WO2022153470A1 - 情報処理装置、及び情報処理方法 - Google Patents
情報処理装置、及び情報処理方法 Download PDFInfo
- Publication number
- WO2022153470A1 WO2022153470A1 PCT/JP2021/001143 JP2021001143W WO2022153470A1 WO 2022153470 A1 WO2022153470 A1 WO 2022153470A1 JP 2021001143 W JP2021001143 W JP 2021001143W WO 2022153470 A1 WO2022153470 A1 WO 2022153470A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- unit
- formula
- steve
- hypothetical
- search
- Prior art date
Links
- 230000010365 information processing Effects 0.000 title claims abstract description 142
- 238000003672 processing method Methods 0.000 title claims description 3
- 238000010276 construction Methods 0.000 claims abstract description 88
- 230000014509 gene expression Effects 0.000 claims abstract description 37
- 238000000605 extraction Methods 0.000 claims description 87
- 238000006243 chemical reaction Methods 0.000 claims description 65
- 238000004364 calculation method Methods 0.000 claims description 21
- 239000000284 extract Substances 0.000 claims description 18
- 238000010586 diagram Methods 0.000 description 38
- 235000019800 disodium phosphate Nutrition 0.000 description 21
- 238000007726 management method Methods 0.000 description 12
- 230000006870 function Effects 0.000 description 8
- 230000000694 effects Effects 0.000 description 6
- 238000000034 method Methods 0.000 description 6
- 230000004048 modification Effects 0.000 description 6
- 238000012986 modification Methods 0.000 description 6
- 102100036848 C-C motif chemokine 20 Human genes 0.000 description 3
- 230000007812 deficiency Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 239000013598 vector Substances 0.000 description 3
- 230000004044 response Effects 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 101000713099 Homo sapiens C-C motif chemokine 20 Proteins 0.000 description 1
- 101000760620 Homo sapiens Cell adhesion molecule 1 Proteins 0.000 description 1
- 101000911772 Homo sapiens Hsc70-interacting protein Proteins 0.000 description 1
- 101001139126 Homo sapiens Krueppel-like factor 6 Proteins 0.000 description 1
- 101000710013 Homo sapiens Reversion-inducing cysteine-rich protein with Kazal motifs Proteins 0.000 description 1
- 101000661816 Homo sapiens Suppression of tumorigenicity 18 protein Proteins 0.000 description 1
- 101000661807 Homo sapiens Suppressor of tumorigenicity 14 protein Proteins 0.000 description 1
- 101100018027 Pisum sativum HSP70 gene Proteins 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 108090000237 interleukin-24 Proteins 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
- G06N5/041—Abduction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
- G06N5/025—Extracting rules from data
Definitions
- the disclosed technology relates to an information processing device and an information processing method.
- the question answering device described in Patent Document 1 that responds to a question from a user asks the user back in order to obtain information necessary for reaching an answer that can be given to the user.
- the question-answering device cannot obtain the above-mentioned necessary information.
- the question answering device could not give an answer to the user because it could not reach the above answer.
- the purpose of this disclosure technique is to give an answer to the user even if the information necessary to reach the answer cannot be obtained.
- the information processing apparatus includes a construction unit that constructs an output formula based on an inference rule related to a search formula in which the text of a question is described by predicate logic, and the output. Includes a generation unit that generates the text of the answer from the formula.
- the information processing device it is possible to give an answer to the user even if the information necessary for reaching the answer cannot be obtained.
- the configuration of the information processing apparatus 1 of the first embodiment is shown. It is a block diagram which shows the operation of the information processing apparatus 1 of Embodiment 1. It is a flowchart which shows the operation of the information processing apparatus 1 of Embodiment 1. It is a functional block diagram of the information processing apparatus 2 of Embodiment 2. It is a block diagram which shows the operation of the information processing apparatus 2 of Embodiment 2. It is a flowchart which shows the operation of the information processing apparatus 2 of Embodiment 2. It is a functional block diagram of the information processing apparatus 3 of Embodiment 3. It is a functional block diagram of the processing unit 38 of Embodiment 3.
- FIG. 1 is a functional block diagram of the information processing device 1 of the first embodiment.
- the information processing device 1 of the first embodiment is used, for example, in a dialogue system. More specifically, the information processing apparatus 1 of the first embodiment outputs an answer sentence (hereinafter, answer sentence) KB for a sentence (hereinafter, referred to as "question sentence”) SB interactively input by the user. As shown in FIG. 1, the input unit 11, the conversion unit 12, the extraction unit 13, the storage unit 14, the construction unit 15, the generation unit 16, and the output unit 17 are included.
