 Number of steps: 0 Total 8

  The constituent elements (steps A to D) are merely names, and are not particularly limited to the matters specifying the invention. If the component is put into the case component, the difference between the methods will be conspicuous.

  From the above-described examination, components (processes or steps) appearing in the method invention may not be included in the case components. Moreover, you may make it count by including the component which comes out in the invention of a thing in a case component.

  Here, as a specific process for excluding components (processes and steps) appearing in the invention of the method from the case components, the wording considered as the name of the invention in the preprocessing step S11 includes “method”. "~ Step and (O)", "~ Step and (O)", "~ Process and (O)", "Function and (O)", "~ Process and (O)", "~ Extract processing and ()) and flag them. Thereby, in the invention of the method, it is possible to perform processing so that these steps and the like are not included in the case components. Alternatively, as an alternative to the above “and ()”, if “,” is hit, a flag may be set. Alternatively, “,” + “step / stage / process / function / process / process” + “as” may be extracted and flagged.

(4) Other Case Component Extraction Rules In this case component extraction step S14, a flag may be set for “any” or “any” extracted as a noun. Next, when there are a plurality of words that are directly related to “any” or “any”, the preceding noun phrase is excluded from the case components to be extracted. That is, when it is described as “A member or B member”, the noun phrase depending on any of them is “A member, B member”, of which the preceding A member is Flag to exclude from case components to be extracted. Since “any” or “any” is flagged, only “B member” is extracted as the case component in this example. However, the total number of case components is 1, and accurate writing component extraction is realized.
In addition, the process of setting a flag every time the preceding word is a noun or a noun phrase is directly performed on “any” or “any” and thereafter.

  Further, in this case component extraction step S14, a noun phrase including a noun + “target / free / possible / impossible / impossible / you” + (“na”, “ni”, “no”, “to”), or these A flag may be set for a sentence in which hiragana is added to the noun phrase.

  For example, by setting a flag for a noun phrase or wording such as “retractable”, “fitable”, “non-slidable”, etc., it can be processed so that these are not extracted as case components. .

  In the case component extraction step S14, when a word composed of “noun or noun phrase” + “A” + “noun or noun phrase” is extracted, A = “up to”, “to”, “to”, In either case of “de” or “ya”, individual “nouns or noun phrases” separated by “A” are extracted separately as case components. Further, when A = “out”, each “noun or noun phrase” separated by “A” is extracted as one case component. When processing is performed with a flag set, a flag may be set for nouns or noun phrases positioned before and after A. As a result, for example, when there are words such as “consistency with the authentication card” and “distance to the user interface”, these can be extracted substantially as one case component.

  In the case component extraction step S14, a pronoun consisting of any one of “this”, “it”, “that”, “which”, “here”, “there”, “that”, “where” should be extracted. It may be included in the case component. That is, these pronouns may be processed so as not to set a flag. This is because even if they are pronouns, they function as case components as described above, and should be included in case components. Thereby, case component extraction with high accuracy is possible.

  On the other hand, adverbs (“Kou”, “So”, “Oh”, “How”) and conjunctions (“Kon”, “That”, “That”, “Which”) are extracted. You may make it exclude from the case component which should be.

  In the case component extraction step S14, the noun form of the verb in which the verb is connected to “koto” may be extracted as the case component. That is, the words displayed as “verb” + “that” are extracted as one case component only by this. For example, “to be operated” → “operated” is a verb, but “that” constitutes a noun phrase. Noun phrases become case components. “Elapsed” is a noun form and is counted. In other words, the noun form of the verb that connects the verb to “Koto” is extracted as a case component.

  In this way, it is determined whether or not the case component to be extracted is included in the above-described predetermined rule, and processing for extracting a component that conforms to the predetermined rule as the case component is performed. .

  Moreover, in case component extraction step S14 mentioned above, it is not limited to the case where all the rules are performed, Arbitrary one rule may be performed.

  In the case component extraction step S14, case components may be extracted from noun phrases obtained by removing “features” in “characteristic of” from the character string. In other words, the “feature” in “characterized by” is a noun but is not extracted as a case component. On the other hand, if necessary, the “feature” in “feature with” may be extracted as a case component. It goes without saying that which extraction policy should be set can be freely set on the user side.

Concept hierarchy identification process In the case component extraction program to which the present invention is applied, a concept hierarchy identification step may be further provided in step S15. A detailed process in this concept hierarchy specifying step will be described below.

  In the concept hierarchy specifying step S15, a concept dictionary in which noun concepts are stored in a hierarchy from a higher concept to a lower concept is referred to.

  The superordinate concept here refers to the case where the word A indicates a more general, more generic, and more abstract thing including the word B. The subordinate concept refers to a part that is more specific, more individual, and more specific among those indicated by the word B. For example, in the example of the elastic body shown in FIG. 4A, the elastic body (word A) refers to a more general, more generic, more abstract material including a spring or rubber (word B). Therefore, an elastic body is a superordinate concept to a spring or rubber. On the other hand, the spring or rubber (word B) refers to a more specific, more specific, more specific part of the elastic body (word A). For this reason, it can be said that a spring or rubber is a superordinate concept to an elastic body.

