WO2016024367A1 - Autonomous knowledge enhancement device - Google Patents

Autonomous knowledge enhancement device Download PDF

Info

Publication number
WO2016024367A1
WO2016024367A1 PCT/JP2014/080004 JP2014080004W WO2016024367A1 WO 2016024367 A1 WO2016024367 A1 WO 2016024367A1 JP 2014080004 W JP2014080004 W JP 2014080004W WO 2016024367 A1 WO2016024367 A1 WO 2016024367A1
Authority
WO
WIPO (PCT)
Prior art keywords
pattern
patterns
knowledge
information
language
Prior art date
Application number
PCT/JP2014/080004
Other languages
French (fr)
Japanese (ja)
Inventor
洋彰 宮崎
Original Assignee
洋彰 宮崎
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Family has litigation
First worldwide family litigation filed litigation Critical https://patents.darts-ip.com/?family=52492424&utm_source=google_patent&utm_medium=platform_link&utm_campaign=public_patent_search&patent=WO2016024367(A1) "Global patent litigation dataset” by Darts-ip is licensed under a Creative Commons Attribution 4.0 International License.
Application filed by 洋彰 宮崎 filed Critical 洋彰 宮崎
Publication of WO2016024367A1 publication Critical patent/WO2016024367A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N99/00—Subject matter not provided for in other groups of this subclass

Definitions

  • the present invention relates to artificial intelligence that evaluates the meaning, novelty, authenticity, logic validity, and the like of input linguistic information, autonomously acquires knowledge, and improves intelligence for solving problems.
  • the machine When the machine is caused to operate, it is realized by incorporating and executing a program created in advance by a program language in a computer mounted on the machine.
  • a program is created so that the corresponding operation is executed when a preset condition is detected. If the condition detection and the corresponding operation are not appropriate, modify the program installed on the computer.
  • the meaning, authenticity, and validity of the input information are evaluated, the validity of the logic is evaluated, the knowledge is constructed autonomously, the question is presented if there is an unknown point, and the answer is given when a human etc. answers the question.
  • the machine When the machine is caused to operate, it is realized by incorporating and executing a program created in advance by a program language in a computer mounted on the machine.
  • a program is created so that the corresponding operation is executed when a preset condition is detected. If the condition detection and the corresponding operation are not appropriate, modify the program installed on the computer.
  • the machine When the machine is caused to operate, it is realized by incorporating and executing a program created in advance by a program language in a computer mounted on the machine. A program is created so that the corresponding operation is executed when a preset condition is detected. If the condition detection and the corresponding operation are not appropriate, modify the program installed on the computer.
  • B, C1, C2, D1, and D2 by patterns and defining the connection relationship between the patterns, it is possible to express human thought patterns expressed by conditional logic as transitions between patterns.
  • the autonomous knowledge improvement apparatus expresses human thinking as a pattern, and expresses thinking transition as pattern to pattern change.
  • the patterns and the connections between the patterns are recorded on the pattern recorder, the corresponding patterns are called up according to the situation, and the corresponding processing is performed.
  • the machine can implement similar thought patterns and behavior patterns. Since the change of the thought pattern can be realized by changing the pattern recorded in the pattern recorder and the connection relationship between the patterns, it is not necessary to change the program which is essential in the prior art.
  • the autonomous knowledge improvement apparatus records human thinking as patterns and connection relations between patterns from the input linguistic information, and constructs knowledge. In addition, the novelty of the input linguistic information, consistency / inconsistency with the already recorded knowledge, the validity of the logic, etc.
  • the autonomous knowledge improvement apparatus converts language information, image information and speech information into patterns. Since the words of linguistic information are converted into corresponding patterns, and the words are expressed as a combination of words, the word pattern can be expressed as a combination of word patterns. Furthermore, since a sentence can be expressed as a combination of words, a sentence pattern can be expressed as a combination of word patterns. Image information and audio information are also converted into patterns. The transformed image information is identified and linked to a linguistic pattern indicative of the corresponding object. The converted speech information is also identified as words, words or sounds. It is associated with a linguistic pattern indicating corresponding words, words or sounds. Human thinking is expressed by words, words, sentences, and sentences, but the pattern is considered to be typed to some extent.
  • conditional logic can be expressed programmatically as follows.
  • the patterns and the connections between the patterns are recorded on the pattern recorder, the corresponding patterns are called up according to the situation, and the corresponding processing is performed.
  • the machine can implement similar thought patterns and behavior patterns. Since the change of the thought pattern can be realized by changing the pattern recorded in the pattern recorder and the connection relationship between the patterns, it is not necessary to change the program which is essential in the prior art.
  • the autonomous knowledge improvement apparatus records human thinking as patterns and connection relations between patterns from the input linguistic information, and constructs knowledge. In addition, the novelty of the input linguistic information, consistency / inconsistency with the already recorded knowledge, the validity of the logic, etc.
  • the autonomous knowledge improvement apparatus records human thinking as patterns and connection relations between patterns from the input linguistic information, and constructs knowledge.
  • the novelty of the input linguistic information, consistency / inconsistency with the already recorded knowledge, the validity of the logic, etc. are evaluated, and if it is judged that it is useful information, it is sequentially added to the knowledge. .
  • the autonomous knowledge improvement apparatus can accumulate useful knowledge and realize improvement of knowledge.
  • the autonomous knowledge improvement device evaluates the matching / mismatch between the input information and the already recorded knowledge, and the validity of the logic, if the mismatch or the jump of the logic is detected, the input information Present questions about the basis of whether or not it is correct.
  • the autonomous knowledge improvement apparatus records human thinking as patterns and connection relations between patterns from the input linguistic information, and constructs knowledge.
  • the novelty of the input linguistic information, consistency / inconsistency with the already recorded knowledge, the validity of the logic, etc. are evaluated, and if it is judged that it is useful information, it is sequentially added to the knowledge. .
  • the autonomous knowledge improvement apparatus can accumulate useful knowledge and realize improvement of knowledge.
  • the patterns generated by the autonomous knowledge enhancing device can be monitored sequentially. If the generated pattern is not appropriate and correction is required, the pattern can be maintained and managed in an appropriate state by performing correction of the pattern and connection between the patterns from outside. [Means for Solving the Problems] (Corresponding to Claim 5)
  • the autonomous knowledge improvement apparatus records human thinking as patterns and connection relations between patterns from the input linguistic information, and constructs knowledge.
  • the novelty of the input linguistic information, consistency / inconsistency with the already recorded knowledge, the validity of the logic, etc. are evaluated, and if it is judged that it is useful information, it is sequentially added to the knowledge. .
  • the autonomous knowledge improvement apparatus can accumulate useful knowledge and realize improvement of knowledge.
  • the relationship between the patterns generated by the autonomous knowledge improvement device has a function to generalize. This function can be generalized if it can be widely interpreted rather than limitingly showing the relationship between a certain pattern and a pattern. As a concrete example, consider the following statement. Example: Hanako was pleased with her grades.
  • the autonomous knowledge improvement apparatus records human thinking as patterns and connection relations between patterns from the input linguistic information, and constructs knowledge.
  • the novelty of the input linguistic information, consistency / inconsistency with the already recorded knowledge, the validity of the logic, etc. are evaluated, and if it is judged that it is useful information, it is sequentially added to the knowledge. .
  • the autonomous knowledge improvement apparatus can accumulate useful knowledge and realize improvement of knowledge.
  • the pattern generated by the autonomous knowledge improvement device is converted into a control output by pattern reverse conversion to implement control of the machine.
  • FIG. 1 is a diagram showing the configuration of an autonomous knowledge improvement apparatus according to an embodiment of the present invention.
  • reference numeral 1 denotes a pattern converter which converts language information into a pattern.
  • Reference numeral 2 denotes a pattern recorder which records patterns and connection between patterns.
  • a pattern holder 3 holds a pattern for processing.
  • a controller 4 controls the pattern holder.
  • An inter-pattern processor 5 processes the relationship between patterns.
  • a pattern converter 1 converts language information into patterns.
  • FIG. 7 shows a configuration example of the pattern converter.
  • Linguistic information identifies words, words, and the arrangement is converted by syntactic analysis. After being organized into subjects, predicates, etc., they are converted into patterns, and language patterns are generated. Here, the elements of the pattern are expressed by “ON”, “OFF” or “1” “0”. Although other expressions may be used, this expression is adopted due to the simplicity of the correlation processing.
  • the language pattern is input to the pattern holder 3 by way of the controller 4.
  • the third pattern holder temporarily holds a pattern for processing.
  • the input pattern and the processed pattern are recorded on two pattern recorders.
  • the connection between patterns is also recorded on the two pattern recorders.
  • the inter-pattern processor 5 performs inter-pattern processing.
  • the pattern input from the pattern holder 3 is processed, and the processed result is output to the pattern holder 3.
  • the controller of 4 controls the operation of the pattern holder of 3, the pattern recorder of 2, and the processor between 5 patterns.
  • the controller 4 calls the processing flow from the pattern recorder 2 according to the situation, and performs processing of the pattern.
  • the pattern being processed is temporarily stored in the 3 pattern holder.
  • the patterns stored in the pattern holder 3 are output to the pattern recorder 2 and the inter-pattern processor 5 to perform recording and pattern processing.
  • the patterns recorded on the two pattern recorders are recalled to the three pattern holders as appropriate. Further, the output of the 5 inter-pattern processor is also output to the 3 pattern holder and is temporarily held.
  • the pattern is processed by the processing flow recorded in advance.
  • the processed patterns are optionally recorded on two pattern recorders to build knowledge.
  • the connection relationship between patterns is set by a human or the like from the outside in the initial phase (learning phase).
  • the process flow is set to operate autonomously when knowledge has accumulated to a certain extent.
  • the device autonomously evaluates the linguistic information (novelty, authenticity, logic validity) and builds up the knowledge autonomously.
  • FIG. 8 shows an operation example of the connection relation between the language pattern and the language pattern as the connection relation between the recording modules stored inside the pattern recorder.
  • the language pattern A and the language pattern B are recorded in the recording module in the pattern recorder.
  • the settings of language pattern A and language pattern B are set from the pattern holder. Each set pattern propagates a signal line and is input to the pattern recorder. Among the recording modules in the pattern recorder, a vacant module is selected, and the language pattern A and the language pattern B are recorded.
  • setting of the connection relationship between patterns will be described. First, the language pattern A is set in the pattern holder to excite the pattern A.
  • the language pattern B is set in another module of the pattern holder and the pattern B is excited.
  • the inter-pattern connection generation is enabled, the inter-pattern connection is generated from the recording module of the language pattern A to the recording module of the language pattern B.
  • the inter-pattern connection is generated, when the language pattern A is excited next, the language pattern B continues to be excited.
  • logic: A ⁇ B can be configured.
  • FIG. 9 shows the method of construction of logic: A ⁇ B using a construction diagram. It can be seen that a connection relation is generated from language pattern A to language pattern B in FIG.
  • FIG. 10 represents an operation example of the conditional processing as a connection relationship between recording modules stored inside the pattern recording device.
  • the condition (A) set when carrying out the process P is searched from the state recording area.
  • the state of the condition (A) is recorded in advance in the state recording area.
  • the search result of the condition (A) is stored in the area of the search result (B) of the pattern holder.
  • the search result (B) is compared with the preset conditions. If (B) is (C1), then (D1) is performed. If (B) is (C2), then (D2) is performed.
  • the determined process (D1 or D2) is stored in the process result storage.
  • FIG. 12 shows an operation of generating an answer to a question.
  • the questions are expressed as search keys.
  • the search key is composed of target sentence components.
  • a subject and a predicate are specified, and "when: When?" Is executed is searched.
  • Related information is searched from the subject and the predicate using "S1" + "V1" as a search key.
  • the patterns recorded in the pattern recorder when a pattern corresponding to “S1” + “V1” is found and a pattern corresponding to “When?” Exists, it is stored as an answer plan in the area of the search result. In the example of FIG. 11, it can be seen that “When 1” is stored as an answer in the search result for the question “When?”.
  • FIG. 13 shows an operation example of generating an answer to a question using a configuration diagram.
  • the pattern corresponding to the question set in the pattern holder is input to the answer generator for the question in the inter-pattern processor.
  • the answer generator searches for relevant patterns in the pattern recorder using a search key corresponding to the question.
  • the related patterns in the pattern recorder are collated, and "When 1" is stored as an answer in the answer storage area in response to the "When?" Question.
  • FIG. 14 shows an operation example of searching for synonyms.
  • the language pattern A1 and the language pattern A2 are synonyms.
  • the language pattern C1 indicates that the input language pattern has a synonym relation.
  • the connection relationship between the language patterns is set so that the language pattern C1 is excited when the language pattern A1 and the language pattern A2 are excited.
  • the connection relationship between the language patterns is set so that the language pattern A2 is excited when the language pattern A1 and the language pattern C1 are excited.
  • connection relationship between the language patterns is set so that the language pattern A1 is excited when the language pattern A2 and the language pattern C1 are excited.
  • connection between language patterns is defined in this way, when the language pattern A1 and the language pattern C1 are input, the language pattern A2 is excited.
  • the synonym of A1 is A2.
  • the synonym of A2 is A1.
  • the relationship between the patterns is described as an example, but the relationship between the patterns indicates the opposite meaning, indicates the similar meaning, indicates the related meaning, etc. It is possible to define.
  • FIG. 15 shows an operation example of the match / mismatch detection using a configuration diagram. It is assumed that the language pattern A1 and the language pattern A2 have the same meaning relationship. This can be defined by setting up the connections between the recording modules of the pattern recorder.
  • the language pattern A1 and the language pattern A2 input to the pattern holder are input to the match / mismatch detector of the inter-pattern processor.
  • the language pattern input to the inter-pattern processor is output to the pattern recorder, and the relationship defined in the pattern recorder is excited.
  • a language pattern indicating the relationship between the excited patterns is again input to the inter-pattern processing matching / mismatch detector.
  • a language pattern indicating this relationship is stored in the area holding the relationship of the pattern holder. In this way, it is possible to answer the matching / mismatch between the patterns.
  • FIG. 16 shows an operation example of novelty detection.
  • a language pattern When a language pattern is input, the associated language pattern recorded in the pattern recorder is retrieved. First, a language pattern showing the same meaning as the input language pattern is searched, and a search key for searching the language pattern related to the input language information is searched using these searched language patterns. It is generated. Since this search key envelops the language which shows the same meaning as the input language, and an equivalent meaning, a related language pattern will be searched at the meaning level with the input language information.
  • the language pattern of the existing information thus retrieved and the language pattern of the input information are compared at the semantic level. In the drawing, when the existing information has a signal but the input information does not have a signal, it means that the existing information has information but the input information has no information.
  • the existing information has a signal and the input information also has a signal, it means that the information indicated in the input information already has information in the existing information. If the existing information has no signal, but the input information has a signal, it means that the input information includes information that is not included in the existing information. In other words, there is novelty. Thus, novelty can be detected by comparing the language patterns of the existing related information with the input information at the semantic level.
  • FIG. 17 shows an operation example of novelty detection using a configuration diagram.
  • information When information is input, it is output to the novelty detector of the inter-pattern processor.
  • a search key for searching related information is generated by searching for a language pattern having the same meaning and equivalent meaning from the input language pattern, and a search of related information is performed.
  • the search results are stored in the area of the pattern holder.
  • a comparison is performed between a language pattern having the same or equivalent meaning as the input language pattern and a language pattern having the same or equivalent meaning as the retrieved related information.
  • a result of comparison when it is detected that there is a language pattern not existing information in the language pattern of the input information, it is considered that there is novelty, and the result is stored in the area of the pattern holder.
  • the novelty of the information thus input can be detected.
  • FIG. 2 is a diagram showing the configuration of an autonomous knowledge improvement apparatus according to an embodiment of the present invention.
  • reference numeral 1 denotes a pattern converter for converting language information, image information and speech information into a pattern.
  • Reference numeral 2 denotes a pattern recorder which records patterns and connection between patterns.
  • a pattern holder 3 holds a pattern for processing.
  • a controller 4 controls the pattern holder.
  • An inter-pattern processor 5 processes the relationship between patterns.
  • the pattern converter 1 converts language information, image information and speech information into patterns.
  • FIG. 7 shows a configuration example of the pattern converter.
  • Linguistic information identifies words, words, and the arrangement is converted by syntactic analysis. After being organized into subjects, predicates, etc., they are converted into patterns, and language patterns are generated. Here, the elements of the pattern are expressed by “ON”, “OFF” or “1” “0”. Although other expressions may be used, this expression is adopted due to the simplicity of the correlation processing.
  • Image information and audio information are also converted into patterns. The transformed image information is identified and linked to a linguistic pattern indicative of the corresponding object.
  • the transformed speech information is also identified as a word, word or sound and is associated with a linguistic pattern indicating the corresponding word, word or sound.
  • the language pattern is input to the pattern holder 3 by way of the controller 4.
  • the third pattern holder temporarily holds a pattern for processing.
  • the input pattern and the processed pattern are recorded on two pattern recorders.
  • the connection between patterns is also recorded on the two pattern recorders.
  • the inter-pattern processor 5 performs inter-pattern processing.
  • the pattern input from the pattern holder 3 is processed, and the processed result is output to the pattern holder 3.
  • the controller of 4 controls the operation of the pattern holder of 3, the pattern recorder of 2, and the processor between 5 patterns.
  • the controller 4 calls the processing flow from the pattern recorder 2 according to the situation, and performs processing of the pattern.
  • the pattern being processed is temporarily stored in the 3 pattern holder.
  • the patterns stored in the pattern holder 3 are output to the pattern recorder 2 and the inter-pattern processor 5 to perform recording and pattern processing.
  • the patterns recorded on the two pattern recorders are called to the three pattern holders as appropriate. Further, the output of the 5 inter-pattern processor is also output to the 3 pattern holder and is temporarily held.
  • the pattern is processed by the processing flow recorded in advance.
  • the processed patterns are optionally recorded on two pattern recorders to build knowledge.
  • the connection relationship between patterns is set by a human or the like from the outside in the initial phase (learning phase).
  • the process flow is set to operate autonomously when knowledge has accumulated to a certain extent. At this stage, the device autonomously evaluates the linguistic information (novelty, authenticity, logic validity) and builds up the knowledge autonomously. [Third Embodiment of the Invention] (corresponds to claim 3)
  • FIG. 3 is a diagram showing the configuration of an autonomous knowledge improvement apparatus according to an embodiment of the present invention.
  • reference numeral 1 denotes a pattern converter which converts language information into a pattern.
  • Reference numeral 2 denotes a pattern recorder which records patterns and connection between patterns.
  • a pattern holder 3 holds a pattern for processing.
  • a controller 4 controls the pattern holder.
  • An inter-pattern processor 5 processes the relationship between patterns.
  • a pattern converter 1 converts language information into patterns.
  • the language pattern is input to the pattern holder 3 by way of the controller 4.
  • the third pattern holder temporarily holds a pattern for processing.
  • the input pattern and the processed pattern are recorded on two pattern recorders.
  • the connection between patterns is also recorded on the two pattern recorders.
  • the inter-pattern processor 5 performs inter-pattern processing.
  • the pattern input from the pattern holder 3 is processed, and the processed result is output to the pattern holder 3.
  • the controller of 4 controls the operation of the pattern holder of 3, the pattern recorder of 2, and the processor between 5 patterns.
  • the controller 4 calls the processing flow from the pattern recorder 2 according to the situation, and performs processing of the pattern.
  • the pattern being processed is temporarily stored in the 3 pattern holder.
  • the patterns stored in the pattern holder 3 are output to the pattern recorder 2 and the inter-pattern processor 5 to perform recording and pattern processing.
  • the patterns recorded on the two pattern recorders are recalled to the three pattern holders as appropriate. Further, the output of the 5 inter-pattern processor is also output to the 3 pattern holder and is temporarily held.
  • the pattern is processed by the processing flow recorded in advance.
  • the processed patterns are optionally recorded on two pattern recorders to build knowledge.
  • the connection relationship between patterns is set by a human or the like from the outside in the initial phase (learning phase).
  • the process flow is set to operate autonomously when knowledge has accumulated to a certain extent. At this stage, the device autonomously evaluates the linguistic information (novelty, authenticity, logic validity) and builds up the knowledge autonomously.
  • the autonomous knowledge improvement device evaluates the matching / mismatch between the input information and the already recorded knowledge, and the validity of the logic, if the mismatch or the jump of the logic is detected, the input information Present questions about the basis of whether or not it is correct.
  • related information recorded using the input information as a search key is searched.
  • the matching / inconsistency check of the retrieved related information and the input information is carried out, and if an inconsistency is detected, a question is presented as to whether the entered information is correct.
  • many knowledge (a truth, a fact, a rule, a custom, a common sense, etc.) are recorded on the pattern recorder of the autonomous knowledge improvement device.
  • the entered information is checked against these knowledge to see if there is any inconsistency with the knowledge already recorded. If there is no inconsistency or relevance between the input information and the information already recorded, it is judged that the logic is low and a question is presented as to whether the input information is correct. When inconsistencies are detected between the temporarily input information and the recorded knowledge, we will compare the grounds. By checking the basis of the input information and the related information with the set of language patterns constituting the basis generated above, it is possible to compare which is appropriate as the basis. If no match is found when matching against the linguistic patterns that make up the basis, the basis will not be explained in the past.
  • FIG. 18 shows an example of logic validity evaluation.
  • a ⁇ B is recorded as knowledge on the pattern recording device.
  • a ⁇ C is input as new information.
  • the new information retrieves relevant information recorded in the pattern recorder. Since the information A ⁇ B is already stored as knowledge in the pattern recorder, this information is compared with the newly input information AAC. At this time, the information A ⁇ C is judged as new information, and is the input information (A ⁇ C) correct? The question is presented.
  • FIG. 19 shows an operation example when a human or the like answers the presented question.
  • the autonomous knowledge improvement device derives (A ⁇ B ⁇ C) from the new information (B ⁇ C) and the knowledge (A ⁇ B) already recorded, and recognizes that (A ⁇ C) is appropriate.
  • a ⁇ B ⁇ C new information
  • a ⁇ B knowledge already recorded
  • FIG. 4 is a diagram showing the configuration of an autonomous knowledge improvement apparatus according to an embodiment of the present invention.
  • reference numeral 1 denotes a pattern converter which converts language information into a pattern.
  • Reference numeral 2 denotes a pattern recorder which records patterns and connection between patterns.
  • a pattern holder 3 holds a pattern for processing.
  • a controller 4 controls the pattern holder.
  • An inter-pattern processor 5 processes the relationship between patterns.
  • a pattern converter 1 converts language information into patterns.
  • the language pattern is input to the pattern holder 3 by way of the controller 4.
  • the third pattern holder temporarily holds a pattern for processing.
  • the input pattern and the processed pattern are recorded on two pattern recorders.
  • the connection between patterns is also recorded on the two pattern recorders.
  • the inter-pattern processor 5 performs inter-pattern processing.
  • the pattern input from the pattern holder 3 is processed, and the processed result is output to the pattern holder 3.
  • the controller of 4 controls the operation of the pattern holder of 3, the pattern recorder of 2, and the processor between 5 patterns.
  • the controller 4 calls the processing flow from the pattern recorder 2 according to the situation, and performs processing of the pattern.
  • the pattern being processed is temporarily stored in the 3 pattern holder.
  • the patterns stored in the pattern holder 3 are output to the pattern recorder 2 and the inter-pattern processor 5 to perform recording and pattern processing.
  • the patterns recorded on the two pattern recorders are recalled to the three pattern holders as appropriate. Further, the output of the 5 inter-pattern processor is also output to the 3 pattern holder and is temporarily held.
  • the pattern is processed by the processing flow recorded in advance.
  • the processed patterns are optionally recorded on two pattern recorders to build knowledge.
  • the connection relationship between patterns is set by a human or the like from the outside in the initial phase (learning phase).
  • the process flow is set to operate autonomously when knowledge has accumulated to a certain extent. At this stage, the device autonomously evaluates the linguistic information (novelty, authenticity, logic validity) and builds up the knowledge autonomously.
  • patterns generated by the autonomous knowledge improvement device can be monitored one after another. If the generated pattern is not appropriate and correction is required, the pattern can be maintained and managed in an appropriate state by performing correction of the pattern and connection between the patterns from outside.
  • FIG. 5 is a diagram showing the configuration of an autonomous knowledge improvement apparatus according to an embodiment of the present invention.
  • reference numeral 1 denotes a pattern converter which converts language information into a pattern.
  • Reference numeral 2 denotes a pattern recorder which records patterns and connection between patterns.
  • a pattern holder 3 holds a pattern for processing.
  • a controller 4 controls the pattern holder.
  • An inter-pattern processor 5 processes the relationship between patterns.
  • 6 is a knowledge generalization unit that generalizes language patterns and generalizes knowledge.
  • a pattern converter 1 converts language information into patterns.
  • the language pattern is input to the pattern holder 3 by way of the controller 4.
  • the third pattern holder temporarily holds a pattern for processing.
  • the input pattern and the processed pattern are recorded on two pattern recorders.
  • the connection between patterns is also recorded on the two pattern recorders.
  • the inter-pattern processor 5 performs inter-pattern processing.
  • the pattern input from the pattern holder 3 is processed, and the processed result is output to the pattern holder 3.
  • the controller of 4 controls the operation of the pattern holder of 3, the pattern recorder of 2, and the processor between 5 patterns.
  • the controller 4 calls the processing flow from the pattern recorder 2 according to the situation, and performs processing of the pattern.
  • the pattern being processed is temporarily stored in the 3 pattern holder.
  • the patterns stored in the pattern holder 3 are output to the pattern recorder 2 and the inter-pattern processor 5 to perform recording and pattern processing.
  • the patterns recorded on the two pattern recorders are recalled to the three pattern holders as appropriate. Further, the output of the 5 inter-pattern processor is also output to the 3 pattern holder and is temporarily held.
  • the pattern is processed by the processing flow recorded in advance.
  • the processed patterns are optionally recorded on two pattern recorders to build knowledge.
  • the connection relationship between patterns is set by a human or the like from the outside in the initial phase (learning phase).
  • the process flow is set to operate autonomously when knowledge has accumulated to a certain extent. At this stage, the device autonomously evaluates the linguistic information (novelty, authenticity, logic validity) and builds up the knowledge autonomously.
  • the relationship between the patterns generated by the autonomous knowledge improvement device has a function to generalize. This function can be generalized if it can be widely interpreted rather than limitingly showing the relationship between a certain pattern and a pattern.
  • This function can be generalized if it can be widely interpreted rather than limitingly showing the relationship between a certain pattern and a pattern.
  • FIG. 6 is a diagram showing the configuration of an autonomous knowledge improvement apparatus according to an embodiment of the present invention.
  • reference numeral 1 denotes a pattern converter which converts language information into a pattern.
  • Reference numeral 2 denotes a pattern recorder which records patterns and connection between patterns.
  • a pattern holder 3 holds a pattern for processing.
  • a controller 4 controls the pattern holder.
  • An inter-pattern processor 5 processes the relationship between patterns.
  • the reference numeral 7 denotes a pattern inverse transformer which performs inverse transformation of the language pattern and generates a control output.
  • a pattern converter 1 converts language information into patterns.
  • the language pattern is input to the pattern holder 3 by way of the controller 4.
  • the third pattern holder temporarily holds a pattern for processing.
  • the input pattern and the processed pattern are recorded on two pattern recorders.
  • the connection between patterns is also recorded on the two pattern recorders.
  • the inter-pattern processor 5 performs inter-pattern processing.
  • the pattern input from the pattern holder 3 is processed, and the processed result is output to the pattern holder 3.
  • the controller of 4 controls the operation of the pattern holder of 3, the pattern recorder of 2, and the processor between 5 patterns.
  • the controller 4 calls the processing flow from the pattern recorder 2 according to the situation, and performs processing of the pattern.
  • the pattern being processed is temporarily stored in the 3 pattern holder.
  • the patterns stored in the pattern holder 3 are output to the pattern recorder 2 and the inter-pattern processor 5 to perform recording and pattern processing.
  • the patterns recorded on the two pattern recorders are called to the three pattern holders as appropriate. Further, the output of the 5 inter-pattern processor is also output to the 3 pattern holder and is temporarily held.
  • the pattern is processed by the processing flow recorded in advance.
  • the processed patterns are optionally recorded on two pattern recorders to build knowledge.
  • the connection relationship between patterns is set by a human or the like from the outside in the initial phase (learning phase).
  • the process flow is set to operate autonomously when knowledge has accumulated to a certain extent.
  • the device autonomously evaluates the linguistic information (novelty, authenticity, logic validity) and builds up the knowledge autonomously.
  • the pattern generated by the autonomous knowledge enhancing device is converted into a control output by the pattern reverse converter 7 and control of the machine can be implemented.
  • a pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders;
  • the inter-pattern processor which processes the relationship between patterns, can sequentially learn human thinking as a pattern, and can execute an operation according to the situation as learned.
  • a pattern converter for converting language information, image information and voice information into a pattern
  • a pattern recorder for recording the pattern and the relationship between the patterns
  • a pattern holder for holding the pattern for processing
  • a pattern holder for holding the pattern for processing
  • a controller that controls and an inter-pattern processor that processes the relationship between patterns can sequentially learn human thinking as a pattern, and can execute an operation according to a situation as learned.
  • a pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders;
  • the inter-pattern processor which processes the relationship between patterns, can sequentially learn human thinking as a pattern, and can execute an operation according to the situation as learned. In addition, it is possible to evaluate newly acquired information (newness, authenticity, value, etc.) by the knowledge built in the machine, and build up systematic knowledge inside the machine.
  • a pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders;
  • the inter-pattern processor which processes the relationship between patterns, can sequentially learn human thinking as a pattern, and can execute an operation according to the situation as learned.
  • a pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders;
  • the inter-pattern processor which processes the relationship between patterns, can sequentially learn human thinking as a pattern, and can execute an operation according to the situation as learned.
  • a pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders;
  • the inter-pattern processor which processes the relationship between patterns, can sequentially learn human thinking as a pattern, and can execute an operation according to the situation as learned.
  • to evaluate newly acquired information (newness, authenticity, value, etc.) by the knowledge built in the machine construct systematic knowledge inside the machine, and make the machine implement control based on the knowledge Is possible.
  • pattern converter 2 pattern recorder 3 pattern holder 4 controller 5 inter-pattern processor 6 knowledge generalization device 7 pattern inverse converter

