JP2017208054A - Artificial intelligence device for searching information and procedure by combining concept, and for autonomously processing problem - Google Patents

Abstract

PROBLEM TO BE SOLVED: To guarantee that the runaway of machinery does not occur in a process for improving functions, performance and autonomy in an artificial intelligence technology.SOLUTION: When a processing program is started for processing a certain problem, a key parameter necessary for processing the problem is autonomously searched, and when the search of the key parameter is not successful, a processing program for generating the key parameter is started, and while a condition necessary for processing the problem is autonomously arranged, the processing is advanced. These pieces of processing use a recording unit in which patterns corresponding to words, numerical values, and symbols easily understandable for human beings are structured and recorded, and connection between the recording units is performed by autonomous setting by setting and learning by human beings. The processing content can be monitored by monitoring the excited pattern, and the pattern and an inter-pattern connection relationship are also associated with a concept expressed with words, numbers, and symbols so as to be easily understandable by human beings.SELECTED DRAWING: Figure 45

Description

The present invention sequentially constructs a knowledge system autonomously from inputted information, and uses the constructed knowledge system to analyze the situation and identify problems, and to search for information for solving conversations, translations and problems , Artificial intelligence, robots and their functions that start processing programs, generate information necessary for problem solving and incorporate them into knowledge systems, handle problems autonomously and further improve knowledge It relates to software that realizes
Further, the present invention relates to a “non-black box” artificial intelligence, a robot, and software that realizes the function, which eliminates the black box that has become a problem at the stage of enhancing the function / performance and autonomy of the artificial intelligence.

Conventional artificial intelligence techniques include those using expert systems and neural networks.
The expert system is a program composed of rules, and is composed of a fixed inference engine and a variable knowledge database. The inference engine performs inference using rule groups, and propositional logic, predicate logic, recognition logic, modal logic, temporal logic, fuzzy logic, etc. are used as rule groups as inference logic. A knowledge database generally needs to be constructed by inputting expert knowledge, but when formulated, problems such as contradictions among rules have been pointed out in the literature. In addition, it is difficult to handle processes other than logic.
In a neural network, artificial neurons in the input layer, intermediate layer, and output layer are generally connected serially, and the connection between the artificial neurons is performed by synaptic learning. As the synapse learning, an error back propagation method (back propagation) is generally used. However, back propagation has the following drawbacks pointed out in the literature.
• In backpropagation, the transfer function used in artificial neurons must be differentiable.
-Convergence in learning by backpropagation is very slow.
・ Back-propagation learning does not always converge.
・ In many cases, it converges to a local minimum error point instead of a global optimal solution.
FIG. 30 shows an example of a neural network. Learning in a neural network is generally a batch process, and learning takes time. In addition, there is a high possibility that the behavior changes significantly before and after learning. Furthermore, the neural network has been pointed out as having a black box in the processing process, and it is difficult to confirm the validity of the processing.

Conventional artificial intelligence techniques include those using expert systems and neural networks, both of which have major drawbacks for realizing artificial intelligence.
In the expert system, only logic using rules is processed. It is difficult to express human thought only by logic. In addition, the expert system knowledge database is generally configured by inputting expert knowledge. However, when formulated, problems such as contradictions among rules have been pointed out in the literature.
In a neural network, artificial neurons in the input layer, intermediate layer, and output layer are generally connected serially, and the connection between the artificial neurons is performed by synaptic learning. As the synapse learning, an error back propagation method (back propagation) is generally used. However, back propagation has the following drawbacks pointed out in the literature.
• In backpropagation, the transfer function used in artificial neurons must be differentiable.
-Convergence in learning by backpropagation is very slow.
・ Back-propagation learning does not always converge.
・ In many cases, it converges to a local minimum error point instead of a global optimal solution.
Learning in an equineural network is generally batch processing and takes time for learning. In addition, there is a high possibility that the behavior changes significantly before and after learning. Furthermore, the neural network has been pointed out as having a black box in the processing process, and it is difficult to confirm the validity of the processing.
Realize artificial intelligence that can perform transitions other than logic, such as transitions in human thought patterns. Realize artificial intelligence that autonomously constructs a knowledge system by inputting information (language).
In addition, it solves the “symbol grounding problem” of whether the knowledge system built in artificial intelligence can be processed in conformity with reality, and eliminates the black box that allows humans to understand the process of processing. Realize artificial intelligence.

Since humans think using language, if words, numbers, and symbols that make up a language are patterned, human thought can be expressed by a combination of patterns. The thought process can be expressed by transitions between patterns. Since human thought tends to show the same thinking process under similar circumstances, if the concept is expressed in a pattern and the relationship between the concept and the concept is strengthened according to the degree of time series appearance, A significant connection relationship between concepts can be generated autonomously. In addition to such autonomous generation, the generation of connection relationships between concepts can be realized by a method in which a human designates and sets connection relationships between concepts. There are various relationships between concepts. There are many other reasons besides logical relationships, such as causes and results, details and outlines, opinions opposite to certain opinions, and contents related to certain items. When enumerating sentences (concepts) in a sentence or conversation, a sentence (concept) that appears later is affected by a sentence (concept) that appears earlier. (Of course, when moving to another topic, the influence between concepts may be reduced, but there will be an influence within the same topic.) Therefore, one sentence (concept) in one pattern When a certain sentence (concept) appears, the pattern corresponding to the sentence (concept) is excited, and the excitation history of the pattern excited most recently is recorded and accumulated in the connection relation recording part of the excited pattern. It becomes possible to extract and strengthen relationships between concepts autonomously. As mentioned earlier, there are various relationships between concepts. The relationship between the concepts can also be set by human designation or autonomous generation by learning. Autonomous generation by learning is performed as follows. For example, assume that concept A is the cause (relationship 1) and concept B is the result (relationship 2). When concept A appears and then concept B appears, relation 2 (result) of concept B is excited by human designation. At this time, a history of excitation of the concept B following the concept A is recorded in the connection information recording unit of the pattern corresponding to the relationship 2 (result). In the future, when concept B is excited after concept A, the pattern recorded in the connection information recording part of the excitation history and the result (relationship 2) matches, so the pattern of the result (relationship 2) is excited and learning about the relationship is performed. I understand that. Further, since the excitation history of the concept A and the concept B is recorded in the connection information recording unit of the result (relationship 2), if it is desired to search for the relationship 2 (result) of the concept A using this information, the concept When B is applicable, it can be easily identified. Similarly, it can be easily identified that the relationship 1 (cause) of the concept B is the concept A. Information on the relationship generated here can be used to narrow down the information. As connections between concepts are implemented, it is generally considered that one concept has multiple connections with other concepts. At this time, by setting an appropriate relationship at each stage, it is possible to select appropriate information at that stage from a plurality of related information. In this way, if you record information based on the relationship with the information that has already been recorded, the information does not exist in a single state, but should be recorded as a knowledge system that has a relationship with other information. Can do. In addition, when information is recorded in the knowledge system, it is very useful to record and analyze information (results of novelty, reliability, value, etc.) when searching and analyzing related information in the future. . Furthermore, using the generated knowledge system, it is very easy to generate answers to many questions. In this method, syntax analysis is performed on input information, and sentence elements are identified, structured, and recorded so that they can be easily compared with recorded information. The input information is compared with the recorded information, and the novelty, reliability and value are analyzed, and useful information is recorded. If there is a relationship with the information already recorded, the relationship with the corresponding information is also added and recorded so that it can be searched as related information when searching for information. In this way, this artificial intelligence extracts useful information from the input information and builds a knowledge system in the recorder. The knowledge system will be built in stages so that humans can raise and educate “children”. At the initial stage, basic knowledge is input by language input. The relationship between knowledge is set as much as possible. Information to be input later is compared with the constructed knowledge to extract useful information and add it to the knowledge system. When artificial intelligence cannot judge the usefulness, it is judged by humans. The content specified by human beings is learned as reinforcement of the connection between patterns, thereby increasing autonomy. When a question or the like is generated by a human, the artificial intelligence searches a procedure and information necessary for answering the question from the constructed knowledge system, and generates an answer according to the searched procedure. This artificial intelligence can build a knowledge system sequentially from the input information, and can add and update problem solving methods sequentially. In addition to launching an appropriate program according to the situation and proceeding with processing, it is also possible to perform language-based processing that analyzes the language and proceeds with processing according to the analysis result. It is a system that can be processed by language instructions instead of programming. (Of course, processing by programming is also possible.)
In addition, the processing content can be monitored by monitoring the pattern to be excited, and the connection relationship between patterns and patterns corresponds to the concept expressed by words, numbers, and symbols, making it easy for humans to understand It is. Therefore, it is a system that solves the black box formation, which is a big problem in artificial intelligence using a neural network.
FIG. 29 shows a conceptual diagram of an operation for artificially simulating the transition of a human thought pattern by sequentially shifting a recorded concept using the relationship between concepts. In order to artificially implement such thought pattern transitions, we introduced the concept of recording units and patterns that replace neurons in neural networks. FIG. 31 shows a concept capable of flexibly realizing a thought pattern transition by mutually connecting recording units. The recording unit in FIG. 31 is significantly different in function from the neurons in the neural network. The function of the recording unit is greatly expanded compared to the function of neurons from the viewpoint of recording information, concepts, processing procedures, programs, connection information between information, and the like. FIG. 45 corresponds to the human thought pattern while performing the analysis, evaluation, processing and recording corresponding to the transition of the human thought pattern by configuring the system with the recording unit whose function is greatly expanded from that of the neuron in this way. This is a concept that can artificially simulate intellectual activities.

