JP5190192B2 - SEARCH DEVICE, SEARCH METHOD, AND PROGRAM - Google Patents

Description

  The present invention relates to a search device, a search method, and a program.

  Conventionally, in the search for similar character strings, similar expansion character strings applied according to the similarity criterion are simply derived without omission and matching is performed with perfect matching. For example, in the case of the similarity criterion “arbitrary arbitrary character strings are in the relationship of FIG. 4” and the question name “NS”, the similar expansion character string “Inu S, HNS, MS,... X3x2-1 = 71.

  In such a conventional method, there are an enormous number of similar expansion character strings, and comparisons are made many times. Therefore, the calculation efficiency is very poor, and the memory usage is enormous.

  On the other hand, a method for searching for a similar character string using an automaton is disclosed (for example, see Patent Document 1).

JP-A-8-339378

  In the similar character string search method using the automaton, the calculation efficiency and the memory efficiency are good, but the rule is difficult to create and the created rule is difficult to understand.

  The present invention has been made in view of such problems, and an object of the present invention is to provide a search device, a search method, and a program with good system maintainability as well as calculation efficiency and memory efficiency.

Therefore, the search device of the present invention includes a search target data string holding unit that stores a search target data string, a search data string acquisition unit that acquires a search data string, and a search data string acquired by the search data string acquisition unit On the other hand, a setting means for setting a range of the number of times of using an object defining a similar relationship as to whether or not the search data sequence and the search target data sequence have a similar relationship according to a predetermined rule And the search data string acquired by the search data string acquisition means and the search object data string held by the search object data string holding means within the range of the number of times set by the setting means as the search processing When, will have a, a search unit for comparing whether the defined similarity relation with the object, the searching means, in said setting means Within the specified number of times, using the object, the unit data constituting the search data string and the unit data constituting the search target data string are compared for each unit data and do not match. If unit data exists, the unit is used to compare whether the unit data that do not match is in a similar relationship defined in the object within the range of the number of times set by the setting unit. It is characterized by that.

  By adopting such a configuration, a range of the number of times the object is used is set according to a predetermined rule, and the search data string and the search target data string are compared within this range. In addition to the memory efficiency, it is possible to provide a search device with good system maintainability.

  The object corresponds to, for example, a similarity relationship table described later.

  Further, the present invention may be a search method and a program.

  According to the present invention, it is possible to provide a search device or the like having good calculation efficiency, memory efficiency, and system maintainability.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

<Embodiment 1>
A hardware configuration of an example of the information processing apparatus 1 is shown in FIG. FIG. 1 is a hardware configuration diagram of an example of the information processing apparatus 1.

  As shown in FIG. 1, the information processing apparatus 1 includes, as a hardware configuration, an input device 11, a display device 12, a recording medium drive device 13, a ROM (Read Only Memory) 15, and a RAM (Random Access Memory). ) 16, at least one CPU (Central Processing Unit) 17, an interface device 18, and an HD (Hard Disk) 19.

  The input device 11 includes a keyboard and a mouse that are operated by an operator (or user) of the information processing apparatus 1, and is used to input various operation information and the like to the information processing apparatus 1. The display device 12 includes a display used by an operator of the information processing device 1 and is used to display various types of information (or screens).

  The interface device 18 is an interface that connects the information processing device 1 to a network or the like. A program related to a similarity search function of the information processing apparatus 1 described later or a flowchart related to a similarity search described later is provided to the information processing apparatus 1 by a recording medium 14 such as a CD-ROM, or downloaded through a network or the like Is done. The recording medium 14 is set in the recording medium drive device 13, and the program is installed from the recording medium 14 to the HD 19 via the recording medium drive device 13.

  The ROM 15 records a program that is read first when the information processing apparatus 1 is powered on. The RAM 16 is a main memory of the information processing apparatus 1. The CPU 17 reads out the program from the HD 19 as necessary, stores it in the RAM 16, and executes the program to provide all or part of the functions related to the similar search described later, or to perform the similar search described later. A flowchart relating to the function is executed. In addition to the program, the HD 19 stores, for example, rules to be described later, the number of applicable similarity tables corresponding to the rules, a threshold, a search target character string, a similarity relationship table, and the like.

