JP6417359B2 - Claim parsing configuration method - Google Patents

Description

  The present invention relates to a syntax analysis configuration method of patent specifications, and more particularly to a syntax analysis configuration method of claims.

The scope of the patent application claims defines the technical features of the patent application according to a plurality of claims. Each claim defines a preface and contained composition, respectively, and when the composition has a primary element noun and a secondary element noun , a correspondence between the primary element noun and the secondary element noun is defined, The correspondence relationship includes a connection relationship, a subordinate relationship, a positional relationship, and the like. The technical features of one patent application are defined by the introductory text, the configuration, and the correspondence between the configurations in each claim.

  Therefore, if the introduction, configuration, and correspondence between the configurations in each claim can be effectively analyzed and stored in the database, the introduction and / or configuration can be obtained by a user searching the database. The patent application corresponding to the correspondence between the configuration and / or the configuration can be searched, and the patent search can be achieved efficiently.

  However, the patent database of a single country is quite large, and in terms of the patent database of the US Patent and Trademark Office (USPTO), the number of published patent publications has reached about 5.12 million, and other published publications are also The number has reached 39.96 million. Each patent application has approximately tens to hundreds of claims, so the total number of claims exceeds 100 million. In order to avoid waste of processing time due to miscellaneous and unnecessary process processing with respect to such a vast database, how to search for an efficient analysis method of claims is an issue. From the viewpoint of efficiency, if one process is increased for each single claim, 70 million extra processes are executed, which is a waste of time. In addition, in the automation process, it is necessary to re-inspect the result after each process in the automation process by the program. As a result, the accuracy of the program is confirmed, and processing of consecutive mistakes caused by previous process processing mistakes is avoided. Therefore, even if the number of processes increases by one, unnecessary process processing occurs and waste due to human inspection occurs.

  According to Taiwan Patent Nos. I347561 and I400621, when the connection term table is automatically analyzed using the connection term table, the connection term table includes: Consists of, composed by, includes the following steps, and so on. However, it is not easy to include all connection terms, and it is not easy to create the connection term table. Since the possibility of leaking is very high, complexity when referring to the table and miscellaneous human inspections occur. Also, the two patents are to analyze the subsequent part of the connection term (referred to as body part in the two patents) and further analyze it. Analyzing is not very useful for claim analysis, so the extra analyzed material increases the complexity of storage processing in another table and increases other data memory space and data access time. End up. In addition, since the two patents are intentionally separated from the introduction and the body part, it is necessary to store the introduction and the body part which are separated in the other data memories even in this state. Further, since the detailed analysis of the introduction is not provided for the two patents, only the description of the entire introduction can be provided, so that no further support for the search at the time of patent analysis can be provided.

  The present invention provides a more efficient and simple direct patent application parsing method in memory space and access time, thereby reducing the memory space in the conventional miscellaneous and inefficient patent analysis method, It is possible to avoid waste due to access time and human inspection.

  Therefore, the present invention has been made in view of the above problems, and an object of the present invention is to provide an efficient, simple, and direct syntax analysis configuration method in the memory space and access time. As a result, it is possible to avoid waste due to memory space, access time, human inspection, etc. with the conventional and inefficient patent analysis method.

The syntactic analysis configuration method of the claims of the present invention that can achieve the above object is a syntactic analysis configuration method of the claims that automatically reads the claims and automatically executes the following steps. ,
An independent claim and dependent claim determination mechanism for determining whether each claim of the claims is an independent claim or a dependent claim by an independent claim and dependent claim determination mechanism;
A range separation step of separating each claim of the claim by the order number of the claim and a dot symbol (.);
For a claim separated from the claim separation step, a prefetching step for extracting a preface with a colon;
An independent term composition analysis step of decomposing a main element noun and a secondary element noun in the definition after the colon with respect to the independent term determined from the independent term and the dependent term determining step;
Against subclaims it is determined from the independent claims and the dependent claims determination process, the gist of the main elements nouns and secondary elements nouns in the definition after the comma in that it includes a structure analysis step of decomposing the dependent claims It is what has.