- the "conversion unit 12" corresponds to the "acquisition unit” and the “conversion unit”
- the "extraction unit 13” corresponds to the "extraction unit”
- the "construction unit 15” corresponds to the "construction unit”.
- the "generation unit 16" corresponds to the "generation unit”.
- conversion unit 12 corresponds to the "construction unit”
- extraction unit 13 corresponds to the "construction unit”
- generation unit 16 corresponds to the "generation unit”.
- the "question text” corresponds to the "question text”
- the "answer text” corresponds to the "answer text”.
- the question text SB and the answer text KB are written in natural language.
- Question sentence SB and answer sentence KB are broad concepts including sentences.
- the question sentence SB and the answer sentence KB may be composed of a plurality of sentences or may be composed of one sentence.
- logical formulas such as “resign (Steve, Company ABC)” are atomic formulas.
- the relationship between two atomic formulas such as “resign (X, Y) ⁇ get (X, money)", that is, the relationship with "the former is the latter” is an inference rule.
- the former is the "hypothesis part” and the latter is the "result part”.
- the input unit 11 is used by a user (not shown) to input a question sentence SB into the information processing device 1.
- the question sentence SB is, for example, an affirmative sentence "Steve resigns Company ABC.” For asking a question related to "Steve resigns from company ABC”.
- the conversion unit 12 converts the question sentence SB written in natural language into the search logical formula KR under the predicate logic.
- the conversion unit 12 converts, for example, the question sentence SB “Steve resigns Company ABC.” Acquired via the input unit 11 into the search logical formula KR “resign (Steve, Company ABC)”.
- the search logical formula KR is an atomic logical formula used by the extraction unit 13 to search the storage unit 14.
- the extraction unit 13 searches the storage unit 14 based on the search logical formula KR. As a result, the extraction unit 13 extracts the inference rule SK related to the search logical expression KR.
- the extraction unit 13 extracts, for example, the inference rule SK “resign (X, Y) ⁇ get (X, money)” related to the search logical expression KR “resign (Steve, Company ABC)”.
- the inference engine SE is used for the extraction unit 13 to perform the search and extraction in the storage unit 14.
- the extraction unit 13 also collates the search logical formula KR with the inference rule SK to refer to the solution of the variable common between the search logical formula KR and the inference rule SK (hereinafter, the solution of the variable is simply referred to as a “solution”. ) Is derived.
- the extraction unit 13 has, for example, the search logical expression KR "resign (Steve, Company ABC)" and the hypothetical unit “resign (X, Y)" of the inference rule SK "resign (X, Y) ⁇ get (X, money)".
- the storage unit 14 stores the inference rule SK in advance as knowledge.
- the construction unit 15 constructs the output logical formula SR from the inference rule SK.
- the output logical formula SR is an atomic logical formula used by the generation unit 16 to generate the answer sentence KB written in natural language.
- the generation unit 16 generates the answer sentence KB in natural language from the output formula SR described in the predicate logic.
- the generation unit 16 generates the answer sentence KB “Steve gets money.” Based on the output logical formula SR “get (Steve, money)”, for example.
- the output unit 17 outputs the answer sentence KB to the user.
- the output unit 17 outputs, for example, the answer sentence KB “Steve gets money.” To the user.
- FIG. 2 shows the configuration of the information processing device 1 of the first embodiment.
- the information processing device 1 includes an input unit N, a processor P, an output unit S, a memory M, and a storage medium K.
- the input unit N is composed of, for example, a microphone, a keyboard, a mouse, and a touch panel.
- the processor P is, for example, a CPU (Central Processing Unit), which is the core of a well-known computer that operates hardware according to software.
- the output unit S is composed of, for example, a speaker, a liquid crystal monitor, and a printer.
- the memory M is composed of, for example, a DRAM (Dynamic Random Access Memory) and a SRAM (Static Random Access Memory).
- the storage medium K is composed of, for example, a hard disk drive (HDD: Hard Disk Drive), a solid state drive (SSD: Solid State Drive), and a ROM (Read Only Memory).
- the storage medium K stores the program PR and the database DB.
- the program PR is a group of instructions that defines the content of processing to be executed by the processor P.
- the database DB stores, for example, the above-mentioned inference rule SK.
- the processor P is stored in the storage medium K while the input unit N and the output unit S communicate with the user as the input unit 11 and the output unit 17.