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

This autonomous intelligence enhancement device converts language information into patterns. The autonomous intelligence enhancement device learns human thinking patterns from input language information or the like, and generates similar thinking patterns to a series of thinking patterns of a human for certain situations. For language patterns corresponding to human thinking patterns, connection relationships defining the connections to language patterns corresponding to subsequently recalled thinking patterns are established, and meaningful connection relationships are constructed between recorded language patterns and the language patterns. Knowledge is autonomously expanded by evaluating the value of input information on the basis of previously-recorded knowledge, and recording information determined to have a high value. Although in the initial stages, knowledge, relationships between pieces of knowledge, and problem solving methods, etc. are taught by a human, in subsequent stages, an accurate assessment of situations and the execution of actions are enabled by autonomously determining the value of information to construct knowledge by a machine itself and by constructing knowledge at the same level as or exceeding human knowledge.

Description

自律型知識向上装置Autonomous knowledge improvement device
　この発明は入力された言語情報の意味、新規性、真偽、論理の妥当性等を評価し、自律的に知識を獲得し、問題解決のための知能を向上させる人工知能に関するものである。 The present invention relates to artificial intelligence that evaluates the meaning, novelty, authenticity, logic validity, and the like of input linguistic information, autonomously acquires knowledge, and improves intelligence for solving problems.
　従来の人工知能は、あらかじめプログラムされた手順に従った行動を行う。プログラムで設定された処理以外の実施は困難であり、入力された言語情報により機械自ら処理の改善および高度化を行うことはできない。 Conventional artificial intelligence acts in accordance with pre-programmed procedures. It is difficult to carry out the processing other than the processing set by the program, and it is impossible to improve and enhance the processing of the machine by the input linguistic information.
　従来の自動機械、ロボット等の知能機械はある状況に対する対応は、あらかじめプログラムされた手順に従った行動を行う。プログラムは人間が設計し、機械に搭載された計算機に入力する必要があり、開発に多大な時間を要する等のデメリットがあった。また、入力された言語情報により機械自ら自律的に知識を獲得し処理を改善、高度化させていくことは困難であった。
［従来の技術］（請求項１に対応） Conventional intelligent machines such as automatic machines and robots perform actions according to a pre-programmed procedure in response to a certain situation. The program needs to be designed by a human and input to a computer mounted on a machine, which has a disadvantage of requiring a lot of time for development. Moreover, it was difficult for the machine to autonomously acquire knowledge and improve processing and sophistication based on the input linguistic information.
[Prior Art] (Corresponding to Claim 1)
　機械に動作を行わせる場合、機械に搭載した計算機にあらかじめプログラム言語により作成されたプログラムを組込み、実行することにより実現する。あらかじめ設定された条件が検出されると対応する動作が実行されるようにプログラムを作成する。条件の検出および対応する動作が適切でなければ、計算機にインストールしたプログラムを修正する。条件および対応する動作を人間の思考に対応するパターンとして設定し、パターンからパターンへの変化により動作を実行する機械は従来無い。また入力した情報の意味、真偽、論理の妥当性を評価し、自律的に知識を獲得し、知能を向上させる機械は従来無い。
［従来の技術］（請求項２に対応） When the machine is caused to operate, it is realized by incorporating and executing a program created in advance by a program language in a computer mounted on the machine. A program is created so that the corresponding operation is executed when a preset condition is detected. If the condition detection and the corresponding operation are not appropriate, modify the program installed on the computer. There is conventionally no machine that sets conditions and corresponding actions as patterns corresponding to human thinking, and executes actions by changing from pattern to pattern. Also, there is no conventional machine that evaluates the meaning, authenticity, and validity of the input information, autonomously acquires knowledge, and improves intelligence.
[Prior Art] (Corresponding to Claim 2)
　画像情報、音声情報および人間から得た情報により機械に動作を行わせる場合、機械に搭載した計算機にあらかじめプログラム言語により作成されたプログラムを組込み、実行することにより実現する。あらかじめ設定された条件が検出されると対応する動作が実行されるようにプログラムを作成する。条件の検出および対応する動作が適切でなければ、計算機にインストールしたプログラムを修正する。条件および対応する動作を人間の思考に対応するパターンとして設定し、パターンからパターンへの変化により動作を実行する機械は従来無い。また入力した情報の意味、真偽、論理の妥当性を評価し、自律的に知識を獲得し、知能を向上させる機械は従来無い。
［従来の技術］（請求項３に対応） When making a machine operate by image information, voice information, and information obtained from human beings, it is realized by incorporating and executing a program created in advance in a program language in a computer mounted on the machine. A program is created so that the corresponding operation is executed when a preset condition is detected. If the condition detection and the corresponding operation are not appropriate, modify the program installed on the computer. There is conventionally no machine that sets conditions and corresponding actions as patterns corresponding to human thinking, and executes actions by changing from pattern to pattern. Also, there is no conventional machine that evaluates the meaning, authenticity, and validity of the input information, autonomously acquires knowledge, and improves intelligence.
[Prior Art] (Corresponding to Claim 3)
　機械に動作を行わせる場合、機械に搭載した計算機にあらかじめプログラム言語により作成されたプログラムを組込み、実行することにより実現する。あらかじめ設定された条件が検出されると対応する動作が実行されるようにプログラムを作成する。条件の検出および対応する動作が適切でなければ、計算機にインストールしたプログラムを修正する。条件および対応する動作を人間の思考に対応するパターンとして設定し、パターンからパターンへの変化により動作を実行する機械は従来無い。また入力した情報の意味、真偽、論理の妥当性を評価し、自律的に知識を構築し、不明な点があれば質問を提示し、質問に対し人間等が回答した場合はその回答を元に知識を更新していく機械は従来無い。
［従来の技術］（請求項４に対応） When the machine is caused to operate, it is realized by incorporating and executing a program created in advance by a program language in a computer mounted on the machine. A program is created so that the corresponding operation is executed when a preset condition is detected. If the condition detection and the corresponding operation are not appropriate, modify the program installed on the computer. There is conventionally no machine that sets conditions and corresponding actions as patterns corresponding to human thinking, and executes actions by changing from pattern to pattern. Also, the meaning, authenticity, and validity of the input information are evaluated, the validity of the logic is evaluated, the knowledge is constructed autonomously, the question is presented if there is an unknown point, and the answer is given when a human etc. answers the question. There is no conventional machine that updates knowledge to the original.
[Prior Art] (Corresponding to Claim 4)
　機械に動作を行わせる場合、機械に搭載した計算機にあらかじめプログラム言語により作成されたプログラムを組込み、実行することにより実現する。あらかじめ設定された条件が検出されると対応する動作が実行されるようにプログラムを作成する。条件の検出および対応する動作が適切でなければ、計算機にインストールしたプログラムを修正する。条件および対応する動作を人間の思考に対応するパターンとして設定し、パターンからパターンへの変化により動作を実行する機械は従来無い。また入力した情報の意味、真偽、論理の妥当性を評価し、自律的に知識を構築し、構築した知識に修正が必要な場合は外部より知識の変更を実施する機械は従来無い。
［従来の技術］（請求項５に対応） When the machine is caused to operate, it is realized by incorporating and executing a program created in advance by a program language in a computer mounted on the machine. A program is created so that the corresponding operation is executed when a preset condition is detected. If the condition detection and the corresponding operation are not appropriate, modify the program installed on the computer. There is conventionally no machine that sets conditions and corresponding actions as patterns corresponding to human thinking, and executes actions by changing from pattern to pattern. In addition, there is no conventional machine that evaluates the meaning, authenticity, and logic validity of the input information, autonomously constructs knowledge, and performs modification of knowledge from the outside when correction is necessary for the constructed knowledge.
[Prior art] (corresponding to claim 5)
　機械に動作を行わせる場合、機械に搭載した計算機にあらかじめプログラム言語により作成されたプログラムを組込み、実行することにより実現する。あらかじめ設定された条件が検出されると対応する動作が実行されるようにプログラムを作成する。条件の検出および対応する動作が適切でなければ、計算機にインストールしたプログラムを修正する。条件および対応する動作を人間の思考に対応するパターンとして設定し、パターンからパターンへの変化により動作を実行する機械は従来無い。また獲得した知識、問題解決手法等の一般化を行い、類似の問題を解決する機械は従来無い。
［従来の技術］（請求項６に対応） When the machine is caused to operate, it is realized by incorporating and executing a program created in advance by a program language in a computer mounted on the machine. A program is created so that the corresponding operation is executed when a preset condition is detected. If the condition detection and the corresponding operation are not appropriate, modify the program installed on the computer. There is conventionally no machine that sets conditions and corresponding actions as patterns corresponding to human thinking, and executes actions by changing from pattern to pattern. There is also no machine that solves similar problems by generalizing acquired knowledge and problem solving methods.
[Prior art] (corresponding to claim 6)
　機械に動作を行わせる場合、機械に搭載した計算機にあらかじめプログラム言語により作成されたプログラムを組込み、実行することにより実現する。あらかじめ設定された条件が検出されると対応する動作が実行されるようにプログラムを作成する。条件の検出および対応する動作が適切でなければ、計算機にインストールしたプログラムを修正する。条件および対応する動作を人間の思考に対応するパターンとして設定し、パターンからパターンへの変化により動作を実行する機械は従来無い。また入力した情報の意味、真偽、論理の妥当性を評価し、自律的に知識を獲得し、知能を向上さ、制御出力を生成する機械は従来無い。
［発明が解決しようとする課題］（請求項１に対応） When the machine is caused to operate, it is realized by incorporating and executing a program created in advance by a program language in a computer mounted on the machine. A program is created so that the corresponding operation is executed when a preset condition is detected. If the condition detection and the corresponding operation are not appropriate, modify the program installed on the computer. There is conventionally no machine that sets conditions and corresponding actions as patterns corresponding to human thinking, and executes actions by changing from pattern to pattern. In addition, there is no conventional machine that evaluates the meaning, authenticity, and logic validity of the input information, autonomously acquires knowledge, improves intelligence, and generates control output.
[Problems to be solved by the invention] (corresponds to claim 1)
　従来は機械に動作を行わせる場合、計算機にあらかじめプログラムを設定する必要があった。入力した情報から状況を判断するプログラム、個々の条件に応じ機械に動作させるプログラムを作成し、機械に搭載した計算機にインストールし実行する必要があった。プログラムは専用のプログラム言語により作成する必要があり、開発に多大な時間を要する等のデメリットがあった。条件の検出および対応する動作が適切でなければ、計算機にインストールしたプログラムを人間が修正する必要があり、修正に多大な時間を要する等のデメリットがあった。
また、機械に構築した知識により新規に獲得した情報を評価（新規性、信憑性、価値等）することは困難であり、機械内部に系統立った知識を構築することは従来困難であった。
［発明が解決しようとする課題］（請求項２に対応） Conventionally, when making a machine operate, it has been necessary to set a program in advance in a computer. It was necessary to create a program for judging the situation from the input information and a program to operate the machine according to each condition, and install and execute it on a computer mounted on the machine. The program needs to be created in a dedicated programming language, and has the disadvantage of requiring a lot of time for development. If the condition detection and the corresponding operation are not appropriate, the program installed in the computer needs to be corrected by a human, and there is a disadvantage that the correction takes a lot of time.
Moreover, it is difficult to evaluate (newly, credibility, value, etc.) newly acquired information by the knowledge built in the machine, and it has been conventionally difficult to build up systematic knowledge inside the machine.
[Problems to be Solved by the Invention] (Corresponding to Claim 2)
　従来は機械に動作を行わせる場合、計算機にあらかじめプログラムを設定する必要があった。画像情報、音声情報および言語情報から状況を判断するプログラム、個々の条件に応じ機械に動作させるプログラムを作成し、機械に搭載した計算機にインストールし実行する必要があつた。プログラムは専用のプログラム言語により作成する必要があり、開発に多大な時間を要する等のデメリットがあった。条件の検出および対応する動作が適切でなければ、計算機にインストールしたプログラムを人間が修正する必要があり、修正に多大な時間を要する等のデメリットがあった。
また、機械に構築した知識により新規に獲得した情報を評価（新規性、信憑性、価値等）することは困難であり、機械内部に系統立った知識を構築することは従来困難であった。
［発明が解決しようとする課題］（請求項３に対応） Conventionally, when making a machine operate, it has been necessary to set a program in advance in a computer. It is necessary to create a program to judge the situation from image information, voice information and language information, and a program to operate the machine according to each condition, and install and execute it on a computer mounted on the machine. The program needs to be created in a dedicated programming language, and has the disadvantage of requiring a lot of time for development. If the condition detection and the corresponding operation are not appropriate, the program installed in the computer needs to be corrected by a human, and there is a disadvantage that the correction takes a lot of time.
Moreover, it is difficult to evaluate (newly, credibility, value, etc.) newly acquired information by the knowledge built in the machine, and it has been conventionally difficult to build up systematic knowledge inside the machine.
[Problems to be Solved by the Invention] (Corresponding to Claim 3)
　従来は機械に動作を行わせる場合、計算機にあらかじめプログラムを設定する必要があった。入力した情報から状況を判断するプログラム、個々の条件に応じ機械に動作させるプログラムを作成し、機械に搭載した計算機にインストールし実行する必要があった。プログラムは専用のプログラム言語により作成する必要があり、開発に多大な時間を要する等のデメリットがあった。条件の検出および対応する動作が適切でなければ、計算機にインストールしたプログラムを人間が修正する必要があり、修正に多大な時間を要する等のデメリットがあった。
また、機械に構築した知識により新規に獲得した情報を評価（新規性、信憑性、価値等）することは困難であり、機械内部に系統立った知識を構築することは困難であった。
さらに機械に新規に獲得した情報と機械内部に構築した知識との間にギャップ（不整合、論理の飛躍等）があった場合には疑問を呈し、人間等に対し確認するような動作の実行は従来困難であった。
［発明が解決しようとする課題］（請求項４に対応） Conventionally, when making a machine operate, it has been necessary to set a program in advance in a computer. It was necessary to create a program for judging the situation from the input information and a program to operate the machine according to each condition, and install and execute it on a computer mounted on the machine. The program needs to be created in a dedicated programming language, and has the disadvantage of requiring a lot of time for development. If the condition detection and the corresponding operation are not appropriate, the program installed in the computer needs to be corrected by a human, and there is a disadvantage that the correction takes a lot of time.
In addition, it is difficult to evaluate (newness, credibility, value, etc.) newly acquired information by the knowledge built in the machine, and it is difficult to build up systematic knowledge inside the machine.
In addition, if there is a gap (mismatch, jump in logic, etc.) between the newly acquired information in the machine and the knowledge built in the machine, an action will be made to ask questions and confirm against humans etc. Was difficult in the past.
[Problems to be Solved by the Invention] (Corresponding to Claim 4)
　従来は機械に動作を行わせる場合、計算機にあらかじめプログラムを設定する必要があった。入力した情報から状況を判断するプログラム、個々の条件に応じ機械に動作させるプログラムを作成し、機械に搭載した計算機にインストールし実行する必要があった。プログラムは専用のプログラム言語により作成する必要があり、開発に多大な時間を要する等のデメリットがあった。条件の検出および対応する動作が適切でなければ、計算機にインストールしたプログラムを人間が修正する必要があり、修正に多大な時間を要する等のデメリットがあった。
プログラムの修正等を施すことなく、機械内部に構築された知識を、パターンおよびパターン間の接続を変更することにより、動作を容易に変更することは従来困難であった。
［発明が解決しようとする課題］（請求項５に対応） Conventionally, when making a machine operate, it has been necessary to set a program in advance in a computer. It was necessary to create a program for judging the situation from the input information and a program to operate the machine according to each condition, and install and execute it on a computer mounted on the machine. The program needs to be created in a dedicated programming language, and has the disadvantage of requiring a lot of time for development. If the condition detection and the corresponding operation are not appropriate, the program installed in the computer needs to be corrected by a human, and there is a disadvantage that the correction takes a lot of time.
Conventionally, it has been difficult to easily change the operation by changing the pattern and the connection between the patterns, without modifying the program and changing the knowledge built inside the machine.
[Problems to be Solved by the Invention] (Corresponding to Claim 5)
　従来は機械に動作を行わせる場合、計算機にあらかじめプログラムを設定する必要があった。入力した情報から状況を判断するプログラム、個々の条件に応じ機械に動作させるプログラムを作成し、機械に搭載した計算機にインストールし実行する必要があった。プログラムは専用のプログラム言語により作成する必要があり、開発に多大な時間を要する等のデメリットがあった。条件の検出および対応する動作が適切でなければ、計算機にインストールしたプログラムを人間が修正する必要があり、修正に多大な時間を要する等のデメリットがあった。
問題または課題に対しての回答も、特定の問題または課題に限定されており、少し問題または課題が変更されると忽ち回答ができなくなる等の問題があった。
［発明が解決しようとする課題］（請求項６に対応） Conventionally, when making a machine operate, it has been necessary to set a program in advance in a computer. It was necessary to create a program for judging the situation from the input information and a program to operate the machine according to each condition, and install and execute it on a computer mounted on the machine. The program needs to be created in a dedicated programming language, and has the disadvantage of requiring a lot of time for development. If the condition detection and the corresponding operation are not appropriate, the program installed in the computer needs to be corrected by a human, and there is a disadvantage that the correction takes a lot of time.
The answers to the problems or issues are also limited to the specific problems or issues, and there is a problem that if the problems or issues are changed a little, they can not be answered.
[Problems to be Solved by the Invention] (Corresponding to Claim 6)
　従来は機械に動作を行わせる場合、計算機にあらかじめプログラムを設定する必要があった。入力した情報から状況を判断するプログラム、個々の条件に応じ機械に動作させるプログラムを作成し、機械に搭載した計算機にインストールし実行する必要があった。プログラムは専用のプログラム言語により作成する必要があり、開発に多大な時間を要する等のデメリットがあった。条件の検出および対応する動作が適切でなければ、計算機にインストールしたプログラムを人間が修正する必要があり、修正に多大な時間を要する等のデメリットがあった。
また、機械に構築した知識により新規に獲得した情報を評価（新規性、信憑性、価値等）することは困難であり、機械内部に系統立った知識を構築し、制御出力を生成することは従来困難であった。
［課題を解決するための手段］（請求項１に対応） Conventionally, when making a machine operate, it has been necessary to set a program in advance in a computer. It was necessary to create a program for judging the situation from the input information and a program to operate the machine according to each condition, and install and execute it on a computer mounted on the machine. The program needs to be created in a dedicated programming language, and has the disadvantage of requiring a lot of time for development. If the condition detection and the corresponding operation are not appropriate, the program installed in the computer needs to be corrected by a human, and there is a disadvantage that the correction takes a lot of time.