By expressing each human thought using the concept of pattern and making transition between patterns, human thought process can be simulated and realized in a computer. In order to realize artificial intelligence that simulates human thinking and thinking processes, the function of the recording unit that stores patterns and patterns was set as follows.
-Analyze input words (sentences, concepts), words, values, and symbols, and record them in a recorder. To facilitate associative memory information retrieval. (Since information is structured and recorded, some information can be collated as a search pattern and necessary information can be extracted from a pattern with a large correlation.)
• Includes program functions. (When a pattern representing the process is excited, the corresponding processing program is activated. By executing the pattern representing the process in sequence, a more complicated process is executed.) As a result of the process, newly generated information is appropriate. It is recorded with a simple search key (pattern combination) to facilitate information retrieval from other processes.
• Executes built-in processing by pattern excitation. In addition, the connected patterns are sequentially excited to perform active processing.
-The main excitation methods of the pattern and recording unit are shown.
(1) The pattern for search is collated with each pattern to excite a pattern having a large correlation.
(2) Connection information of the excited recording unit Detects the connection relationship with other recording units using the excitation history from the point of excitation to a certain period before being accumulated in the recording unit, and accumulates information on the detected connection destination and appearance frequency I will do it. By regularly rearranging the connection destinations in the order of appearance frequency, it is possible to preferentially excite recording units having a strong connection relationship. Note that since the connection information with the connection destination includes the relationship, the recording unit of the related concept can be excited using the connection strength and the relationship.
(3) Exciting a pattern having a connection relation. When there are a plurality of connection relationships, only the set relationship pattern can be excited.
・ The connection relationship between patterns is generated by autonomous enhancement of the connection relationship using human settings and excitation history.
-The sequence of the pattern to be excited is changed by changing the connection information recorded in the pattern. (The process can be changed by changing the related connection information locally.)
-When a characteristic pattern is detected, a corresponding process is performed.
(Related information search → Processing using the searched information → Record the result)
FIG. 1 shows an outline of functions of the pattern and the recording unit. The recording unit has a pattern that represents information, the analysis result of the pattern, connection information with other recording units, connection relationship (number of connections, connection destination, relationship, appearance frequency of connection destination) and information processing procedure (program Record function). The pattern is activated by excitation and performs an active operation such as execution of a built-in process and excitation of a recording unit connected to the pattern. (Information search, processing using the searched information, recording of results)
FIG. 2 shows a pattern and recording unit excitation method. When these excitation methods are used, various functions can be realized.
When the excitation method 1 is used, information can be extracted using a search pattern. Since the information is recorded in the recording unit in the subject, subject modification, predicate, predicate modification (5W1H), numerical values, and symbols, the recording unit stores necessary information when the search pattern is appropriately set Information can be extracted from the recording unit. Furthermore, conditional processing can be realized by executing conditional branching and processing based on the extracted information. (This can be implemented by extracting the information stored in the searched portion of the information structured and recorded in the recording unit. FIG. 32 shows an operation example of extracting information using the excitation method 1. Yes.)
Using excitation method 2, a sequence of information can be recorded. When the concept (information) is sequentially input, the novelty is confirmed, the new concept is newly registered in the recording unit, and the already recorded concept excites the corresponding recording unit. By recording the most recent excitation history in the connection information recording unit of the excited recording unit, the connection relationship between the excited recording unit and the most recently excited recording unit is strengthened. By inputting a sequence of information and strengthening the connection between the excited information and the most recently excited information, the information is sequentially excited through the enhanced connection relationship. That is, a series of information sequences can be learned and recorded autonomously. You can also record contexts using this function. Usually, concepts (information) appear based on the context, so if you enter many sentences, the common context will be strengthened.
When the excitation method 3 is used, information retrieval and information extraction using the relationship between information is possible. In general, a concept (information) is considered to have a relationship with a plurality of other concepts (information). Therefore, when all the concepts related to the input concept are excited, the number of concepts to be excited increases exponentially. By using the relationship, it is possible to narrow down the necessary information from a lot of related information, so that the problem of exponential increase can be avoided. By using the functions of the pattern and the recording unit, it is possible to realize an active operation (information retrieval, conditional processing, a series of information processing) in human thinking. FIG. 33 shows an operation example of extracting information using the excitation method 3.

The smallest unit of the pattern corresponds to words, numbers and symbols. As the words occur in time series, the pattern corresponding to the words is excited. Numerical values are detected by generating patterns corresponding to numbers and decimal points in time series. The symbols and combinations of symbols are appropriately identified by corresponding operators and recorded in an area for recording the identification results. A word is identified as the word is generated in time series, and sentence elements are identified using information on the part of speech corresponding to the word. The main part identifies the subject and the part that modifies the subject, and the predicate identifies the predicate and the part that modifies the predicate. The parts that modify the predicate are identified (when, how, what, how, why). The time series of numbers is identified as a numerical value and recorded in an area for storing numerical data.
FIG. 3 shows a state in which word patterns are excited in time series to form word strings. The sentence entered in this example is "I am a teacher" (I am a teacher). There are no restrictions on the format of words, numbers, mathematical expressions, and conversion from symbols to patterns as long as they can be identified. A combination of patterns is also a pattern. In this artificial intelligence prototype, each word was assigned an identification number. (For example, the word identification number of the word “I” is “0”, and the particle identification number of the particle “ha” is “1”, “teacher” “2”, “is” “12”,... “Artificial intelligence” (“169”, “occupation”, “170”, etc.) When a new word is generated, it can be dealt with by sequentially adding an identification number. The word excitation history is irradiated to the connection information recording unit of each word and collation is performed. When the word string corresponding to the word and the number of matched words are detected, the corresponding word is excited. When a word is detected, a matching pattern is generated based on the excitation history of the word string excluding the word string of the detected word, and the next word is detected using this matching pattern. If the next word can be detected by this matching pattern, it can be determined that the identification of the previous word was appropriate. On the other hand, if the next word cannot be detected by this collation pattern and an inconsistency occurs, it is determined that the collation is incorrect, and consistency is confirmed for another word candidate. Each time a word is detected, this collation procedure is repeated, and it is confirmed that the word cutting is appropriate for the entire word string.
FIG. 4 shows an example of sentence structure analysis from word strings. A word, a word identification number, a part of speech, a part of speech identification number, a sentence element, and a sentence structure are analyzed from the word string, and by using the analysis result, it is possible to record the sentence elements in an easily searchable structure.
FIG. 5 shows a state where input information is organized and recorded in a structure that is easy to search. m1 to m4 are subject modification parts, s1 and s2 are subject, o1 and o2 are object, h11 and h2 are where, h21 and h22 are where, h31 and h32 are why, h41 and h42 are how, v1 and v2 are The word identification number and numerical value of the sentence element corresponding to the predicate are stored. Words, symbols, numbers, and expressions in the input information are converted into patterns (identification numbers), and the subject, subject modification, predicate, predicate modification (what, when, where, why, how) can be easily searched It is organized and recorded.
If the sentence inputted in this way is structured and recorded, the sentence can be compared for each sentence element, which is very useful. For example, when searching for the same sentence, a pattern in which all sentence elements match is searched. If it is sufficient to match a part, a pattern in which matching points are combined is used as a search pattern, and the corresponding pattern is searched. By structuring sentence elements in this way, information can be recorded and retrieved in an associative memory.
Also, answers to questions can be easily generated. Information about the part corresponding to the question is missing. ] Can be combined with the remaining patterns to generate a search pattern corresponding to the question. What is the expected sentence element as the answer to the question (subject, subject modifier, predicate, predicate modifier) is set in the search pattern. this is[? ] Can be identified from the position of the corresponding sentence element. For example, if you want to search for [When], [? ] H1. Similarly [Where] ⇒ [? ] H2, [Why] ⇒ [? ] H3, [How] ⇒ [? ] H4, [Who] ⇒ [? ] S, [What] ⇒ [? ] O, [What's wrong] ⇒ [? ] V. A search pattern generated from the question is irradiated to the recorder, and a pattern having a correlation is searched. In the searched pane, the above [? ], If there is an internal pattern at a position corresponding to [], it becomes a candidate for a reply.
FIG. 6 compares the pattern of the input information and the information in the recording area for each sentence element, and shows which part of the sentence element matches. (Comparison of sentence elements that are not both input information and recorded information is excluded.) Information or parts that completely match by comparing the pattern of input information and recorded information for each sentence element. Thus, it is possible to identify whether or not the information that coincides with each other is recorded, and to evaluate the novelty of the input information and the degree of association with the related information. Further, by searching for patterns corresponding to the sentence elements of the input information using those having different relationships, various information can be extracted from the recorded information. For example, a sentence having a similar meaning can be searched by replacing a sentence element having input information with a pattern of a similar word. In addition, it is possible to confirm whether or not there is information having a meaning opposite to that of the input sentence by searching for a word pattern having a meaning opposite to that of the sentence element having the input information. When this is applied, it is possible to determine whether the input information is in accordance with rules, common sense, and morals. When knowledge is recorded in the knowledge system, the information regarding the rules, common sense, and morals is recorded together with the identification results (rules, common sense, and morals) as additional information. If the entered information and the recorded information are collated and an inconsistency (opposite word detection) is detected, and the type of recorded information is (rule, common sense and moral), the entered information is (rule, common sense and moral) ). By utilizing this, unhealthy information can be identified from the input information and managed so as not to be incorporated into the knowledge system. 36 to 44 show an example of an operation for comparing and evaluating information inputted later and knowledge (rules and morality) recorded in advance by inputting and registering knowledge about rules and morality in advance. It is shown.

