JP2021140282A - Program for extracting case component from english patent specification - Google Patents

Abstract

To accurately extract case components defined in documents converted to digital data including English patent specifications and the like.SOLUTION: A computer is caused to execute a character string extraction step of extracting a character string from description in a range of a patent claim in English patent specifications converted to digital data and a case component extraction step of extracting noun phrases from the character string extracted in the character string extraction step and extracting case components from the extracted noun phrases. In the case component extraction step, when a noun phrase modifying a verb or a noun phrase modified by a verb is described in each of a preceding stage and a following stage and the noun phrases in the preceding stage and in the following stage are consistent with each other, such processing is performed that description of the noun phrase modifying a verb or the noun phrase modified by a verb in the following stage is excluded from case components to be extracted.SELECTED DRAWING: Figure 3

Description

The present invention relates to a case component extraction program from an English patent specification that extracts a case component from a character string described in an English patent specification that has been converted into electronic data.

Conventionally, the purpose has been to check patent specifications in a short time, and in particular, even when the number of claims is kept as small as possible, it is easier to include the description of such an intermediate concept, and the scope of rights and the scope of rights Patent Document 1 discloses a specification analysis display device that can be checked in consideration of the patent rate, and further, the direction itself of the description of the specification itself can be easily checked including an intermediate concept. There is.

Further, Patent Document 2 discloses a patent specification analysis display device capable of counting and displaying the degree of limitation of an invention defined in the claims in an English patent specification.

Japanese Unexamined Patent Publication No. 2006-155151 Japanese Unexamined Patent Publication No. 2014-222371

However, according to the disclosure technique of Patent Document 2, there is a place where a method that can be easily and accurately realized in actually extracting a case component as a degree of limitation of an invention from an English patent specification has not been studied. there were.

Therefore, the present invention has been devised in view of the above-mentioned problems, and the case components defined in electronic data-converted documents such as English patent specifications can be easily extracted with high accuracy. It is an object of the present invention to provide a case component extraction program from a patent specification in English that can be used.

The first invention is a character string extraction step for extracting a character string from the description of the scope of a patent claim in an English patent specification converted into electronic data, and a noun phrase for the character string extracted in the above character string extraction step. It also has a case component extraction step that extracts a case component from the extracted noun phrase, and in the case component extraction step, the verb, verb, to When there is a description of a verb, present part, or past part, the verb, verb, to verb, present part, which depends on the same noun phrase described later in the description, and which the noun phrase depends on. An English patent specification characterized by having a computer perform processing to exclude the same noun phrase described in the latter part of the above from the case components to be extracted when the meanings of the past verbs match. It is a noun phrase extraction program from.

The second invention is a character string extraction step for extracting a character string from the description of the scope of a patent claim in an English patent specification converted into electronic data, and a noun phrase for the character string extracted in the above character string extraction step. It has a case component extraction step that extracts a case component from the extracted noun phrase, and in the above case component extraction step, character strings corresponding to the object and the complement of the fifth sentence pattern are described in the first stage and the second stage, respectively. If they match each other, let the computer perform a process to exclude the combination of the object and the noun corresponding to the noun described in the latter part from the case components to be extracted. It is a case component extraction program from a featured English patent specification.

In the third invention, in the first case component extraction step, when the character strings corresponding to the object and the complement are described in the relation of the predicate of the subject in the first stage or the second stage, they match each other. In this case, it is characterized in that the combination of the object and the character string corresponding to the complement described in the latter part is excluded from the case components to be extracted.

In the fourth invention, in the first to third inventions, in the above case component extraction step, in determining whether or not they match each other, the identity of each verb is not determined up to the commonality of the active voice or the passive voice. It is characterized in that the judgment is based only on the commonality of the verbs.

In the first to third inventions, the fifth invention determines the identity of each other's nouns or noun phrases in a, the, each, both, in the determination of whether or not they match each other in the above case component extraction step. When only said is different, it is determined that they are the same as each other.

According to the sixth invention, in the first to third inventions, in the above case component extraction step, when "configured to verb phrase" is applied from behind to a noun phrase in determining whether or not they match each other. It is characterized in that the past part of the verb is regarded as being dependent on the noun phrase from behind, and the identity is judged.

In the seventh invention, in the first to third inventions, in the above-mentioned case component extraction step, each character string in the first stage and the second stage, which is a target for determining whether or not they match each other, is when, if, in case, in, respectively. When a conditional clause consisting of any of the case, while, only, providing, or whether + noun phrase + verb is involved, if the dependent conditional clause differs between the first and second stages, they are not identical to each other. It is characterized in that it is judged to be.

According to the eighth invention, in the first to third inventions, in the above case component extraction step, in determining whether or not they match each other, the verbs are identified without determining the identity of each other's verbs up to the commonality of auxiliary verbs. It is characterized in that the judgment is based only on the commonality of.

In the present invention having the above-described configuration, it is possible to easily and accurately extract the case component as the degree of limitation of the invention defined in the claims in the English patent specification.

It is a figure for demonstrating the case component. It is another figure for demonstrating the case component. It is a flowchart which shows the processing operation step of the case component extraction program to which this invention is applied.

Hereinafter, as a mode for carrying out the present invention, a case component extraction program for extracting a case component from a character string described in an English patent specification will be described in detail with reference to the drawings.

In the following example, the case where the case component is extracted from the description of the claims in the English patent specification will be described, but the present invention is not limited to this, and the present invention is applied to any other document. Of course, it may be done.