  Hereinafter, an example of a functional configuration of the information processing apparatus 1 including the CPU 17, the RAM 16, the HD 19, a program, and the like is illustrated in FIG. FIG. 2 is a functional configuration diagram (part 1) of an example of the information processing apparatus 1. As shown in FIG. 2, the information processing apparatus 1 includes a search character string acquisition unit 21, a setting unit 22, a rule holding unit 23, a search unit 24, and a search target character string holding unit 25 as functional configurations. The similarity relationship table holding unit 26 and the search result aggregation unit 27 are included.

  The search character string acquisition unit 21 acquires a search character string. The search character string acquisition unit 21 acquires, as a search character string, for example, a character string that the user inputs on the search screen or the like. The setting unit 22 can apply the similarity relationship table as shown in FIG. 3 according to the rules held in the rule holding unit 23 for the search character string acquired by the search character string acquisition unit 21. And set a threshold. FIG. 3 is a diagram (part 1) illustrating an example of the applicable number of times and threshold values of the similarity relationship table. Here, the applicable number of times of the similarity relationship table represents the maximum value (maximum number of times) at which the similarity relationship table can be used (referenced), and the threshold value must be the similarity relationship table (reference must be made). It can be said that it represents the minimum value (minimum number of times).

  The rule holding unit 23 holds the rules and the applicable number of times and threshold values of the similarity relationship table as shown in FIG. 3 corresponding to the rules. Here, the rule is, for example, “one sound or two sounds are in the relationship of the similarity relationship table A” or the like. The rules are described in, for example, a file.

  There may be a plurality of rules or a single rule. When there are a plurality of rules, for example, the information processing apparatus 1 automatically selects one or a plurality of rules from the plurality of rules according to the search condition selected (or input) by the user on the search screen or the like. A rule may be selected and applied to the search character string, or the user may be able to select one or more rules on the search screen or the like.

  The search unit 24 compares the search character string with the search target character string held in the search target character string holding unit 25 within the range of the applicable number of times and the threshold value of the similarity relationship table set in the setting unit 22. And output the search target character string. Details of the comparison processing in the search unit 24 will be described with reference to FIG.

  The search target character string holding unit 25 holds a search target character string. Here, the search target character string is, for example, a trademark (a trademark of a character string) in the case of a trademark name search. The search target character string holding unit 25 holds the search target character string in, for example, a trie structure.

  The similarity relationship table holding unit 26 holds a similarity relationship table as shown in FIG. FIG. 4 is a diagram (part 1) illustrating an example of the similarity relationship table.

  The search result aggregating unit 27 aggregates the results of the search process in the search unit 24 (for example, merging duplicate data) and outputs the search results (search result list) to the search result screen or the like.

FIG. 5 is a flowchart illustrating an example of the comparison process. The comparison process shown in FIG. 5 is a comparison process between one search character string and one search target character string.
In step S10, the search unit 24 resets the similarity relationship table application counter (sets it to zero).

  Subsequently, in step S11, the search unit 24 sets the position of the target character (target character position) in the search character string and the search target character string to the top. In step S <b> 12, the search unit 24 takes out the target character position between the search character string and the search target character string. Hereafter, the character of the search character string at the target character position extracted by the search unit 24 in step S12 is referred to as a search character. Hereinafter, the character in the search target character string at the target character position extracted by the search unit 24 in step S12 is referred to as a search target character.

In step S13, the search unit 24 compares the search character and the search target character without applying the similarity relationship table. Here, the search unit 24 expresses the type of the search character and the type of the search target character by the position of the bit that is 1 in the bit string, and determines whether or not the search character matches the search target character. Compare using bit operations.
For example,
A: 0001, A: 0010, C: 0100, D: 1000,
The four types of characters are represented by 4 bits and a or i or c (= a | i | c) is expressed as a logical sum "0111", and the search unit 24 determines whether the search character matches the search target character. Whether or not is compared using a bit operation. For simplification of description, a character is represented by 4 bits, but a larger number of bits may be used to represent more types of characters. Of course, the number of bits and the number of types need not be the same.
That is, when the search character is “0001” and the search target character is “0001”, the search unit 24 performs a bit product operation,
0001 & 0001 = 0001 ≠ 0
And calculate. When the search character is “0001” and the search target character is “0010”, the search unit 24 performs a bit product operation,
0001 & 0010 = 0000 = 0
And calculate. It is also possible to compare whether or not the search character string has a matching character in the search target character string using a bit product operation. When the search character obtained by applying the similarity table to the search character is “A”, “I”, or “C” “0111” and the search target character is “A” “0001”, the search unit 24 performs a bit product operation,
0111 & 0001 = 0001 ≠ 0
And calculate. As described above, even if there are a plurality of characters to be compared, the comparison with one search character can be performed by a single bit product operation, so that the calculation efficiency is very good.