  As one aspect of the present invention, it is preferable that the preface extracting step extracts a main preface and a sub preface of the preface.

As one aspect of the present invention, it is preferable that the independent component configuration analysis step and / or the dependent component configuration analysis step analyze a correspondence relationship between the main element noun and the sub- element noun .

  As one aspect of the present invention, it is preferable to take out the main preface and the sub preface of the preface by using a colon and a comma.

As an aspect of the present invention, it is preferable that the method further includes an independent term storage step of correspondingly storing the introduction of the independent term and all the main element nouns and secondary element nouns of the independent term in the independent term storage table.

As one aspect of the present invention, the introduction of a dependent claim, all the main element nouns and sub- element nouns of the dependent claim, and all the main element nouns and sub- element nouns of the dependent claim that are subordinate to an independent claim preferably further comprises manner and dependent Kouki憶step of storing in the subclaims storage table, the.

  As one aspect of the present invention, a search step of searching the independent term storage table and / or the dependent term storage table and retrieving patent specification data belonging to the independent term storage table and / or the dependent term storage table; It is preferable that it is further included.

  As one aspect of the present invention, it is preferable that the correspondence relationship includes a connection relationship and / or a dependency relationship and / or a positional relationship.

As one aspect of the present invention, the independent component configuration analysis step and / or the dependent component configuration analysis step does not further separate the definition after the colon, and directly defines the definition after the colon. it is preferable to decompose the main elements nouns and secondary elements nouns corresponding to each of said main elements noun.

  The claimed parsing configuration method of the present invention can provide a more efficient, simple and direct claim parsing configuration method in memory space and access time, thereby improving It is possible to avoid waste due to memory space, access time, human inspection, etc. by the inefficient patent analysis method.

It is a figure which shows the process of the syntax analysis structure method of the claim concerning the embodiment of this invention. FIG. 7 is a sample diagram showing the structure of an independent claim in the claim syntax analysis configuration method according to the embodiment of the present invention; 3 is a logic table created by a part-of-speech table according to the present invention. It is a figure which shows the syntax analysis structure method of the claim concerning another embodiment of this invention. It is a figure which shows the syntax analysis structure method of the claim concerning another embodiment of this invention. It is a figure which shows the syntax analysis structure method of the claim concerning another embodiment of this invention.

  In the following, an embodiment of the present invention will be described with reference to FIGS. The corresponding description is merely an example of the embodiment of the present invention, and does not limit the embodiment of the present invention.

As shown in FIG. 1, it is a diagram showing the steps of the syntax analysis configuration method according to the claims of the present invention. The method for synthesizing and parsing a claim according to the present invention is a method for synthesizing and analyzing a claim, wherein the claim is automatically read (step 11), and the following steps are automatically executed. And an independent claim and dependent claim determining step for determining whether each claim in the claims is an independent claim or a dependent claim by the dependent claim judging mechanism (step 12), and the claim order number and dot symbol (.) A range separation step of separating each claim of the range of the claims (step 13), and a step of extracting a preface with a colon for each of the claims separated from the range separation step of the claim (step) 14), the relative independent claim and has been independent claim determined from the dependent claims determination process, a configuration process of analysis of the major elements nouns and secondary elements noun decomposing independent claims in the definition after the colon (step 15 ), Independent claim and dependent claim judgment It has a gist in that it includes a subordinate component structure analysis step (step 16) for decomposing the main element noun and subelement noun in the definition after the comma with respect to the subordinate term determined from the step It is.

  The step of automatically reading the claims of the patent (step 11) is to read the claims that are recognized or stored in units of the claims by a file reading method. The scope of the claims is constituted by a plurality of claims. Each structured claim is the goal of constructing the syntax analysis of the present invention. In addition, the claims distinguish between independent and dependent claims. In the present invention, the importance of each claim can be distinguished by distinguishing the claims from independent claims or dependent claims. The membership structure of a single independent term has an implicitly higher importance than the membership structure of a dependent term.