About the theory of the number of case components

A case component has been proposed as the minimum extraction unit in a quantification method for quantifying and quantifying the breadth of the technical scope of a patented invention.

The claims are described in the so-called requirement enumeration method, which is defined by enumerating each component A, B, C as "○○ device (method) including ~ A, ~ B, and C". A verb phrase is assigned to each of the components A, B, and C, assuming that they are. That is, when each component A, B, and C is the subject, a verb phrase is assigned to each of them, and a sentence consisting of the subject and the predicate can be established. When forming a sentence as a predicate, the verbs in these verb phrases basically determine what kind of noun phrase combination to take in order to realize the movement, state, and relationship that they represent. If the function of selectively requesting the combination of noun phrases necessary for the formation of a sentence according to the categorical form of the verb meaning that the verb has is tentatively called the case control of the verb, the verb The noun phrase required as a component of a sentence can be said to be a case component for supplementing a verb.

To explain by way of example, it is assumed that the component "signal generation means" is defined by the following B-1).

B-1) "Signal generation means for generating a drive signal in response to a user's request, and ..."

At this time, the verb phrase "generate a drive signal in response to a user's request" depends on the signal generation means. In order to realize and complete the action of the verb "generate" in this verb phrase, the function of requesting the noun phrases "user's request (according to)" and "drive signal (o)" is the rule of thumb. , These noun phrases are case components. Then, noun phrases such as "user's request (according to)" and "drive signal (o)" have similar relational meanings such as "operation start condition" and "object" for the verb "generate", respectively. Is a case. What is formed by the lexical meaning of the noun phrases that realize this case component and the case as the relational meaning of those noun phrases should be realized by this verb "generate". 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 the "drive signal" must be generated as the target. Because they have to be, these two are conditional numbers for the realization of the proposition by the verb.

As shown in FIG. 1, it schematically represents a case where the number of conditions that must be cleared in order to realize a proposition by such a verb is small. Since the number of conditions for starting an action by a verb is reduced, the possibility of realizing the proposition is improved. On the other hand, FIG. 2 schematically shows a case where the number of conditions that must be cleared in order to realize a proposition by a verb increases, and in such a case, the possibility that the proposition can be realized decreases. ..

As the number of conditions increases, the number of steps for determining whether or not the conditions are satisfied increases before the operation is actually started. Even considering that there are some differences in the probabilities of satisfying each condition, the verb phrase is affected as the number of judgment steps of this condition increases, in other words, as the case component that defines the condition increases. It means that the possibility of falling under component A is reduced, and the technical scope is narrowed by the reduced possibility. On the contrary, when the number of case components is small, it means that there is a high possibility that the verb phrase corresponds to the component A, which indicates that the technical range is expanded accordingly. ..

In this way, the number of case components (number of conditions) controls the possibility of starting an action by a verb, and by extension, the possibility of realizing a proposition, which affects the breadth and narrowness of the technical scope. By counting the number of case components included in this verb phrase, using the minimum extraction unit of the conversion method as the case component, it is considered that quantification according to the wide technical range can be realized.

In the above-mentioned example of B-1), the two case components "according to the user's request" and "driving signal" which are related to the verb "generate" for realizing the proposition of the component "signal generating means". Because of the existence of, the number of case components is 2. At this time, if the case component "in response to the user's request" that defines the operation start condition of the verb "generate" does not exist, the number of case components becomes 1, and the number of case components becomes 1, regardless of the presence or absence of the user's request. Since the "drive signal" can be "generated", the possibility that the proposition of the "signal generation means" can be realized increases. This suggests that the component "signal generation means" can increase the possibility of proving the identity of the technical elements of the alleged infringement product, and the increased possibility is the alleged infringement. It is considered that it corresponds to the ease of capturing the patented invention for the product, and by extension, the wide technical range of the patented invention. Therefore, by expressing the propositional realizability of the components through the number of case components, it is considered that more optimal quantification and quantification can be achieved according to the wide technical range.

The scope of claims is described by the so-called requirement enumeration method, which is defined by enumerating each component A, B, and C as "○○ device (method) including ~ A, ~ B, and C". Assuming that, the number of case components is obtained for each of the constituent elements A, B, and C, and the total is taken as the number of case components of the XX device as a patented invention. Assuming that 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 composed of these is assumed. The number of components is 6, which is the total number of components.

When actually obtaining 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 are related to the verb "generate" are used. Although it will be extracted, the actual extraction work of this case component will be carried out using morphemes such as "accordingly" and "o" as markers.

Table 1 shows an example of a morpheme to be referred to as a mark when extracting a case component.

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

C-1) "Reflective means that totally reflects the luminous flux whose spot diameter is controlled by the first lens by the reflector, and ..."

For example, the verb "total internal reflection" is applied to the component "reflection means" that constitutes the scope of claims. As a case that defines the operation start condition of the verb "total internal reflection", "luminous flux (o)" and "reflector (by)" exist, and the number of case components 2 must be counted first. Can be done. However, when focusing on one of the case components, "luminous flux," it can be seen that the verb phrase "the spot diameter was controlled by the first lens" is further involved. Rather than defining every luminous flux as an object of "total reflection", defining a luminous flux with a condition as an object of "total reflection" is technically limited and of the component "reflection means". Proposition feasibility is reduced. Therefore, it is necessary to correct the decrease in the feasibility of this proposition as the number of case components.