  In step S14, the search unit 24 determines whether the search character matches the search target character based on the result of the bit product operation in step S13. If the result of the bit product operation is not 0, the search unit 24 determines that the search character matches the search target character (YES in step S14), and proceeds to step S19. On the other hand, when the bit product operation result = 0, the search unit 24 determines that the search character does not match the search target character (NO in step S14), and proceeds to step S15.

  In step S15, the search unit 24 determines whether or not the number of times that the similarity relationship table A can be applied is larger than the similarity relationship table application counter. If the search unit 24 determines that the applicable number of similar relationship table A is greater than the similar relationship table application counter (YES in step S15), the search unit 24 proceeds to step S16. On the other hand, when determining that the applicable number of similar relationship table A is not greater than the similarity relationship table application counter (NO in step S15), the search unit 24 ends the process illustrated in FIG.

  In step S16, the search unit 24 compares the search character and the search target character by applying, for example, the similarity relationship table A. In step S <b> 17, the search unit 24 determines whether the search character and the search target character are in a similar relationship. When the search unit 24 determines that the search character and the search target character are in a similar relationship (YES in step S17), the search unit 24 proceeds to step S18. On the other hand, when the search unit 24 determines that the search character and the search target character are not in a similar relationship (NO in step S17), the process illustrated in FIG. 5 ends.

  In step S18, the search unit 24 increments the similarity relationship table application counter by one. In step S19, the search unit 24 determines whether there is a next character in the search character string. If search unit 24 determines that there is a next character in the search character string (YES in step S19), it proceeds to step S20. On the other hand, when determining that there is no next character in the search character string (NO in step S19), search unit 24 proceeds to step S21.

  In step S20, the search unit 24 advances the attention character position of the search character string and the search target character string by one (one character). Then, the search unit 24 returns to step S12.

  On the other hand, in step S21, the search unit 24 determines whether or not the similarity relationship table application counter is greater than or equal to a threshold value. When the search unit 24 determines that the similarity relationship table application counter is equal to or greater than the threshold (YES in step S21), the search unit 24 proceeds to step S22. On the other hand, when the search unit 24 determines that the similarity relationship table application counter is not equal to or greater than the threshold (NO in step S22), the search unit 24 ends the process illustrated in FIG.

  In step S <b> 22, the search unit 24 outputs the target search target character string to the search result aggregation unit 27 as a character string similar to the search character string, that is, as a search target character string hit in the search.

  In addition, when it determined with NO in determination of step S15, step S17, and step S21, it demonstrated so that the process shown in FIG. 5 might be complete | finished, but the search part 24 does not complete | finish a process as it is, for example, A search result or the like indicating that the target search target character string is not a character string similar to the search character string may be output to the search result aggregation unit 27.

  Next, the comparison process performed by the search unit 24 will be described using a more specific example. Here, it is assumed that the search unit 24 compares the search character string “Aiweo” with the search target character string “Ae Yueo”.

  As described above, in step S10, the search unit 24 resets the similarity relationship table application counter (sets it to zero). Subsequently, in step S11, the search unit 24 sets the attention character position of the search character string and the search target character string to the top. In step S <b> 12, the search unit 24 takes out the target character position between the search character string and the search target character string. That is, first, the search unit 24 extracts “a” that is the first character of the search character string as the search character, and extracts “a” that is the first character of the search target character string as the search target character.