As FIG. 2 shows, it is a sample diagram illustrating the structure of the independent claim. The claim is referred to as a preface, generally having a colon after the preface, and synthesizing a main element noun or a plurality of main element nouns that the preface has in the colon. In FIG. 2, the main element noun called “the first connecting mechanism” has a secondary element noun called “the elastic member” and is called the “second connecting mechanism”. The primary element noun has a secondary element noun called “stretching member”. Then, the secondary element noun called the “expandable member” and the main element noun called the “first series contact mechanism” generate a correspondence with the subordinate relationship called “having”. Then, the secondary element noun called the “stretching member” and the main element noun called the “second articulation mechanism” also generate a correspondence with a subordinate relationship called “having”.

An object of the present invention is more "Introduction" in each claim of the claims with high efficiency, "the main elements noun", "secondary element noun", "correspondence between the main elements nouns and incidental elements noun" is By being decomposed and analyzed, the user can search the claims and search for a more accurate test result to obtain a more accurate search analysis result.

  In the independent claim and dependent claim determination step (step 12) in FIG. 1, it is determined by the independent claim and dependent claim determination mechanism whether each claim in the claims is an independent claim or a dependent claim. The independent and dependent claim determination mechanism includes “Arabic numerals”, “.”, “Claims”, “claims”, and “kind”. The decision logic first determines that the claim starts with `` number '', the continuation is ``. '', The continuation is `` kind '', and is an independent term, the next is `` number '', the continuation is ``. '', If the continuation is “Claims”, it is determined as a dependent claim. Since this determination method is very simple, there is no need for a table, and it is possible to execute the determination in a character combination mode and increase the determination speed.

  The claim separation step (step 13) in FIG. 1 is to separate each claim of the claim by a claim order number and a dot symbol (.). The dot symbol (.) Is the same as an English period in an English claim. However, the Chinese claim is different from Chinese punctuation (.). Therefore, for the sake of easy understanding, the “.” Symbol is referred to as a “dot symbol” in this specification. In this process, the character range of each claim is determined by simply separating the starting number of the claim or the directly connected “dot symbol (.)”, And not using the punctuation here. In the conventional separation process using punctuation, it is necessary to read out characters from all claims, so that more time is required than in the past. Therefore, since the logic method of the decomposition performed by the number of the start of the direct claim or the “dot symbol (.)” Directly connected can be directly used in the SQL language, further time can be saved. Further, it may be separated using a line feed symbol, a number number in the claim or a subsequent “dot symbol (.)”, So that the number at the beginning of the claim and the “dot symbol (.) ", It is possible to avoid erroneous determination in any number format when separating only by" ".

  The order between the independent claim and dependent claim determination step (step 12) and the claim separation step (step 13) can be exchanged, and the order of the independent claim and dependent claim determination step (step 12) is always It is not necessarily before the claim separation step (step 13).

  In the introduction fetching step (step 14) of FIG. 1, the introduction is taken out with a colon for each claim separated from the claim dividing step. The present invention uses the "colon" directly to extract the preface and does not require a table. Conventionally, a table is used, and the table is a so-called “connection term” table, and the connection term is constituted by, for example, includes, includes, includes, and is composed of innumerable numbers. The preparation of the table is very complicated and is very likely to leak. The creation of a table with such a wide range of possibilities is impractical, the execution on a personal computer is not efficient, and the inspection time for the result after the execution is long, so the efficiency is low. In the present invention, sufficient efficiency can be obtained by extracting the preface directly using "colon".