In such a case, in the verb phrase "the spot diameter is controlled by the first lens", which is related to "light beam", the two case components "first lens" (the first lens) are governed by the verb "control". Because) ”and“ spot diameter () ”exist, the number of case components 2 is added. As a result, C-1) has a total number of case components of 4.

The description C-1) of the scope of claims may be defined as C-2) below depending on the specification creator.

C-2) "A luminous flux control means for controlling the spot diameter of a luminous flux by a first lens, and a reflecting means for totally reflecting such a luminous flux by a reflector."

The technical range formed by the "luminous flux control means" and the "reflection means" in C-2) is substantially the same as that in C-2). In C-2), the content of the verb phrase related to "luminous flux" in C-1) is redefined as a proposition to be realized by "luminous flux control means". Actually, even in C-1), some means or members are used to create the defined "luminous flux", but the "luminous flux control means" is not defined. .. Similarly, when the number of case components is counted for this C-2), the number of case components is 2 for the luminous flux control means and the number of case components is 2 for the reflection means, resulting in a total of 4 case components, which is the same result as C-1). Become.

As described above, since the breadth of the technical range is counted through the number of case components required to realize the proposition of the component, the description method and the number of words are different from each other, but substantially. C-1) and C-2) having the same technical range can be expressed by the same number of case components. If the minimum extraction unit is a component, C-1) is 1, C-2) is 2, and if the minimum extraction unit is a word, C-1) is 7, and C-2) is 9. In view of the above, it is considered that this number of case components can be the minimum extraction unit of a more suitable quantification method.

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

C-3) "A light beam control means for controlling the spot diameter of a light flux by the first lens, and a reflection means for totally reflecting a light beam whose spot diameter is controlled by the light beam control means by a reflecting plate."

The underlined 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 underlined part is a proposition that has already been realized by "controlling the spot diameter of the luminous flux" that is related to the "luminous flux control means", and the underlined description is intentionally made. What I put in is just a confirmation so that there is no doubt in interpretation about the "luminous flux" as the object to be totally reflected by the "reflection means". Therefore, if the description of the underlined portion is counted in the same manner when calculating the number of case components, the content of the underlined portion will be double-counted, resulting in deterioration of measurement accuracy.

For this reason, it is possible to eliminate the deviation in the number of case components due to the difference in the description order and description method of the claims by paying attention not to count the underlined description that has already been executed as a proposition. Actually, is the proposition realized by the description of "controlling the spot diameter of the luminous flux" in the first stage and the description of the underlined part in the second stage ("the light flux whose spot diameter is controlled by the above-mentioned luminous flux control means") are the same? It is necessary to judge whether or not.

The quantification method using the minimum extraction unit as the case component is particularly useful when judging the identity of a proposition. Unlike words, case components have the property of being dominated by verbs, so the same case component means that in addition to the content of the case component, the verb that governs it must also be the same. There is. If both the case component and the verb that governs it are the same, the proposition realized by this is also the same. Therefore, it is possible to identify whether or not the propositions are the same through the determination of the identity of the case components, and by extension, it is possible to easily identify whether or not the propositions are double counts. In addition, the difference in the description method between the specification creators does not affect the obtained numerical value.

In the above-mentioned example, the case where the description of the claims is described in the constituent requirement enumeration type is taken as an example, but the description can be performed in the same manner even if it is described in another format. For example, when the description is in the Jepson type, the description of "written", which is a known part, is also counted by the same method. At this time, if the counting method described in "Note" is not known, the Jepson type may be rewritten to the constituent requirement enumeration type before counting. By the way, the method of rewriting from the Jepson type to the constituent requirement enumeration type has been well known.

Further, when the claims are described in a written format, they can be counted by the same method. For example, when C-2) is written in a writing type, it becomes as shown in C-4) below.

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

Compared with C-2), C-4) described in this writing type is the same except that the words "luminous flux control means" and "reflection means" are omitted. .. "Luminous flux control means" and "reflection means" are subjects and are not counted, but only count noun phrases related to this subject. The number of components does not differ.

Extraction of Case Component from English Patent Specification Based on the above-mentioned theory of case component, the case component is extracted from the description of the scope of claims in the electronically converted English patent specification. When extracting case components from an English patent specification, it is based on the following process. What is hereinafter referred to as "count" means that all of them are extracted as case components.

The process of extracting case components in this English patent specification is composed of the flowchart shown in FIG. The case component extraction process is composed of a character string extraction step S11, a word inclusion step S12, a noun phrase extraction step S13, and a case component extraction step S14.

The operation of each step will be described in detail below.

First, in the character string extraction step S11, a character string is extracted from the description of the claims in the electronically converted English patent specification. In this character string extraction step S11, the extracted character string is normalized.

In the normalization process, when there is a space between characters in a word, the space is filled and the characters are connected to each other. Further, in this normalization process, if a quotation mark is attached to the extracted character string, the process of deleting the quotation mark is performed. In the normalization process, if one word or one word is connected by a delimiter (",", ":", ";") in parentheses, the parentheses and their contents are deleted. .. Further, in this normalization process, if the extracted character string contains "・", a process of deleting the "・" is performed. In the normalization process, full-width characters are replaced with half-width characters.

In the character string extraction step S11, the name of the invention is further extracted. In such a case, the noun phrase at the beginning of the character string should be specified as the name of the invention, and the specified invention name and the wording consisting of the or said + (name of the invention) + comprising should be extracted. Exclude from.