In addition, it is difficult to evaluate newly acquired information (newness, authenticity, value, etc.) by the knowledge built in the machine, and it is necessary to build up systematic knowledge inside the machine and generate the control output. It was difficult conventionally.
[Means for Solving the Problems] (Corresponding to Claim 1)
　この発明における自律型知識向上装置は言語情報をパターンに変換する。言語情報の語は対応するパターンに変換され、単語は語の組合せで表現されるので、単語のパターンは語のパターンの組合せとして表現できる。さらに、文は単語の組合せで表現できるので、文のパターンは単語のパターンの組合せとして表現できる。
　人間の思考は語、単語、文および文章によって表現されるが、そのパターンはある程度型が決まっていると考えられる。人間は状況に応じ、判断し行動するが、その思考パターン個々の詳細を見てみると、条件付き論理で表現することが可能である。
条件つき論理をプログラム的に表現すると、下記のようになる。
　条件付き論理のプログラム的表現
ＩＦ（Ｂ＝Ｃ１）Ｄ１、ＩＦ（Ｂ＝Ｃ２）Ｄ２
上記の意味としては下記となる。
「ＢがＣ１のとき」「Ｄ１を実行せよ！」、「ＢがＣ２のとき」「Ｄ２を実行せよ！」
少し具体的な例で示すと、
Ｂ＝天気、Ｃ１＝晴れ、Ｃ２＝雨、Ｄ１＝ハイキングに行く、Ｄ２＝映画に行く
と置くと、「天気が晴れならハイキングに行く。」、「天気が雨なら映画に行く。」とうい思考を表現することができる。Ｂ、Ｃ１、Ｃ２、Ｄ１、Ｄ２をパターンで表現し、パターン間の接続関係を定義することにより、条件付き論理で表現された人間の思考パターンをパターン間の遷移として表現することができる。
　この発明における自律型知識向上装置は、人間の思考をパターンとして表現し、思考の遷移をパターンからパターンの変化として表現する。パターンおよびパターン間の接続関係はパターン記録器に記録され、状況に応じて対応するパターンが呼び出され、対応する処理が実行される。人間の典型的な思考パターン、行動パターン等を予め登録しておくことにより、機械も同様の思考パターン、行動パターンを実施することができる。
　思考パターンの変更はパターン記録器に記録したパターンおよびパターン間の接続関係を変更することにより実現できるため、従来技術では必須であったプログラムの変更は不要である。
　この発明における自律型知識向上装置は入力された言語情報から人間の思考をパターンおよびパターン間の接続関係として記録し、知識を構築していく。また入力された言語情報の新規性、既に記録している知識との整合・不整合、論理の妥当性等を評価し、有益な情報であると判断した場合には知識に逐次追加していく。これにより、自律型知識向上装置は有益な知識を蓄積し、知識の向上を実現することができる。
［課題を解決するための手段］（請求項２に対応） The autonomous knowledge improvement apparatus in the present invention converts linguistic information into a pattern. Since the words of linguistic information are converted into corresponding patterns, and the words are expressed as a combination of words, the word pattern can be expressed as a combination of word patterns. Furthermore, since a sentence can be expressed as a combination of words, a sentence pattern can be expressed as a combination of word patterns.
Human thinking is expressed by words, words, sentences, and sentences, but the pattern is considered to be typed to some extent. Humans judge and act according to the situation, but looking at the details of each thought pattern, it is possible to express with conditional logic.
The conditional logic can be expressed programmatically as follows.
Programmatic representation of conditional logic IF (B = C1) D1, IF (B = C2) D2
The above meaning is as follows.
“When B is C1” “Run D1!” “When B is C2” “Run D2!”
As a little more concrete example,
B = Weather, C1 = Sunny, C2 = Rain, D1 = Hiking, D2 = Going to the movie, "If the weather is fine, go hiking.""If the weather is rain, go to the movie." I can express my thoughts. By expressing B, C1, C2, D1, and D2 by patterns and defining the connection relationship between the patterns, it is possible to express human thought patterns expressed by conditional logic as transitions between patterns.
The autonomous knowledge improvement apparatus according to the present invention expresses human thinking as a pattern, and expresses thinking transition as pattern to pattern change. The patterns and the connections between the patterns are recorded on the pattern recorder, the corresponding patterns are called up according to the situation, and the corresponding processing is performed. By previously registering typical human thought patterns, behavior patterns, etc., the machine can implement similar thought patterns and behavior patterns.
Since the change of the thought pattern can be realized by changing the pattern recorded in the pattern recorder and the connection relationship between the patterns, it is not necessary to change the program which is essential in the prior art.
The autonomous knowledge improvement apparatus according to the present invention records human thinking as patterns and connection relations between patterns from the input linguistic information, and constructs knowledge. In addition, the novelty of the input linguistic information, consistency / inconsistency with the already recorded knowledge, the validity of the logic, etc. are evaluated, and if it is judged that it is useful information, it is sequentially added to the knowledge. . As a result, the autonomous knowledge improvement apparatus can accumulate useful knowledge and realize improvement of knowledge.
[Means for Solving the Problems] (Corresponding to Claim 2)
　この発明における自律型知識向上装置は言語情報、画像情報および音声情報をパターンに変換する。言語情報の語は対応するパターンに変換され、単語は語の組合せで表現されるので、単語のパターンは語のパターンの組合せとして表現できる。さらに、文は単語の組合せで表現できるので、文のパターンは単語のパターンの組合せとして表現できる。
画像情報および音声情報もパターンに変換する。変換された画像情報は識別され、対応する対象を示す言語パターンに結びつけられる。変換された音声情報も語、単語または音として識別され。対応する語、単語または音を示す言語パターンと結びつけられる。
　人間の思考は語、単語、文および文章によって表現されるが、そのパターンはある程度型が決まっていると考えられる。人間は状況に応じ、判断し行動するが、その思考パターン個々の詳細を見てみると、条件付き論理で表現することが可能である。
条件つき論理をプログラム的に表現すると、下記のようになる。
　条件付き論理のプログラム的表現
ＩＦ（Ｂ＝Ｃ１）Ｄ１、ＩＦ（Ｂ＝Ｃ２）Ｄ２
上記の意味としては下記となる。
「ＢがＣ１のとき」「Ｄ１を実行せよ！」、「ＢがＣ２のとき」「Ｄ２を実行せよ！」
少し具体的な例で示すと、
Ｂ＝天気、Ｃ１＝晴れ、Ｃ２＝雨、Ｄ１＝ハイキングに行く、Ｄ２＝映画に行く
と置くと、「天気が晴れならハイキングに行く。」、「天気が雨なら映画に行く。」とうい思考を表現することができる。Ｂ、Ｃ１、Ｃ２、Ｄ１、Ｄ２をパターンで表現し、パターン間の接続関係を定義することにより、条件付き論理で表現された人間の思考パターンをパターン間の遷移として表現することができる。
画像情報から晴れを示す画像パターンが検出されると、「晴れ」を示す言語パターンが励起される。画像情報から雨を示す画像パターンが検出されると、「雨」を示す言語パターンが励起される。検出された「晴れ」または「雨」を示す言語パターンは、それぞれのパターンにより「ハイキングに行く」または「映画にいく」を励起することになる。
　この発明における自律型知識向上装置は、人間の思考をパターンとして表現し、思考の遷移をパターンからパターンの変化として表現する。パターンおよびパターン間の接続関係はパターン記録器に記録され、状況に応じて対応するパターンが呼び出され、対応する処理が実行される。人間の典型的な思考パターン、行動パターン等を予め登録しておくことにより、機械も同様の思考パターン、行動パターンを実施することができる。
　思考パターンの変更はパターン記録器に記録したパターンおよびパターン間の接続関係を変更することにより実現できるため、従来技術では必須であったプログラムの変更は不要である。
　この発明における自律型知識向上装置は入力された言語情報から人間の思考をパターンおよびパターン間の接続関係として記録し、知識を構築していく。また入力された言語情報の新規性、既に記録している知識との整合・不整合、論理の妥当性等を評価し、有益な情報であると判断した場合には知識に逐次追加していく。これにより、自律型知識向上装置は有益な知識を蓄積し、知識の向上を実現することができる。
［課題を解決するための手段］（請求項３に対応） The autonomous knowledge improvement apparatus according to the present invention converts language information, image information and speech information into patterns. Since the words of linguistic information are converted into corresponding patterns, and the words are expressed as a combination of words, the word pattern can be expressed as a combination of word patterns. Furthermore, since a sentence can be expressed as a combination of words, a sentence pattern can be expressed as a combination of word patterns.
Image information and audio information are also converted into patterns. The transformed image information is identified and linked to a linguistic pattern indicative of the corresponding object. The converted speech information is also identified as words, words or sounds. It is associated with a linguistic pattern indicating corresponding words, words or sounds.
Human thinking is expressed by words, words, sentences, and sentences, but the pattern is considered to be typed to some extent. Humans judge and act according to the situation, but looking at the details of each thought pattern, it is possible to express with conditional logic.
The conditional logic can be expressed programmatically as follows.
Programmatic representation of conditional logic IF (B = C1) D1, IF (B = C2) D2
The above meaning is as follows.
“When B is C1” “Run D1!” “When B is C2” “Run D2!”
As a little more concrete example,
B = Weather, C1 = Sunny, C2 = Rain, D1 = Hiking, D2 = Going to the movie, "If the weather is fine, go hiking.""If the weather is rain, go to the movie." I can express my thoughts. By expressing B, C1, C2, D1, and D2 by patterns and defining the connection relationship between the patterns, it is possible to express human thought patterns expressed by conditional logic as transitions between patterns.
When an image pattern indicating "clear" is detected from the image information, a language pattern indicating "clear" is excited. When an image pattern indicating rain is detected from the image information, a language pattern indicating "rain" is excited. The detected "clean" or "rain" language patterns will excite "go hiking" or "go to movie" with each pattern.
The autonomous knowledge improvement apparatus according to the present invention expresses human thinking as a pattern, and expresses thinking transition as pattern to pattern change. The patterns and the connections between the patterns are recorded on the pattern recorder, the corresponding patterns are called up according to the situation, and the corresponding processing is performed. By previously registering typical human thought patterns, behavior patterns, etc., the machine can implement similar thought patterns and behavior patterns.
Since the change of the thought pattern can be realized by changing the pattern recorded in the pattern recorder and the connection relationship between the patterns, it is not necessary to change the program which is essential in the prior art.
The autonomous knowledge improvement apparatus according to the present invention records human thinking as patterns and connection relations between patterns from the input linguistic information, and constructs knowledge. In addition, the novelty of the input linguistic information, consistency / inconsistency with the already recorded knowledge, the validity of the logic, etc. are evaluated, and if it is judged that it is useful information, it is sequentially added to the knowledge. . As a result, the autonomous knowledge improvement apparatus can accumulate useful knowledge and realize improvement of knowledge.
[Means for Solving the Problems] (Corresponding to Claim 3)
　この発明における自律型知識向上装置は入力された言語情報から人間の思考をパターンおよびパターン間の接続関係として記録し、知識を構築していく。また入力された言語情報の新規性、既に記録している知識との整合・不整合、論理の妥当性等を評価し、有益な情報であると判断した場合には知識に逐次追加していく。これにより、自律型知識向上装置は有益な知識を蓄積し、知識の向上を実現することができる。
　また、自律型知識向上装置は入力情報と既に記録している知識との整合・不整合、論理の妥当性を評価した際に、不整合または論理の飛躍が検出されると、入力された情報が正しいか否か、正しいならその根拠について質問を提示する。既に記録している知識との整合・不整合の確認は、入力された情報を検索キーとして記録されている関連情報が検索される。検索された関連情報と入力情報の整合・不整合の確認が実施され、不整合が検出された場合には、入力された情報が正しいのか質問を提示する。
　また自律型知識向上装置のパターン記録器には数々の知識（真実、事実、規則、慣習、常識等）が記録されている。入力された情報はこれらの知識と照合され、既に記録されている知識との間で不整合が無いか確認される。入力された情報と既に記録されている情報の間で不整合または関連性が全く無い場合は、論理の妥当性が低いと判断し、入力された情報が正しいのか質問を提示する。仮に入力した情報と記録している知識との間に不整合が検出された時に、その根拠について比較することにする。入力した情報および関連情報の根拠を上記において生成した根拠を構成する言語パターンの集合と照合することにより、どちらが根拠として妥当であるか比較することができる。
根拠を構成する言語パターンと照合した時に、照合するものが検出されなかった場合、その根拠は過去に説明されていないことになる。つまり、根拠を構成する言語パターンの集合と照合した時に、照合が多いほど根拠の妥当性が高く、照合が少ないほど、根拠の妥当性が低いことになる。
　自律型知識向上装置が提示した質問に対し、人間等が回答すると、その回答内容は追加情報として入力され、再度評価される。追加された情報により、論理の飛躍が解消されると、新しく知識が追加、更新される。
［課題を解決するための手段］（請求項４に対応） The autonomous knowledge improvement apparatus according to the present invention records human thinking as patterns and connection relations between patterns from the input linguistic information, and constructs knowledge. In addition, the novelty of the input linguistic information, consistency / inconsistency with the already recorded knowledge, the validity of the logic, etc. are evaluated, and if it is judged that it is useful information, it is sequentially added to the knowledge. . As a result, the autonomous knowledge improvement apparatus can accumulate useful knowledge and realize improvement of knowledge.
In addition, when the autonomous knowledge improvement device evaluates the matching / mismatch between the input information and the already recorded knowledge, and the validity of the logic, if the mismatch or the jump of the logic is detected, the input information Present questions about the basis of whether or not it is correct. In the confirmation of the matching / mismatch with the already recorded knowledge, related information recorded using the input information as a search key is searched. The matching / inconsistency check of the retrieved related information and the input information is carried out, and if an inconsistency is detected, a question is presented as to whether the entered information is correct.
Moreover, many knowledge (a truth, a fact, a rule, a custom, a common sense, etc.) are recorded on the pattern recorder of the autonomous knowledge improvement device. The entered information is checked against these knowledge to see if there is any inconsistency with the knowledge already recorded. If there is no inconsistency or relevance between the input information and the information already recorded, it is judged that the logic is low and a question is presented as to whether the input information is correct. When inconsistencies are detected between the temporarily input information and the recorded knowledge, we will compare the grounds. By checking the basis of the input information and the related information with the set of language patterns constituting the basis generated above, it is possible to compare which is appropriate as the basis.
If no match is found when matching against the linguistic patterns that make up the basis, the basis will not be explained in the past. That is, when collated with the set of linguistic patterns that constitute the rationale, the more the collation, the higher the rationale of the rationale, and the smaller the collation, the lower the validity of the rationale.
When a person or the like answers the question presented by the autonomous knowledge improvement device, the contents of the answer are input as additional information and evaluated again. The added information causes new knowledge to be added and updated when the logic leap is resolved.
[Means for Solving the Problems] (Corresponding to Claim 4)
　この発明における自律型知識向上装置は入力された言語情報から人間の思考をパターンおよびパターン間の接続関係として記録し、知識を構築していく。また入力された言語情報の新規性、既に記録している知識との整合・不整合、論理の妥当性等を評価し、有益な情報であると判断した場合には知識に逐次追加していく。これにより、自律型知識向上装置は有益な知識を蓄積し、知識の向上を実現することができる。
　自律型知識向上装置で生成するパターンは逐次モニターすることが可能である。生成されたパターンが適切でなく、修正が必要な場合には外部より、パターンの修正およびパターン間の接続の修正を行うことにより、適切な状態にパターンを維持・管理することができる。
［課題を解決するための手段］（請求項５に対応） The autonomous knowledge improvement apparatus according to the present invention records human thinking as patterns and connection relations between patterns from the input linguistic information, and constructs knowledge. In addition, the novelty of the input linguistic information, consistency / inconsistency with the already recorded knowledge, the validity of the logic, etc. are evaluated, and if it is judged that it is useful information, it is sequentially added to the knowledge. . As a result, the autonomous knowledge improvement apparatus can accumulate useful knowledge and realize improvement of knowledge.
The patterns generated by the autonomous knowledge enhancing device can be monitored sequentially. If the generated pattern is not appropriate and correction is required, the pattern can be maintained and managed in an appropriate state by performing correction of the pattern and connection between the patterns from outside.
[Means for Solving the Problems] (Corresponding to Claim 5)
　この発明における自律型知識向上装置は入力された言語情報から人間の思考をパターンおよびパターン間の接続関係として記録し、知識を構築していく。また入力された言語情報の新規性、既に記録している知識との整合・不整合、論理の妥当性等を評価し、有益な情報であると判断した場合には知識に逐次追加していく。これにより、自律型知識向上装置は有益な知識を蓄積し、知識の向上を実現することができる。
　自律型知識向上装置で生成するパターン間の関係は一般化する機能を有している。あるパターンとパターンの関係がある事象を限定的に示すのではなく、広く拡大解釈できる場合には本機能により一般化が可能である。
　具体的な例として、下記の文を考える。
例：花子さんは成績が上がったので喜んだ。
　「花子さんは」＋「成績が上がったので」＋「喜んだ。」
この文の場合、具体的には「花子さん」が喜んだのであるが、「花子さん」に限らず、成績が上がると喜ぶと考えられるため、「成績が上がったので」＋「喜んだ。」は一般性を有すると考えられる。このような場合、「花子さん」の代わりに「Ｘさん」と置き換えて、一般化する。ある時に「Ｘさん」は成績が上がったとすると、「Ｘさん」は「喜んだ」と推測することが可能である。このように、具体的なもののみを記録するのでは無く、ある程度一般化したものを記録することにより、知識として「慣習」または「常識」に対応するものを構築することができる。
［課題を解決するための手段］（請求項６に対応） The autonomous knowledge improvement apparatus according to the present invention records human thinking as patterns and connection relations between patterns from the input linguistic information, and constructs knowledge. In addition, the novelty of the input linguistic information, consistency / inconsistency with the already recorded knowledge, the validity of the logic, etc. are evaluated, and if it is judged that it is useful information, it is sequentially added to the knowledge. . As a result, the autonomous knowledge improvement apparatus can accumulate useful knowledge and realize improvement of knowledge.
The relationship between the patterns generated by the autonomous knowledge improvement device has a function to generalize. This function can be generalized if it can be widely interpreted rather than limitingly showing the relationship between a certain pattern and a pattern.
As a concrete example, consider the following statement.
Example: Hanako was pleased with her grades.
"Hanako" + "Because I got a grade" + "I am delighted."
In this case, "Hanako" is particularly happy, but not only "Hanako", but I think he will be happy if his grades go up, so "I got my grades up" + "I'm happy. Is considered to have generality. In such a case, replace with "Mr. X" instead of "Hanako" and generalize. If "Mr. X" gets up at one time, it is possible to guess that "Mr. X" is "willing". As described above, by recording something which is generalized to a certain degree instead of recording only specific things, it is possible to construct a thing corresponding to "custom" or "common sense" as knowledge.
[Means for Solving the Problems] (Corresponding to Claim 6)
　この発明における自律型知識向上装置は入力された言語情報から人間の思考をパターンおよびパターン間の接続関係として記録し、知識を構築していく。また入力された言語情報の新規性、既に記録している知識との整合・不整合、論理の妥当性等を評価し、有益な情報であると判断した場合には知識に逐次追加していく。これにより、自律型知識向上装置は有益な知識を蓄積し、知識の向上を実現することができる。