  Similarly, as shown in FIG. 4B, a communication device is a superordinate concept of a mobile terminal or a fixed phone, whereas a mobile terminal or a fixed phone is a subordinate concept of a communication device. A mobile terminal is a superordinate concept for smartphones, mobile phones, and tablet terminals, while a smartphone, a mobile phone, and tablet terminals are subordinate concepts for mobile terminals.

  As described above, various words are arranged and stored in the concept dictionary from the upper concept to the lower concept. In such a concept dictionary, what is hierarchized from a superordinate concept to a subordinate concept is hereinafter referred to as a concept hierarchy.

  In FIG. 4A, the elastic body is in a higher hierarchy than the rubber or spring, and is in a lower hierarchy than the rubber, spring, or elastic body. In FIG. 4B, the communication device is in a higher hierarchy than the mobile terminal and the fixed phone, and the mobile terminal and fixed phone are in the lower hierarchy than the communication device. The mobile terminal is in a higher hierarchy than the smartphone, the mobile phone, and the tablet terminal, and the smartphone, the mobile phone, and the tablet terminal are in a lower hierarchy than the mobile terminal.

  In the concept hierarchy specifying step S15, the noun concept hierarchy included in each case component extracted in the case component extracting step S14 is specified. For example, when the phrase “sent from a mobile terminal” is extracted, refer to the conceptual dictionary described above for the conceptual hierarchy of the noun “mobile terminal” included in “from mobile terminal” as a noun phrase. Identify. If the phrase “sent from a tablet-type terminal” is described, the conceptual hierarchy of the noun “tablet-type terminal” is identified with reference to the concept dictionary. A lower hierarchy is identified.

  Note that the present invention is not limited to the case where the noun concept hierarchy of all case components included in the extracted wording is specified, and the noun concept hierarchy of some case components may be specified. Further, if the case component noun is not included in the concept dictionary, the concept hierarchy need not be specified, or an appropriate concept hierarchy may be assigned.

  Any known dictionary in which the concept hierarchy is defined may be used.

  In the concept hierarchy specifying step S15, after specifying the concept hierarchy of the nouns included in each case component, the process proceeds to the post-processing step S16.

  In post-processing step S16, the number of case components extracted in case component extraction step S14 and the noun concept hierarchy of each case component specified in concept hierarchy specifying step S15 are output or stored. In post-processing step S16, the evaluation value may be calculated based on the number of extracted case components and the noun conceptual hierarchy in each case component.

  The evaluation value may be calculated based on any method that uses the number of case components extracted as explanatory variables and the noun concept hierarchy of each case component. Also, it is not limited to using the concept hierarchy of all case component nouns included in the extracted wording as explanatory variables, but may use the concept hierarchy of some case component nouns as explanatory variables. Good.

  When calculating the evaluation value, the concept hierarchy may be digitized and included in the arithmetic expression. For example, FIG. 5 (a) is an example in which concept hierarchies are described from 1 to 3 from the upper concept to the lower concept for food and in FIG. As shown in FIG. 5 (a), it may be set such that the first hierarchy has the largest value and the third hierarchy has the smallest value from the first hierarchy that is the higher concept to the third hierarchy that is the lower concept. However, as shown in FIG. 5B, the value may be set so that the value is the smallest in the first layer and the value is largest in the third layer. Also, the actual allocation value may be any change rate from the upper layer to the lower layer.

  The concept hierarchy of nouns of case components included in the wording extracted in this way is digitized, and an evaluation value is calculated based on this. When calculating the evaluation value, for example, for one claim, if the number of case components is 24 and the conceptual hierarchy average of nouns of each case component included in the extracted wording is 2.1, output it as it is. You may output as an evaluation value. Further, the number of case components and the concept hierarchy average of nouns of each case component may be calculated based on any calculation formula and output as an evaluation value.

  In general, the broader the concept, the wider the patent, and the lower the concept, the narrower the patent. For this reason, the evaluation value may be increased as the conceptual hierarchy becomes higher and the evaluation value may be reduced as the conceptual hierarchy becomes lower. That is, an evaluation value that broadens the patent as the concept hierarchy becomes higher, and an evaluation value that becomes narrower as the concept hierarchy becomes lower. calculate.

  For example, the smaller the number of case components, the wider the right, and the larger the number, the narrower the right. For this reason, as shown in FIG. 5B, the value is made smaller for the higher concept and the value is made larger by the lower concept. Then, by multiplying the number of case components and the numerical value assigned to the concept hierarchy, for example, the size of the grammatical structural units and the concept hierarchy of each noun are taken into account and the breadth of the patent is evaluated. It becomes possible to do.

  In the case of the sentence “with the rotation angle signal detecting means for detecting the rotation angle signal output from the output shaft of the internal combustion engine” as shown in the example of Table 5, the internal components, the output shaft, the rotation angle signal are included in the case components. The noun is included. It is specified what concept hierarchy the internal combustion engine belongs to, and what concept hierarchy the output shaft belongs to. As a result, when the higher concept is assigned so that the score is lower, the “internal combustion engine” has 2 points, the “output shaft” has 3 points, and the “rotation angle signal” does not exist in the concept dictionary. An average of 2.2 points in the hierarchy will be assigned.