  In step S13, the search unit 24 uses a bit product operation to compare the search character string and the search target character string without applying the similarity relationship table. In step S14, the search unit 24 determines whether or not they match. Now, since both match with "a", the search part 24 progresses to step S19 and determines whether the next character is in a search character string. In this case, since there is a character next, the search unit 24 proceeds to step S20, advances the attention character position by one, and proceeds to step S12.

  In step S12, the search unit 24 extracts “i”, which is the second character of the search character string, as the search character, and extracts “d”, which is the second character of the search target character string, as the search target character. . In step S13, the search unit 24 uses a bit product operation to compare the search character string and the search target character string without applying the similarity relationship table. In step S14, the search unit 24 determines whether or not they match by the comparison process using the above-described bits. Since the search character is “a” and the search target character is “D”, the search unit 24 proceeds to step S15 and determines whether the applicable number of similar relationship table A is larger than the similar relationship table application counter. Determine whether. In the present example, since the number of applicable times of the similarity relationship table A is 2 and the similarity relationship table application counter is 0, the search unit 24 proceeds to step S16, and the search character and the search target character are converted to the similarity relationship table A. Apply and compare.

  Referring to the similarity relationship table A, since “i” and “d” are in a similar relationship, the search unit 24 determines that they are similar by the comparison process using the above-described bit, and proceeds to step S18, where the similarity relationship The table application counter is incremented by 1 to 1. Then, the search unit 24 proceeds to step S19, and determines whether or not there is a next character in the search character string. In this case, since there is a character next, the search unit 24 proceeds to step S20, advances the attention character position by one, and proceeds to step S12.

  In step S12, the search unit 24 extracts “c”, which is the third character of the search character string, as a search character, and extracts “yu”, which is the third character of the search target character string, as a search target character. . In step S13, the search unit 24 uses a bit product operation to compare the search character string and the search target character string without applying the similarity relationship table. In step S14, the search unit 24 determines whether or not they match. Since the search character is “C” and the search target character is “Y”, the search unit 24 proceeds to step S15, and whether or not the applicable number of similar relationship table A is larger than the similar relationship table application counter. Determine whether. In the present example, since the number of applicable times of the similarity relationship table A is 2 and the similarity relationship table application counter is 1, the search unit 24 proceeds to step S16, and the search character and the search target character are converted to the similarity relationship table A. Apply and compare.

  Referring to the similarity relationship table A, since “U” and “Yu” are in a similar relationship, the search unit 24 determines that they are similar, and proceeds to step S18 to increment the similarity relationship table application counter by one and set 2 And Then, the search unit 24 proceeds to step S19, and determines whether or not there is a next character in the search character string. In this case, since there is a character next, the search unit 24 proceeds to step S20, advances the attention character position by one, and proceeds to step S12.

  In step S12, the search unit 24 extracts “d”, which is the fourth character of the search character string, as the search character, and extracts “d”, which is the fourth character of the search target character string, as the search target character. .

  In step S13, the search unit 24 uses a bit product operation to compare the search character string and the search target character string without applying the similarity relationship table. In step S14, the search unit 24 determines whether or not they match. Now, since both match with "d", the search part 24 progresses to step S19 and determines whether the next character exists in a search character string. In this case, since there is a character next, the search unit 24 proceeds to step S20, advances the attention character position by one, and proceeds to step S12.

  In step S12, the search unit 24 extracts “o”, which is the fifth character of the search character string, as the search character, and extracts “o”, which is the fifth character of the search target character string, as the search target character. .

  In step S13, the search unit 24 uses a bit product operation to compare the search character string and the search target character string without applying the similarity relationship table. In step S14, the search unit 24 determines whether or not they match. Now, since both match with “O”, the search unit 24 proceeds to step S19, and determines whether or not there is a next character in the search character string. In this case, since there is no next character, the search unit 24 proceeds to step S21, and determines whether or not the similarity relationship table application counter is equal to or greater than a threshold value.

  In this case, since the similarity relationship table application counter is 2 and the threshold value is 1, the search unit 24 determines that the similarity relationship table application counter is equal to or greater than the threshold value, and proceeds to step S22 to search for the search target character string “AEEUEO "Is output to the search result aggregating unit 27 as a character string similar to the search character string" Iueo ", that is, as a search target character string hit in the search.