　自律型知識向上装置で生成したパターンはパターン逆変換により制御出力に変換され、機械の制御を実施する。 The autonomous knowledge improvement apparatus according to the present invention records human thinking as patterns and connection relations between patterns from the input linguistic information, and constructs knowledge. In addition, the novelty of the input linguistic information, consistency / inconsistency with the already recorded knowledge, the validity of the logic, etc. are evaluated, and if it is judged that it is useful information, it is sequentially added to the knowledge. . As a result, the autonomous knowledge improvement apparatus can accumulate useful knowledge and realize improvement of knowledge.
The pattern generated by the autonomous knowledge improvement device is converted into a control output by pattern reverse conversion to implement control of the machine.
［発明の実施の形態１］（請求項１に対応） [First embodiment of the invention] (corresponds to claim 1)
　図１はこの発明の一実施例における自律型知識向上装置の構成を示した図である。
図１において１は言語情報をパターンに変換するパターン変換器である。２はパターンおよびパターン間の接続関係を記録するパターン記録器である。３は処理を行うためにパターンを保持するパターン保持器である。４はパターン保持器を制御する制御器である。５はパターン間の関係を処理するパターン間処理器である。 FIG. 1 is a diagram showing the configuration of an autonomous knowledge improvement apparatus according to an embodiment of the present invention.
In FIG. 1, reference numeral 1 denotes a pattern converter which converts language information into a pattern. Reference numeral 2 denotes a pattern recorder which records patterns and connection between patterns. A pattern holder 3 holds a pattern for processing. A controller 4 controls the pattern holder. An inter-pattern processor 5 processes the relationship between patterns.
　次に動作について説明する。
１のパターン変換器は言語情報をパターンに変換する。
図７にパターン変換器の構成例を示す。言語情報は語、単語が識別され、構文分析により配置が変換される。主語、述語等に整理された後、パターンに変換され、言語パターンが生成される。ここではパターンの要素を「ＯＮ」「ＯＦＦ」または「１」「０」で表現することにする。他の表現でも良いが相関処理の簡便性により本表現を採用している。
言語パターンは４の制御器を経由して３のパターン保持器に入力される。３のパターン保持器では処理のためにパターンが一時的に保持される。入力されたパターンおよび処理されたパターンは２のパターン記録器に記録される。パターン間の接続関係も２のパターン記録器に記録される。５のパターン間処理器はパターン間の処理を行う。３のパターン保持器から入力されたパターンを処理し、処理した結果を３のパターン保持器に出力する。４の制御器は３のパターン保持器、２のパターン記録器および５のパターン間処理器の動作を制御する。４の制御器は状況に応じて２のパターン記録器から処理フローを呼び出し、パターンの処理を行っていく。処理中のパターンは一時的に３のパターン保持器に格納される。３のパターン保持器に格納されたパターンは２のパターン記録器および５のパターン間処理器に出力され、記録およびパターンの処理が行われる。２のパターン記録器に記録されたパターンは適宜、３のパターン保持器に呼び出される。また、５のパターン間処理器の出力も３のパターン保持器に出力され、一時的に保持される。
　パターンは予め記録された処理フローにより、処理されていく。処理されたパターンは適宜２のパターン記録器に記録され、知識が構築される。パターン間の接続関係は初期のフェーズ（学習フェーズ）においては人間等が外部より設定していく。知識が或る程度、蓄積した段階で、処理フローを自律的に動作させるように設定する。この段階では装置が自律的に言語情報の評価（新規性、真偽、論理の妥当性）を実施し、知識を自律的に構築していく。 Next, the operation will be described.
A pattern converter 1 converts language information into patterns.
FIG. 7 shows a configuration example of the pattern converter. Linguistic information identifies words, words, and the arrangement is converted by syntactic analysis. After being organized into subjects, predicates, etc., they are converted into patterns, and language patterns are generated. Here, the elements of the pattern are expressed by “ON”, “OFF” or “1” “0”. Although other expressions may be used, this expression is adopted due to the simplicity of the correlation processing.
The language pattern is input to the pattern holder 3 by way of the controller 4. The third pattern holder temporarily holds a pattern for processing. The input pattern and the processed pattern are recorded on two pattern recorders. The connection between patterns is also recorded on the two pattern recorders. The inter-pattern processor 5 performs inter-pattern processing. The pattern input from the pattern holder 3 is processed, and the processed result is output to the pattern holder 3. The controller of 4 controls the operation of the pattern holder of 3, the pattern recorder of 2, and the processor between 5 patterns. The controller 4 calls the processing flow from the pattern recorder 2 according to the situation, and performs processing of the pattern. The pattern being processed is temporarily stored in the 3 pattern holder. The patterns stored in the pattern holder 3 are output to the pattern recorder 2 and the inter-pattern processor 5 to perform recording and pattern processing. The patterns recorded on the two pattern recorders are recalled to the three pattern holders as appropriate. Further, the output of the 5 inter-pattern processor is also output to the 3 pattern holder and is temporarily held.
The pattern is processed by the processing flow recorded in advance. The processed patterns are optionally recorded on two pattern recorders to build knowledge. The connection relationship between patterns is set by a human or the like from the outside in the initial phase (learning phase). The process flow is set to operate autonomously when knowledge has accumulated to a certain extent. At this stage, the device autonomously evaluates the linguistic information (novelty, authenticity, logic validity) and builds up the knowledge autonomously.
　次に言語パターンと言語パターンの結合関係の動作例について説明する。
図８は言語パターンと言語パターンの結合関係の動作例をパターン記録器の内部に格納されている記録モジュール間の接続関係として表現したものである。
図８においてパターン記録器内の記録モジュールに言語パターンＡおよび言語パターンＢを記録する。言語パターンＡおよび言語パターンＢの設定はパターン保持器からそれぞれ設定する。設定された各パターンは信号ラインを伝搬し、パターン記録器に入力される。パターン記録器内の記録モジュールの内、空きモジュールが選択され、言語パターンＡおよび言語パターンＢが記録される。
　次にパターン間接続関係の設定について説明する。まず、パターン保持器に言語パターンＡを設定しパターンＡを励起させる。次にパターン保持器の別モジュールに言語パターンＢを設定しパターンＢを励起させる。この時、パターン間結合生成に関しイネーブル状態にしておくと、言語パターンＡの記録モジュールから言語パターンＢの記録モジュールに向けてパターン間接続が生成される。
　パターン間接続が生成されると、次に言語パターンＡが励起すると、引き続き言語パターンＢが励起されるように動作する。これにより論理：Ａ⇒Ｂが構成できる。 Next, an operation example of the language pattern and the connection relation of the language pattern will be described.
FIG. 8 shows an operation example of the connection relation between the language pattern and the language pattern as the connection relation between the recording modules stored inside the pattern recorder.
In FIG. 8, the language pattern A and the language pattern B are recorded in the recording module in the pattern recorder. The settings of language pattern A and language pattern B are set from the pattern holder. Each set pattern propagates a signal line and is input to the pattern recorder. Among the recording modules in the pattern recorder, a vacant module is selected, and the language pattern A and the language pattern B are recorded.
Next, setting of the connection relationship between patterns will be described. First, the language pattern A is set in the pattern holder to excite the pattern A. Next, the language pattern B is set in another module of the pattern holder and the pattern B is excited. At this time, when the inter-pattern connection generation is enabled, the inter-pattern connection is generated from the recording module of the language pattern A to the recording module of the language pattern B.
When the inter-pattern connection is generated, when the language pattern A is excited next, the language pattern B continues to be excited. Thus, logic: A⇒B can be configured.
　図９は論理：Ａ⇒Ｂの構成方法につき構成図を使用して示したものである。
図９において言語パターンＡから言語パターンＢに向けて接続関係が生成される様子が分かる。 FIG. 9 shows the method of construction of logic: A⇒B using a construction diagram.
It can be seen that a connection relation is generated from language pattern A to language pattern B in FIG.
　次に条件付処理の動作例について説明する。
図１０は条件付処理の動作例をパターン記録器の内部に格納されている記録モジュール間の接続関係として表現したものである。
図１０において処理Ｐを設定するとパターン記録器の記録モジュールの内、処理Ｐに対応するモジュールが励起する。処理Ｐを実施する上で設定した条件（Ａ）が何か状態記録領域から検索する。条件（Ａ）の状態については予め状態記録領域に記録しておく。条件（Ａ）の検索結果はパターン保持器の検索結果（Ｂ）の領域に格納する。検索結果（Ｂ）と予め設定した条件との照合を行う。もし（Ｂ）が（Ｃ１）なら（Ｄ１）を実施する。もし（Ｂ）が（Ｃ２）なら（Ｄ２）を実施する。照合の結果、決定した処理（Ｄ１またはＤ２）を処理結果格納器に格納する。 Next, an operation example of the conditional process will be described.
FIG. 10 represents an operation example of the conditional processing as a connection relationship between recording modules stored inside the pattern recording device.
When the process P is set in FIG. 10, among the recording modules of the pattern recorder, the module corresponding to the process P is excited. The condition (A) set when carrying out the process P is searched from the state recording area. The state of the condition (A) is recorded in advance in the state recording area. The search result of the condition (A) is stored in the area of the search result (B) of the pattern holder. The search result (B) is compared with the preset conditions. If (B) is (C1), then (D1) is performed. If (B) is (C2), then (D2) is performed. As a result of the collation, the determined process (D1 or D2) is stored in the process result storage.
　図１１は条件付処理：ＩＦ（Ｂ＝Ｃ１）Ｄ１，ＩＦ（Ｂ＝Ｃ２）Ｄ２の構成方法につき構成図を使用して示したものである。
図１１において検索結果（Ｂ）の値（Ｂ＝Ｃ１）により、処理としてＤ１が選択される様子が分かる。 FIG. 11 is a diagram showing a configuration method of the conditional processing: IF (B = C1) D1 and IF (B = C2) D2.
In FIG. 11, the value (B = C1) of the search result (B) indicates how D1 is selected as the process.
　次に質問に対する回答生成の動作例について説明する。
図１２は質問に対して回答を生成する動作について示したものである。質問は検索キーとして表現される。検索キーは対象とする文の構成要素で構成される。図１２の例では主語および述語を指定し、「いつ：Ｗｈｅｎ？」実施されたかを検索する。主語および述語から「Ｓ１」＋「Ｖ１」を検索キーとして関連情報を検索する。パターン記録器に記録されたパターンの内、「Ｓ１」＋「Ｖ１」と照合し、かつ「Ｗｈｅｎ？」に対応するパターンが存在する時、回答案として検索結果の領域に格納する。図１１の例では「Ｗｈｅｎ？」の質問に対して「Ｗｈｅｎ１」が回答として検索結果に格納されることが分かる。 Next, an operation example of generating an answer to a question will be described.
FIG. 12 shows an operation of generating an answer to a question. The questions are expressed as search keys. The search key is composed of target sentence components. In the example of FIG. 12, a subject and a predicate are specified, and "when: When?" Is executed is searched. Related information is searched from the subject and the predicate using "S1" + "V1" as a search key. Among the patterns recorded in the pattern recorder, when a pattern corresponding to “S1” + “V1” is found and a pattern corresponding to “When?” Exists, it is stored as an answer plan in the area of the search result. In the example of FIG. 11, it can be seen that “When 1” is stored as an answer in the search result for the question “When?”.
　図１３は質問に対する回答生成の動作例につき構成図を使用して示したものである。
パターン保持器に設定された質問に対応するパターンがパターン間処理器の内、質問に対する回答生成器に入力される。回答生成器は質問に対応する検索キーを使用して、パターン記録器内の関連パターンを検索する。パターン記録器内の関連パターンが照合し、「Ｗｈｅｎ？」の質問に対して「Ｗｈｅｎ１」が回答として回答格納領域に格納される。 FIG. 13 shows an operation example of generating an answer to a question using a configuration diagram.
The pattern corresponding to the question set in the pattern holder is input to the answer generator for the question in the inter-pattern processor. The answer generator searches for relevant patterns in the pattern recorder using a search key corresponding to the question. The related patterns in the pattern recorder are collated, and "When 1" is stored as an answer in the answer storage area in response to the "When?" Question.
　次に同意語の検索動作例について説明する。
図１４は同意語を検索する動作例について示したものである。
図１４において言語パターンＡ１と言語パターンＡ２は同意語であるとする。言語パターンＣ１は入力された言語パターンが同意語の関係にあることを示すとする。この時、言語パターンＡ１と言語パターンＡ２が励起すると言語パターンＣ１が励起するように言語パターン間の接続関係を設定しておく。このように言語パターン間の接続を定義しておくと、言語パターンＡ１と言語パターンＡ２を入力すると、その関係は同意語であることを回答することができる。次に言語パターンＡ１と言語パターンＣ１が励起すると言語パターンＡ２が励起するように言語パターン間の接続関係を設定する。また言語パターンＡ２と言語パターンＣ１が励起すると言語パターンＡ１が励起するように言語パターン間の接続関係を設定する。このように言語パターン間の接続を定義しておくと、言語パターンＡ１と言語パターンＣ１を入力すると、言語パターンＡ２が励起する。これはＡ１の同意語はＡ２であることを回答することができる。同様にＡ２の同意語はＡ１であることを回答することができる。
　このではパターン間の関係が同意語であることを例に説明したが、パターン間の関係として反対の意味を示すもの、類似の意味を示すもの、関連する意味を示すもの等、数々の関係を定義することが可能である。 Next, a search operation example of synonyms will be described.
FIG. 14 shows an operation example of searching for synonyms.
In FIG. 14, it is assumed that the language pattern A1 and the language pattern A2 are synonyms. It is assumed that the language pattern C1 indicates that the input language pattern has a synonym relation. At this time, the connection relationship between the language patterns is set so that the language pattern C1 is excited when the language pattern A1 and the language pattern A2 are excited. By defining the connection between language patterns in this manner, it is possible to reply that the relationship is a synonym when the language pattern A1 and the language pattern A2 are input. Next, the connection relationship between the language patterns is set so that the language pattern A2 is excited when the language pattern A1 and the language pattern C1 are excited. Further, the connection relationship between the language patterns is set so that the language pattern A1 is excited when the language pattern A2 and the language pattern C1 are excited. When the connection between language patterns is defined in this way, when the language pattern A1 and the language pattern C1 are input, the language pattern A2 is excited. This can be answered that the synonym of A1 is A2. Similarly, it can be answered that the synonym of A2 is A1.
In this example, the relationship between the patterns is described as an example, but the relationship between the patterns indicates the opposite meaning, indicates the similar meaning, indicates the related meaning, etc. It is possible to define.
　図１５は整合・不整合検出の動作例につき構成図を使用して示したものである。
言語パターンＡ１と言語パターンＡ２は同じ意味の関係にあるとする。このことはパターン記録器の記録モジュール間の接続を設定することにより定義することができる。パターン保持器に入力された言語パターンＡ１および言語パターンＡ２はパターン間処理器の整合・不整合検出器に入力される。パターン間処理器に入力された言語パターンはパターン記録器に出力され、パターン記録器に定義された関係が励起される。励起されたパターン間の関係を示す言語パターンは再度、パターン間処理の整合・不整合検出器に入力される。この関係を示す言語パターンはパターン保持器の関係を保持する領域に格納される。このように、パターン間の整合・不整合を回答することができる。 FIG. 15 shows an operation example of the match / mismatch detection using a configuration diagram.
It is assumed that the language pattern A1 and the language pattern A2 have the same meaning relationship. This can be defined by setting up the connections between the recording modules of the pattern recorder. The language pattern A1 and the language pattern A2 input to the pattern holder are input to the match / mismatch detector of the inter-pattern processor. The language pattern input to the inter-pattern processor is output to the pattern recorder, and the relationship defined in the pattern recorder is excited. A language pattern indicating the relationship between the excited patterns is again input to the inter-pattern processing matching / mismatch detector. A language pattern indicating this relationship is stored in the area holding the relationship of the pattern holder. In this way, it is possible to answer the matching / mismatch between the patterns.
　図１６は新規性検出の動作例について示したものである。
言語パターンが入力されると、パターン記録器に記録されている関連する言語パターンが検索される。まず、入力された言語パターンと同じ意味、同等の意味を示す言語パターンが検索され、これらの検索された言語パターンを使用して、入力された言語情報と関連する言語パターンを検索する検索キーが生成される。この検索キーは入力した言語と同じ意味、同等の意味を示す言語を包絡しているため、入力された言語情報と意味レベルで関連する言語パターンが検索されることになる。このように検索された既存情報の言語パターンと入力情報の言語パターンが意味レベルで比較されることになる。
図において既存情報には信号を有するが、入力情報には信号を有しない場合は、既存情報には情報を有しているが、入力情報には情報が無いことを意味する。既存情報に信号を有し、入力情報にも信号を有する場合は、入力情報に示された情報は、既に既存情報にも情報を有していることを意味する。既存情報には信号を有しないが、入力情報には信号を有する場合は、入力情報には既存情報には無い情報が含まれていることを意味する。つまり新規性があることになる。このように既存の関連情報と入力された情報の言語パターンを意味レベルで比較することにより新規性を検出することができる。 FIG. 16 shows an operation example of novelty detection.
When a language pattern is input, the associated language pattern recorded in the pattern recorder is retrieved. First, a language pattern showing the same meaning as the input language pattern is searched, and a search key for searching the language pattern related to the input language information is searched using these searched language patterns. It is generated. Since this search key envelops the language which shows the same meaning as the input language, and an equivalent meaning, a related language pattern will be searched at the meaning level with the input language information. The language pattern of the existing information thus retrieved and the language pattern of the input information are compared at the semantic level.
In the drawing, when the existing information has a signal but the input information does not have a signal, it means that the existing information has information but the input information has no information. If the existing information has a signal and the input information also has a signal, it means that the information indicated in the input information already has information in the existing information. If the existing information has no signal, but the input information has a signal, it means that the input information includes information that is not included in the existing information. In other words, there is novelty. Thus, novelty can be detected by comparing the language patterns of the existing related information with the input information at the semantic level.
　図１７は新規性検出の動作例につき構成図を使用して示したものである。
情報が入力されるとパターン間処理器の新規性検出器に出力される。新規性検出器では入力された言語パターンから同じ意味、同等の意味を有する言語パターンを検索することにより、関連情報を検索するための検索キーが生成され、関連情報の検索が実施される。検索結果はパターン保持器の領域に格納される。さらに入力された言語パターンと同じ、または同等の意味を有する言語パターンと検索された関連情報と同じ、または同等の意味を有する言語パターンの比較が実施される。比較の結果、入力された情報の言語パターンに既存情報に無い言語パターンが有ることが検出されると、新規性が有ることになり、その結果をパターン保持器の領域に格納する。このように入力された情報の新規性を検出することができる。
［発明の実施の形態２］（請求項２に対応） FIG. 17 shows an operation example of novelty detection using a configuration diagram.
When information is input, it is output to the novelty detector of the inter-pattern processor. In the novelty detector, a search key for searching related information is generated by searching for a language pattern having the same meaning and equivalent meaning from the input language pattern, and a search of related information is performed. The search results are stored in the area of the pattern holder. Furthermore, a comparison is performed between a language pattern having the same or equivalent meaning as the input language pattern and a language pattern having the same or equivalent meaning as the retrieved related information. As a result of comparison, when it is detected that there is a language pattern not existing information in the language pattern of the input information, it is considered that there is novelty, and the result is stored in the area of the pattern holder. The novelty of the information thus input can be detected.
[Second Embodiment of the Invention] (Corresponding to Claim 2)
　図２はこの発明の一実施例における自律型知識向上装置の構成を示した図である。