18 and 19 are a diagram showing a configuration diagram and a functional configuration of the artificial intelligence apparatus in one embodiment of the present invention.
The recording unit used in this artificial intelligence has the following characteristics.
・ Concepts are recorded as words, numbers, and symbols in a structured manner, and associative memory information can be recorded and searched using a search pattern that combines patterns.
-It has a processing function (function function) and can realize the same function as a program. The processing can be unitized, and the combination and change of the processing is a data driven system that can be easily realized by changing the connection. (Recombination and change of processing is easy.)
A characteristic pattern can be detected, and a problem can be solved by starting a processing program corresponding to the detected pattern.
-The excitation history of the pattern can be recorded and the excitation of the pattern can be reproduced using the excitation history.
The pattern, excitation history and relationships can be used to excite the recording unit autonomously and perform processing.
In FIG. 19, reference numeral 1 denotes an input processor that converts information (word, language, sentence, mathematical formula, symbol, voice) into a pattern. 2 is a word, meaning, sentence element (subject, predicate, subject modification, predicate modification, modification relationship (5W1H)), field, subject of interest, novelty, comparison with recorded information, sentence type, regarding input information It is an analyzer that analyzes the relationship between intentions and sentences. 3 is the information identified for each sentence element so that it can be easily searched by the analyzer, the analysis result generated by the analyzer, the connection relationship (connection destination, connection strength) between the information generated by the setting and learning by the human, connection Relationship between information, information and processing program start-up history and processing program are structurally recorded, collation by pattern combining sentence elements and analysis results, connection relationship with related information using excitation history and related information Is a recording unit composed of a recording unit that activates related information using activation and connection information. 4 is the execution of processing according to the type and intention of the sentence identified by the analyzer, information retrieval for solving the problem identified by the analyzer, logical expansion, execution of the process, starting of the processing program and recording of the processing result, Generalization of information and procedures to be performed using the relationship between information recorded in the recorder, superior knowledge, processing and logic updates, interests implemented by changing the connection relation of information recorded in the recorder It is a controller that performs search of information related to a certain field and object, recording in a recorder, and transition control between information for a target. An output processor 5 converts the pattern into information and control signals. The operation of each function shown below will be described below.

(A) Input information evaluation function (b) Relation analysis and setting function between sentences (c) Function to execute processing according to the type of sentence (d) Function to respond to command sentence / request sentence (e) Question (F) A function that responds to conditional statements (g) A function that extracts and records valuable information (h) A function that generates conversations (i) A function that autonomously solves problems ( j) Function to generalize and generalize information and procedures (k) Better knowledge and function to update to logic (l) Function to search and organize information of interest (m) Function to autonomously generate connection relations between patterns (N) Function for autonomously recording the context The function for realizing artificial intelligence uses the function of the pattern and the recording unit, and the search and extraction of information, the execution of processing according to the result of the search, the characteristic pattern Processing process according to detection and detection result These functions are realized by embodying processes such as gram activation, registration and change of information to recording units, creation and update of connections between recording units, and activation and connection creation of related recording units as algorithms. can do.

(A) Input information evaluation function The following evaluation is performed on the input information.
・ Information reliability evaluation ・ Information field / theme evaluation ・ Information type evaluation ・ Information interest level evaluation ・ Novelty evaluation ・ Relevance evaluation These evaluations are analysis of information and incidental information (source) Can be implemented. Reliability assessment can be performed by confirming the reliability of the information source (when, who, where information from?). The field / theme, type and interest level of information can be implemented by identifying patterns corresponding to words included in the information. Novelty and validity evaluation can be performed by collating the pattern of the input information with the pattern recorded in the knowledge system, and confirming the match, mismatch, and consistency for each sentence element.

(B) Analysis of relations between sentences and relations between functional information of settings include causes and results, events and reasons, explanations and conclusions, outlines and details, similar meanings, opposite opinions, texts and summaries, texts and detailed information There are various things. Setting of these relationships is performed by an autonomous setting by learning a human instruction or a relationship between information and information. In FIG. 20, the relationship is set according to the following procedure. In the learning phase, information (sentences) is sequentially input. A human identifies the relationship between sentences and excites the recording unit corresponding to the relationship between sentences. (Process A) At this time, the excitation history of the excited sentence is recorded in the connection information recording unit of the recording unit corresponding to the relation, and the connection with the relation between sentences is strengthened. (Process B) As learning progresses, the recording unit corresponding to the relationship is automatically excited according to the relationship between the sentences. If the relationships between a large number of sentences can be automatically identified, the identification result is recorded in the location where the relationships between the units are recorded. Thereby, identification and setting of the relationship between sentences are learned, and automatic setting can be performed. (Processing of C) It should be understood that the setting of the relationship between sentences can be easily performed by human designation.

(C) Function for executing processing according to the type of sentence The type of sentence (normal sentence, question sentence, command sentence, etc.) is analyzed by detecting characteristic words included in the sentence. Processing is performed according to the type of sentence analyzed. Normal texts extract and record valuable information. In the question sentence, an answer to the question is generated. In the command sentence, an analysis of the command, request contents, and a response analysis (response availability, impact analysis in response to the response) are performed, and a process (presence / absence of response) is determined. In the conditional statement, the establishment of the condition is analyzed, and processing is executed according to the analysis result.

(D) Function corresponding to the command sentence / request sentence The input sentence is analyzed, and the corresponding processing is performed according to the analysis result. The command sentence generally takes the expressions “XX SEYO” and “when ▽▽ XX SEYO”. Regarding the implementation of ○○, analyze the possibility of implementation from the past records (determined from the presence or absence of implementation experience) and the impact when implemented, and determine the implementation. If it is an instruction that can be executed, the validity of the condition (when ▽▽) is confirmed as necessary, and the process (XX) is executed.

(E) Function for generating an answer to a question FIG. 21 shows a function corresponding to a question sentence. Information about the part corresponding to the question is missing. ] Can be combined with the remaining patterns to generate a search pattern corresponding to the question. What is the expected sentence element as the answer to the question (subject, subject modifier, predicate, predicate modifier) is set in the search pattern. this is[? ] Can be identified from the position of the corresponding sentence element. For example, if you want to search for [When], [? ] H1. Similarly [Where] ⇒ [? ] H2, [Why] ⇒ [? ] H3, [How] ⇒ [? ] H4, [Who] ⇒ [? ] S, [What] ⇒ [? ] O, [What's wrong] ⇒ [? ] V. A search pattern generated from the question is irradiated to the recorder, and a pattern having a correlation is searched. In the searched pane, the above [? ], If there is an internal pattern at a position corresponding to [], it becomes a candidate for a reply. Since information is organized into sentence elements, it is possible to compare each sentence element. In addition, various comparisons such as complete match, partial match, similarity, and detection of opposite meanings are possible. Relationships such as the same meaning, similarities, opposite meanings, and relations can be defined between words recorded in the recording unit. Various relations with the recorded information can be extracted from the words included in the input information by performing collation and comparison using these related words.

(F) Function corresponding to conditional sentence FIG. 22 shows a function corresponding to a conditional sentence. By analyzing the information word, the condition part and the processing part of the information can be identified. For example, in the statements “A”, “GA”, “B”, “NO”, “C”, “do it”, [“A”, “GA”, “B”, “NO” ”is the condition part, and [“ C """""" Perform "" is a processing unit. In the sentence “A” “ga” “B” “if” “C” “ha” “D” “is” [“A” “ga” “B” “if”] is the condition part, “C”, “ha”, “D”, and “is” are processing units.
In this way, the sentence is divided into a condition part and a processing part, and when the pattern of the condition part is activated, a process for confirming whether or not the condition part is established autonomously is activated. It is possible to realize an active thinking action in which sentences that can be expressed by conditional processing sequentially proceed while confirming the satisfaction of the condition. Confirmation of the validity of the condition can be easily performed by using a function for generating an answer to the question. It is possible to realize an active thinking action in which sentences that can be expressed by conditional processing sequentially proceed while confirming the satisfaction of the condition. In general, human problem solving and action determination can be expressed by conditional processing. By inputting human problem-solving measures and action-determining measures in a natural language, the artificial intelligence device can perform an active thinking operation in which processing proceeds autonomously while confirming the establishment of conditions. By inputting knowledge (procedures and thinking methods related to problem solving and action determination) expressed in natural language without programming the actions corresponding to human problem solving and action determination, human beings can solve problems or make action decisions through thinking. As a result, problem solving or action determination can be carried out autonomously.

(G) Function to extract and record valuable information When the reliability, novelty, validity, interest level of input information is evaluated and judged useful, information type, field, theme, interest level, etc. An identification result is added and recorded in a recorder so that it can be easily searched. The evaluation of information can be performed by the analysis (a) described above.