For example, if "A data transfer device for transferring the image data input from an image processing device to an electronic device, said data transfer device comprising " is described from the beginning, "A data transfer device" is the name of the invention. Identify as. Then, the name of the present invention is excluded from the case components to be extracted. Similarly, the "data transfer device" corresponding to the title of the invention in the wording consisting of said + (name of the invention) + comprising in the underlined part is also excluded from the case components to be extracted. Even if the underlined part contains a wording that partially matches the name of the invention, such as "said device comprising", it is also excluded from the case components to be extracted.

Further, in the character string extraction step S11, the sentence is made into a phrase. In the phrase formation of this sentence, the part of speech of each word is acquired by performing morphological analysis on the actually extracted character string. Next, by performing a syntactic analysis on this character string, the document tree structure when the character string is made into one sentence is obtained. Since the document tree structure differs depending on the analyzer, the analysis error by the analyzer is eliminated by acquiring the document tree structure in this phase. Further, in the phrase-making process of a sentence, chunking processing is performed to make a phrase by putting together the phrase structure.

Further, in the character string extraction step S11, each phrase is converted into a part of speech and a dependency structure that are easy to handle in the case component extraction. In such a case, the target compound word group is registered in the compound word master in advance. Then, it is compared with the idioms registered in this idiom master, and if the extracted character string corresponds to this registered idiom, this is included as one phrase and the idiom flag is set. Then, the phrase with the compound word flag is converted into an easy-to-use part of speech and dependency structure. Idioms including verbs are excluded from such processing operations.

As a rule for extracting case components, there is a rule for excluding nouns that make up English idioms from the case components to be extracted. For example, for in association with, in response to, words such as association and response are included in the English idiom and can be interpreted as nouns, but these should be excluded from the case components to be extracted. Perform processing. Actually, what constitutes such a compound word is registered in the database, it is confirmed whether or not the registered compound word is included in the extracted character string, and if it is included, if it is included. , Perform processing to exclude from the case components to be extracted.

By setting the compound word flag in this way, there is also an advantage that it can be easily determined whether or not it should be excluded from the case component.

Further, in the character string extraction step S11, the convenience of the subsequent processing operation may be improved by setting a flag for the verb and the causative verb that take the fifth sentence pattern. The fifth sentence pattern is an English sentence consisting of subject (S) + verb (V) + object (O) + complement (C), and the object and complement may point to the same content, or the complement is the purpose. In some cases, they complement the meaning of the word.

In this step S11, the following processing is performed based on the "adverb rule" as the case component extraction rule.

Adverb rules The adverb rules are rules that do not count adverbs that are directly related to the verb, for example, ~ freely, ~ possible, ~ impossible, ~ impossible, and so on. In English, adverbs such as temporarily are not included in the case component. Adverbs often end with (~ ly), but adverbs may be identified by extracting them using text mining technology. Such an adverb is extracted in the character string extraction step S11, and a process of deleting the adverb is also performed. At the time of deleting the adverbs, if a plurality of adverbs are separated by conjunctions such as "," and are continuous, the continuous adverbs including such conjunctions are deleted at once.

After completing such a character string extraction step S11, the process proceeds to the word inclusion step S12.

In the word inclusion step S12, the character string analyzed in the character string extraction step S11 is subjected to a process of including the association expression as one phrase. The types of complex expressions mentioned here are coronary + noun (the data), noun + noun, adjective + noun, present part (~ ing) + noun, past part (~ ed) + noun, "at least, not less". "than, any of, either of, any" + quantitative expression + noun plural, "at least, not less than, any of, either of, any" + quantitative expression + noun plural + "between, among" + noun group And so on. This adnominal expression also includes numbers, units, parentheses, a list of calculation symbols, and the like.

Such a noun phrase is often constructed as a noun phrase, but in extracting this noun phrase, the character string is decomposed into English words by performing morphological analysis and parsing. Since these individual English words may constitute a noun phrase, in such a case, such an English word is extracted as a noun phrase. By the way, a noun phrase is not limited to the case of grammatically constructing a phrase, but is a concept including a noun consisting of one English word. In addition to individual nouns, articles may be included and extracted as noun phrases, for example, "the data". Further, a comprehensive word such as a noun + a noun may be extracted as a noun phrase, such as "the image data" or "an optical disc".

By the way, this noun phrase may include a progressive form of the verb (current participle), such as "light receiving areas" or "the management information writing means", or a past participle, such as "a fixed range". May be included, but these are also extracted as noun phrases using morpheme analysis techniques and syntactic analysis techniques.

By extracting such a noun phrase, it is possible to correspond to the "adjective rule" which is one of the case component extraction rules.

Rules for adjectives The rules for adjectives are that adjectives related to nouns and modifiers that modify nouns are considered as one with nouns and are not counted independently. For example, as the adjectives shown in the underlined part below, "the big handle of ..." and " information that can be input by ..." are both adjectives that are related to the noun, so think of them as one with the noun. , Collectively make one component.

Also, "at least, not less than, any of, either of, any" + quantitative expression + noun plural, "at least, not less than, any of, either of, any" + quantitative expression + noun plural + " By extracting "between, among" + noun group, it is possible to count "any one of ..." as a nominative component.

Further, in the word inclusion step S12, the following may be included as one phrase based on the "rule of" case "".

"In case" rule
It is assumed that there is an adverbial clause consisting of "when, if, in case, while, only, providing, whether, in a case when" + noun phrase (S) + verb (V).