In the configuration analysis step (step 15) of the independent term in FIG. 1, the definition after the colon is decomposed into the main component noun and the secondary component noun for the independent term determined from the independent term and the dependent term determining step. To do. As shown in FIG. 2, in this process, the noun after “one” (a / an in English) connected after the colon is recognized as the start of one main element noun , and how far the main element noun ends. As for processing logic, the present invention will be described with reference to a parsing configuration of an English claim as an example, and for a parsing configuration of a Chinese claim, “noun table”, “modification term” (for example: “Ha”, “ga”) or “corresponding terms” (for example: “join”, “provided”, etc.), but the explanation is omitted here. In the syntax analysis configuration method of the above-mentioned English claims, a regular syntax analysis configuration is advanced based on various parts-of-speech tables, and details are as follows.

  The part of speech table includes a noun prefix table, a relative pronoun table, a dependent verb table, a first participle table, a second participle (Ving) table, a preposition table, an interruption table, and an exception table. . The first participle table includes a table of two types of participle such as Ved and Ving. FIG. 3 shows a logic table created by the part of speech table. The following is a syntactic analysis composition with the text: “a circuit board provided with edges extending in a longitudinal direction of said circuit board;” according to the logic table of FIG.

When reading the sentence "" a circuit board provided with edges extending in a longitudinal direction of said circuit board; ", the letters in""are the read letters, and the noun" a "is the" noun prefix table " As shown in the column with “1” in the first column of the third column, the reading continues (ie, the operation of 1+);
Next, read “a“ circuit ”board provided with edges extending in a longitudinal direction of said circuit board; Since this “circuit” does not belong to any table, continue reading.
Next, read "a circuit" board "provided with edges extending in a longitudinal direction of said circuit board; since this" board "does not belong to any table, continue reading.
Then read "a circuit board" provided "with edges extending in a longitudinal direction of said circuit board;" because this "provided" belongs to the first participle table, that is, the first participle in the participle form of the verb It is. Therefore, according to the logic table, it should proceed to the processing of the column with the item number “5” to “10”. The processing of the six types of columns has one common point. That is, by selecting as a breakpoint in the first participle and subsequently executing reading, it is determined which column processing is to be selected next;
Next, read “a circuit board provided“ with ”edges extending in a longitudinal direction of said circuit board;” Since this “with” belongs to the preposition table, the item number is “6” according to the logic table. Should proceed to the next column. The separation point determined by the first participle in the column processing is displayed as “established”, is selected as the separation point by the first participle, and the part before the separation point, that is, “a circuit board” is selected as a main element noun. Should be treated as At this time, since the main element noun is already identified, the break point determined by the original first participle is canceled, and at this time, there is no break point, so reading is continued;
Next, read "a circuit board provided with" edges "extending in a longitudinal direction of said circuit board;", because this "edges" ends with "s", so in the noun prefix table "with", "having" The characters connected after "has", "have" and ending with "s" belong to the noun prefix table and continue to read;
Next, "a circuit board provided with edges" extending "in a longitudinal direction of said circuit board;" is read. Since this "extending" belongs to the first participle table, the item number is " It should proceed to the processing of the column “5” to “10”. Select “extending” as a break point, and at this time, it is necessary to judge by the reading lexical to be connected next;
Next, read “a circuit board provided with edges extending“ in ”a longitudinal direction of said circuit board;” Since this “in” belongs to the preposition table, it matches the column “6” according to the logic table. In order to do this, select as a breakpoint in the extending, extract "arms" as a secondary element noun , and continue reading;
Next, read “a circuit board provided with edges extending in“ a ”longitudinal direction of said circuit board;” Since “a” belongs to the “noun prefix table” and appears in the second time, it is judged. Is recognized as the beginning of secondary element nouns . Therefore, the subelement noun is started to be read starting from “a”. Since the logic to read is the same as the reading of the main element noun , the logic may be cited;
Next, read “a circuit board provided with edges extending in a“ longitudinal ”direction of said circuit board;” Since “longitudinal” does not belong to any of the tables, continue reading.
Next, read “a circuit board provided with edges extending in a longitudinal“ direction ”of said circuit board;” Since “direction” does not belong to any table, continue reading.
Next, “a circuit board provided with edges extending in a longitudinal direction“ of ”said circuit board;” is read. Since “of” belongs to the preposition table, the item number is “11” according to the logic table. Should proceed to the processing of the column “13”. The processing of the four types of columns has one common point. That is, by selecting the preposition as a breakpoint and continuing to read, it determines which column processing to select next;
Next, read “a circuit board provided with edges extending in a longitudinal direction of“ said ”circuit board;” Since the “said” belongs to the “noun prefix table”, the item number is “11”. By selecting the processing of the sequence of, i.e., by establishing the preposition (in) as a determined breakpoint, the characters between the "said" in the "noun prefix table" from the back of the preposition (in) Take out as a secondary element noun , continue to read, and at this time the occurrence of the “said”, the “said” has already appeared, so based on this, the noun that appeared after “said” Classify and abandon as “Exception Handling Table”. That is, the noun behind “said” is not separated. ;
Next, “a circuit board provided with edges extending in a longitudinal direction of said“ circuit ”board;” is read. Since the “circuit” does not belong to any table, the reading is continued.
Next, “a circuit board provided with edges extending in a longitudinal direction of said circuit“ board ”;” is read. Since the “board” does not belong to any table, the reading is continued;
Next, read “a circuit board provided with edges extending in a longitudinal direction of said circuit board“; ”.” Since the “;” belongs to the interruption table, it can be interrupted. Perform element noun parsing configuration.