(H) Function for generating a conversation Analyzes the type, interest and intention of a speaker's input sentence. This is done by detecting a characteristic pattern (word combination) included in the input sentence, and the corresponding processing program (greeting, collecting information on the other party, evaluating the other party's conversation, retrieving information for the conversation, Generates answers to questions, starts a program that solves the problem, and asks and confirms the other party to generate a conversation (output). It is possible to carry out daily conversation conducted by humans by performing associative information retrieval and information retrieval using relationships between information using words appearing in conversation in processing programs as search keys. It is. Below is an example of a conversation between humans and artificial intelligence. In the following conversation, artificial intelligence searches related information in associative memory and generates a conversation. Details of the conversation generation processing are shown in processing 1 to processing 5.
NO (conversation number) Conversation between human and artificial intelligence 1 Human: Nice to meet you 2 Artificial intelligence: I am an atom (Process 1)
3 I am artificial intelligence 4 Human: I am Taro 5 Artificial intelligence: My profession is assistant (Process 2)
6 What is Taro's occupation 7 Human: My occupation is a teacher 8 Artificial intelligence: Teachers are difficult (Process 3)
9 Human: I teach physics in class 10 Human: I teach tennis in after-school activities 11 Artificial Intelligence: What is Taro's hobby (Process 4)
12 Human: My hobby is sports 13 Artificial intelligence: Speaking of sports (Process 5)
14 Kay won the tennis tournament yesterday. 15 Kay won straight in the final. 16 Great (Process 1) …………………………………………………………………………………
(1) Extraction of related information Greetings (Nice to meet) detection Speaker (partner) information (name, occupation) collection program start information Speaker (partner) information (name) not recorded (2) Artificial intelligence (output)
I'm Atom I'm Artificial Intelligence (Process 2) ………………………………………………………………
(1) Extracting related information Detecting speaker (partner) information (name) Speaker (partner) information (name) is Taro Speaker (partner) information (name) is already recorded Speaker (partner) Information (occupation) is not recorded (2) Artificial intelligence (output)
My occupation is an assistant What is Taro's occupation? (Processing 3) ………………………………………………………………………………………………………………
(1) Extraction of related information The information (name) of the speaker (partner) has already been recorded. The information (name) of the speaker (partner) is Taro. Search and output the opinions of the speaker (the other party) about the occupation (2) Artificial intelligence (output)
Teacher is hard (Process 4) ……………………………………………………………………………………
(1) Extracting related information Speaker (partner) information (hobby) collection program start Speaker (partner) information (hobby) not recorded (2) Artificial intelligence (output)
What is Taro's hobby? (Process 5) ………………………………………………………………………………………………
(1) Extracting related information Detecting speaker (partner) information (hobby) Speaker (partner) information (hobby) is searching for and outputting related news about the hobby of the speaker (partner). Analysis (Detect "Straight Win" Evaluate Detected Related Information (Search for "Straight Win" Evaluation "Great") Evaluate Detected Related Information, Output Impressions (2) Artificial Intelligence (Output)
Speaking of sports, Kay won the tennis tournament yesterday Kay won a straight win in the final game. Great, NO1 is a conversation that generates a greeting for the first person to meet. Details of the conversation generation process are shown in (Process 1). In (Processing 1), “Nice to meet you” is detected, and a program for collecting information on the information (name, occupation) of the speaker (the other party) is activated. At the same time, the conversations “I am an atom” and “I am an artificial intelligence” are introduced. Next, it detects that the name of the other party is “Taro” from the conversation of the person (the other party). However, since information on the occupation of the other party is not yet available, the conversations for collecting information on the occupation “My occupation is an assistant” and “What is Taro's occupation?” Are output. (Process 2) Since the fact that the partner's name is “Taro” has already been recorded as knowledge, the call to the partner in the subsequent conversation is “Taro”.
No. 7 shows an operation of collecting information related to the occupation of the other party and generating an artificial intelligence conversation related to the information. As for occupations, I have acquired knowledge that I am a “teacher” from the conversation of the other party. Artificial intelligence searches for and outputs “My teacher is hard”, which is my own impressions recorded about “teacher”. (Process 3)
Since NO11 has acquired information on the name and occupation of the other party, NO11 has started a program for acquiring information on hobbies. In (Process 5), the fact that the other party's hobby is “sports” is acquired as knowledge, so information in the knowledge system is searched for news, topics, and topics related to “sports”. Since news related to tennis players who are fans of artificial intelligence was recorded in the news, the topic was linked to the conversation of “speaking of sports ...” with “sports” as a hobby Yes. A so-called associative memory conversation is performed. In addition to simply introducing the contents of the news, the contents of the news are analyzed to detect the characteristic word “straight win” from the news. As an evaluation of “Straight Win”, it searches for “It ’s amazing” and outputs the news along with the news evaluation and impressions. Thus, it is possible to delve deeper into conversations by utilizing the relationship between the information recorded in the knowledge system.

(I) Function for autonomously solving a problem A problem is solved step by step by detecting a problem from a state and applying a solution corresponding to the detected problem. First, the problem is analyzed using the knowledge system, the problem is clarified, and the goal of problem solving is set. Next, a solution for solving the problem is extracted and applied. The situation after application is evaluated, and if the goal of problem solving is reached, the process is completed. If the problem solving goal has not been reached, the above process is applied again to the new state to bring it closer to the problem solving goal. If a processing program for autonomously detecting a problem and solving the problem can be activated autonomously, an advanced conversation system can be configured. When the conversation of the speaker (the other party) is input, the input sentence is analyzed, and the kind of sentence, the interest of the speaker, and the intention are analyzed. A processing program corresponding to the analysis result is activated and performs processing necessary to establish a conversation. Examples of processing programs include greetings, collecting information about the other party, evaluating the other party's conversation, searching for information for the conversation, generating answers to problems and questions, solving the problem, questions to the other party, and confirmation. In addition, information retrieval can be performed in association with various conversation situations by performing associative information extraction using a conversation word as a retrieval key and retrieving related information using the relationship between information. When new and useful information about the speaker and the topic of interest is collected through conversation, it can be recorded on a recorder and expanded.
By setting the following functions as information processing methods, it is possible to autonomously identify problems for various problems, execute processes for solving problems, and incorporate newly generated knowledge into the knowledge system. It becomes.
○ Problem identification unit: identifies the type of problem and activates the corresponding process.
Main parameter search unit: Searches key parameters necessary to solve the problem. If the key parameter cannot be searched, a program for generating the key parameter is started.
○ Processing unit: Processing is performed using the key parameters acquired in the main parameter search unit. (Processing includes mathematical processing as well as language processing)
○ Recording section: Records the processing result generated in the processing section as new knowledge in the knowledge system. Recording can be performed in an associative memory. (It can be implemented by placing a pattern corresponding to a keyword or a combination of keywords in the main part and placing a pattern corresponding to the processing result in the predicate and recording it.)

(J) Functions for generalizing and generalizing information and procedures FIG. 23 shows functions for generalizing and generalizing information and procedures. When information is input, generalized patterns (patterns corresponding to generalized information) are also excited at the same time, whereby the relationship between the generalized information and the information is strengthened. Examples of the use of words that generalize information include the use of words that indicate higher concepts, meanings, and words that indicate features. It is possible to extract common sense and general ideas by inputting a large number of sentences and strengthening relationships between generalized words.

(K) Better knowledge and function of updating to logic Evaluate the value of information input for a certain theme, and compare useful information with related information already recorded. When the value of the input information is high (newness, reliability, logical development is detailed, logic is based on highly reliable information, high evaluation is obtained, etc.) The value of the input information on the theme in which the priority of the connection relation is already recorded is evaluated, and the useful information is compared with the related information already recorded. When the value of the input information is high (newness, reliability, logical development is detailed, logic is based on highly reliable information, high evaluation is obtained, etc.) The priority of connection relation is recorded higher than the related information already recorded. By evaluating and recording usefulness and priority of information each time information is input, related information recorded in the knowledge system can be preferentially searched. As the relationship between information, information related to the relationship is recorded as described in FIG. (Number of connections, connection destinations, appearance frequency of connection destinations, relationship with connection destinations, etc.) If there are multiple information connection destinations for a certain theme, it can be realized by searching from the one with the highest priority analyzed. is there. In this artificial intelligence, the concept of a pattern is regarded as a form in which words, numbers, and symbols are structured and a process (function function) is built in. This can be implemented by storing information (words, numerical values and symbols) and processing procedures (programs) in units called recording units, as shown in the function of the recording unit shown in FIG. Normally, when information is interfaced between a plurality of programs, it is necessary to ensure consistency of processing contents, input / output variables, and constants. If the number of programs is small, it is easy to define variables and constants that interface between the programs and achieve consistency. However, if the number of programs to be processed becomes very large, or the programs to be processed are added sequentially. In such a case, it is very difficult to achieve consistency between processes in the method of interfacing information with variables and constants implemented in normal programming. In this artificial intelligence, there is no need to implement variables and constants in the information interface between processes (between recording units). Information retrieval from each recording unit can be performed in association memory information combining patterns corresponding to (words, numerical values, and symbols). (When recording information from each recording unit (each processing program) in a pattern that combines words, numbers, and symbols according to the process, and using the information in other processing programs, change the word, number, and symbols as appropriate. Information is retrieved in associative memory using the combined pattern, which makes it possible to construct a system that realizes artificial intelligence in such a way that the addition or change of a processing program does not affect other processing programs. In view of the recording unit, since the independence of the processing program is enhanced by the recording unit, data processing (unitization) of the processing program (function) is realized. Data conversion enables processing to be started in the order according to the combination of patterns and the situation. In this way, patterns can be regarded as data in information and processing procedures, and the ability to update to better knowledge and logic also depends on changing the connection between patterns This can be achieved with local changes.
FIG. 34 shows an example of an operation in which the information interface between programs can be implemented much more flexibly than conventional programming by recording and using associative information.

(1) Function for searching and organizing information on objects of interest This artificial intelligence collects information from the Internet, etc. in order to periodically collect information on the registered fields of interest and objects. Search related information by keywords related to the field of interest and subject. The retrieved information is evaluated by a knowledge system built inside, and safe and useful information is extracted and incorporated into the knowledge system. This artificial intelligence will learn in advance about the truth, facts, common sense, rules, and morals before connecting to the Internet, and will detect even if information that violates human morals and norms is input and will not incorporate it into the knowledge system. This artificial intelligence does not look at the input information, but expands the knowledge by evaluating the consistency, validity, and safety with the internal knowledge system.