  The “output shaft of the internal combustion engine” is one component, but 2.5 points, which is the average of the two points of the internal combustion engine and the three points of the output shaft, is assigned, and 2.2 points are assigned to the “rotation angle signal”. In this way, each case component is weighted according to the concept hierarchy specified for each noun phrase contained therein. Then, the evaluation value may be calculated based on the sum of the weighted case components. In the above example, the total score (4.7 points) can be used as the evaluation value.

  By doing this for one claim, find the breadth of the patent taking into account both the grammatical structural unit size and the noun conceptual hierarchy of all the limiting elements in one claim. Is possible.

  Note that the weighting method, the degree of weighting, and how the evaluation value is calculated from each weighted case component may be any value using existing statistical analysis techniques.

  Note that the concept dictionary to be used may be one in which noun concepts are stored in a hierarchy for each technical field. In the description of the patent specification, there are many cases where technical terms are used for each technical field, and therefore, a concept hierarchy is arranged in advance for each technical field. Then, a concept dictionary corresponding to the technical field of the patent specification to be extracted is referred to. At this time, it may be arranged in advance in the scope of the claims in the patent specification, or may be analyzed and arranged for the whole sentence including the embodiment.

  For example, when the patent from which the case component is to be extracted is a communication equipment related technology, the concept dictionary in which the technical field is hierarchized is referred to. In the concept dictionary, words as shown in FIG. 4B, for example, corresponding to communication devices are hierarchized. By using this, more accurate hierarchization can be realized. In particular, a patent specification may have a completely different terminology if the technical field is slightly different, so it is desirable to use a concept dictionary arranged for each technical field corresponding to the patent to be extracted.

  Further, the concept dictionaries that are conceptually hierarchized for each technical field may be normalized for each technical field. For example, when the description of past patent specifications with the technical field as the population is analyzed using text mining technology, the statistical distribution of the words used from the superordinate concept to the subordinate concept may be obtained. it can. Based on the distribution, some normalization processing may be performed from the superordinate concept to the subordinate concept, and this may be reflected in the evaluation value. As for the normalization processing, any known method may be used as long as it is performed on the distribution obtained by using the text mining analysis based on the past population of each technical field.

  The concept dictionary is not only implemented in the program according to the present invention, but also uses a concept dictionary newly received via a network, or a concept dictionary newly acquired from other software or another database. Of course, it may be.

  Through the process described above, it is possible to evaluate the breadth of patents by taking into account both the size relationship of structural units in the grammar and the concept hierarchy of each noun.

Frequency rank identification process The present invention is not limited to the above-described embodiment. In the above-described embodiment, as an alternative to the conceptual hierarchy specifying step S15, a frequency rank specifying step S17 may be provided as shown in FIG.

  This frequency rank specifying step specifies the frequency ranks of nouns included in each case component. The frequency rank obtains information on the appearance frequency of nouns by analyzing the description of the patent specification in advance with a conventional patent group as a population. And the rank of the application frequency of the noun contained in each case component extracted in case component extraction step S14 is specified with reference to the information regarding the appearance frequency of the noun.

  The rank of appearance frequency is the frequency of the wording in each specification of a patent group filed in a certain period (which may be analyzed for any column such as claims and full description). It is the rank which shows whether it has appeared in. For example, in the case of patents in the communication technology field, if there are 100,000 "communication devices", 60,000 "mobile terminals", and 30,000 "tablet terminals" in individual terms, The “communication device” has the highest appearance frequency rank, the next highest appearance frequency rank is the “mobile terminal”, and the lowest appearance frequency rank is the “tablet terminal”. That is, the rank of the appearance frequency is higher as the appearance frequency is higher. Such appearance frequencies are analyzed in advance, and the ranks are ranked from high to low. The number of ranks for ranking and the rules for ranking may be determined based on any rule on the assumption that the ranks are hierarchized according to the appearance frequency. Information on the appearance frequency and its rank may be stored in a database in advance and attached to the program.

  In other words, the appearance frequency rank is based on the concept of document-frequency, which is considered to be uncharacteristic in the population, in terms of many patents in the population. In other words, a high appearance frequency (rank) means a general term that is not so characteristic because it is a commonly used word. On the other hand, those having a low appearance frequency (rank) are rare as words used in the specification, and are characterized only by being limited to the words.

  In other words, based on the concept of document-frequency, words with a low appearance frequency (rank) in a population of multiple patents become narrower and narrower concepts. Conversely, words with a high appearance frequency (rank) I think that it will be a broader concept in generalized terms. In the frequency rank specifying step S17, the rank of the appearance frequency of nouns included in each case component extracted in the case component extracting step S14 is specified. For example, when the phrase “sent from a mobile terminal” is extracted, the appearance frequency of the above-mentioned noun is the frequency of appearance of the noun “mobile terminal” included in “from the mobile terminal” as the noun phrase. Identify and refer to information about. If the phrase “sent from a tablet-type terminal” is described, the rank of the appearance frequency of the noun “tablet-type terminal” is identified with reference to the information on the appearance frequency. It is identified that it is lower than the “terminal”. In other words, since the tablet type terminal is more characteristic than the communication device and the portable terminal, it is limited in that amount and is located in the lower hierarchy.