In addition, when the applicable number of similar relationship table A is not larger than the similar relationship table application counter, it is not necessary to perform comparison processing for all the search character strings. For example, a case will be described in which the search target character string is “Ae Yule” and other conditions are the same as those described above. The processing up to the third character position of interest in the search character string is omitted because it is the same processing as described above. The attention character of the search character string moves to the fourth position, and in step S14, the search unit 24 determines whether or not they match by the comparison process using the above-described bits. Since the search character is “c” and the search target character is “le”, the search unit 24 proceeds to step S15 and determines whether the applicable number of similar relationship table A is greater than the similar relationship table application counter. Determine whether. Since the similarity relationship table A is applied and compared in the second and third search character strings, the number of applicable times of the similarity relationship table A is 2, and the similarity relationship table application counter is 2. Therefore, if the search unit 24 determines that the number of times that the similarity relationship table A can be applied is not larger than the similarity relationship table application counter (NO in step S15), the processing illustrated in FIG.
In other words, the character string that is not similar to the search character string “Aiweo” is selected as the search target character string “Aeulere” without executing the process of comparing the remaining fourth and fifth search character strings (steps 16 and 13). Can be determined as follows.

  As described above, according to the present embodiment as described above, the comparison time (or search time) is expressed by expressing the type of sound by the position of the bit that becomes 1 in the bit string and performing the comparison using the bit product operation. Can be shortened. That is, calculation efficiency can be improved.

  Further, as described above, according to the present embodiment, the applicable number of similar relationship tables and the threshold value are set according to the rule, and the comparison is performed within this range, so that unnecessary comparison processing may not be performed. The comparison time (or search time) can be shortened. That is, calculation efficiency can be improved. In addition, since it is not necessary to perform useless similar expansion, memory efficiency can be improved. Further, compared to the method using an automaton, the user can easily set rules and understand the rules, so that the maintenance is good.

  Further, for the sake of simplification of explanation, the comparison with one search target character string is given as an example. However, the comparison with all the plurality of search target character strings held in the search target character string holding unit 25 is as follows. Needless to say, the present application is applicable. For example, if the search target character string holding unit 25 holds the search target character string in a trie structure, backtracking is used, and if the search target character string holding unit 25 holds the search target character string in a list format, the processing shown in FIG. 5 is repeated. As a result, it is possible to compare with all the search target character strings held in the search target character string holding unit 25 and output the result to the search result aggregation unit 27.

<Embodiment 2>
FIG. 6 is a functional configuration diagram (part 2) of an example of the information processing apparatus 1. As illustrated in FIG. 6, the information processing apparatus 1 includes a search character string acquisition unit 21, a setting unit 22, a rule storage unit 23, a search unit 24, and a search target character string storage unit 25 as functional configurations. The similarity relation table holding unit 26, the search result aggregation unit 27, and the rule operation unit 28 are included.

Since the functional configuration other than the rule operation unit 28 is the same as that of the first embodiment, the description thereof is omitted in this embodiment.
The rule operation unit 28 determines the rule held in the rule holding unit 23, the applicable number of times of the similarity table corresponding to the rule, the threshold value, and the like in accordance with a user operation using the input device 11, the display device 12, and the like. The rule holding unit 23 sets a new rule, the applicable number of times of the similarity table corresponding to the rule, a threshold value, and the like.

  For example, in the first embodiment, as an example, the rule “1 to 2 sounds are in the relationship of the similarity relationship table A” has been described as an example. When it is desired to change to “A relationship”, the rule operation unit 28 is described in a file held in the rule holding unit 23 according to a user operation using the input device 11, the display device 12, or the like. “1 to 2 sounds are in the relationship in the similarity relationship table A” is changed to “0 to 2 sounds are in the relationship in the similarity relationship table A”. Further, the rule operation unit 28 changes the applicable number of times and the threshold threshold value of the similarity relationship table as shown in FIG. 3 held in the rule holding unit 23 from 1 to 0 corresponding to the rule. By changing in this way, a search target character string that completely matches the search character string is also stored in the search result aggregating unit 27 as a character string similar to the search character string, that is, as a search target character string that has been hit by the search. It can be changed to output specifications.