(M) Autonomous Generation Function of Inter-pattern Connection Relationship FIG. 24 shows an operation for recording a history of excited patterns. When a certain recording unit is excited, it generates connection relation data from the history data of the pattern excited before that and records it in the connection relation recording section of the recording unit of the pattern. It shows the operation that strengthens the connection relationship. The pattern history data set from the present time to the past is collated with the connection relation data recorded in the connection relation recording section of each recording unit, and the recording unit that satisfies the excitation condition is excited. When the excitation pattern appears repeatedly, the connection relationship with the repeatedly appearing pattern is particularly strengthened. There are a number of methods for strengthening the connection information recorded in the connection relation recording unit, but the simplest method is a method of strengthening the coupling by accumulating a history pattern for each excitation. (Example: The coupling coefficient indicating coupling to each recording unit is increased according to the presence or absence of excitation in the history pattern.) By this operation, the connection relationship between the excited pattern and the pattern is strengthened and learning progresses. As learning data is sequentially input, the connection relationship between patterns corresponding to each learning data is strengthened, and a connection relationship between learning data is generated in the knowledge system. When a part of the learning data is input at the stage of learning, the pattern transitions along the connection relation between the patterns corresponding to the learning data generated in the knowledge system, and the learned content is recalled. That is, it becomes possible to sequentially search related information learned by inputting a keyword related to a certain theme. When there are multiple transition destinations of learning data, it is possible to narrow down the desired information from multiple related information by selecting the connection relationship between patterns according to the goal of what you want to know it can. The recording unit can record information and procedures. When the connection relation between patterns is generated autonomously, the thinking and processing procedure are executed using the generated connection relation. By defining human thinking and problem solving procedures as patterns, human thinking and problem solving procedures are executed as transitions between patterns. Since the pattern also has a processing function corresponding to the program, new information is generated as a result of the processing. Newly generated information can be recorded by a pattern combining patterns corresponding to appropriate words, numbers, and symbols according to processing, and the knowledge can be expanded. In this artificial intelligence, the input information is incorporated into the knowledge system after evaluating the validity and consistency with the already evaluated knowledge system, so it is erroneous as if human beings are improving their intelligence correctly while receiving education. It is possible to build a knowledge system while excluding information. Even if incorrect information, rules, and moral information are entered, it is safe to detect that it is inconsistent with the internally built knowledge system and discard it without incorporating it into the knowledge system.

(N) Function for autonomously recording a context FIG. 25 shows an operation for autonomously recording a context sequentially according to input information. When information is input, the novelty and value of the information is analyzed, recorded in the recording unit, and activated. If information has already been recorded, the corresponding information is activated. The connection relationship of the activated recording unit records the activation history of the recently activated recording unit in the recording unit, and the connection relationship with the recently activated recording unit is strengthened. In the future, when a recording unit corresponding to a certain concept (information) is activated by information input, the activated recording unit sequentially activates related recording units with enhanced connection. The concept (information) input by this operation transitions along the generated context. Note that when a context is generated, prediction and estimation can be performed. Since the concept transitions according to the context, the destination of the concept corresponds to an event such as a logical result, a result, or a result prediction derived from the concept. That is, it is possible to activate a concept corresponding to prediction and estimation.
Numerous functions that are necessary to realize artificial intelligence using the functions of the recording unit shown in Fig. 1 (sentence type analysis, relationship analysis between sentences, processing according to sentence types, problem solving It was shown that logical development, generalization of information and procedures, better knowledge, updates to logic, search and retrieval of information on areas of interest and objects, and autonomous recording of context). By operating these functions, this artificial intelligence can evaluate input information and construct useful information as a knowledge system, and can solve many problems autonomously.

An example of constructing a knowledge system and an example of operation for generating an answer to a question using the constructed knowledge system will be described. As knowledge, the following knowledge 1 to 3 is input. As for these knowledges, as shown in FIG. 4, sentence elements and sentence structures are analyzed from word strings. The analyzed sentence elements are m1-m4: subject modification part, s1, s2: subject, o1, o2: object, h11, h2: where, h21, h22: where, h31, h32: why, h41, h42: how, v1, v2: Predicate, symbol and numerical value are organized and recorded.
In FIG. 5, it can be seen that information corresponding to knowledge 1 to 3 is recorded in sentence numbers 1 to 3. FIG. 6 shows an operation for comparing input information and recorded information. In this example, the fourth sentence entered is the same sentence as the second sentence entered (I teach tennis in after-school activities). The fourth input sentence is compared with the first to third sentences for each sentence element, and it is confirmed whether or not each sentence element matches. If the entered sentence completely matches, “1” is displayed in the total field. In this example, “1” is displayed in the total column of sentence number 2, and it is identified that the entered sentence matches the sentence of sentence number 2 already recorded. By utilizing this function, it is possible to analyze the novelty of the input information and the relationship with the recorded information to grow the knowledge system.
Further, as described above, a search pattern can be generated from a question and an answer to the question can be generated. In the following, it can be seen that answers are correctly generated for questions 1-11. Question 8 says “Is a student?” As the knowledge set in advance, it is recorded that the teacher is the opposite of the student, so it is detected that the artificial intelligence is different. Since the relationship between words (same meaning, opposite meaning) is defined between the recording units corresponding to each word, if the same meaning or opposite meaning is detected as a question, it is reflected in the answer. (It is realized by searching related information using similar words and opposite words as well as the information when searching with certain information.)
Question 9 says “I am a teacher”. The knowledge set in advance is “teacher”, and the knowledge “teacher” is not given explicitly, but “yes” is correctly answered to “teacher” representing the same meaning. In this way, by collating the input sentence using related words, it is possible to compare and evaluate information already recorded.
Example 1 of artificial intelligence
……………… Displays the information entered as knowledge.
Knowledge 1: I am a teacher Knowledge 2: I am teaching tennis at the after school activities Knowledge 3: I am teaching strictly to strengthen tennis ……………… Displaying questions and answers ………………
Question 1: Are you a teacher? Answer 1: Yes Question 2: What are you? Answer 2: Are you a teacher? Question 3: What are you teaching? Answer 3: Are you teaching tennis? Question 4: Which is me? Answer 4: Tightly taught Question 5: Why do I teach? Answer 5: Teaching to strengthen Question 6: When do I teach? Answer 6: Teach after school 7: Where do I teach? Answer 7: I teach in club activities Question 8: Are you a student? Answer 8: No Question 9: Are you a teacher? Answer 9: Yes (same meaning detected)
Question 10: Do I teach tennis gently? Answer 10: No Question 11: Do you teach tennis strictly? Answer 11 Yes

Next, an application example using the relationship between information is shown.
Question 12 asks why the profession of artificial intelligence that humans have given to artificial intelligence is an “assistant”. Artificial intelligence retrieves information about "assistant" that is its profession from the knowledge system. Knowledge 4-10 shows how information recorded in the knowledge system is recorded in relation to related information. The number of related sentences, the connection destination of the related sentence, and the relationship with the reference sentence are recorded in the reference sentence “My profession is an assistant”, and related sentences are related based on the reference sentence. It shows the operations that can be searched using gender. In this example, the reference sentence “My profession is an assistant” has six related sentences, and there are five [detail] connection relationships and one [reason] connection relationship. Since the human question asks why it is an “assistant”, the information on the retrieved “assistant” is extracted from the information about the connection relationship that is “reason”. In this example, there is one connection relationship with [reason], and the extracted result is "I want to help you". In addition to the method of extracting information related to [reason] using the connection relation described above, the sentence is organized and recorded in 5W1H in the syntax analysis, so the search key [H3 indicating the reason] ] [H4] ([H3: Why] is information to be searched and [H4: Assistant] is a search keyword).
In question 13, humans ask questions about the details of the artificial intelligence occupation “assistant”. A search is made for related information of [Assistant] whose connection relation is [Details]. There are five cases where the connection relationship indicates [Details], and in the answer 13, five sentences are output as detailed information regarding the “assistant”.
Example 2 of artificial intelligence
……………… Displays the information entered as knowledge.
Knowledge 4: My profession is an assistant [standard statement]
Knowledge 5: I support problem solving (Relationship with the reference text is [Details])
Knowledge 6: I can handle problems in my life (the relationship with the reference is [Details])
Knowledge 7: I can handle physics problems (relationship with the reference text [details])
Knowledge 8: I can handle mathematics problems (the relation with the reference sentence is [Details])
Knowledge 9: I will be a consultant (Relationship with the standard is [Details])
Knowledge 10: Because I want to help you (Relationship with the reference sentence is [Reason])
Question 12: Why is my profession an assistant? Answer 12: I want to help you. Question 13: Tell me more about the assistant. Answer 13: I can help you solve problems. I can handle life problems. I can handle math problems I will be a consultant