図２において１は言語情報、画像情報および音声情報をパターンに変換するパターン変換器である。２はパターンおよびパターン間の接続関係を記録するパターン記録器である。３は処理を行うためにパターンを保持するパターン保持器である。４はパターン保持器を制御する制御器である。５はパターン間の関係を処理するパターン間処理器である。 FIG. 2 is a diagram showing the configuration of an autonomous knowledge improvement apparatus according to an embodiment of the present invention.
In FIG. 2, reference numeral 1 denotes a pattern converter for converting language information, image information and speech information into a pattern. Reference numeral 2 denotes a pattern recorder which records patterns and connection between patterns. A pattern holder 3 holds a pattern for processing. A controller 4 controls the pattern holder. An inter-pattern processor 5 processes the relationship between patterns.
　次に動作について説明する。
１のパターン変換器は言語情報、画像情報および音声情報をパターンに変換する。
図７にパターン変換器の構成例を示す。言語情報は語、単語が識別され、構文分析により配置が変換される。主語、述語等に整理された後、パターンに変換され、言語パターンが生成される。ここではパターンの要素を「ＯＮ」「ＯＦＦ」または「１」「０」で表現することにする。他の表現でも良いが相関処理の簡便性により本表現を採用している。
画像情報および音声情報もパターンに変換される。変換された画像情報は識別され、対応する対象を示す言語パターンに結びつけられる。変換された音声情報も語、単語または音として識別され、対応する語、単語または音を示す言語パターンと結びつけられる。
言語パターンは４の制御器を経由して３のパターン保持器に入力される。３のパターン保持器では処理のためにパターンが一時的に保持される。入力されたパターンおよび処理されたパターンは２のパターン記録器に記録される。パターン間の接続関係も２のパターン記録器に記録される。５のパターン間処理器はパターン間の処理を行う。３のパターン保持器から入力されたパターンを処理し、処理した結果を３のパターン保持器に出力する。４の制御器は３のパターン保持器、２のパターン記録器および５のパターン間処理器の動作を制御する。４の制御器は状況に応じて２のパターン記録器から処理フローを呼び出し、パターンの処理を行っていく。処理中のパターンは一時的に３のパターン保持器に格納される。３のパターン保持器に格納されたパターンは２のパターン記録器および５のパターン間処理器に出力され、記録およびパターンの処理が行われる。２のパターン記録器に記録されたパターンは適宜、３のパターン保持器に呼び出される。また、５のパターン間処理器の出力も３のパターン保持器に出力され、一時的に保持される。
　パターンは予め記録された処理フローにより、処理されていく。処理されたパターンは適宜２のパターン記録器に記録され、知識が構築される。パターン間の接続関係は初期のフェーズ（学習フェーズ）においては人間等が外部より設定していく。知識が或る程度、蓄積した段階で、処理フローを自律的に動作させるように設定する。この段階では装置が自律的に言語情報の評価（新規性、真偽、論理の妥当性）を実施し、知識を自律的に構築していく。
［発明の実施の形態３］（請求項３に対応） Next, the operation will be described.
The pattern converter 1 converts language information, image information and speech information into patterns.
FIG. 7 shows a configuration example of the pattern converter. Linguistic information identifies words, words, and the arrangement is converted by syntactic analysis. After being organized into subjects, predicates, etc., they are converted into patterns, and language patterns are generated. Here, the elements of the pattern are expressed by “ON”, “OFF” or “1” “0”. Although other expressions may be used, this expression is adopted due to the simplicity of the correlation processing.
Image information and audio information are also converted into patterns. The transformed image information is identified and linked to a linguistic pattern indicative of the corresponding object. The transformed speech information is also identified as a word, word or sound and is associated with a linguistic pattern indicating the corresponding word, word or sound.
The language pattern is input to the pattern holder 3 by way of the controller 4. The third pattern holder temporarily holds a pattern for processing. The input pattern and the processed pattern are recorded on two pattern recorders. The connection between patterns is also recorded on the two pattern recorders. The inter-pattern processor 5 performs inter-pattern processing. The pattern input from the pattern holder 3 is processed, and the processed result is output to the pattern holder 3. The controller of 4 controls the operation of the pattern holder of 3, the pattern recorder of 2, and the processor between 5 patterns. The controller 4 calls the processing flow from the pattern recorder 2 according to the situation, and performs processing of the pattern. The pattern being processed is temporarily stored in the 3 pattern holder. The patterns stored in the pattern holder 3 are output to the pattern recorder 2 and the inter-pattern processor 5 to perform recording and pattern processing. The patterns recorded on the two pattern recorders are called to the three pattern holders as appropriate. Further, the output of the 5 inter-pattern processor is also output to the 3 pattern holder and is temporarily held.
The pattern is processed by the processing flow recorded in advance. The processed patterns are optionally recorded on two pattern recorders to build knowledge. The connection relationship between patterns is set by a human or the like from the outside in the initial phase (learning phase). The process flow is set to operate autonomously when knowledge has accumulated to a certain extent. At this stage, the device autonomously evaluates the linguistic information (novelty, authenticity, logic validity) and builds up the knowledge autonomously.
[Third Embodiment of the Invention] (corresponds to claim 3)
　図３はこの発明の一実施例における自律型知識向上装置の構成を示した図である。
図３において１は言語情報をパターンに変換するパターン変換器である。２はパターンおよびパターン間の接続関係を記録するパターン記録器である。３は処理を行うためにパターンを保持するパターン保持器である。４はパターン保持器を制御する制御器である。５はパターン間の関係を処理するパターン間処理器である。 FIG. 3 is a diagram showing the configuration of an autonomous knowledge improvement apparatus according to an embodiment of the present invention.
In FIG. 3, reference numeral 1 denotes a pattern converter which converts language information into a pattern. Reference numeral 2 denotes a pattern recorder which records patterns and connection between patterns. A pattern holder 3 holds a pattern for processing. A controller 4 controls the pattern holder. An inter-pattern processor 5 processes the relationship between patterns.
　次に動作について説明する。
１のパターン変換器は言語情報をパターンに変換する。
言語パターンは４の制御器を経由して３のパターン保持器に入力される。３のパターン保持器では処理のためにパターンが一時的に保持される。入力されたパターンおよび処理されたパターンは２のパターン記録器に記録される。パターン間の接続関係も２のパターン記録器に記録される。５のパターン間処理器はパターン間の処理を行う。３のパターン保持器から入力されたパターンを処理し、処理した結果を３のパターン保持器に出力する。４の制御器は３のパターン保持器、２のパターン記録器および５のパターン間処理器の動作を制御する。４の制御器は状況に応じて２のパターン記録器から処理フローを呼び出し、パターンの処理を行っていく。処理中のパターンは一時的に３のパターン保持器に格納される。３のパターン保持器に格納されたパターンは２のパターン記録器および５のパターン間処理器に出力され、記録およびパターンの処理が行われる。２のパターン記録器に記録されたパターンは適宜、３のパターン保持器に呼び出される。また、５のパターン間処理器の出力も３のパターン保持器に出力され、一時的に保持される。
　パターンは予め記録された処理フローにより、処理されていく。処理されたパターンは適宜２のパターン記録器に記録され、知識が構築される。パターン間の接続関係は初期のフェーズ（学習フェーズ）においては人間等が外部より設定していく。知識が或る程度、蓄積した段階で、処理フローを自律的に動作させるように設定する。この段階では装置が自律的に言語情報の評価（新規性、真偽、論理の妥当性）を実施し、知識を自律的に構築していく。
また、自律型知識向上装置は入力情報と既に記録している知識との整合・不整合、論理の妥当性を評価した際に、不整合または論理の飛躍が検出されると、入力された情報が正しいか否か、正しいならその根拠について質問を提示する。既に記録している知識との整合・不整合の確認は、入力された情報を検索キーとして記録されている関連情報が検索される。検索された関連情報と入力情報の整合・不整合の確認が実施され、不整合が検出された場合には、入力された情報が正しいのか質問を提示する。
　また自律型知識向上装置のパターン記録器には数々の知識（真実、事実、規則、慣習、常識等）が記録されている。入力された情報はこれらの知識と照合され、既に記録されている知識との間で不整合が無いか確認される。入力された情報と既に記録されている情報の間で不整合または関連性が全く無い場合は、論理の妥当性が低いと判断し、入力された情報が正しいのか質問を提示する。仮に入力した情報と記録している知識との間に不整合が検出された時に、その根拠について比較することにする。入力した情報および関連情報の根拠を上記において生成した根拠を構成する言語パターンの集合と照合することにより、どちらが根拠として妥当であるか比較することができる。
根拠を構成する言語パターンと照合した時に、照合するものが検出されなかった場合、その根拠は過去に説明されていないことになる。つまり、根拠を構成する言語パターンの集合と照合した時に、照合が多いほど根拠の妥当性が高く、照合が少ないほど、根拠の妥当性が低いことになる。
図１８は論理の妥当性評価の例について示したものである。
例ではパターン記録器に知識としてＡ⇒Ｂが記録されているとする。この時、新しい情報としてＡ⇒Ｃが入力されたとする。新しい情報はパターン記録器に記録されている関連情報を検索する。パターン記録器には既に知識としてＡ⇒Ｂという情報が格納されているので、この情報と新規入力されたＡ⇒Ｃという情報が比較される。この時、Ａ⇒Ｃという情報は新規の情報と判断され、入力された情報（Ａ⇒Ｃ）は正しいのか？という質問が提示される。
　自律型知識向上装置が提示した質問に対し、人間等が回答すると、その回答内容は追加情報として入力され、再度評価される。追加された情報により、論理の飛躍が解消されると、新しく知識が追加、更新される。
図１９は提示した質問に対し、人間等が回答した場合の動作例について示したものである。人間系が回答としてＡ⇒Ｃの根拠としてＢ⇒Ｃであることを新しい知識として示す。自律型知識向上装置はこの新しい情報（Ｂ⇒Ｃ）と既に記録している知識（Ａ⇒Ｂ）から（Ａ⇒Ｂ⇒Ｃ）を導き、（Ａ⇒Ｃ）が妥当であることを認識する。このように新規の情報の妥当性が確認されると、新しい知識としてパターン記録器に蓄積される。新規入力された情報は逐次、妥当性の確認を実施し、妥当性が確認された情報は知識として蓄積され、知識の拡大が実現される。
［発明の実施の形態４］（請求項４に対応） Next, the operation will be described.
A pattern converter 1 converts language information into patterns.
The language pattern is input to the pattern holder 3 by way of the controller 4. The third pattern holder temporarily holds a pattern for processing. The input pattern and the processed pattern are recorded on two pattern recorders. The connection between patterns is also recorded on the two pattern recorders. The inter-pattern processor 5 performs inter-pattern processing. The pattern input from the pattern holder 3 is processed, and the processed result is output to the pattern holder 3. The controller of 4 controls the operation of the pattern holder of 3, the pattern recorder of 2, and the processor between 5 patterns. The controller 4 calls the processing flow from the pattern recorder 2 according to the situation, and performs processing of the pattern. The pattern being processed is temporarily stored in the 3 pattern holder. The patterns stored in the pattern holder 3 are output to the pattern recorder 2 and the inter-pattern processor 5 to perform recording and pattern processing. The patterns recorded on the two pattern recorders are recalled to the three pattern holders as appropriate. Further, the output of the 5 inter-pattern processor is also output to the 3 pattern holder and is temporarily held.
The pattern is processed by the processing flow recorded in advance. The processed patterns are optionally recorded on two pattern recorders to build knowledge. The connection relationship between patterns is set by a human or the like from the outside in the initial phase (learning phase). The process flow is set to operate autonomously when knowledge has accumulated to a certain extent. At this stage, the device autonomously evaluates the linguistic information (novelty, authenticity, logic validity) and builds up the knowledge autonomously.
In addition, when the autonomous knowledge improvement device evaluates the matching / mismatch between the input information and the already recorded knowledge, and the validity of the logic, if the mismatch or the jump of the logic is detected, the input information Present questions about the basis of whether or not it is correct. In the confirmation of the matching / mismatch with the already recorded knowledge, related information recorded using the input information as a search key is searched. The matching / inconsistency check of the retrieved related information and the input information is carried out, and if an inconsistency is detected, a question is presented as to whether the entered information is correct.
Moreover, many knowledge (a truth, a fact, a rule, a custom, a common sense, etc.) are recorded on the pattern recorder of the autonomous knowledge improvement device. The entered information is checked against these knowledge to see if there is any inconsistency with the knowledge already recorded. If there is no inconsistency or relevance between the input information and the information already recorded, it is judged that the logic is low and a question is presented as to whether the input information is correct. When inconsistencies are detected between the temporarily input information and the recorded knowledge, we will compare the grounds. By checking the basis of the input information and the related information with the set of language patterns constituting the basis generated above, it is possible to compare which is appropriate as the basis.
If no match is found when matching against the linguistic patterns that make up the basis, the basis will not be explained in the past. That is, when collated with the set of linguistic patterns that constitute the rationale, the more the collation, the higher the rationale of the rationale, and the smaller the collation, the lower the validity of the rationale.
FIG. 18 shows an example of logic validity evaluation.
In the example, it is assumed that A⇒B is recorded as knowledge on the pattern recording device. At this time, it is assumed that A⇒C is input as new information. The new information retrieves relevant information recorded in the pattern recorder. Since the information A 格納 B is already stored as knowledge in the pattern recorder, this information is compared with the newly input information AAC. At this time, the information A⇒C is judged as new information, and is the input information (A⇒C) correct? The question is presented.
When a person or the like answers the question presented by the autonomous knowledge improvement device, the contents of the answer are input as additional information and evaluated again. The added information causes new knowledge to be added and updated when the logic leap is resolved.
FIG. 19 shows an operation example when a human or the like answers the presented question. We show as a new knowledge that the human system is B⇒C as the basis of A⇒C as an answer. The autonomous knowledge improvement device derives (A⇒B⇒C) from the new information (B⇒C) and the knowledge (A⇒B) already recorded, and recognizes that (A⇒C) is appropriate. . When the validity of the new information is confirmed in this way, it is stored in the pattern recorder as new knowledge. The newly input information is sequentially validated, and the validated information is accumulated as knowledge, and the expansion of knowledge is realized.
[Fourth embodiment of the invention] (corresponds to claim 4)
　図４はこの発明の一実施例における自律型知識向上装置の構成を示した図である。
図４において１は言語情報をパターンに変換するパターン変換器である。２はパターンおよびパターン間の接続関係を記録するパターン記録器である。３は処理を行うためにパターンを保持するパターン保持器である。４はパターン保持器を制御する制御器である。５はパターン間の関係を処理するパターン間処理器である。 FIG. 4 is a diagram showing the configuration of an autonomous knowledge improvement apparatus according to an embodiment of the present invention.
In FIG. 4, reference numeral 1 denotes a pattern converter which converts language information into a pattern. Reference numeral 2 denotes a pattern recorder which records patterns and connection between patterns. A pattern holder 3 holds a pattern for processing. A controller 4 controls the pattern holder. An inter-pattern processor 5 processes the relationship between patterns.
　次に動作について説明する。
１のパターン変換器は言語情報をパターンに変換する。
言語パターンは４の制御器を経由して３のパターン保持器に入力される。３のパターン保持器では処理のためにパターンが一時的に保持される。入力されたパターンおよび処理されたパターンは２のパターン記録器に記録される。パターン間の接続関係も２のパターン記録器に記録される。５のパターン間処理器はパターン間の処理を行う。３のパターン保持器から入力されたパターンを処理し、処理した結果を３のパターン保持器に出力する。４の制御器は３のパターン保持器、２のパターン記録器および５のパターン間処理器の動作を制御する。４の制御器は状況に応じて２のパターン記録器から処理フローを呼び出し、パターンの処理を行っていく。処理中のパターンは一時的に３のパターン保持器に格納される。３のパターン保持器に格納されたパターンは２のパターン記録器および５のパターン間処理器に出力され、記録およびパターンの処理が行われる。２のパターン記録器に記録されたパターンは適宜、３のパターン保持器に呼び出される。また、５のパターン間処理器の出力も３のパターン保持器に出力され、一時的に保持される。
　パターンは予め記録された処理フローにより、処理されていく。処理されたパターンは適宜２のパターン記録器に記録され、知識が構築される。パターン間の接続関係は初期のフェーズ（学習フェーズ）においては人間等が外部より設定していく。知識が或る程度、蓄積した段階で、処理フローを自律的に動作させるように設定する。この段階では装置が自律的に言語情報の評価（新規性、真偽、論理の妥当性）を実施し、知識を自律的に構築していく。
　また、自律型知識向上装置で生成するパターンは逐次モニターすることが可能である。生成されたパターンが適切でなく、修正が必要な場合には外部より、パターンの修正およびパターン間の接続の修正を行うことにより、適切な状態にパターンを維持・管理することができる。
［発明の実施の形態５］（請求項５に対応） Next, the operation will be described.
A pattern converter 1 converts language information into patterns.
The language pattern is input to the pattern holder 3 by way of the controller 4. The third pattern holder temporarily holds a pattern for processing. The input pattern and the processed pattern are recorded on two pattern recorders. The connection between patterns is also recorded on the two pattern recorders. The inter-pattern processor 5 performs inter-pattern processing. The pattern input from the pattern holder 3 is processed, and the processed result is output to the pattern holder 3. The controller of 4 controls the operation of the pattern holder of 3, the pattern recorder of 2, and the processor between 5 patterns. The controller 4 calls the processing flow from the pattern recorder 2 according to the situation, and performs processing of the pattern. The pattern being processed is temporarily stored in the 3 pattern holder. The patterns stored in the pattern holder 3 are output to the pattern recorder 2 and the inter-pattern processor 5 to perform recording and pattern processing. The patterns recorded on the two pattern recorders are recalled to the three pattern holders as appropriate. Further, the output of the 5 inter-pattern processor is also output to the 3 pattern holder and is temporarily held.
The pattern is processed by the processing flow recorded in advance. The processed patterns are optionally recorded on two pattern recorders to build knowledge. The connection relationship between patterns is set by a human or the like from the outside in the initial phase (learning phase). The process flow is set to operate autonomously when knowledge has accumulated to a certain extent. At this stage, the device autonomously evaluates the linguistic information (novelty, authenticity, logic validity) and builds up the knowledge autonomously.
Also, patterns generated by the autonomous knowledge improvement device can be monitored one after another. If the generated pattern is not appropriate and correction is required, the pattern can be maintained and managed in an appropriate state by performing correction of the pattern and connection between the patterns from outside.
[Fifth embodiment of the invention] (corresponds to claim 5)
　図５はこの発明の一実施例における自律型知識向上装置の構成を示した図である。
図５において１は言語情報をパターンに変換するパターン変換器である。２はパターンおよびパターン間の接続関係を記録するパターン記録器である。３は処理を行うためにパターンを保持するパターン保持器である。４はパターン保持器を制御する制御器である。５はパターン間の関係を処理するパターン間処理器である。６は言語パターンの汎用化を行い知識の一般化を実施する知識汎用化器である。 FIG. 5 is a diagram showing the configuration of an autonomous knowledge improvement apparatus according to an embodiment of the present invention.
In FIG. 5, reference numeral 1 denotes a pattern converter which converts language information into a pattern. Reference numeral 2 denotes a pattern recorder which records patterns and connection between patterns. A pattern holder 3 holds a pattern for processing. A controller 4 controls the pattern holder. An inter-pattern processor 5 processes the relationship between patterns. 6 is a knowledge generalization unit that generalizes language patterns and generalizes knowledge.
　次に動作について説明する。
１のパターン変換器は言語情報をパターンに変換する。
言語パターンは４の制御器を経由して３のパターン保持器に入力される。３のパターン保持器では処理のためにパターンが一時的に保持される。入力されたパターンおよび処理されたパターンは２のパターン記録器に記録される。パターン間の接続関係も２のパターン記録器に記録される。５のパターン間処理器はパターン間の処理を行う。３のパターン保持器から入力されたパターンを処理し、処理した結果を３のパターン保持器に出力する。４の制御器は３のパターン保持器、２のパターン記録器および５のパターン間処理器の動作を制御する。４の制御器は状況に応じて２のパターン記録器から処理フローを呼び出し、パターンの処理を行っていく。処理中のパターンは一時的に３のパターン保持器に格納される。３のパターン保持器に格納されたパターンは２のパターン記録器および５のパターン間処理器に出力され、記録およびパターンの処理が行われる。２のパターン記録器に記録されたパターンは適宜、３のパターン保持器に呼び出される。また、５のパターン間処理器の出力も３のパターン保持器に出力され、一時的に保持される。