In the subordinate component analysis step (step 16) in FIG. 1, the main element noun and subelement noun are decomposed from the definition after the comma with respect to the subordinate term determined from the independent term and subordinate term determining step. . In this case, the main element nouns and sub element nouns are decomposed at a comma as a boundary, and there is no need to create specific tables such as: The efficiency of the configuration can be increased. Since the logic method used in the decomposition of the main element noun and sub- element noun of the subordinate term is the same as the logic method used in the main element noun and sub- element noun of the independent item, the description is omitted here.

  Furthermore, the present invention distinguishes the main and sub-introductions from the introduction. The distinction method is executed based on a specific noun or a combination with an exceptional specific noun. For example, before the preposition with is defined as the main preface and after that is defined as the sub-preface. The combination of specific nouns and prepositions and verbal nouns and nouns is the main preface, and other prefaces are defined as sub-prefaces. Alternatively, the main preface and the sub preface are taken out by using the logic method used in the independent component configuration analysis step (step 15) or in the dependent component configuration analysis step (step 16). Although there are an infinite number of implementation methods, those who are familiar with the technology in the field can obtain many methods according to the above, but the description thereof is omitted here.

Further, according to the syntax analysis composition method of the claims of the embodiment of the present invention, the correspondence relationship between the main element nouns and the sub- element nouns is analyzed in the structure analysis process of the independent claim and / or the structure analysis process of the dependent claims To do. The correspondence is searched for the sentence position of the extracted main element noun and the extracted sub- element noun , and the first participle table or the second participle (Ving) table between the main element noun position and the sub- element noun position Select a word belonging to as a correspondence. Taking "a circuit board provided with edges extending in a longitudinal direction" of "said circuit board;" as an example, "a circuit board" is a primary element noun and "edges" is a secondary element noun. “Provided” belongs to the first participle table, so “provided” is a correspondence between the main element noun position and the sub- element noun position. The correspondence relationship includes a connection relationship and / or a dependency relationship and / or a positional relationship. The connection relationship includes connection, connection, and the like; the dependency relationship includes, includes, and the like; the positional relationship includes configuration, configuration, and the like.

  As shown in FIG. 2, in the syntactic analysis configuration method of the claims of the embodiment of the present invention, the main preface and the sub preface of the preface are directly extracted by using a colon and a comma. Since no table needs to be created, access time and human inspection time after personal computer processing can be saved.