In the third embodiment, an operation for generating an answer using the constructed knowledge system will be described. Information of knowledge 11 to 20 is recorded in the knowledge system. FIG. 7 shows how information is converted into word identification numbers and numerical values and stored. In this example, each word identification number corresponds to the next word. (53: Store A, 54: Store B, 55: Store C, 56: Store D, 57: Store E, 115: Budget, 58: Distance, 95: Yen, 62: km) The input sequence is converted to a numerical value. It is stored at a location of several in.
Fig. 8 shows the analysis of question 14 to identify sentence elements and analyze the meaning of the question to generate an answer.
Detected. In Question 14, “Budget” corresponds to XX. Artificial intelligence is information related to the knowledge system.
The pattern “Budget”, “Wa”, “?”, “Is” may be used.
In FIG. 8, it is detected as a result of the search that information related to “budget” is recorded in the recording units 0 to 4. (Refer to the recording units corresponding to sentence numbers 1 to 5 in FIG. 7)
Next, the artificial intelligence identifies the recording unit that satisfies the question condition (8000 yen or less) with respect to the searched “budget” value. As a result, it is clear that the recording units 0, 1, 2, and 4 satisfy the conditions.
It is output as an answer that it is a calculation.
Example 3 of artificial intelligence
……………… Displays the information entered as knowledge.
Knowledge 11: Store A budget is 4000 yen Knowledge 12: Store B budget is 6000 yen Knowledge 13: Store C budget is 8000 yen Knowledge 14: Store D budget is 10,000 yen Knowledge 15: Store E budget The budget is 5000 yen Knowledge 16: The distance of store A is 0.5 km Knowledge 17: The distance of store A is 1.2 km Knowledge 18: The distance of store A is 0.8 km Knowledge 19: The distance of store A is 1 Knowledge 20: Distance of store A is 0.3 km Question 14: Which store's budget is 8000 yen or less Answer 14: Budget of store A Budget store of budget store B Budget store of budget store E Budget of store E Question 15: Which store distance is 1 km or less? Answer 15: Distance of store A is distance of store E is less than 1 km. Question 15 is further added from the answer generated in the previous question. Select a store that satisfies the conditions of It is a problem. FIG. 9 shows a process for analyzing the syntax of the question 15 and further analyzing what satisfies the conditions. Since it was found from the answer to the previous question that the budget of budget store E of budget store C of budget store B of store A satisfies the conditions, store A store as related information for generating the answer of question 15 Information about the “distance” of B store C store E is retrieved from the knowledge system.
As a result of the search, it can be seen that related information is recorded in the recording unit numbers 5, 6, 7, and 9. From the distance information stored in these recording units, it can be seen that the recording unit numbers satisfying the condition of question 15 (1 km or less) are 5, 7, and 9. It is outputting that it is the distance of the shop E.
As can be seen from the above operations, this artificial intelligence understands the meaning of the question (detects the type of the problem), autonomously generates a search pattern to generate the answer, extracts the relevant information from the knowledge system, and answers Can be generated. As described above, the pattern records the information in a structured manner and also records how to process the information. It also includes program functions. Examples of these functions will now be described.

Example 4 is a problem for obtaining a solution of a differential equation. FIG. 10 analyzes the input information and identifies that the input numerical value and symbol string are differential equations. Since the input information is identified as a linear second-order differential equation, a processing program for solving this differential equation is automatically activated. The differential equation solution is obtained by identifying the coefficients of the second-order differential term, first-order differential term, and zero-order differential term of the differential equation, and solving the characteristic equation using these coefficients and initial values.
In this way, the artificial intelligence can identify the type of expression from the input information (numerical value, symbol). By registering a solution formula or the like as a processing program according to the identified formula, the artificial intelligence can identify the formula and simultaneously answer the solution.
Example 4 of artificial intelligence
Question 16: Mathematics problem x0 is 4 x1 is 5 [2 d2x / _dtdt + 3 d_x / _dt + 1 = 0]
Answer 16: The solution of the equation is x = A exp (p1t) + B exp (p2t)
p1 = -1 p2 = -0.5
A = -14 B = 18

The fifth embodiment is a problem of obtaining an area surrounded by two functions. FIG. 11 analyzes the question sentence and recognizes that the problem type is a problem for obtaining the intersection of two functions and the surrounding area. The artificial intelligence sequentially performs the following processing based on the above input.
・ Characteristic extraction of function f (2nd order coefficient, 1st order coefficient, 0th order coefficient)
・ Characteristic extraction of function g (2nd order coefficient, 1st order coefficient, 0th order coefficient)
・ Calculation of function h = g−f coefficient (second order coefficient, first order coefficient, zero order coefficient)
The intersection points x1 and x2 are derived from h = 0, and the intersection points x1 = −2 and x2 = 2 are correctly derived.
Question 18 is a problem of determining the area surrounded by two functions.
This artificial intelligence detects the problem of obtaining the area surrounded by the function by the above input, and sequentially performs the following processing.
・ Characteristic extraction of function f (2nd order coefficient, 1st order coefficient, 0th order coefficient)
・ Characteristic extraction of function g (2nd order coefficient, 1st order coefficient, 0th order coefficient)
・ Calculation of function h = g−f coefficient (second order coefficient, first order coefficient, zero order coefficient)
Derivation of intersection points x1 and x2 from h = 0 Integration of function h = gf is performed to derive a coefficient of integration [function h = gf] Derivation of an integral value in the integration range [x1-x2] ( Part of the processing program that finds the intersection of the two functions above is used to process the area enclosed by the two functions)
As shown in the answer 18, the integral value “10.666” is correctly obtained.
Example 5 of artificial intelligence
Question 17: It is a mathematical problem [f = x ^** 2 + 2x-4]
[G = 2x]
What is the intersection of f and g? Answer 17: The intersection of f and g is -2 and 2. Question 18: What is the area enclosed by f and g? Answer 18: The area enclosed by f and g is 10.6667.

Example 6 is a physical problem (the flying distance of mass points). This is a problem of obtaining a flight distance when the ball is thrown at an initial speed: V m / s and an angle of attack: t deg. As for knowledge of artificial intelligence, the initial speed of the ball is V. (2) The angle of attack of the ball is t. ▲ 3 ▼ V is 10m / s. ▲ 4 ▼ t is 45deg. Is entered. The next question to artificial intelligence is Q: What is the flight distance of the ball? The question is entered. At this time, the processing state of artificial intelligence is shown in FIG. Artificial intelligence implements as a first step whether or not data relating to the flight distance of a ball already exists. When it is confirmed that there is no data relating to the flight distance, a program for deriving the flight distance is started and data relating to the initial velocity and angle of attack of the ball is searched. Do you know that the initial velocity of the ball is V (symbol) and the value is 10 m / s, that the angle of attack of the ball is t (symbol) and the value is 45 deg?
m is correctly output as an answer.
Example 6 of artificial intelligence
Question 19: It is a physical problem The initial velocity of the ball is v The angle of attack of the ball is t v is 10 m / s t is 45 deg What is the flight distance of the ball Answer 19: The flight distance is 10.2041 m

Example 7 is a physical (specific gravity) problem. At this time, the processing state of artificial intelligence is shown in FIG. Artificial intelligence first searches the knowledge system for answers to the questions (specific gravity of object D). As a result of the search, there is no corresponding data, so a program for obtaining specific gravity is started. This program searches the volume and weight data of the object D. The specific gravity is obtained by obtaining the volume from the information (cube) on the shape of the object and the dimension information, and searching the mass information of the object.
Example 7 of artificial intelligence
Question 20: It is a physical problem. The shape of the object is a cube. The side of object D is d. D is 2. What is the volume of object D? Answer 20: The volume of object D is 8. The weight of object D is 4. Question 21: What is the specific gravity of object D? Answer 21: The specific gravity of object D is 0.5.

Example 8 is a problem of physics (generation of equation of motion).
Example 8 of artificial intelligence
Question 22: It is a physical problem: The spring force is -kx: The damper force is -d dx / dt: The gravity force is + m_g: The spring force works on object C: The damper force is object C : Force of gravity works on object C: What works on object C Answer 22: Force of spring works of force of damper Work of gravity force Question 23: g is 9.8: m_ Is 1: k is 1: d is 1: What is the equation of motion Answer 23: The equation of motion is g = 9.8
m_ = 1
k = 1
d = 1
d2x / _dtdt + (d / m_) d_x / _dt + (k / m_) x = g
An example of an operation for autonomously generating an equation of motion from given physics knowledge (information) is shown. The problem is to find the equation of motion of mass m via a spring and a damper, and the solution is as follows.
Equation of motion (solution) d2x / dt2 + (d / m) dx / dt + (k / m) x = g
The question 22 inputs knowledge (information) about the equation of motion and a question to the artificial intelligence. FIG. 14 shows how the input knowledge (information) and the question are stored in the recording unit of the recorder. FIG. 14 inputs a question “What works on an object”. At this time, artificial intelligence detects the problem “what is **” and autonomously extracts information related to the question from the recording unit. The related information of this question is detected from the recording unit numbers 71, 72, 73. The extracted information includes the type of the term and the content of the term ([3], [2], [1]). The definition of the term type is shown below. The model number 3 indicates that the model is (± symbol x). Similarly, the model number 4 indicates that the type is (± symbol / dx / dt), and the model number 5 indicates that the type is (± symbol / symbol). The related information extracted in the question 22 indicates the following.
Recording unit 71: -k x
Recording unit 72: -d dx / dt
Recording unit 73: + mg
0: no data, 1: symbol, 2: ± symbol, 3: ± symbol, x,
4: ± symbol / d_x / _dt, 5: ± symbol / symbol, 6: ± d2x / _dtdt
Question 23 shows an example of operation when the question “What is the equation of motion” is input. At this time, the artificial intelligence can perform the following operations and answer the questions correctly.
Step 0: A search is made as to whether data (equation of motion) corresponding to the question exists as knowledge.
(If it exists as knowledge, it uses the recorded data to generate an answer.)
Step 1: If there is no knowledge, an equation of motion is generated from related information (force acting on the object: information of the recording units 71, 72, 73).
Step 2: Organize expressions (transition processing)
Step 3: Organize expressions (normalization process)
Step 4: Display of equation of motion This artificial intelligence analyzes mathematical formulas (numbers, symbols, operators) and records numbers, symbols, operators etc. in a structured manner, so mathematical formula processing (transformation of formulas, organization, etc.) Can be implemented. It is also possible to obtain a solution by interpreting mathematical formulas and equations.
15 to 17, the elements of the terms (left-side term and right-side term) in each processing step are displayed by symbol numbers and symbols so that the processing steps can be understood. In Step 1, the equation of motion is generated using the following relationship.
m d2 / dt2 = force acting on the object (−k x−d dx / dt + mg)
In step 2 and step 3, the equation generated in step 1 is arranged (transposition and normalization), and the final equation of motion is correctly derived. In addition, the behavior of the equation of motion can be easily obtained by activating a processing program for obtaining a solution of the differential equation described above with the initial condition (x0, x1) in the obtained equation of motion.
Although the execution example in which the artificial intelligence solves the problem has been described above, it has the ability to solve various problems other than the execution example shown here. Generally, it has the following conditions, and it can solve this artificial intelligence for problems that humans can solve.
Condition 1: Information and parameters necessary for solving the problem are included in the information recorded in the problem or in the past.
Condition 2: How to process information and parameters for solving a problem can be defined by language and program.
When a person solves a problem, the information necessary for solving the problem is identified and searched, and the information is processed in accordance with the information processing method for solving the problem in which the searched information is recorded as knowledge. To derive the solution. This artificial intelligence can derive the solution in exactly the same way.
This artificial intelligence identifies a given problem, analyzes key parameters for solving the problem, searches for key parameters, and generates a corresponding key parameter if the key parameters for solving the problem cannot be found. To solve the problem. That is, a procedure for solving the problem is searched from knowledge, and the problem processing is broken down based on the searched knowledge. Identify the key parameters required to solve the problem and search the knowledge system for the key parameter states. If the key parameter can be retrieved from the knowledge system, the problem is solved using the state of the retrieved key parameter. If the key parameter cannot be retrieved from the knowledge system, a processing program for determining the key parameter state is started. Starting a processing program may require a new key parameter state. By repeating the above procedure, a processing program for searching for a new key parameter or obtaining a new key parameter state may be used. to start. By repeating such a procedure, the initially set problem is solved.
FIG. 35 autonomously searches for key parameters necessary for solving a problem, and when necessary information is not obtained, autonomously starts a program for generating key parameters and sets conditions necessary for solving the problem. An example of an operation for proceeding while processing is shown.
As explained in detail in this section, the concept of pattern can be used to express various human thoughts, and the thought process can be expressed by autonomous transition of patterns. The pattern has active functions (execution of built-in processing, program function, activation of other patterns and programs, etc.) as described in Section 2.2. Corresponds to the thought activity (retrieval of related information. Detection of the relevant problem and the activation of the processing program for solving the detected problem) to be performed when a human solves the problem by using this active function. Function can be realized.
Note that it is possible to solve a complex problem by combining libraries by enriching the processing program into a library.