Here, the verb (V) in the above formula is defined as an adverbial clause verb because it is a verb within the adverbial clause. At this time, if the object or adverbial phrase related to the verb in the adverbial clause is described in the upper part, it will be double-counted if it is counted again, so it is not counted as a case component. On the other hand, the object or adverbial phrase related to the verb in the adverbial clause is counted as another case component if it is not described in the upper part.

Adverb If the intra-clause verb (V) is a be verb that takes a complement, it is "when, if, in case, while, only, providing, whether, in a case when" + noun phrase (S) + be verb + complement. .. In such a case, the part from the adverb clause starting with when etc. to the complement is counted as a nominative component.

For the noun clause "that, whether, whether or not" + noun phrase (S) + verb (V), which is the object of the verb (V), the object and adverb phrase related to the verb in the noun clause are from this. However, if there is a description in the upper row, if this is counted again, it will be a double count, so it is not counted as a case component. On the other hand, the object or adverbial phrase related to the verb in the adverbial clause is counted as another case component if it is not described in the upper part.

In the case of a be verb whose verb (V) takes a complement, these one phrase is counted as a nominative component in "that, ..." + noun phrase (S) + be verb + complement.

That is, as a rule when the subject and the verb follow immediately after the when etc., when the time etc. + the subject + the verb are used, these are collectively regarded as one case component. For example, for "~ when the wiper arm is located", this when the wiper arm is located is collectively regarded as a nominative component. In addition, if other words such as "at a predetermined position" follow, the presence or absence of the case component will be determined individually. For example, for "when the wiper arm is located at a predetermined position and a relative position signal", "when the wiper arm is located", "predetermined position", and "a relative position signal" are extracted as case components, respectively. .. Similarly, for "when the management information writing means writes updated management information to a page of the flash memory", the one-class component, "updated management information", "a page of", includes up to when the management information writing means writes. the flash memory "Extract each as a case component.

Further, in the word inclusion step S12, the following phrases may be comprehensively counted as a nominative component based on the following "rules of morphemes".

Morpheme rules In the morpheme rules, for example, "B related to A", "B related to A", "B in A", "B to be A", "B to be A", "B" in Japanese. "B in A", "B in A", and "B as A" are similarly counted as one case component for A and B even when they are written in English. ..

・ Noun phrase + noun connection prefix (of, in, for, or, within, about) + noun phrase ・ Noun phrase + noun connection prefix (of, in, for, or, within, about) + noun phrase + noun Inter-connection prefix (of, in, for, or, within, about) + noun phrase ・ Noun phrase + "between, among, from between, from among" + nouns taken by noun phrases are plural If it is a predicate included in a compound word, or if the noun phrase following in, for is "current part (~ ing) + noun phrase" or "past part (~ ed) + noun phrase", the above-mentioned counting rule Exclude from.

Next, the process proceeds to the noun phrase extraction step S13. In this noun phrase extraction step S13, in the character string extraction step S11, this is extracted on the premise that the adnominal expression is included as one phrase, and these one phrase is counted as one case component.

Next, the process proceeds to the case component extraction step S14. In this case component extraction step S14, a processing operation for extracting a case component is executed based on a rule explaining how.

First, the wording regarded as the name of the invention, "the ・ said" + "the name of the invention or a part thereof" + "the said" corresponding to comprising + "the name of the invention or a part thereof" is excluded from the case components. Perform processing.

For example, if "A data transfer device for transferring the image data input from an image processing device to an electronic device, said data transfer device comprising " is described from the beginning, "A data transfer device" is the name of the invention. Identify as. Then, the name of the present invention is excluded from the case components to be extracted. Similarly, the "data transfer device" corresponding to the title of the invention in the wording consisting of said + (name of the invention) + comprising in the underlined part is also excluded from the case components to be extracted. Even if the underlined part contains a wording that partially matches the name of the invention, such as "said device comprising", it is also excluded from the case components to be extracted.

Further, in the case component extraction step S14, the case component is extracted based on the following dual definition rule.

Double-defined rule If a noun phrase related to a verb is duplicated, that noun phrase is excluded from the case component. This is based on the double-defined rule in the case component extraction rule.

In the double definition rule, the following processing operations are performed.
1) Sequentially extract verbs consisting of the passive voice. This is because the verb consisting of this passive voice may have a double definition.
2) Identify whether the active voice (including infinitive), progressive tense, or passive voice of the verb consisting of the extracted passive voice is already defined in the upper part.
3) When it is identified that any of the active voice, progressive form, and passive voice of the verb has already been defined, the nomenclature that depends on any of the active voice, progressive form, and passive voice of the identified verb is specified. Or, identify a nomenclature that indicates the subject to which the active, progressive, or passive voice of the identified verb is involved.
4) It is determined whether or not at least a part of the noun phrase specified above and the noun phrase that is applied to the extracted passive verb via by or in are the same.
5) If it is determined in 4) that at least a part of these is the same, the noun phrase related to the verb consisting of the extracted passive voice is excluded from the case components to be extracted. On the other hand, when it is determined in 4) that at least a part of these is not the same, the noun phrase related to the extracted passive verb is included in the case component to be extracted.

The above processing operation will be described with reference to specific examples.
For example, "a graphic memory for storing the image data;
a control circuit for managing the image data stored in said graphic memory, "
Is "an image memory that stores image data and a control circuit that controls the image data stored in the image memory" in Japanese.
Here, "stored in the image memory" is an operation corresponding to the proposition already realized in the upper stage. Therefore, if these are also extracted as case components, they will be extracted twice corresponding to the same proposition with the upper row. Therefore, this "stored in the image memory" is excluded from the case components to be extracted.