　パターンは予め記録された処理フローにより、処理されていく。処理されたパターンは適宜２のパターン記録器に記録され、知識が構築される。パターン間の接続関係は初期のフェーズ（学習フェーズ）においては人間等が外部より設定していく。知識が或る程度、蓄積した段階で、処理フローを自律的に動作させるように設定する。この段階では装置が自律的に言語情報の評価（新規性、真偽、論理の妥当性）を実施し、知識を自律的に構築していく。
　自律型知識向上装置で生成するパターン間の関係は一般化する機能を有している。あるパターンとパターンの関係がある事象を限定的に示すのではなく、広く拡大解釈できる場合には本機能により一般化が可能である。
　具体的な例として、下記の文を考える。
例：花子さんは成績が上がったので喜んだ。
　「花子さんは」＋「成績が上がったので」＋「喜んだ。」
この文の場合、具体的には「花子さん」が喜んだのであるが、「花子さん」に限らず、成績が上がると喜ぶと考えられるため、「成績が上がったので」＋「喜んだ。」は一般性を有すると考えられる。このような場合、「花子さん」の代わりに「Ｘさん」と置き換えて、一般化する。ある時に「Ｘさん」は成績が上がったとすると、「Ｘさん」は「喜んだ」と推測することが可能である。このように、具体的なもののみを記録するのでは無く、ある程度一般化したものを記録することにより、知識として「慣習」または「常識」に対応するものを構築することができる。
図５の６知識汎用化器は特定の主語、目的語等に限らず、別の名詞等の言語に置き換えても意味が通じる場合には、汎用化を実施し、「慣習」または「常識」に対応する知識として記録する。このように言語パターンの汎用化を行うことにより、全く同一の言語情報でなくても、「慣習」または「常識」に対応する知識が関連情報として検索することができる。このように特定の問題または課題に対してのみに回答できるのではなく、少々変更された問題または課題に対しても回答を生成することができる。
［発明の実施の形態６］（請求項６に対応） Next, the operation will be described.
A pattern converter 1 converts language information into patterns.
The language pattern is input to the pattern holder 3 by way of the controller 4. The third pattern holder temporarily holds a pattern for processing. The input pattern and the processed pattern are recorded on two pattern recorders. The connection between patterns is also recorded on the two pattern recorders. The inter-pattern processor 5 performs inter-pattern processing. The pattern input from the pattern holder 3 is processed, and the processed result is output to the pattern holder 3. The controller of 4 controls the operation of the pattern holder of 3, the pattern recorder of 2, and the processor between 5 patterns. The controller 4 calls the processing flow from the pattern recorder 2 according to the situation, and performs processing of the pattern. The pattern being processed is temporarily stored in the 3 pattern holder. The patterns stored in the pattern holder 3 are output to the pattern recorder 2 and the inter-pattern processor 5 to perform recording and pattern processing. The patterns recorded on the two pattern recorders are recalled to the three pattern holders as appropriate. Further, the output of the 5 inter-pattern processor is also output to the 3 pattern holder and is temporarily held.
The pattern is processed by the processing flow recorded in advance. The processed patterns are optionally recorded on two pattern recorders to build knowledge. The connection relationship between patterns is set by a human or the like from the outside in the initial phase (learning phase). The process flow is set to operate autonomously when knowledge has accumulated to a certain extent. At this stage, the device autonomously evaluates the linguistic information (novelty, authenticity, logic validity) and builds up the knowledge autonomously.
The relationship between the patterns generated by the autonomous knowledge improvement device has a function to generalize. This function can be generalized if it can be widely interpreted rather than limitingly showing the relationship between a certain pattern and a pattern.
As a concrete example, consider the following statement.
Example: Hanako was pleased with her grades.
"Hanako" + "Because I got a grade" + "I am delighted."
In this case, "Hanako" is particularly happy, but not only "Hanako", but I think he will be happy if his grades go up, so "I got my grades up" + "I'm happy. Is considered to have generality. In such a case, replace with "Mr. X" instead of "Hanako" and generalize. If "Mr. X" gets up at one time, it is possible to guess that "Mr. X" is "willing". As described above, by recording something which is generalized to a certain degree instead of recording only specific things, it is possible to construct a thing corresponding to "custom" or "common sense" as knowledge.
The 6-knowledge generalization unit shown in FIG. 5 is not limited to a specific subject, object, etc., but if the meaning is the same even if it is replaced with a language such as another noun, generalization is performed, and "custom" or "common sense" Record as knowledge corresponding to By generalizing the language pattern in this manner, knowledge corresponding to “custom” or “common sense” can be retrieved as related information even if the language information is not exactly the same. Thus, not only responses to specific problems or issues can be answered, but responses can also be generated for slightly changed problems or issues.
[Sixth Embodiment of the Invention] (corresponding to claim 6)
　図６はこの発明の一実施例における自律型知識向上装置の構成を示した図である。
図６において１は言語情報をパターンに変換するパターン変換器である。２はパターンおよびパターン間の接続関係を記録するパターン記録器である。３は処理を行うためにパターンを保持するパターン保持器である。４はパターン保持器を制御する制御器である。５はパターン間の関係を処理するパターン間処理器である。７は言語パターンの逆変換を実施し制御出力を生成するパターン逆変換器である。 FIG. 6 is a diagram showing the configuration of an autonomous knowledge improvement apparatus according to an embodiment of the present invention.
In FIG. 6, reference numeral 1 denotes a pattern converter which converts language information into a pattern. Reference numeral 2 denotes a pattern recorder which records patterns and connection between patterns. A pattern holder 3 holds a pattern for processing. A controller 4 controls the pattern holder. An inter-pattern processor 5 processes the relationship between patterns. The reference numeral 7 denotes a pattern inverse transformer which performs inverse transformation of the language pattern and generates a control output.
　次に動作について説明する。
　次に動作について説明する。
１のパターン変換器は言語情報をパターンに変換する。
言語パターンは４の制御器を経由して３のパターン保持器に入力される。３のパターン保持器では処理のためにパターンが一時的に保持される。入力されたパターンおよび処理されたパターンは２のパターン記録器に記録される。パターン間の接続関係も２のパターン記録器に記録される。５のパターン間処理器はパターン間の処理を行う。３のパターン保持器から入力されたパターンを処理し、処理した結果を３のパターン保持器に出力する。４の制御器は３のパターン保持器、２のパターン記録器および５のパターン間処理器の動作を制御する。４の制御器は状況に応じて２のパターン記録器から処理フローを呼び出し、パターンの処理を行っていく。処理中のパターンは一時的に３のパターン保持器に格納される。３のパターン保持器に格納されたパターンは２のパターン記録器および５のパターン間処理器に出力され、記録およびパターンの処理が行われる。２のパターン記録器に記録されたパターンは適宜、３のパターン保持器に呼び出される。また、５のパターン間処理器の出力も３のパターン保持器に出力され、一時的に保持される。
　パターンは予め記録された処理フローにより、処理されていく。処理されたパターンは適宜２のパターン記録器に記録され、知識が構築される。パターン間の接続関係は初期のフェーズ（学習フェーズ）においては人間等が外部より設定していく。知識が或る程度、蓄積した段階で、処理フローを自律的に動作させるように設定する。この段階では装置が自律的に言語情報の評価（新規性、真偽、論理の妥当性）を実施し、知識を自律的に構築していく。
　自律型知識向上装置で生成されたパターンは７のパターン逆変換器により制御出力に変換され、機械の制御を実施することができる。
［発明の効果１］（請求項１に対応） Next, the operation will be described.
Next, the operation will be described.
A pattern converter 1 converts language information into patterns.
The language pattern is input to the pattern holder 3 by way of the controller 4. The third pattern holder temporarily holds a pattern for processing. The input pattern and the processed pattern are recorded on two pattern recorders. The connection between patterns is also recorded on the two pattern recorders. The inter-pattern processor 5 performs inter-pattern processing. The pattern input from the pattern holder 3 is processed, and the processed result is output to the pattern holder 3. The controller of 4 controls the operation of the pattern holder of 3, the pattern recorder of 2, and the processor between 5 patterns. The controller 4 calls the processing flow from the pattern recorder 2 according to the situation, and performs processing of the pattern. The pattern being processed is temporarily stored in the 3 pattern holder. The patterns stored in the pattern holder 3 are output to the pattern recorder 2 and the inter-pattern processor 5 to perform recording and pattern processing. The patterns recorded on the two pattern recorders are called to the three pattern holders as appropriate. Further, the output of the 5 inter-pattern processor is also output to the 3 pattern holder and is temporarily held.
The pattern is processed by the processing flow recorded in advance. The processed patterns are optionally recorded on two pattern recorders to build knowledge. The connection relationship between patterns is set by a human or the like from the outside in the initial phase (learning phase). The process flow is set to operate autonomously when knowledge has accumulated to a certain extent. At this stage, the device autonomously evaluates the linguistic information (novelty, authenticity, logic validity) and builds up the knowledge autonomously.
The pattern generated by the autonomous knowledge enhancing device is converted into a control output by the pattern reverse converter 7 and control of the machine can be implemented.
[Advantage 1 of the Invention] (Corresponding to Claim 1)
　第１の発明によれば機械に動作を行わせる場合、機械に搭載した計算機に逐次、人間がプログラムを設定する必要がない。言語情報をパターンに変換するパターン変換器と、パターンおよびパターン間の関係を記録するパターン記録器と、処理を行うためにパターンを保持するパターン保持器と、パターン保持器を制御する制御器と、パターン間の関係を処理するパターン間処理器により、人間の思考をパターンとして逐次、学習し、状況に応じた動作が学習した通りに実行することができる。また、機械に構築した知識により新規に獲得した情報を評価（新規性、信憑性、価値等）し、機械内部に系統立った知識を構築することが可能である。
［発明の効果２］（請求項２に対応） According to the first aspect of the invention, when the machine is caused to operate, it is not necessary for a human to set a program one after another on a computer mounted on the machine. A pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders; The inter-pattern processor, which processes the relationship between patterns, can sequentially learn human thinking as a pattern, and can execute an operation according to the situation as learned. In addition, it is possible to evaluate newly acquired information (newness, authenticity, value, etc.) by the knowledge built in the machine, and build up systematic knowledge inside the machine.
[Advantage 2 of the Invention] (Corresponding to Claim 2)
第２の発明によれば機械に動作を行わせる場合、機械に搭載した計算機に逐次、人間がプログラムを設定する必要がない。言語情報、画像情報および音声情報をパターンに変換するパターン変換器と、パターンおよびパターン間の関係を記録するパターン記録器と、処理を行うためにパターンを保持するパターン保持器と、パターン保持器を制御する制御器と、パターン間の関係を処理するパターン間処理器により、人間の思考をパターンとして逐次、学習し、状況に応じた動作が学習した通りに実行することができる。また、機械に構築した知識により新規に獲得した情報を評価（新規性、信憑性、価値等）し、機械内部に系統立った知識を構築することが可能である。
［発明の効果３］（請求項３に対応） According to the second aspect of the invention, when the machine is caused to operate, it is not necessary for a human to set a program one after another on a computer mounted on the machine. A pattern converter for converting language information, image information and voice information into a pattern, a pattern recorder for recording the pattern and the relationship between the patterns, a pattern holder for holding the pattern for processing, and a pattern holder A controller that controls and an inter-pattern processor that processes the relationship between patterns can sequentially learn human thinking as a pattern, and can execute an operation according to a situation as learned. In addition, it is possible to evaluate newly acquired information (newness, authenticity, value, etc.) by the knowledge built in the machine, and build up systematic knowledge inside the machine.
[Advantage 3 of the Invention] (Corresponding to Claim 3)
　第３の発明によれば機械に動作を行わせる場合、機械に搭載した計算機に逐次、人間がプログラムを設定する必要がない。言語情報をパターンに変換するパターン変換器と、パターンおよびパターン間の関係を記録するパターン記録器と、処理を行うためにパターンを保持するパターン保持器と、パターン保持器を制御する制御器と、パターン間の関係を処理するパターン間処理器により、人間の思考をパターンとして逐次、学習し、状況に応じた動作が学習した通りに実行することができる。また、機械に構築した知識により新規に獲得した情報を評価（新規性、信憑性、価値等）し、機械内部に系統立った知識を構築することが可能である。さらに機械に新規に獲得した情報と機械内部に構築した知識との間にギャップ（不整合、論理の飛躍等）があった場合には疑問を呈し、人間等に対し確認するような動作の実行が可能である。
［発明の効果４］（請求項４に対応） According to the third aspect of the invention, when the machine is caused to operate, it is not necessary for a human to set a program one by one on the computer mounted on the machine. A pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders; The inter-pattern processor, which processes the relationship between patterns, can sequentially learn human thinking as a pattern, and can execute an operation according to the situation as learned. In addition, it is possible to evaluate newly acquired information (newness, authenticity, value, etc.) by the knowledge built in the machine, and build up systematic knowledge inside the machine. In addition, if there is a gap (mismatch, jump in logic, etc.) between the newly acquired information in the machine and the knowledge built in the machine, an action will be made to ask questions and confirm against humans etc. Is possible.
[Advantage of the Invention 4] (Corresponding to Claim 4)
　第４の発明によれば機械に動作を行わせる場合、機械に搭載した計算機に逐次、人間がプログラムを設定する必要がない。言語情報をパターンに変換するパターン変換器と、パターンおよびパターン間の関係を記録するパターン記録器と、処理を行うためにパターンを保持するパターン保持器と、パターン保持器を制御する制御器と、パターン間の関係を処理するパターン間処理器により、人間の思考をパターンとして逐次、学習し、状況に応じた動作が学習した通りに実行することができる。また、機械に構築した知識により新規に獲得した情報を評価（新規性、信憑性、価値等）し、機械内部に系統立った知識を構築することが可能である。プログラムの修正等を施すことなく、機械内部に構築された知識を、パターンおよびパターン間の接続を変更することにより、動作を容易に変更することが可能である。
［発明の効果５］（請求項５に対応） According to the fourth aspect of the invention, when the machine is caused to operate, it is not necessary for a human to set a program one by one on the computer mounted on the machine. A pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders; The inter-pattern processor, which processes the relationship between patterns, can sequentially learn human thinking as a pattern, and can execute an operation according to the situation as learned. In addition, it is possible to evaluate newly acquired information (newness, authenticity, value, etc.) by the knowledge built in the machine, and build up systematic knowledge inside the machine. It is possible to easily change the operation of the knowledge built inside the machine by changing the pattern and the connection between the patterns without modifying the program or the like.
[Advantage of the Invention 5] (Corresponding to Claim 5)
　第５の発明によれば機械に動作を行わせる場合、機械に搭載した計算機に逐次、人間がプログラムを設定する必要がない。言語情報をパターンに変換するパターン変換器と、パターンおよびパターン間の関係を記録するパターン記録器と、処理を行うためにパターンを保持するパターン保持器と、パターン保持器を制御する制御器と、パターン間の関係を処理するパターン間処理器により、人間の思考をパターンとして逐次、学習し、状況に応じた動作が学習した通りに実行することができる。また、機械に構築した知識により新規に獲得した情報を評価（新規性、信憑性、価値等）し、機械内部に系統立った知識を構築することが可能である。問題または課題に対しての回答も、汎用化により知識として蓄積しているため、特定の問題または課題に限定されず、問題または課題が少々変更されても回答を生成することが可能である。
［発明の効果６］（請求項６に対応） According to the fifth aspect of the invention, when the machine is caused to operate, it is not necessary for a human to set a program one by one on the computer mounted on the machine. A pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders; The inter-pattern processor, which processes the relationship between patterns, can sequentially learn human thinking as a pattern, and can execute an operation according to the situation as learned. In addition, it is possible to evaluate newly acquired information (newness, authenticity, value, etc.) by the knowledge built in the machine, and build up systematic knowledge inside the machine. Since answers to problems or issues are also accumulated as knowledge by generalization, they are not limited to specific issues or problems, and it is possible to generate answers even if the issues or problems are slightly changed.
[Aspect 6 of the Invention] (Corresponding to Claim 6)
　第６の発明によれば機械に動作を行わせる場合、機械に搭載した計算機に逐次、人間がプログラムを設定する必要がない。言語情報をパターンに変換するパターン変換器と、パターンおよびパターン間の関係を記録するパターン記録器と、処理を行うためにパターンを保持するパターン保持器と、パターン保持器を制御する制御器と、パターン間の関係を処理するパターン間処理器により、人間の思考をパターンとして逐次、学習し、状況に応じた動作が学習した通りに実行することができる。また、機械に構築した知識により新規に獲得した情報を評価（新規性、信憑性、価値等）し、機械内部に系統立った知識を構築し、その知識に基づいた制御を機械に実施させることが可能である。 According to the sixth aspect of the invention, when the machine is caused to operate, it is not necessary for a human to set a program one after another on a computer mounted on the machine. A pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders; The inter-pattern processor, which processes the relationship between patterns, can sequentially learn human thinking as a pattern, and can execute an operation according to the situation as learned. In addition, to evaluate newly acquired information (newness, authenticity, value, etc.) by the knowledge built in the machine, construct systematic knowledge inside the machine, and make the machine implement control based on the knowledge Is possible.
自律型知能向上装置の構成例（請求項１対応）Configuration Example of Autonomous Intelligence Improvement Device (Claim 1) 自律型知能向上装置の構成例（請求項２対応）Configuration Example of Autonomous Intelligence Improvement Device (Claim 2) 自律型知能向上装置の構成例（請求項３対応）Configuration Example of Autonomous Intelligence Improvement Device (Claim 3) 自律型知能向上装置の構成例（請求項４対応）Configuration Example of Autonomous Intelligence Improvement Device (Claim 4) 自律型知能向上装置の構成例（請求項５対応）Configuration Example of Autonomous Intelligence Improvement Device (Claim 5) 自律型知能向上装置の構成例（請求項６対応）Configuration Example of Autonomous Intelligence Improvement Device (Claim 6) パターン変換器の構成例Configuration example of pattern converter 言語パターンと言語パターンの結合関係の動作例（Ａ⇒Ｂ）（１／２）Example of operation of the connection relation between language patterns and language patterns (A B B) (1/2) 言語パターンと言語パターンの結合関係の動作例（Ａ⇒Ｂ）（２／２）Operation example of language pattern and language pattern connection relation (A (B) (2/2) 条件付処理の動作例（ＩＦ（Ｂ＝Ｃ１）Ｄ１，ＩＦ（Ｂ＝Ｃ２）Ｄ２）（１／２）Operation example of conditional processing (IF (B = C1) D1, IF (B = C2) D2) (1/2) 条件付処理の動作例（ＩＦ（Ｂ＝Ｃ１）Ｄ１，ＩＦ（Ｂ＝Ｃ２）Ｄ２）（２／２）Operation example of conditional processing (IF (B = C1) D1, IF (B = C2) D2) (2/2) 質問に対する回答生成（１／２）Answer generation for questions (1/2) 質問に対する回答生成（２／２）Answer generation for questions (2/2) 同意語の検索動作例Synonym search operation example 整合・不整合検出の動作例Example of operation of matching / mismatch detection 新規性検出の動作例（１／２）Operation example of novelty detection (1/2) 新規性検出の動作例（２／２）Operation example of novelty detection (2/2) 論理の妥当性評価の例（１／２）Example of logic validity evaluation (1/2) 論理の妥当性評価の例（２／２）Example of logic validity evaluation (2/2)
１　　パターン変換器
２　　パターン記録器
３　　パターン保持器
４　　制御器
５　　パターン間処理器
６　　知識汎用化器
７　　パターン逆変換器 1 pattern converter 2 pattern recorder 3 pattern holder 4 controller 5 inter-pattern processor 6 knowledge generalization device 7 pattern inverse converter