  As described above, according to the present embodiment, as described above, the user can more easily set and change the applicable number of rules and the similarity table corresponding to the rule, the threshold value, and the like. That is, maintainability can be improved.

<Other embodiments>
In the first embodiment described above, the description has been given by taking “one sound or two sounds are in the relationship of the similarity relationship table A” as an example of the rule. However, the rule may be, for example, “one or two sounds are in the relationship of the similarity relationship table A, the similarity relationship table B, or the similarity relationship table C”. In such a case, the search unit 24 applies the similarity relationship table A, the similarity relationship table B, and the similarity relationship table C to determine whether or not the search character and the search target character are similar in step S16 of FIG. And compare. At this time, if a priority is added to each table, the search unit 24 performs comparison using the similarity relationship table according to the priority.

  In such a case, the applicable number of similar relationship tables may be set for each table, or one value may be set for all tables. Similarly, the threshold value may be set for each table, or one value may be set for all tables.

  In the above-described embodiment, in step S22 of FIG. 5, the search unit 24 searches the target search target character string as a character string similar to the search character string, that is, as a search target character string hit in the search. The description was made so that the result is output to the result aggregation unit 27.

  However, the search unit 24 may output the search target character string to the search result aggregation unit 27 together with the similarity relationship table application counter in step S22 of FIG. By doing so, the search result aggregating unit 27 can perform weighting according to the similarity relationship table application counter and output the search result according to this weighting.

  For example, it can be said that a search target character string having a similarity relationship table application counter of 1 is more similar to a search character string than a search target character string having a similarity relationship table application counter of 2. Therefore, the search result aggregating unit 27 weights the search target character string whose similarity relation table application counter is 1 than the search target character string whose similarity relation table application counter is 2, for example, the weight is higher (more similar) The search result character strings may be output in order from the search target character string.

  By doing so, the user can obtain the search target character strings in the order of search target character strings that are more similar to the search character string, for example.

  In the above-described embodiment, an example in which the number of times that the similarity table can be applied is set for the search character string has been described. However, as shown in FIG. 7 or FIG. 8, each character (one character) constituting the search character string. May be set to the number of times the similarity table can be applied, or the similarity table may be set for each character set (that is, two characters, three characters,...) Constituting the search character string. The applicable number of times may be set. However, when setting the number of times that the similarity table can be applied to each character set constituting the search character string, the comparison between the search character string and the search target character string is not a single character but a set of each character. (That is, two letters, three letters, ...).

  In the above-described embodiment, one similar character corresponding to one applicable character is one character. However, as shown in FIG. 9, the similar character corresponding to one applicable character is a character string composed of a plurality of characters. Also good.

  Further, in the above-described embodiment, description has been made using an example in which a table (similarity relationship table) is used to indicate a similarity relationship between characters (or sound of characters), but the implementation of the present invention is limited to the table. For example, it may be a file (a file in which a similar relationship is described) or the like.

  In the above-described embodiment, the search character string and the search target character string have been described using an example in which the character strings are compared from the top of each other. For example, the comparison may be performed from the end of each character string or from a specific arbitrary place.

  In the above-described embodiment, the search character string acquisition unit 21 has been described to acquire, as a search character string, a character string to be searched that is input by a user on a search screen or the like. It is not limited to that. The search character string acquisition unit 21 may acquire, for example, a character string read by an OCR (Optical Character Reader) as a search target. In this case, for example, the information processing apparatus 1 corresponds to the image forming apparatus or the like, and may read the character string by itself, or may acquire the character string from the image forming apparatus that has read the character string. The OCR reads, for example, an underlined character string as a search character string.

  In the above-described first embodiment, the rule is explained as an example of “one sound or two sounds are in the relationship of the similarity relationship table A”, so that the search is a similarity search. It is not limited to the similarity search, but an arrangement between symbols according to a certain rule may be searched. For example, if the first and second amino acid sequences from the beginning of a part of the DNA sequence “A (adenine) G (guanine) C (cytosine) T (thymine)” are the same, they have the same characteristics. If it is desired to search for DNA having the same characteristics under these conditions, the application relation table shown in FIG. 10, the applicable number of times, and the threshold value are set and processed in the same manner as in the above-described embodiment, so that “AGGT”, “AGCA”, “ It becomes possible to search for DNA sequences having the same characteristics such as “AGTT”. The application relationship table here is a table having the same role as the similarity relationship table (FIG. 4) in the first embodiment.