The same applies to English claims.

First, in 1), the verb "stored" consisting of the passive voice is sequentially extracted. Next, in 2), it is identified whether or not any of the active voice, progressive tense, and passive voice of the extracted passive verb is described in the upper row. In the above example, it is identified that storing, which is the progressive form of the store, is described.

Next, we moved to 3), and since storing, which is the progressive form of the store, was described, the noun phrase that was related to this was specified, or the active voice, progressive tense, or passive voice of the identified verb was involved. Identify the noun phrase that indicates the receiving subject. In the above example, "a graphic memory" and "the image data" are specified.

Next, move to 4), and the noun phrase "said graphic" is applied to the specified noun phrases "a graphic memory" and "the image data" and the extracted passive verb "stored" via by or in. Determine if at least part of the memory "is the same. In this example, it is determined that the "graphic memory" is at least partially identical.

Next, we move to 5), and since it is judged that at least a part is the same in 4), the noun phrase "said graphic memory" that depends on the verb consisting of the extracted passive voice should be extracted. Performs processing to exclude from the case component.

On the other hand, if it is determined in 4) that at least some of these are not the same, a process is performed in which the noun phrase related to the extracted passive verb is included in the case component to be extracted.

On the other hand, in 2), when any of the active, progressive, and passive voices of the store is not described, or in 4), the noun phrase "said graphic memory" is assigned to "stored" via by or in. If there is no noun phrase that is at least partly the same as the noun phrase, the noun phrase "said graphic memory" that is related to the verb consisting of the extracted passive voice is included in the case component to be extracted.

When processing the description of the following claims based on the rule of double definition, the store, which is the active voice of stored, is extracted, and the memory unit following stored in is memory unit optionally and at least a part of it. Since they are the same, the memory unit following stored in is excluded from the case components to be extracted.
"a memory unit appropriately to store a program composed of a plurality of instructions; and a processor accurately to fetch each instruction in turn from the program stored in the memory unit,"

It should be noted that the processes 1) to 5) described above are not necessarily limited to the case where they are performed in this order, and it goes without saying that some of the processes may be changed or some processes may be performed at the same time. ..

The double definition rule also makes the same judgment in the following cases.

When there is a description of a verb, a verb, a to verb, a present part, or a past part that is related to a noun phrase, and is related to a noun phrase, it is related to the same noun phrase described later in the description. If the meanings of the verb, verb, to verb, current part, and past part that the noun phrase is related to match, the same noun phrase described in the latter part of the above is excluded from the case components to be extracted. It may be processed as follows.

For example, as shown in the following example, it is assumed that the noun phrase "a map" is related to the present participle "generating" in the component "map generation means" defined in the previous paragraph. At this time, the same noun phrase as this noun phrase "map" is defined in the latter part, and generated is involved in the noun phrase "map" in the subsequent part.
"Map generation means for generating a map ・ ・ ・ ・ ・ the map generated by the map generation means ・ ・ ・

In such a case, since the meaning and content of the present participle that the noun phrase in the first sentence is related to and the past participle that is related to the same noun match, the same noun phrase " Exclude "map" from the case components to be extracted.

In addition, it is assumed that there is a description of a [noun phrase] that depends on a verb or a [noun phrase] that depends on a verb in the first and second stages, respectively. At this time, if the [noun phrase] that depends on the verb or the [noun phrase] that depends on the verb match, which are described in the first and second stages, it depends on the verb that is described in the second stage [ Exclude [noun phrases] or [noun phrases] that are dependent on verbs from the case components that should be counted.

In addition, "that", "which", "where in" + noun phrase (S) + verb / verb are described in the first and second stages, respectively, and if they match each other, "that" and "that" and "that" are described in the second stage. "which" "where in" + noun phrase (S) + verbs / nouns are excluded from the case components to be counted. At this time, check the identity of the noun phrase (S) + verb / verb, and if they are the same, exclude the noun phrase (S) + verb / verb described in the latter part from the case components to be counted. You may.

In addition, when O and C of the fifth sentence pattern are described in the first and second stages, respectively, and if the combination of O and C matches each other between the first and second stages, the O and C described in the second stage are used. The combination of C may be excluded from the case components to be counted. Further, when O and C are described in the relation of the predicate of the subject in the first stage or the second stage, they may be included in the judgment target of identity.

In the above-mentioned judgment of the degree of agreement between the first stage and the second stage, the identity may be judged including the judgment of whether the verb is active or passive, and conversely, whether the verb is active or not. , It may be judged that they are the same as long as the verbs are common, without considering whether or not they are passive. That is, in determining whether or not they match each other, the identity of each other's verbs may be determined based only on the commonality of the verbs without determining the commonality of the active voice or the passive voice. Further, when the relative pronoun is the subject, the noun phrase of the reference destination may be regarded as the subject and the identity may be judged. Further, when determining the identity, if only a, the, each, both, said are different, it may be determined that they are the same as each other. If the "configured to verb phrase" is dependent on the noun phrase from behind, the identity may be determined by assuming that the past part of the verb is dependent on the noun phrase from behind. At the time of determination of identity, a verb having "comprise" as its original form may be excluded from this determination of identity.