As shown in FIG. 4, the syntax analysis configuration method according to the embodiment of the present invention further includes an independent term storing step (step 17). The independent term storage step stores the introduction of the independent term and all main element nouns and subelement nouns of the independent term in the independent term storage table correspondingly.

Further, as shown in FIG. 5, the syntax analysis configuration method according to the claims of the embodiment of the present invention further includes a dependent term storage step (step 18). The dependent term storing step corresponds to the introduction of the dependent claim, all the main element nouns and sub element nouns of the dependent claim, and all the main element nouns and sub element nouns of the dependent claim subordinate to the independent claim. Are stored in the dependent term storage table.

  As shown in FIG. 6, the syntax analysis configuration method according to the embodiment of the present invention further includes a search step (step 19). The retrieval step retrieves the independent term storage table and / or the dependent term storage table, and extracts patent specification data belonging to the independent term storage table and / or the dependent term storage table.

The unique, fast, and efficient reason for the syntactic analysis configuration method of the claims of the embodiment of the present invention is that the configuration analysis step of the independent claim and / or the configuration analysis step of the dependent claim of the present invention is performed. This is because the definition after the colon is not further divided, but the definition after the colon is directly decomposed into all main element nouns and sub- element nouns corresponding to the respective main element nouns .

  While the preferred embodiments of the present invention have been disclosed above, as may be appreciated by those skilled in the art, they are not intended to limit the invention in any way. Various adjustments and modifications can be made without departing from the spirit and scope of the present invention. Accordingly, the scope of the claims of the present invention should be construed broadly including such changes and modifications.

Claims (6)

A claim parsing configuration method for automatically reading a claim and automatically executing the following steps:
An independent claim and dependent claim determination mechanism for determining whether each claim of the claims is an independent claim or a dependent claim by an independent claim and dependent claim determination mechanism;
A range separation step of separating each claim of the claim by a claim order number and a dot symbol;
For each claim separated from the claim separation step, an introductory step for extracting the preface with a colon;
An independent term structure analysis step of decomposing element nouns in the definition after the colon with respect to the independent term determined from the independent term and the dependent term determining step;
A subordinate component structure analysis step of decomposing element nouns in the definition after the comma for the subordinate term determined from the independent and subordinate term determining steps;
Including
In the configuration analysis step of the independent claim and the configuration analysis step of the dependent claim, the claim analyzed in each claim sentence in the claims based on the logic of a predetermined English sentence structure. Due to the position of the noun prefix in the sentence, the part-of-speech of the word after the noun prefix, and the order of the words with the particular part-of-speech after the noun prefix, the position of the noun prefix immediately follows the noun prefix Identifying a lexical phrase that is in front of a word having a specific part of speech that matches the logic of the predetermined English sentence structure and has an arrangement order as an element noun, and decomposes the element noun identified in the claim sentence A parsing configuration method according to claim.   2. The syntactic analysis configuration method according to claim 1, wherein the configuration analysis step of the independent claim and / or the configuration analysis step of the dependent claim analyzes a correspondence relationship between the element nouns. .   2. The syntactic analysis of claim 1 further comprising: an independent term storage step of correspondingly storing an introductory term of the independent term and all element nouns of the independent term in an independent term storage table. Configuration method. And a dependent term storage step of correspondingly storing in the dependent term storage table all the element nouns of the dependent claims dependent on the introductory sentence of the dependent claims, all of the element nouns of the dependent claims, and all of the dependent claims. The syntactic analysis construction method according to claim 1, characterized in that it is characterized in that: A search step of searching the independent term storage table and / or the dependent term storage table and retrieving patent specification data belonging to the independent term storage table and / or the dependent term storage table, A method of constructing a syntax analysis according to claim 1. 3. The parsing configuration method according to claim 2, wherein the correspondence relation includes a connection relation and / or a subordinate relation and / or a positional relation.

Images