  In addition, it is not limited to specifying the ranks of appearance frequency of nouns of all case components included in the extracted wording, and is to specify the ranks of appearance frequency of nouns of some case components. Also good. Further, if the case component noun is not included in the appearance frequency hierarchy, the appearance frequency rank need not be specified, or a suitable appearance frequency rank may be assigned.

  In the frequency rank specifying step S15, the rank of the appearance frequency of nouns included in each case component is specified, and then the process proceeds to the post-processing step S16.

  In post-processing step S16, the number of case components extracted in case component extraction step S14 and the rank of appearance frequency of nouns in each case component specified in frequency rank specifying step S15 are output or stored. In post-processing step S16, the evaluation value may be calculated based on the number of extracted case components and the rank of the appearance frequency of nouns in each case component.

  The evaluation value may be calculated based on any method that uses the number of case components extracted as explanatory variables and the rank of the appearance frequency of nouns in each case component. In addition, it is not limited to the case where the ranks of the appearance frequencies of all case components included in the extracted wording are used as explanatory variables, but the ranks of the appearance frequencies of some case components are used as explanatory variables. It may be.

  When calculating the evaluation value, the rank of the appearance frequency may be digitized and included in the arithmetic expression. As for the rank of appearance frequency, the numerical value may be set higher from the higher level, or the numerical value may be set lower from the higher level. Further, regarding the rank of appearance frequency, the actual allocation value may be any change rate from the upper layer to the lower layer.

  The rank of the appearance frequency of nouns of case components included in the wording extracted in this way is digitized, and the evaluation value is calculated based on this. When calculating the evaluation value, for example, for one claim, if the number of case components is 24 and the rank average of the appearance frequency of each case component included in the extracted wording is 2.1, it is output as it is. And may be output as an evaluation value. Further, the number of case components and the rank average of the appearance frequency of nouns of each case component may be calculated based on any calculation formula and output as an evaluation value.

  The higher the frequency of appearance, the lower the degree of limitation in general terms, and the closer the term of the higher concept, the broader the patent. The lower the frequency of appearance, the higher the degree of limitation with more limited special terms and the lower level. Suppose that it is close to the terminology of the concept and the breadth of the patent is narrowed. Therefore, the evaluation value is calculated so that the width of the patent becomes wider as the appearance frequency rank increases, and the evaluation value is calculated so that the width of the patent becomes narrower as the appearance frequency rank becomes lower. It may be.

  Also, on a case-by-case basis, the evaluation value is calculated so that the patent width becomes narrower as the appearance frequency rank becomes higher, and the patent value becomes wider as the appearance frequency rank becomes lower. May be calculated.

  For example, the smaller the number of case components, the wider the right, and the larger the number, the narrower the right. Therefore, as shown in FIG. 5B, the higher the appearance frequency rank, the smaller the value, and the lower the appearance frequency rank, the larger the value. And by multiplying the number of case components and the frequency assigned to the rank of appearance frequency, for example, taking into account both the size relationship of the structural units in the grammar and the concept of each noun, the breadth of the patent It becomes possible to evaluate.

  In the case of the sentence “with the rotation angle signal detecting means for detecting the rotation angle signal output from the output shaft of the internal combustion engine” as shown in the example of Table 5, the internal components, the output shaft, the rotation angle signal are included in the case components. The noun is included. It is specified which rank of appearance rank each internal combustion engine and output shaft belong to. As a result, if the higher the appearance frequency rank, the lower the score, the “internal combustion engine” is assigned 2 points, the “output shaft” is 3 points, and the “rotation angle signal” is 1.2 points. I will do it.

  The “output shaft of the internal combustion engine” is one component, but 2.5 points, which is the average of the two points of the internal combustion engine and the three points of the output shaft, is assigned, and 1.2 points are assigned to the “rotation angle signal”. In this way, each case component is weighted according to the rank of appearance frequency specified for each noun phrase included therein. Then, the evaluation value may be calculated based on the sum of the weighted case components. In the above-described example, the total score (3.7 points) can be used as the evaluation value.

  By doing this for one claim, find the breadth of the patent taking into account both the grammatical structural unit size and the noun conceptual hierarchy of all the limiting elements in one claim. Is possible.

  Note that the weighting method, the degree of weighting, and how the evaluation value is calculated from each weighted case component may be any value using existing statistical analysis techniques.

  Note that the appearance frequency information to be used may be information in which ranks of appearance frequencies of nouns are stored in a hierarchical manner for each technical field. Since the technical specifications are often used for each technical field in the description of the patent specification, the ranks of appearance frequencies for each technical field are arranged in advance, including statistical processing, and hierarchized. Then, the rank information of the appearance frequency corresponding to the technical field of the patent specification to be extracted is referred to. At this time, it may be arranged in advance in the scope of the claims in the patent specification, or may be analyzed and arranged for the whole sentence including the embodiment.