In addition, it is assumed that there is a sentence relating to a conditional clause consisting of "when, if, in case, in the case, while, only, providing, whether" + noun phrase (S) + verb (V). At this time, if a conditional clause that exactly matches the conditional clause is not described in the upper part, the case component that overlaps in the first or second part of the sentence to be applied is counted without being excluded from the count target of the case component. You may. In other words, each character string in the first and second stages, which is the target of determining whether or not they match each other, is either when, if, in case, in the case, while, only, providing, or whether + noun phrase (S). When a conditional clause consisting of + verb (V) is applied, if the conditional clause to be applied is different between the first stage and the second stage, it is judged that they are not the same as each other. As an example of such a case
Noun phrase A + verb B when noun phrase P + verb Q
Noun phrase A + verb B when noun phrase P + verb R
Suppose that In such a case, the conditional clause consisting of "when, if, in case, in the case, while, only, providing, whether" + noun phrase (S) + verb (V) is different from each other in the first and second stages. Therefore, it is judged that they are not the same as each other. On the other hand, if this when clause is common to the first stage and the second stage, it may be determined that the objects for determining the identity of the first stage and the second stage are the same.

Further, when determining the degree of agreement between the first stage and the second stage described above, it may be determined that the verbs are common and only the auxiliary verbs are different, that they are the same, or that they are different. ..

Subject Rule In the case component extraction step S14, the case component extraction may be performed based on the subject rule described below. In this subject rule, when the extracted character string contains the relationship between the subject and the verb or gerund related to it, the wording corresponding to the subject is already described in the upper part. If so, the wording corresponding to the subject is excluded from the case components to be extracted. If the wording corresponding to the subject is not yet described in the upper part, the wording corresponding to the subject is processed so as to be included in the case component to be extracted.

For example, in the case of that (which / similarly) + subject + verb, the subject is judged based on the above-mentioned rule.

For example, in the case of "... are lowered by the drive machine, wherein the drive machine is arranged a rotation axis", the subject "the drive machine" has already appeared in the previous stage. Therefore, "the drive machine" is processed so as to be excluded from the case component. If "the drive machine" as the subject is described as "a drive machine" etc. and the same noun phrase is not described above this (before this), this "" Process so that "a drive machine" is included in the case component.

In this subject rule, the same processing operation may be performed in the case of that (which / herein) + subject + gerund.

When actually performing processing based on the rule of the subject, it may be based on the following processing flow.
1) First, extract the subject.
2) Sort and confirm whether the extracted subject is described before that. As a result, if there is a subject that is 100% consistent with the extracted subject, or if there is a part that includes it, the process proceeds to 3). Other than that, since the extracted subject appears for the first time, the extracted subject is specified as a case component.
3) When shifting to 3) in 2) above, none of them are counted as case components.

That is, it is first determined whether or not the extracted character string contains the relationship between the subject and the verb. As a result, if the extracted character string contains 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 row. If the wording corresponding to the subject is already described in the upper part, the wording corresponding to the subject is excluded from the case components to be extracted, and the wording corresponding to the subject is still described in the upper part. If it is not, it is a new appearance, so the wording corresponding to the subject is included in the case component to be extracted.

This makes it possible to eliminate the absurdity that the larger the number of subjects, the larger the number of case components in the description of the claims having many subjects.

When applying this subject rule, it may be specifically based on the following algorithm.

It is determined whether or not the extracted character string contains the relationship between the subject and the [verb / gerund / to verb] related to it. As a result, when the relationship between the subject and [verb / gerund / to verb] is included, it is determined whether or not the word corresponding to the subject is already described in the upper part. If the wording corresponding to the subject has already been described in the upper part, this subject is excluded from the case components to be extracted. When determining whether or not a wording that matches the wording corresponding to this subject is described in the upper part of the wording, the degree of matching may include not only an exact match but also a partial match.

The partial match here may be regarded as a partial match when only a, the, each, both, said are different. Also, if each, both, said, the is at the beginning of a noun phrase whose subject is a noun phrase that includes a noun phrase other than that word (searched by a suffix match), it may be regarded as a partial match. .. For noun phrases that correspond to the subject, if the noun phrases that are in parallel with the noun phrase are connected by "and" and ",", the noun phrases that are in parallel are comprehensively regarded as the subject. You may do so.

As an alternative to the relationship between the subject and the [verb / gerund / to verb] related to it, the same applies to "that", "which", "where in" + noun phrase (S) + verb / gerund. The subject rule may be applied. The noun phrase (S) following "that", "which", and "where in" is the subject, and in many cases it is the subject in relation to the verbs and gerunds involved in this, but this is also the subject. Determine if a wording that matches the corresponding wording is listed above it.

Component Rule In the case component extraction step S14, the case component extraction may be performed based on the “component rule”. This "rule of component" is a rule that does not count the component of the underlined portion described by the following requirement enumeration method as the number of case components. Adjustments are made to prevent a difference in the number of case components between the requirement enumeration method and the writing method in which no component is described.

・ Requirements enumeration method " ~ A and
~ B arranged in A above,
C , which is connected to B above and consists of ...
A device including D attached to C above. "

In the case of an English patent specification, if the patent sentence has the following structure, the noun phrase at the beginning of each sentence following comprising (comprise), including (include), consisting of (consist of), characterized by is started from the case component. It will be excluded.

Similarly, for noun phrases + comprising (comprise), including (include), consisting of (consist of), characterized by (+ noun phrases), treat each noun phrase so that it is not regarded as the number of case components. It may be.

For noun phrases that are judged to be excluded from the case component under the rules of the components, if the noun phrases that have a parallel relationship with the corresponding noun phrase are connected by "and" and ",", the noun phrase is in a parallel relationship. A certain noun phrase may be comprehensively regarded as a component and processed so as not to be counted as a case component.