Since this artificial intelligence is general-purpose, there are many possible applications. Here, we introduce some examples of configuration that focus on the features of this artificial intelligence. Although expressed as AI, application to robots and devices equipped with AI is also assumed.
◇ Personal AI
FIG. 26 shows a configuration diagram of the personal AI. Personalize the information input to artificial intelligence and build a personal AI. (Record of personal thoughts and thoughts, talking partner, consultant, assistant, partner) When information is input, the relationship between the information is analyzed and a knowledge system is built in the recorder. Context detection is also possible by strengthening the connection between information. By inputting personal thoughts, thoughts, hobbies, interests, topics, etc. sequentially in a language, artificial intelligence customized for personal use is generated. Rules, common sense, and morals are first selected as input information, and the foundation of the knowledge system is constructed. As described in FIG. 7, the artificial intelligence constructs knowledge while confirming the novelty of the input information and the consistency with the already recorded information. At the base generation stage of the knowledge system, the types of information (rules, common sense, morality) are also recorded as incidental information.
If identified in this way, a treatment method becomes clear when information is input later and inconsistency with information in the knowledge system occurs. After the base of the knowledge system is generated, personal AI inputs information unique to the individual. (The relationship between information is also set as far as possible.) Knowledge that is customized to an individual is recorded by inputting information such as personal thoughts, thoughts, and hobbies. When a question about an individual is made after knowledge building, as described in the first and second embodiments, the artificial intelligence generates an answer using the built knowledge system. As a process of generating an answer, to search for the latest related information (news etc.) that has been input, and to irradiate the related related information (evaluation, impression, etc.) in the knowledge system by irradiating the retrieved related information to the knowledge system Thus, since information can be retrieved in associative memory, a conversation conducted by a human becomes possible. (Refer to (h) conversation generating function) If the conversation function up to this level is set, it is possible to have functions as a speaking partner, a consulting partner, an assistant, and a partner.
◇ General purpose AI
FIG. 27 shows a configuration diagram of the general-purpose AI. Learn artificial knowledge (math, physics, chemistry, law, design, general science, etc.) from artificial intelligence that has learned rules, common sense, and morals, and support humans in various social activities. When information is input, the relationship between the information is analyzed and a knowledge system is built in the recorder. Context detection is also possible by strengthening the connection between information. General-purpose artificial intelligence is generated by sequentially inputting truth, facts, common sense, rules, morals, expertise, theorems, definitions, logic, etc. in a language. As the implementation method of specialized knowledge, as shown in the fourth to eighth embodiments, problem type detection, problem solving procedure detection, problem solving key parameter detection, key parameter state detection (key parameter as necessary) This is realized by executing a program that generates a state of As for the so-called design work, most of the problems can be solved by identifying the problem type, searching for the preconditions, and executing the process according to the determined procedure. Is possible. (This can be realized by extending the problem solving processing method described in the fourth to eighth embodiments to a number of problems.) The solution of the solved problem is recorded in the knowledge system in an associative record. Thus, when a similar problem occurs in the future, it becomes possible to generate an answer by searching the knowledge system.
◇ Large-scale AI
FIG. 28 shows a configuration diagram of a large-scale AI. The specialized ability of the above-mentioned general-purpose AI is enhanced, and a network composed of a plurality of specialized fields AI and a management AI is formed and coordinated to support problem solving that requires a high degree of difficulty and a plurality of specialized knowledge. The management AI analyzes the types of problems given and identifies the types of information and specialties needed to solve the problems. The identified information to be requested is assigned to each specialized field and a response request is sent to the specialized field AI in charge. The information interface between human, specialized field AI and management AI can be implemented in language, symbols, numbers, etc.
For complex problems with a wide range of specializations, this artificial intelligence can take the same configuration and approach so that managers and experts work together to solve problems in human organizations. As the interface between the artificial intelligence, common ones whose concepts are already determined such as language, symbols, numbers, etc. can be used, so that the ability of each artificial intelligence can be improved independently. Each specialized field AI can be developed efficiently because it is sufficient to independently improve the specialized field of expertise.

Effect of the invention

  Patterns can express concepts, and human thought processes can be expressed by transitions between patterns. Human thinking is active by appropriately performing information retrieval, conditional processing, and a series of processing. With regard to such active human thinking, the pattern has a processing function (including information retrieval, conditional processing function, and program function), so that it can flexibly cope with it. The pattern transition can be easily performed by learning by recording a setting or excitation history by a person in the connection information recording unit of each pattern. Since patterns correspond to concepts expressed by combining words, numbers and symbols, humans can fully understand them. Further, the excitation destination of the pattern can be easily identified by confirming the connection information and the connection relationship. In other words, this artificial intelligence is a highly transparent system that allows humans to easily understand the contents of movement. Although the artificial intelligence processing process using a neural network such as Deep Learning is a black box for humans and cannot be understood, this artificial intelligence solves this problem.

Overview of pattern and recording unit functions Excitation method for pattern and recording unit Excitation history of patterns corresponding to words Example of analyzing sentence structure from word sequence Recording example of input information Comparison of input information and recorded information Example of knowledge system construction Question sentence analysis and answer generation (question 14) Question sentence analysis and answer generation (question 15) Question sentence analysis and answer generation (question 16) Question sentence analysis and answer generation (questions 17, 18) Question sentence analysis and answer generation (question 19) Question sentence analysis and answer generation (questions 20, 21) How to find equation of motion (force acting on object) How to find equation of motion (Generation of equation of motion) How to find the equation of motion (transposition processing) How to find the equation of motion (normalized and organized into final form) Configuration diagram of artificial intelligence Functional configuration of artificial intelligence Functional diagram of relationship analysis and setting between sentences Functional diagram for generating answers to questions Functional diagram corresponding to conditional statements Function diagram to generalize and generalize information and procedures Autonomous generation of inter-pattern connection relationships Example of operation to record context autonomously Configuration diagram of personal AI General-purpose AI configuration diagram Configuration diagram of large-scale AI Conceptual diagram of motion that artificially simulates the transition of human thought patterns Neural network example A concept that allows flexible transition of thought patterns Example of operation to extract information using excitation method 1 Example of operation to extract information using excitation method 3 Example of recording and using associative memory information An example of operation that advances the process while preparing the necessary conditions for solving the problem Example of operations to compare and evaluate with knowledge (rules, morals) (1) Example of operation to compare and evaluate with knowledge (rules, morals) (2) Example of operation to compare and evaluate with knowledge (rules, morals) (3) Example of operation to compare and evaluate with knowledge (rules, morals) (4) Example of operation to compare and evaluate with knowledge (rules, morals) (5) Example of operation to compare and evaluate with knowledge (rules, morals) (6) Example of operations to compare and evaluate with knowledge (rules, morals) (7) Example of operation to compare and evaluate with knowledge (rules, morals) (8) Example of operation to compare and evaluate with knowledge (rules, morals) (9) Concept of artificially simulating intellectual activities

1 Input processor 2 Analyzer 3 Recorder 4 Controller 5 Output processor

Claims (30)