  For example, when the patent from which the case component is extracted is a communication device-related technology, the information in which the rank of the appearance frequency is hierarchized is referred to in the technical field. In the information regarding the rank of the appearance frequency, for example, the rank of the appearance frequency of words as shown in FIG. By using this, more accurate hierarchization can be realized. In particular, if a technical specification is slightly different in a technical field, the wording used in the specification may be completely different. Therefore, information on the rank of appearance frequency arranged for each technical field corresponding to the patent to be extracted is used. It is desirable.

  In addition, information in which ranks of appearance frequencies are hierarchized for each technical field may be normalized for each technical field. For example, when the description of a past patent specification having the technical field as a population is analyzed using a text mining technique, a statistical distribution from the higher rank to the lower rank of the appearance frequency can be obtained. Based on the distribution, some normalization processing may be performed from the higher rank of the appearance frequency to the lower rank, and this may be reflected in the evaluation value. As for the normalization processing, any known method may be used as long as it is performed on the distribution obtained by using the text mining analysis based on the past population of each technical field. Furthermore, in the present invention, as described above, not only when the appearance frequency is identified from a plurality of patent populations, but also the appearance frequency is identified from one specification that is an analysis target of the case component. May be. Text mining analysis is performed on one patent specification to be analyzed based on the description of the claims and the description of the entire specification. As a result, the number of appearances of the wording used in the one patent specification can be obtained, and the appearance frequency can be obtained therefrom.

  As a result, words frequently appearing in the one patent specification can be considered characteristic in the patent. On the other hand, a word that appears only infrequently in the one patent specification can be considered as a generalized term that is non-characteristic in the patent. In this way, the appearance frequency rank is created by focusing on only one patent specification to be analyzed. And the rank of appearance frequency is specified about the noun in each case component of the claim of analysis object based on the created appearance frequency rank. The subsequent processes are the same as described above.

  In other words, as the appearance frequency rank, words frequently appearing in one patent specification are performed based on the term-frequency concept considered to be characteristic in the one patent specification. In other words, those with a high appearance frequency (rank) are characteristic to the patent, but are not commonly used words. On the other hand, those with a low appearance frequency (rank) are words that are frequently used and are recognized as being highly versatile.

  That is, based on the term-frequency concept, words with low appearance frequency (rank) in one patent specification become a broad concept that is not so limited. Conversely, words with high appearance frequency (rank) are: Think of it as a broader term in generalized terms.

  In this term-frequency, an evaluation value is calculated so that the patent width becomes narrower as the appearance frequency rank increases, and the patent value becomes wider as the appearance frequency rank becomes lower. May be calculated.

  In addition, on a case-by-case basis, the evaluation value is calculated so that the width of the patent increases as the appearance frequency rank increases, and the patent value decreases as the appearance frequency rank becomes lower. May be calculated.

  Furthermore, in the present invention, evaluation may be performed by combining document-frequency and term-frequency. In such a case, the evaluation value is calculated based on the appearance frequency rank detected for a plurality of patents as a population and the appearance frequency rank detected for the analysis target patent.

  Note that the rank information of the appearance frequency described above is not only implemented in the program according to the present invention, but also newly received from the network, newly acquired from other software or another database. Of course, it may be used.

  Through the process described above, it is possible to evaluate the breadth of patents by taking into account both the size relationship of structural units in the grammar and the concept hierarchy of each noun.

Implementation Method of the Present Invention In implementing the present invention, a personal computer (PC) is used and is performed based on a program loaded therein.

  First, a character string described in a specific column (description of claims) in a patent specification converted into electronic data is extracted from the description of the patent specification converted into electronic data. The specific column corresponds to one claim in the scope of claims. At this time, character strings may be extracted only for independent claims (for example, claim 1).

  Next, a case component is extracted from the extracted character string based on the above-described rules. The number of case components may be obtained based on the extracted case components. The number of case components means the number of case components in one claim. In the case of a main claim, the number of case components may be counted as it is from the character string extracted from the claim in which the main claim is defined. In the case of a sub claim, the claim in which the sub claim is defined. The number of case components is counted as it is from the character string extracted from, and the number of case components of the main claim on which the subclaim is dependent is added to this.

  Then, the counted number of case components is displayed, for example, via a display or the like as the degree of limitation of the invention.

  At this time, the case component may be extracted using a morpheme as shown in Table 1 as a mark. Incidentally, Table 1 here is merely an example, and it is needless to say that counting may be performed based on any other morpheme.

  The case component may be extracted based on the case component extraction process described above.

  At this time, the above-mentioned count extracts words that define overlapping actions in the extracted character string, counts the number of duplicate case components included in the extracted word, and counts the total number of case components counted You may make it display the value which subtracted the number of case components of duplication as a limitation degree of invention. This takes into account the double count described above.

  In the present invention, a character string described in a specific column in the patent specification before correction is extracted, and further a character string described in a specific column in the corrected patent specification is extracted and corrected. The total number of components is counted from the character string extracted from the previous patent specification and the character string extracted from the corrected patent specification, respectively. Alternatively, the difference from the total number of components of the corrected patent specification may be obtained and displayed as the correction degree as described above in the claims.