  In the embodiment described above, the search target character string is configured by a character string in which a part or all of the search character string is replaced. However, the embodiment of the present invention is not limited to this, and the search character string is not limited to this. The search target character string may be configured such that an arbitrary number of blanks, characters, symbols, and the like are inserted at arbitrary positions.

  Further, in the above-described embodiment, the description of the similarity search has been made by taking the character string and the symbol as an example. However, the implementation of the present invention is not limited to these, and the search is performed between all data strings such as voice and images. There may be.

As described above, according to each of the above-described embodiments, it is possible to provide a search device, a search method, and a program with good system maintainability as well as calculation efficiency and memory efficiency.
In addition, you may implement combining each embodiment mentioned above arbitrarily.

  The preferred embodiments of the present invention have been described in detail above, but the present invention is not limited to such specific embodiments, and various modifications can be made within the scope of the gist of the present invention described in the claims.・ Change is possible.

2 is a hardware configuration diagram of an example of an information processing apparatus 1. FIG. 2 is a functional configuration diagram (part 1) of an example of an information processing apparatus 1. FIG. FIG. 6 is a diagram (part 1) illustrating an example of the number of applicable times and a threshold value of a similarity relationship table. It is a figure (the 1) which shows an example of a similarity relationship table. It is a flowchart which shows an example of a comparison process. FIG. 3 is a functional configuration diagram (part 2) of an example of the information processing apparatus 1; FIG. 6 is a diagram (part 2) illustrating an example of the number of applicable times and a threshold value of a similarity relationship table. FIG. 13 is a diagram (part 3) illustrating an example of the applicable number of times and threshold values of the similarity relationship table. It is a figure (the 2) which shows an example of a similarity relationship table. It is a figure which shows an example of a similar relationship table, the applicable frequency | count of a similar relationship table, and a threshold value.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Information processing apparatus 11 Input apparatus 12 Display apparatus 13 Recording medium drive apparatus 14 Recording medium 15 ROM
16 RAM
17 CPU
18 Interface device 19 HD
21 Search character string acquisition unit 22 Setting unit 23 Rule storage unit 24 Search unit 25 Search target character string storage unit 26 Similarity relationship table storage unit 27 Search result aggregation unit 28 Rule operation unit

Claims (9)