In the other case component extraction step S14, the following rules may be adopted when counting the case components on a case-by-case basis.

In the case of a noun phrase + "to be passive verb", "to be passive verb" may be excluded from the case component. Further, "each other" may be excluded from the case component. Further, "the step of" following "comprising" may be excluded from the case component. If "in a ~ manner" exists, "a ~ manner" may be excluded from the case component. If "as a result of ~" exists, "as a result of ~" may be excluded from the case component. If "in part" takes an adverbial phrase by itself (without taking the phrase that follows part), then "part" may be excluded from the case component. If "in response to ~" exists, "response" may be excluded from the case component.

It is an adjective that is related to the noun that is the case of the noun phrase to be extracted from the case component, and there are cases where multiple adjectives are in a parallel relationship, that is, the adjectives are connected by and. In such a case, the number of case components is counted according to the number of adjectives. That is, when there are two adjectives A and B related to a noun, the total number of case components consisting of the noun + adjective A and the noun + adjective B is counted. ..

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

First, from the description of the patent specification in English that has been converted into electronic data, the character string described in a specific column (statement of claims) in the patent specification that has been converted into electronic data is extracted. In addition, this specific column corresponds to one claim in the claims.

Next, the case component is extracted from the extracted character string based on the above-mentioned rule. 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, but in the case of a subclaim, the claim in which the subclaim 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 to which the subclaim depends is added to this.

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

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

The present invention may also be applied in a network system. First, on the server side, data in which the number of case components of the patent specification is quantified in advance is stored in the server. Then, when the client side requests transmission of the number of case components for the 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.

If the number of case components has not yet been requested for the patent requested by the client, the server counts the number of case components for the requested patent and sends this to the client side via the network. You may send it.

In the 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-mentioned rule, and finally the case component for one claim is 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 may be output, but also the output data may be added to all kinds of information such as patent maps, graphs, and other evaluation values.

Further, in the present invention, the case component specified for each claim may be displayed on the screen, printed, or converted into data. That is, as shown in the following embodiment, underlining, highlighting, or the like may be performed so that the case component specified for the claim can be visually grasped.

Incidentally, as an alternative to the above-mentioned number of case components, an evaluation value based on the number of case components may be obtained and output, stored, or transmitted.

It should be noted that the present invention is not limited to the extraction of case components described in the claims of the English patent specification, and when extracting case components from any other English document based on the same method. It can also be applied to.

Claims (8)

A character string extraction step for extracting a character string from the description of the claims in an English patent specification converted into electronic data, and a character string extraction step.
It has a case component extraction step that extracts a noun phrase from the extracted character string in the above character string extraction step and extracts a case component from the extracted noun phrase.
In the case component extraction step, when there is a description of a verb, a verb, a to verb, a present part, or a past part that depends on the noun phrase and is related to the noun phrase, it is described later in the description. If the meanings of the verb, verb, to verb, current part, and past part that are related to the same noun phrase are the same, the same noun phrase described in the latter part of the above A case component extraction program from an English patent specification, characterized in that a computer is executed to exclude the case component from the case component to be extracted. A character string extraction step for extracting a character string from the description of the claims in an English patent specification converted into electronic data, and a character string extraction step.
It has a case component extraction step that extracts a noun phrase from the extracted character string in the above character string extraction step and extracts a case component from the extracted noun phrase.
In the above case component extraction step, when the object and the character string corresponding to the complement of the fifth sentence pattern are described in the first stage and the second stage, respectively, if they match each other, the object described in the second stage and the character string are described. A case component extraction program from an English patent specification, characterized in that a computer executes processing so as to exclude a combination of character strings corresponding to complements from the case components to be extracted. In the case component extraction step, when the character strings corresponding to the object and the complement are described in the relation of the predicate of the subject in the first stage or the second stage, if they match each other, the purpose described in the second stage. The case component extraction program from the English patent specification according to claim 1, wherein a combination of a character string corresponding to a word and a complement is processed so as to be excluded from the case components to be extracted. The above case component extraction step is characterized in that, in determining whether or not they match each other, the identity of each other's verbs is determined based only on the commonality of the verbs without determining the commonality of the active voice or the passive voice. The case component extraction program from the English patent specification according to any one of claims 1 to 3. In the above case component extraction step, in determining whether or not they match each other, the identity of each noun or noun phrase is considered to be the same if only a, the, each, both, said are different. A case component extraction program from an English patent specification according to any one of claims 1 to 3, characterized in that determination is made. In the above case component extraction step, when "configured to verb phrase" is related to a noun phrase from the back in determining whether or not they match each other, the past part of the verb is related to the noun phrase from the back. A case component extraction program from an English patent specification according to any one of claims 1 to 3, characterized in that the identity is judged as being received. In the above case component extraction step, each of the first and second character strings to be judged whether or not they match each other is either when, if, in case, in the case, while, only, providing, or +. Claim 1 is characterized in that, when a conditional clause consisting of a noun phrase + a verb is applied, if the conditional clause to be applied is different between the first stage and the second stage, it is determined that they are not the same as each other. A case component extraction program from an English patent specification described in any one of ~ 3. The case component extraction step is characterized in that, in determining whether or not they match each other, the identity of each other's verbs is determined based only on the commonality of the verbs without determining the commonality of the auxiliary verbs. A case component extraction program from an English patent specification according to any one of claims 1 to 3.

Images