  Further, the present invention may be embodied as a device such as a PC on which the above-described analysis program is installed, or a program for causing a PC to execute the program or a recording medium on which the program is recorded. It may be embodied.

  Further, the present invention may be applied to a network system. First, on the server side, data obtained by previously digitizing the number of case components for the patent specification is stored in the server. When the client side requests the number of case components for a desired patent, the number of case components of the patent may be read from the server and sent to the client side via the network.

  In addition, when the number of case components has not yet been obtained for the patent requested for transmission from the client side, the server counts the number of case components for the requested patent and sends this to the client side via the network. You may make it send.

  For extraction of case components, any existing text mining technology, data mining technology, language analysis processing technology, or the like may be used. Then, using these techniques, it is determined whether or not each character string to be analyzed corresponds to a case component based on the above-described rules, and finally case components for one claim are specified. Then, the total number of the specified case components for one claim is obtained, and this is output as the number of case components of the claim.

  In the present invention, not only the number of case components for each claim but also the output data may be added to any information such as a patent map, a graph, and other evaluation values.

Incidentally, an evaluation value based on the following calculation formula may be obtained as an alternative to the number of case components described above, and this may be output or stored.
Evaluation value = {α-log (CN + β)} × γ
Where α = 3-8
β = 15-25
γ = 15-40

  Here, the above formula was derived by fitting based on the relationship between the number of case components and the patent assessment rate. The larger the evaluation value, the smaller the number of case components and the wider the technical range. This is the evaluation value of the patent called Technology Size (registered trademark). The evaluation value (Technology Size (registered trademark)) can be used in patent value evaluation and macro analysis. It can also be used as a comprehensive score in combination with other index values such as the number of citations and the number of requests for browsing.

  The evaluation value is calculated in the case component extraction step S14 or the post-processing step S16. That is, as an alternative to the number of case components described above, processing such as output, display, and transmission of the evaluation value may be performed.

  Further, regarding the evaluation value, if the concept hierarchy of the case component noun included in the word extracted in the concept hierarchy specifying step S15 is digitized, this can also be reflected when the evaluation value is calculated. Technology Size (registered trademark) is a parameter indicating that the larger the right, the wider the right, and the smaller, the narrower the right. For this reason, as shown in FIG. 5A, the value is made smaller for the higher concept and the value is made larger by the lower concept. You may make it reflect it as a correction coefficient in Technology Size (trademark) mentioned above. What coefficients and arithmetic expressions are used may be freely set by the user, or may be freely set by the software producer. In addition to the concept hierarchy, the rank of appearance frequency may be similarly reflected as a correction coefficient.

  In the present invention, the case component specified for each claim may be displayed on a screen, printed, or converted into data. That is, as shown in the following embodiments, underline and highlight display may be performed so that the case components specified for the claims can be visually grasped.

  As described above, according to the present invention, it is possible to display and display in accordance with similar rules in any claims. At this time, the number of case components may be written at the same time.

  Note that the present invention is not limited to the case component extraction described in the claims of the patent specification, and is applicable to the case component extraction for any other document based on the same method. be able to.

  Of course, the present invention is not limited to Japanese but can be applied to other languages including English.

Claims (31)