Search target data string holding means for storing the search target data string;
A search data string acquisition means for acquiring a search data string;
A range of the number of times of using an object that defines a similar relationship as to whether or not the search data sequence and the search target data sequence are similar to the search data sequence acquired by the search data sequence acquisition means Setting means for setting according to a predetermined rule;
As search processing, within the range of the number of times set in the setting means, the search data string acquired in the search data string acquisition means, the search target data string held in the search target data string holding means, Search means for performing a comparison as to whether or not there is a similarity relationship defined for the object;
I have a,
The search means uses the object to compare the unit data constituting the search data string and the unit data constituting the search target data string within the range of the number of times set by the setting means. If there is unit data that does not match when the unit data does not match, the unit data that does not match each other within the range of the number of times set by the setting unit is defined in the object. A search device characterized in that a comparison is made as to whether or not there is any . Search target data string holding means for storing the search target data string;
A search data string acquisition means for acquiring a search data string;
A range of the number of times of using an object that defines a similar relationship as to whether or not the search data sequence and the search target data sequence are similar to the search data sequence acquired by the search data sequence acquisition means Setting means for setting according to a predetermined rule;
As search processing, within the range of the number of times set in the setting means, the search data string acquired in the search data string acquisition means, the search target data string held in the search target data string holding means, Search means for performing a comparison as to whether or not there is a similarity relationship defined for the object;
Have
The setting means uses the table that holds the relationship between unit data and similar unit data related to the search target data that can be applied to the search data string to be output as a search result as an object, and the search data string acquisition means For the search data string acquired in step 1, the range of the number of times the object is used is set according to a predetermined rule,
The search means uses the object to compare the unit data constituting the search data string and the unit data constituting the search target data string within the range of the number of times set by the setting means. If there is unit data that does not match when the unit data does not match, the unit data that does not match each other within the range of the number of times set by the setting unit is defined in the object. A search device characterized in that a comparison is made as to whether or not there is any . Search target data string holding means for storing the search target data string;
A search data string acquisition means for acquiring a search data string;
A range of the number of times of using an object that defines a similar relationship as to whether or not the search data sequence and the search target data sequence are similar to the search data sequence acquired by the search data sequence acquisition means Setting means for setting according to a predetermined rule;
As search processing, within the range of the number of times set in the setting means, the search data string acquired in the search data string acquisition means, the search target data string held in the search target data string holding means, Search means for performing a comparison as to whether or not there is a similarity relationship defined for the object;
Have
The search means uses the object to count the number of comparisons between the unit data constituting the search data string and the unit data constituting the search target data string, and according to the counted number A search device that performs weighting according to the degree of application of the search data string and the search target data string, and outputs the search target data string to be applied to the search data string in accordance with the weighting . The said search means performs the comparison with the unit data which comprises the said search data row | line | column, and the unit data which comprise the said search object data row | line | column using bit product operation, The Claim 1 thru | or 3 characterized by the above-mentioned. The search device according to any one of the above. A plurality of the objects,
The search means is acquired by the search data string acquisition means as a search process using one of the plurality of objects in accordance with the priority order within the range of the number of times set by the setting means. 5. The search device according to claim 1 , wherein the search data string is compared with the search object data string held in the search object data string holding unit . A search method in a search device,
A search data string acquisition stage for acquiring a search data string;
For the search data string acquired in the search data string acquisition step, a range of the number of times to use an object that defines a similar relationship as to whether the search data string and the search target data string are similar A setting stage for setting according to a predetermined rule;
As a search process, within the range of the number of times set in the setting stage, the search data string acquired in the search data string acquisition stage, the search target data string held in the search target data string holding stage, A search stage for comparing whether or not there is a similarity relationship defined for the object;
Including
In the search stage, within the range of the number of times set in the setting stage, using the object, the unit data constituting the search data string and the unit data constituting the search target data string are compared. If there is unmatched unit data for each unit data, the unmatched unit data is defined in the object using the object within the range of the number of times set in the setting step. A search method characterized in that a comparison is made as to whether or not there is any. A search method in a search device,
A search data string acquisition stage for acquiring a search data string;
For the search data string acquired in the search data string acquisition step, a range of the number of times to use an object that defines a similar relationship as to whether the search data string and the search target data string are similar A setting stage for setting according to a predetermined rule;
As a search process, within the range of the number of times set in the setting stage, the search data string acquired in the search data string acquisition stage, the search target data string held in the search target data string holding stage, A search stage for comparing whether or not there is a similarity relationship defined for the object;
Including
In the setting step, the search data string acquisition step is performed using, as an object, a table holding a relationship between the unit data and the similar unit data related to the search target data applicable to the search data string to be output as a search result. For the search data string acquired in step 1, the range of the number of times the object is used is set according to a predetermined rule,
In the search stage, within the range of the number of times set in the setting stage, using the object, the unit data constituting the search data string and the unit data constituting the search target data string are compared. If there is unmatched unit data for each unit data, the unmatched unit data is defined in the object using the object within the range of the number of times set in the setting step. A search method characterized in that a comparison is made as to whether or not there is any. A search method in a search device,
A search data string acquisition stage for acquiring a search data string;
For the search data string acquired in the search data string acquisition step, a range of the number of times to use an object that defines a similar relationship as to whether the search data string and the search target data string are similar A setting stage for setting according to a predetermined rule;
As a search process, within the range of the number of times set in the setting stage, the search data string acquired in the search data string acquisition stage, the search target data string held in the search target data string holding stage, A search stage for comparing whether or not there is a similarity relationship defined for the object;
Including
In the search step, the object is used to count the number of comparisons between the unit data constituting the search data string and the unit data constituting the search target data string, and according to the counted number A search method characterized by weighting the degree of applicability between the search data string and the search object data string, and outputting the search object data string to be applied to the search data string according to the weighting . A program for causing a computer to execute the search method according to any one of claims 6 to 8.

Images