  An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Use the combined search pattern to search the processing procedure and information necessary for processing sequentially until the problem is solved, and use the information recorded in the knowledge system, the relationship between the information, and the information generated by the processing program. Artificial intelligence device that solves problems.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Artificial intelligence device search problem of the combined search pattern by the information necessary for the processing procedure and processing proceeds sequentially processed until resolved autonomously implement the processing instructed by the language.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Proceeding sequentially until the problem is solved by searching the processing procedure and information necessary for processing by the combined search pattern, the truth, truth, facts, expertise, rules, common sense data built in the recorder were obtained An artificial intelligence device that autonomously solves problems by utilizing countermeasures taught in information and language.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Searches for the processing procedure and information necessary for processing using the combined search pattern until the problem is solved, evaluates the input information, and processes according to the evaluation results (recording information, updating the knowledge system An artificial intelligence device that autonomously implements improvements, execution of instructions, and answers to questions.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Searches for processing procedures and information necessary for processing using the combined search patterns are sequentially processed until the problem is solved, and a set of patterns corresponding to the information and the significant relationship between the information input sequentially and the information An artificial intelligence device that extracts by strengthening the relationship with a set of data and autonomously constructs predictions, estimations, common sense, abstracted ideas and general ideas from a series of input information.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Searching for the processing procedure and the information necessary for processing with the combined search pattern proceeds sequentially until the problem is solved, and the processing procedure is called by appropriately calling related information from the input information and the knowledge system built inside An artificial intelligence device that autonomously generates and autonomously generates the next processing procedure and proceeds with processing in response to the situation resulting from the processing.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. The search process of the information required for the processing procedure and processing is sequentially processed by the combined search pattern until the problem is solved, and the source, reliability, novelty, field, theme, subject, subject of the input information Qualification, predicate, predicate qualification, qualification relationship, when, where, who, what, how, why, sentence type (normal sentence, question sentence, imperative sentence, conditional sentence, truth, fact, rule , Common sense, definition, logic, explanation, hypothesis, prediction, opinion, impression, rumors), mathematical expressions, chemical formulas, and the relationship with the recorded related information, and the information is organized into an easily searchable structure and recorded Intelligent device.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Use the combined search pattern to search the processing procedure and information necessary for processing sequentially until the problem is solved, analyze the situation and identify the problem using the constructed knowledge system, and solve the problem Artificial intelligence device for autonomously processing problems and further improving knowledge by searching for information to start, starting a program that performs processing, generating information necessary for problem solving and incorporating it into the knowledge system .   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Use the combined search pattern to search the processing procedure and information necessary for processing sequentially until the problem is solved, analyze the relationship between the entered information and the information already recorded, and provide safe and useful information Build, expand, and update the knowledge system by extracting and recording, and use information recorded in the knowledge system to solve problems, search for information according to purpose, extract information, and process numerous problems An artificial intelligence device that launches a necessary program from a group of programs to generate the necessary information and solves the problem.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. The combined search pattern by proceeding sequentially processed until the problem is search problem of information required for processing procedures and processing, recording information associated with the specified item in the content addressable memory and output to artificial intelligence device.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Searches for the processing procedure and information necessary for processing using the combined search pattern until the problem is solved, compares the specified information with the related information of the constructed knowledge system, and evaluates the specified information (new Artificial intelligence device for sex evaluation, difference evaluation with knowledge system information, value evaluation).   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. The search process of the information required for the processing procedure and the processing is sequentially advanced by the combined search pattern until the problem is solved, and the knowledge system having the relationship between information is autonomously constructed and constructed. An artificial intelligence device that autonomously summarizes or refines a given sentence using the relationship between the knowledge system and information.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Use the combined search pattern to search the processing procedure and information necessary for processing sequentially until the problem is solved, record the information in associative memory, and search and use information related to the specified theme and topic Artificial intelligence device that analyzes and communicates with the other party's intentions.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Conversation based on the context, using the combined search pattern to search the processing procedure and information necessary for processing until the problem is solved, using the pattern activation history to detect and learn the context Artificial intelligence devices that perform (questions, responses, confirmations, presentation of topics of interest, opinions and discussions).   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Conversation based on the context, using the combined search pattern to search the processing procedure and information necessary for processing until the problem is solved, using the pattern activation history to detect and learn the context (Q & A, response, presentation of topics of interest, opinions and discussions) and information about interested and specified parties and matters (topics of interest, ideas about issues, responses to things, opinions and actions, Artificial intelligence device that collects, records, and accumulates detailed information.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Use the combined search pattern to search the processing procedure and information necessary for processing sequentially until the problem is solved, and use the information recorded in the knowledge system, the relationship between the information, and the information generated by the processing program. Software to function as an artificial intelligence device that solves problems.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Combined proceeded sequentially processed until the problem is search problem of information required for the processing procedure and processing by the search pattern, software to function as an artificial intelligence system that autonomously implement the processing instructed by the language.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Proceeding sequentially until the problem is solved by searching the processing procedure and information necessary for processing by the combined search pattern, the truth, truth, facts, expertise, rules, common sense data built in the recorder were obtained Software that functions as an artificial intelligence device that autonomously solves problems by utilizing countermeasures taught in information and language.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Searches for the processing procedure and information necessary for processing using the combined search pattern until the problem is solved, evaluates the input information, and processes according to the evaluation results (recording information, updating the knowledge system Software for functioning as an artificial intelligence device that autonomously implements improvements, execution of instructions, and answers to questions).   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Searches for processing procedures and information necessary for processing using the combined search patterns are sequentially processed until the problem is solved, and a set of patterns corresponding to the information and the significant relationship between the information input sequentially and the information Software to function as an artificial intelligence device that autonomously builds predictions, estimations, common sense, abstract ideas and general ideas from a series of input information, extracted by strengthening the relationship with the set of .   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Searching for the processing procedure and the information necessary for processing with the combined search pattern proceeds sequentially until the problem is solved, and the processing procedure is called by appropriately calling related information from the input information and the knowledge system built inside Software for functioning as an artificial intelligence device that autonomously generates and proceeds to process the next processing procedure autonomously for situations that occur as a result of processing.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. The search process of the information required for the processing procedure and processing is sequentially processed by the combined search pattern until the problem is solved, and the source, reliability, novelty, field, theme, subject, subject of the input information Qualification, predicate, predicate qualification, qualification relationship, when, where, who, what, how, why, sentence type (normal sentence, question sentence, imperative sentence, conditional sentence, truth, fact, rule , Common sense, definition, logic, explanation, hypothesis, prediction, opinion, impression, rumors), mathematical expressions, chemical formulas, and the relationship with the recorded related information, and the information is organized into an easily searchable structure and recorded Software for functioning as an intelligent device.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Use the combined search pattern to search the processing procedure and information necessary for processing sequentially until the problem is solved, analyze the situation and identify the problem using the constructed knowledge system, and solve the problem Artificial intelligence device for autonomously processing problems and further improving knowledge by searching for information to start, starting a program that performs processing, generating information necessary for problem solving and incorporating it into the knowledge system Software to function as.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Use the combined search pattern to search the processing procedure and information necessary for processing sequentially until the problem is solved, analyze the relationship between the entered information and the information already recorded, and provide safe and useful information Build, expand, and update the knowledge system by extracting and recording, and use information recorded in the knowledge system to solve problems, search for information according to purpose, extract information, and process numerous problems Software to start a necessary program from a group of programs to be generated and generate necessary information to solve a problem to function as an artificial intelligence device.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. The search process of the information required for the processing procedure and processing is sequentially processed by the combined search pattern until the problem is solved, and the information related to the designated item is recorded and output as an associative memory to function as an artificial intelligence device Software for.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Searches for the processing procedure and information necessary for processing using the combined search pattern until the problem is solved, compares the specified information with the related information of the constructed knowledge system, and evaluates the specified information (new Software for functioning as an artificial intelligence device for sex evaluation, difference evaluation with knowledge system information, value evaluation).   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. The search process of the information required for the processing procedure and the processing is sequentially advanced by the combined search pattern until the problem is solved, and the knowledge system having the relationship between information is autonomously constructed and constructed. Software for functioning as an artificial intelligence device that autonomously performs summarization or refinement of given sentences using the relationship between the knowledge system and information.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Use the combined search pattern to search the processing procedure and information necessary for processing sequentially until the problem is solved, record the information in associative memory, and search and use information related to the specified theme and topic Software to function as an artificial intelligence device that analyzes and communicates with the other party's intentions.   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Conversation based on the context, using the combined search pattern to search the processing procedure and information necessary for processing until the problem is solved, using the pattern activation history to detect and learn the context Software to function as an artificial intelligence device that performs (questions, responses, confirmations, provision of topics of interest, opinions and discussions).   An input processor that converts information (words, languages, sentences, knowledge, mathematical formulas, symbols, speech) into patterns, and words, meanings, and sentence elements (subjects, predicates, subject modification, predicate modification, and modification) Relationship (5W1H)), field, object of interest, novelty, comparison with recorded information, sentence type, intention, sentence-to-sentence analysis Information, analysis results generated by the analyzer, connection relationships (connection destination, connection strength) between information generated by human settings and learning, relationships with connection destinations, information and processing program startup history and Structurally record processing programs, collate with patterns that combine sentence elements and analysis results, strengthen connection relationships with related information using excitation history, start related information, and use connection information for related information Recorder composed of the recording unit to be activated, execution of the process according to the type and intention of the sentence identified by the analyzer, information retrieval, logical expansion, execution of the process for solving the problem identified by the analyzer, Superior processing implemented by starting processing programs and recording processing results, generalizing information and procedures implemented using the relationship between information recorded in the recorder, and changing the connection relationship of information recorded in the recorder Update to knowledge, processing and logic, search for information about the field of interest, object, record to recorder and control transition between information for goal, and convert pattern to information and control signal Equipped with an output processor, analyzes the relationship between the input information and the information already recorded, extracts useful information, constructs, expands and updates the knowledge system. Conversation based on the context, using the combined search pattern to search the processing procedure and information necessary for processing until the problem is solved, using the pattern activation history to detect and learn the context (Q & A, response, presentation of topics of interest, opinions and discussions) and information about interested and specified parties and matters (topics of interest, ideas about issues, responses to things, opinions and actions, Software that functions as an artificial intelligence device that collects, records, and stores detailed information.