A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step,
A noun phrase extraction step for extracting a noun phrase from the character string including the word that has undergone the word inclusion step;
In the noun phrase extraction step, a case component extraction step for extracting a case component as a noun phrase required for realizing the action of a verb related to each component constituting the claim from the extracted noun phrase; and ,
A concept hierarchy that identifies the noun concept hierarchy included in each case component extracted in the case component extraction step with reference to a concept dictionary in which noun concepts are stored in a hierarchy from upper concepts to lower concepts. A case component extraction program that causes a computer to execute a specific step. Output or store the number of case components extracted in the case component extraction step and the concept hierarchy of nouns in each case component specified in the concept layer specification step, respectively, or the number of case components extracted The case component extraction program according to claim 1, further comprising a post-processing step of outputting or storing an evaluation value calculated based on a noun conceptual hierarchy in each case component. In the post-processing step, an evaluation value is set such that the breadth of the patent becomes wider as the concept hierarchy becomes higher, and the breadth of the patent becomes narrower as the concept hierarchy becomes lower. The case component extraction program according to claim 2, wherein the case component is calculated so as to be an evaluation value. In the concept hierarchy specifying step, the concept dictionary stores a concept of nouns hierarchically for each technical field, and refers to a concept dictionary corresponding to the technical field of the patent specification to be extracted. The case component extraction program of any one of 1-3. 5. The case component extraction program according to claim 4, wherein the concept hierarchy specifying step uses a concept dictionary in which the concept hierarchy is normalized for each technical field. In the post-processing step, each case component is weighted according to a concept hierarchy specified for each noun phrase included therein, and the evaluation value is calculated based on a sum of the weighted case components. The case component extraction program according to claim 2 or 3. The concept hierarchy specifying step uses the concept dictionary newly received via the network, or a concept dictionary newly acquired from other software or another database. The case component extraction program according to item 1. A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step,
A noun phrase extraction step for extracting a noun phrase from the character string including the word that has undergone the word inclusion step;
In the noun phrase extraction step, a case component extraction step for extracting a case component as a noun phrase required for realizing the action of a verb related to each component constituting the claim from the extracted noun phrase; and ,
With reference to the information on the appearance frequency of nouns by analyzing the description of the patent specification in advance using a conventional patent as a population, the frequency of appearance of nouns included in each case component extracted in the case component extraction step is A case component extraction program that causes a computer to execute a frequency rank specifying step for specifying each rank. Output or store the number of case components extracted in the case component extraction step and the frequency rank of nouns in each case component specified in the frequency rank specification step, or the number of case components extracted The case component extraction program according to claim 8, further comprising a post-processing step of outputting or storing an evaluation value calculated based on a noun frequency rank in each case component. In the post-processing step, an evaluation value is set such that the width of the patent becomes wider as the frequency rank becomes higher, and the evaluation becomes such that the width of the patent becomes narrower as the frequency rank becomes lower. The case component extraction program according to claim 9, wherein the case component is calculated so as to be a value. The frequency rank specifying step stores the frequency rank for each technical field, and refers to the frequency rank corresponding to the technical field of the patent specification to be extracted. The case component extraction program according to item 1. The frequency rank specifying step uses the frequency rank normalized for each technical field, or normalized according to the application year of the patent to be analyzed. The case component extraction program according to any one of the preceding claims. In the post-processing step, each of the case components is weighted according to a frequency rank specified for each noun phrase included therein, and the evaluation value is calculated based on a sum of the weighted case components. The case component extraction program according to claim 9 or 10. The frequency rank specifying step uses the frequency rank information newly received via the network, or frequency rank information newly acquired from other software or another database. The case component extraction program according to any one of the preceding claims. A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step,
A noun phrase extraction step for extracting a noun phrase from the character string including the word that has undergone the word inclusion step;
In the noun phrase extraction step, a case component extraction step for extracting a case component as a noun phrase required for realizing the action of a verb related to each component constituting the claim from the extracted noun phrase; and ,
Frequency of identifying the rank of appearance frequency of nouns included in each case component extracted in the case component extraction step with reference to information on appearance frequency of nouns by analyzing description of patent specification to be analyzed A case component extraction program causing a computer to execute a rank specifying step. Output or store the number of case components extracted in the case component extraction step and the frequency rank of nouns in each case component specified in the frequency rank specification step, or the number of case components extracted The case component extraction program according to claim 15, further comprising a post-processing step of outputting or storing an evaluation value calculated based on a noun frequency rank in each case component. In the case component extraction step, the computer is caused to execute “to extract at least as a case component” a noun phrase excluding the “feature” in the character string from the character string. The case component extraction program according to any one of claims 1 to 16. Evaluation value calculation means for calculating the evaluation value by the case component extraction program according to any one of claims 2 to 7 and 9 to 17,
A server that stores at least one of the evaluation value obtained by the evaluation value calculating means and a figure or data based on the evaluation value;
An information transmission system comprising: a transmission unit that reads the evaluation value requested from the client side and a figure or data based on the evaluation value from the server and transmits the read value to the client side via the network. A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step and the above word inclusion step A noun phrase extraction step for extracting a noun phrase from the character string including the word that has undergone
In the noun phrase extraction step, if A and B are defined as alternative noun phrases in the extracted noun phrases, either A or B is defined. A case component extraction program that causes a computer to execute a case component extraction step that excludes from a case component to be extracted and extracts at least the remaining noun phrases as case components. A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step,
A noun phrase extraction step for extracting a noun phrase from the character string including the word that has undergone the word inclusion step;
In extracting a case component as a noun phrase required to realize the action of a verb related to each constituent element constituting a claim from the noun phrase extracted in the noun phrase extracting step, noun + parallel A case component extraction program that causes a computer to execute excluding a word composed of either “ka” or “ ya ” as a particle from case components to be extracted. A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step,