Claims (6)

  1. 　言語情報をパターンに変換するパターン変換器と、パターンおよびパターン間の関係を記録するパターン記録器と、処理を行うためにパターンを保持するパターン保持器と、パターン保持器を制御する制御器と、パターン間の関係を処理するパターン間処理器を備え、入力した言語情報の意味、新規性、真偽、論理の妥当性等を評価し、自律的に知識を獲得し、知能を向上させる人工知能装置。 A pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders; An artificial intelligence that has an inter-pattern processor that processes relationships between patterns, evaluates the meaning, novelty, authenticity, logic validity, etc. of input linguistic information, acquires knowledge autonomously, and improves intelligence apparatus.
  2. 　言語情報、画像情報および音声情報をパターンに変換するパターン変換器と、パターンおよびパターン間の関係を記録するパターン記録器と、処理を行うためにパターンを保持するパターン保持器と、パターン保持器を制御する制御器と、パターン間の関係を処理するパターン間処理器を備え、入力した言語情報の意味、新規性、真偽、論理の妥当性等を評価し、自律的に知識を獲得し、知能を向上させる人工知能装置。 A pattern converter for converting language information, image information and voice information into a pattern, a pattern recorder for recording the pattern and the relationship between the patterns, a pattern holder for holding the pattern for processing, and a pattern holder A controller for controlling and an inter-pattern processor for processing the relationship between patterns to evaluate the meaning, novelty, authenticity, logic validity, etc. of input language information, and autonomously acquire knowledge, Artificial intelligence device that improves intelligence.
  3. 　言語情報をパターンに変換するパターン変換器と、パターンおよびパターン間の関係を記録するパターン記録器と、処理を行うためにパターンを保持するパターン保持器と、パターン保持器を制御する制御器と、パターン間の関係を処理するパターン間処理器を備え、入力した言語情報の意味、新規性、真偽、論理の妥当性等を評価し、自律的に知識を構築し、不明な点があれば質問を提示し、質問に対し人間等が回答した場合はその回答を元に知識を更新していく人工知能装置。 A pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders; It has an inter-pattern processor that processes the relationship between patterns, evaluates the meaning, novelty, authenticity, logic validity, etc. of the input linguistic information, builds knowledge autonomously, and if there is an unknown point An artificial intelligence device that presents a question and updates knowledge based on the answer when a person etc. answers the question.
  4. 言語情報をパターンに変換するパターン変換器と、パターンおよびパターン間の関係を記録するパターン記録器と、処理を行うためにパターンを保持するパターン保持器と、パターン保持器を制御する制御器と、パターン間の関係を処理するパターン間処理器を備え、入力した言語情報の意味、新規性、真偽、論理の妥当性等を評価し、自律的に知識を構築し、構築した知識に修正が必要な場合は外部より知識の変更が可能な人工知能装置。 A pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders; It is equipped with an inter-pattern processor that processes the relationship between patterns, evaluates the meaning, novelty, authenticity, logic validity, etc. of the input linguistic information, autonomously constructs knowledge, and corrects the constructed knowledge An artificial intelligence device that allows external knowledge to be changed if necessary.
  5. 　言語情報をパターンに変換するパターン変換器と、パターンおよびパターン間の関係を記録するパターン記録器と、処理を行うためにパターンを保持するパターン保持器と、パターン保持器を制御する制御器と、パターン間の関係を処理するパターン間処理器と、知識の汎用化を行う知識汎用化器を備え、入力した言語情報の意味、新規性、真偽、論理の妥当性等を評価し、自律的に知識を記録、更新し、さらに獲得した知識、問題解決手法等の一般化を行い、類似の問題を解決する人工知能装置。 A pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders; It has an inter-pattern processor that processes the relationship between patterns, and a knowledge generalization unit that generalizes knowledge, and evaluates the meaning, novelty, authenticity, logic validity, etc. of input language information, and is autonomous An artificial intelligence device that solves similar problems by recording and updating knowledge, generalizing acquired knowledge, and problem solving methods.
  6. 　言語情報をパターンに変換するパターン変換器と、パターンおよびパターン間の関係を記録するパターン記録器と、処理を行うためにパターンを保持するパターン保持器と、パターン保持器を制御する制御器と、パターン間の関係を処理するパターン間処理器と、パターンを逆変換し制御出力を生成するパターン逆変換器を備え、入力した言語情報の意味、新規性、真偽、論理の妥当性等を評価し、自律的に知識を獲得し、知能を向上させ、その知能に基づき機械の制御を行う人工知能装置。 A pattern converter for converting language information into a pattern, a pattern recorder for recording patterns and relationships between patterns, a pattern holder for holding patterns for processing, and a controller for controlling the pattern holders; It is equipped with an inter-pattern processor that processes the relationship between patterns, and a pattern reverse converter that converts the pattern back to generate a control output, and evaluates the meaning, novelty, authenticity, logic validity, etc. of the input language information An artificial intelligence device that autonomously acquires knowledge, improves intelligence, and controls machines based on that intelligence.
PCT/JP2014/080004 2014-08-15 2014-10-30 Autonomous knowledge enhancement device WO2016024367A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2014178513A JP5737642B2 (en) 2014-08-15 2014-08-15 Autonomous knowledge improvement device
JP2014-178513 2014-08-15