- the functions of the conversion unit 12 to the generation unit 16 are realized.
- FIG. 3 is a block diagram showing the operation of the information processing device 1 of the first embodiment.
- FIG. 4 is a flowchart showing the operation of the information processing apparatus 1 of the first embodiment.
- Step ST11 The input unit 11, which is the input unit N of the microphone, keyboard, etc., receives the input of the question sentence SB "Steve resigns Company ABC.” From the user.
- Step ST12 When the question sentence SB "Steve resigns Company ABC.” Is input in step ST11, the input unit 11 delivers the question sentence SB "Steve resigns Company ABC.” To the conversion unit 12.
- Step ST13 In step ST12, when the question sentence SB "Steve resigns Company ABC.” Is delivered from the conversion unit 12, the conversion unit 12, which is the processor P, receives the question sentence SB "Steve resigns Company ABC.” Under the predicate logic. Is converted to the search logical expression KR “resign (Steve, Company ABC)". The conversion unit 12 outputs the search logical formula KR "resign (Steve, Company ABC)" to the extraction unit 13.
- Step ST14 In step ST13, when the search logical formula KR “resign (Steve, CompanyABC)" is output from the conversion unit 12, the extraction unit 13 which is the processor P has the search logical formula KR “resign (Steve, CompanyABC)". , The storage unit 14 which is the storage medium K is searched. As a result, the extraction unit 13 extracts the inference rule SK “resign (X, Y) ⁇ get (X, money)” related to the search logical expression KR “resign (Steve, Company ABC)” by the inference engine SE.
- Step ST15 When the inference rule SK “resign (X, Y) ⁇ get (X, money)” is extracted in step ST14, the extraction unit 13 also uses the search logical expression KR “resign (Steve, Company ABC)”.
- the solution AN “X Steve” is derived by collating with the hypothetical part “resign (X, Y)" of the inference rule SK “resign (X, Y) ⁇ get (X, money)”.
- Step ST17 In step ST16, when the output logical formula SR “get (Steve, money)” is output from the construction unit 15, the generation unit 16 which is the processor P outputs the output logical formula SR “get (Steve, money)”. Generate the answer sentence KB "Steve gets money.” The generation unit 16 outputs the answer sentence KB “Steve gets money.” To the output unit 17.
- Step ST18 In step ST17, when the answer sentence KB “Steve gets money.” Is output from the generation unit 16, the output unit 17, which is the output unit S of the speaker, the LCD monitor, etc., outputs the answer sentence KB “Steve gets money.” ".” Is output to the user.
- the conversion unit 12 converts the question sentence SB “Steve resigns Company ABC.” Input from the input unit 11 into the search logical formula KR “resign (Steve, Company ABC)”.
- the extraction unit 13 searches the storage unit 14 in which the inference rule SK is accumulated based on the search logical formula KR "resign (Steve, Company ABC)" to obtain the search logical formula KR "resign (Steve, Company ABC)".
- the generation unit 16 generates the answer sentence KB “Steve gets money.” From the output logical formula SR “get (Steve, money)”.
- the output unit 17 outputs the answer sentence KB “Steve gets money.” To the user.
- the information processing device 1 needs to obtain the answer sentence KB from the question sentence SB from the user, and more specifically, "resign" in the question sentence SB. (resign) "related information without asking the user to get” money, person, goods? “,” Get or lose? “, Etc. It is possible to give the answer sentence KB to the user.
- the information processing device 1 of the first embodiment can be used in, for example, a machine translation system or a net search system instead of the above-mentioned dialogue system.
- the inference rule SK stored in the storage unit 14 may be added and updated at any time in addition to being stored in advance.
- the storage unit 14 may be constructed in a form connected to the Internet (for example, a centralized management database system or a distributed management database system), or may be constructed in a form not connected to the Internet (for example, a stand-alone database device). You may.
- the recursive neural circuit "Echo State Network” is made to learn the above conversion and generation. Allows the conversion and generation described above to be programmed.
- the inference engine SE is constructed by using the language Prolog, for example, in the following documents.
- the question sentence SB is an "affirmative sentence”
- the question sentence SB is a "question sentence”.
- FIG. 5 is a functional block diagram of the information processing device 2 of the second embodiment.
- the information processing device 2 of the second embodiment has an input unit 21, a conversion unit 22, and an extraction unit 23, similarly to the information processing device 1 of the first embodiment (shown in FIG. 1). , A storage unit 24, a construction unit 25, a generation unit 26, and an output unit 27.