A noun phrase extraction step for extracting a noun phrase from the character string including the word that has undergone the word inclusion step;
In extracting the case component from the noun phrase extracted in the above noun phrase extraction step, the extracted character string is either “to is”, “to is”, “to” or “to”. If there is a relationship between the subject to be defined and the verb that depends on it, and if the wording corresponding to the subject has already been described above, the wording corresponding to the subject Is excluded from the case component to be extracted, and if the word corresponding to the subject is not yet described in the upper stage, the case component to be processed so that the word corresponding to the subject is included in the case component to be extracted A case component extraction program that causes a computer to execute an extraction step. A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step,
A noun phrase extraction step for extracting a noun phrase from the character string including the word that has undergone the word inclusion step;
In extracting the case component as the noun phrase required to realize the action of the verb related to each component constituting the scope of claims from the noun phrase extracted in the noun phrase extracting step, the verbs are sequentially When the identification step of extracting and identifying whether the extracted verb is described in the upper stage, and as a result of the identification in the identification process, the extracted verb has been already described in the upper stage Includes a noun phrase that depends on the identified verb and a noun phrase that indicates a subject that the verb already described in the upper part depends on; a noun phrase that is identified in the above identifying step; The identity determination step for determining the identity between the extracted noun phrase and the noun phrase that depends on the extracted verb and the noun phrase that depends on the extracted verb are the same in the identity determination step. The determined for those, rated component extraction program characterized by executing the rated component extraction step on a computer having a exclude exclusion step from rated component to be extracted it. A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step,
A noun phrase extraction step for extracting a noun phrase as a noun phrase requested to realize the action of a verb relating to each component constituting the claim from the character string including the word that has undergone the word inclusion step When,
In extracting the case component from the noun phrase extracted in the noun phrase extraction step, "any" or "any" extracted as a noun is excluded from the case components to be extracted, and the "any" Or a case component characterized by causing a computer to execute a case component extraction step of excluding a case component to be extracted for a preceding noun phrase when there are a plurality of words directly related to "any" Extraction program. A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step and the above word inclusion step A noun phrase extraction step for extracting a noun phrase as a noun phrase requested to realize the action of a verb related to each component constituting the claim from the character string including the word that has undergone
In extracting the case component from the noun phrase extracted in the noun phrase extraction step, among the extracted noun phrases, the noun + “target / free / possible / impossible / impossible” + (“na” / “ (2), (2), (2), (2), (2), (2), (2), and (2) a case component extraction step of excluding from the case components to be extracted words having hiragana added to these noun phrases. Case component extraction program. A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step,
A noun phrase extraction step for extracting a noun phrase from the character string including the word that has undergone the word inclusion step;
In extracting a case component from the noun phrase extracted in the noun phrase extraction step, when a word composed of “noun or noun phrase” + “A” + “noun or noun phrase” is extracted, A = “until In the case of any of “No”, “To”, “Tono”, “In”, “Ya”, each “noun or noun phrase” separated by “A” is separately used as a case component. A case component extraction program that causes a computer to execute a case component extraction step to be extracted. A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step,
A noun phrase extraction step for extracting a noun phrase from the character string including the word that has undergone the word inclusion step;
In the noun phrase extraction step, when extracting a case component from the noun phrase extracted in the noun phrase extraction step, among the extracted character strings, “this”, “it”, “that”, “which , “Here”, “there”, “that”, “where”, a case component extraction program that causes a computer to execute the inclusion of a pronoun consisting of any one of the case components to be extracted. A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step,
A noun phrase extraction step for extracting a noun phrase from the character string including the word that has undergone the word inclusion step;
In the noun phrase extraction step, in extracting a case component from the noun phrase extracted in the noun phrase extraction step, a case component extraction step of extracting the verb noun form as a case component, connecting the verb to “ko” A case component extraction program that causes a computer to execute and. A preprocessing step of extracting a character string from the description of the claims in the patent specification converted into electronic data;
By performing morphological analysis and syntactic analysis on the character string processed in the above preprocessing step, the character string is decomposed into words, and then noun + noun / noun + “no” + noun / noun + “out of” + noun・ Noun + “Target / Free / Possible” ・ Noun + “To” ・ Noun + “By ・” ・ “At least” + Quantity expression ・ Noun + “Yes” Summarize as a word, or combine "above", "above", "this", "this", and a conjunction with the following word as one word, or noun + "no" + "adjective" + noun and noun + A word inclusion that summarizes either “being, relating to, relating to, becoming, becoming, or being” as one word, or “abbreviated”, “slightly”, “sequential” + a noun as one word Step,
A noun phrase extraction step for extracting a noun phrase from the character string including the word that has undergone the word inclusion step;
In the noun phrase extraction step, a case component as a noun phrase required to realize the action of the verb related to each component constituting the claim is extracted from the noun phrase extracted in the noun phrase extraction step. A case component extraction step to
In the word inclusion step, a case component extraction program that causes a computer to execute “abbreviation”, “slightly”, “sequential” + a noun as one word. In the case component extraction step, the computer is caused to extract at least a noun phrase excluding “feature” in the character string from “characteristic” from the character string as a case component. Item case extraction program of any one of claim | item 19-28. In the case component extraction step, the number of extracted case components is counted and output or stored, and the evaluation value based on the number of case components, or the action of the verb related to each component constituting the claim 20. The computer is caused to output or store the number of case components (hereinafter referred to as CN) as noun phrases required for realizing the above as an evaluation value based on the following calculation formula: The case component extraction program of any one of -29.
Evaluation value = {α-log (CN + β)} × γ
Where α = 3-8
β = 15-25
γ = 15-40 A case component as a noun phrase required for realizing the action of a verb related to each constituent element constituting a claim by the case component extraction program according to any one of claims 19 to 29 is extracted. Case component extraction means;
A server that stores at least one of the number of case components obtained by the case component extraction means or an evaluation value based on the number of case components;
An information transmission system comprising: a transmission unit that reads out the number of case components or an evaluation value of a patent requested from the client side from the server and transmits it to the client side via a network.

Images