Publications (1)

Publication Number Publication Date
WO2016024367A1 true WO2016024367A1 (en) 2016-02-18

Family

ID=52492424

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2014/080004 WO2016024367A1 (en) 2014-08-15 2014-10-30 Autonomous knowledge enhancement device

Country Status (2)

Country Link
JP (1) JP5737642B2 (en)
WO (1) WO2016024367A1 (en)

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5807829B2 (en) * 2015-02-02 2015-11-10 洋彰 宮崎 Autonomous knowledge analyzer
JP5807830B2 (en) * 2015-02-17 2015-11-10 洋彰 宮崎 Autonomous intelligence generator
JP5821142B2 (en) * 2015-03-03 2015-11-24 洋彰 宮崎 Autonomous intelligent system construction machine
JP5807831B2 (en) * 2015-03-27 2015-11-10 洋彰 宮崎 Autonomous problem solving machine
JP5810476B2 (en) * 2015-04-09 2015-11-11 洋彰 宮崎 Autonomous knowledge extractor
JP5828437B2 (en) * 2015-05-07 2015-12-09 洋彰 宮崎 Autonomous processing procedure generation and execution machine
JP5828438B2 (en) * 2015-05-18 2015-12-09 洋彰 宮崎 Autonomous knowledge system construction machine
JP5866721B2 (en) * 2015-05-25 2016-02-17 洋彰 宮崎 Artificial intelligence device
JP5854250B2 (en) * 2015-06-17 2016-02-09 洋彰 宮崎 Artificial intelligence device
JP5854251B2 (en) * 2015-07-08 2016-02-09 洋彰 宮崎 Artificial intelligence device

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH03233632A (en) * 1990-02-09 1991-10-17 Hitachi Ltd Expert system construction supporting method and document retrieving method
JPH0762824B2 (en) * 1987-09-08 1995-07-05 富士通株式会社 Consultation system
JPH10124086A (en) * 1996-10-25 1998-05-15 Meidensha Corp Voice interactive system
JP2008226168A (en) * 2007-03-15 2008-09-25 Omron Corp Cause-and-result inferring device, control program for it, and control method
JP2009251470A (en) * 2008-04-09 2009-10-29 Toyota Motor Corp In-vehicle information system
JP2011115944A (en) * 2000-03-31 2011-06-16 Sony Corp Robot device, robot device action control method, and program

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003080484A (en) * 2001-09-07 2003-03-18 Tomy Co Ltd Action reaction toy

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0762824B2 (en) * 1987-09-08 1995-07-05 富士通株式会社 Consultation system
JPH03233632A (en) * 1990-02-09 1991-10-17 Hitachi Ltd Expert system construction supporting method and document retrieving method
JPH10124086A (en) * 1996-10-25 1998-05-15 Meidensha Corp Voice interactive system
JP2011115944A (en) * 2000-03-31 2011-06-16 Sony Corp Robot device, robot device action control method, and program
JP2008226168A (en) * 2007-03-15 2008-09-25 Omron Corp Cause-and-result inferring device, control program for it, and control method
JP2009251470A (en) * 2008-04-09 2009-10-29 Toyota Motor Corp In-vehicle information system

Also Published As

Publication number Publication date
JP2015028791A (en) 2015-02-12
JP5737642B2 (en) 2015-06-17

Similar Documents

Publication Publication Date Title
WO2016024367A1 (en) Autonomous knowledge enhancement device
US11210590B2 (en) Generating free text representing semantic relationships between linked entities in a knowledge graph
JP5737641B2 (en) Autonomous thinking pattern generator
US20080310718A1 (en) Information Extraction in a Natural Language Understanding System
US11151117B2 (en) Increasing the accuracy of a statement by analyzing the relationships between entities in a knowledge graph
JP6173972B2 (en) Detection device, correction system, detection method and program
WO2016059733A1 (en) Autonomous learning-type knowledge construction machine
JP2015072700A (en) Automatic logic construction machine
WO2016157563A1 (en) Autonomous problem solver
WO2016163039A1 (en) Autonomous knowledge extraction machine
JP5821142B2 (en) Autonomous intelligent system construction machine
Case et al. Strongly Non-U-Shaped Learning Results by General Techniques.
Becerra-Bonache et al. Linguistic models at the crossroads of agents, learning and formal languages
KR20190042490A (en) Voice control system
JP5854251B2 (en) Artificial intelligence device
JP5828437B2 (en) Autonomous processing procedure generation and execution machine
JP5828438B2 (en) Autonomous knowledge system construction machine
JP5821141B2 (en) Autonomous knowledge system construction machine
Meurs et al. Spoken language interpretation: On the use of dynamic bayesian networks for semantic composition
Srinivasan et al. Mode-Directed Neural-Symbolic Modelling
JP5854250B2 (en) Artificial intelligence device
JP5866721B2 (en) Artificial intelligence device
US20240143935A1 (en) Method and system for facilitating an enhanced search-based interactive system
Sethi et al. Autonlu: Detecting, root-causing, and fixing nlu model errors
Anguel et al. Towards models and metamodels co-evolution approach

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14899747

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14899747

Country of ref document: EP

Kind code of ref document: A1