- the input unit 21 corresponds to the input unit 11 of the first embodiment
- the conversion unit 22 corresponds to the conversion unit 12 of the first embodiment
- the extraction unit 23 corresponds to the extraction unit 13 of the first embodiment
- the storage unit 24 corresponds to the storage unit 14 of the first embodiment
- the construction unit 25 corresponds to the construction unit 15 of the first embodiment
- the generation unit 26 corresponds to the generation unit 16 of the first embodiment and outputs.
- the unit 27 corresponds to the output unit 17 of the first embodiment.
- the configuration of the information processing device 2 of the second embodiment is the same as the configuration of the information processing device 1 of the first embodiment (shown in FIG. 2). ⁇ Operation of the second embodiment> The operation of the information processing apparatus 2 of the second embodiment will be described.
- FIG. 6 is a block diagram showing the operation of the information processing device 2 of the second embodiment.
- FIG. 7 is a flowchart showing the operation of the information processing apparatus 2 of the second embodiment.
- Step ST21 The input unit 21 accepts the input of the question sentence SB "Who resigns Company ABC?" Which is a question sentence from the user.
- Step ST22 When the question text SB "Who resigns Company ABC?" Is input in step ST21, the input unit 21 delivers the question text SB "Who resigns Company ABC?" To the conversion unit 22.
- Step ST23 In step ST22, when the question sentence SB "Who resigns Company ABC?" Is delivered from the conversion unit 22, the conversion unit 22 searches for the question sentence SB "Who resigns Company ABC?" Under the predicate logic. Convert to KR “resign (X, Company ABC)". The conversion unit 22 outputs the search logical formula KRresign (X, CompanyABC) ”to the extraction unit 23.
- the conversion unit 22 has an attribute (for example, whether or not it is an interrogative word) of a word (who, resign, etc.) in the interrogative sentence SB “Who resigns Company ABC?”. To analyze. As a result, the conversion unit 22 replaces the question word "who", that is, the word "who” indicating an unknown object with the variable "X”, so that the above-mentioned search logical formula KR "resign (X, Company ABC)" is used. Convert to.
- Step ST24 In step ST23, when the search logical expression KR “resign (X, CompanyABC)" is output from the conversion unit 22, the extraction unit 23 is based on the search logical expression KR “resign (X, CompanyABC)". The storage unit 24 is searched.
- Step ST26 In step ST25, when the output logical formula SR "resign (Steve, Company ABC)" is output from the construction unit 25, the generation unit 26 responds from the output logical formula SR "resign (Steve, Company ABC)". Generate KB “Steve resigns Company ABC.” The generation unit 26 outputs the answer sentence KB “Steve resigns Company ABC.” To the output unit 27.
- Step ST27 In step ST26, when the response sentence KB "Steve resigns Company ABC.” Is output from the generation unit 26, the output unit 27 outputs the answer sentence KB "Steve resigns Company ABC.” To the user.
- the output unit 27 When the answer sentence KB "I don't know.” Is output from the generation unit 26, the output unit 27 outputs the answer sentence KB "I don't know.” To the user in step ST27.
- the information processing device 2 of the second embodiment can be used in, for example, a machine translation system or a net search system instead of the above-mentioned dialogue system.
- the inference rule SK stored in the storage unit 24 may be added and updated at any time in addition to being stored in advance.
- the storage unit 24 may be constructed in a form connected to the Internet (for example, a centralized management database system or a distributed management database system), or may be constructed in a form not connected to the Internet (for example, a stand-alone database device). You may.
- FIG. 8 is a functional block diagram of the information processing apparatus 3 of the third embodiment.
- the information processing device 3 of the third embodiment has an input unit 31, a conversion unit 32, and an extraction unit in substantially the same manner as the information processing device 1 of the first embodiment (shown in FIG. 1). 33, a storage unit 34, a generation unit 36, and an output unit 37 are included.
- the input unit 31 corresponds to the input unit 11 of the first embodiment
- the conversion unit 32 corresponds to the conversion unit 12 of the first embodiment
- the extraction unit 33 corresponds to the extraction unit 13 of the first embodiment
- the storage unit 34 corresponds to the storage unit 14 of the first embodiment
- the generation unit 36 corresponds to the generation unit 16 of the first embodiment
- the output unit 37 corresponds to the output unit 17 of the first embodiment.
- the information processing device 3 of the third embodiment unlike the information processing device 1 of the first embodiment, does not include the construction unit 15, and on the other hand, further includes the processing unit 38.
- FIG. 9 is a functional block diagram of the processing unit 38 of the third embodiment.
- the processing unit 38 has a construction unit 38A, a calculation unit 38B, and a selection unit 38C.
- the "construction unit 38A” corresponds to the "construction unit”
- the "calculation unit 38B” corresponds to the “calculation unit”
- the "selection unit 38C” corresponds to the "selection unit”.
- the construction unit 38A generates a plurality of hypothetical logical formulas HR by substituting the solution AN obtained by collating the search logical formula KR and the inference rule SK into the plurality of inference rule SKs extracted by the extraction unit 33. ..
- the calculation unit 38B calculates a plurality of occurrence probability SPs.
- Each occurrence probability SP is, in detail, the probability that one word included in each hypothetical formula HR occurs.
- the selection unit 38C selects the hypothetical logical formula HR having the largest occurrence probability SP among the plurality of hypothetical logical formulas HR as the output logical formula SR based on the plurality of occurrence probability SPs.
- the configuration of the information processing device 3 of the third embodiment is the same as the configuration of the information processing device 1 of the first embodiment (shown in FIG. 2).
- FIG. 10 is a block diagram showing the operation of the information processing device 3 of the third embodiment.
- FIG. 11 is a flowchart showing the operation of the information processing apparatus 3 of the third embodiment.
- Step ST31 The input unit 31 accepts the input of the question sentence SB "Steve resigns Company ABC.” From the user.
- Step ST32 When the question sentence SB "Steve resigns Company ABC.” Is input in step ST31, the input unit 31 delivers the question sentence SB "Steve resigns Company ABC.” To the conversion unit 32.
- Step ST33 In step ST32, when the question sentence SB "Steve resigns Company ABC.” Is delivered from the conversion unit 32, the conversion unit 32 searches for the question sentence SB "Steve resigns Company ABC.” Under the predicate logic. Convert to KR "resign (Steve, Company ABC)". The conversion unit 32 outputs the search logical formula KR "Steve resigns Company ABC.” To the extraction unit 33.
- Step ST34 In step ST33, when the search logical formula KR "resign (Steve, CompanyABC)" is output from the conversion unit 32, the extraction unit 33 is based on the search logical formula KR "resign (Steve, CompanyABC)". The storage unit 34 is searched.
- the extraction unit 33 uses hypothetical reasoning to symbolize the search logic formula KR “resign (Steve, Company ABC)” and the atomic logic formula “resign (X, Y)”. ) ”Is the“ consequent part ”, and three inference rules SK are extracted.
- the three inference rules SK are the inference rule SK (1) "sick (X) ⁇ resign (X, Y)", the inference rule SK (2) “hate (X, Y) ⁇ resign (X, Y)", and The inference rule SK (3) "old (X) ⁇ resign (X, Y)".
- the extraction unit 33 includes an inference rule SK (1) “sick (X) ⁇ resign (X, Y)”, an inference rule SK (2) “hate (X, Y) ⁇ resign (X, Y)”, and an inference rule SK. (3) Output “old (X) ⁇ resign (X, Y)” to the processing unit 38.
- Step ST35 In step ST34, the inference rule SK (1) “sick (X) ⁇ resign (X, Y)” and the inference rule SK (2) “hate (X, Y) ⁇ resign (X, Y)” are transmitted from the extraction unit 33.
- the construction unit 38A uses the search logical expression KR "resign (Steve, CompanyABC)".
- the construction unit 38A has the hypothetical logical formula HR (1) “sick (Steve)", the hypothetical logical formula HR (2) “hate (Steve, Company ABC”, and the hypothetical logical formula HR (3) "old (Steve)". To build.
- the construction unit 38A calculates the hypothetical logical formula HR (1) "sick (Steve)", the hypothetical logical formula HR (2) “hate (Steve, CompanyABC”", and the hypothetical logical formula HR (3) "old (Steve)". Output to.
- Step ST37 The calculation unit 38B is hypothetical logical formula HR (1) "sick (Steve)", hypothetical logical formula HR (2) “hate (Steve, Company ABC", hypothetical logical formula HR (3) "old (Steve)".
- the occurrence probability SP of the word included in is calculated.
- the calculation unit 38B has hypothetical formula HR (1) "sick (Steve)", hypothetical logical formula HR (2) “hate (Steve, Company ABC", hypothetical logical formula HR (3) “old (Steve)”. , And specifically, the occurrence probability SP of any of the subject, predicate, object, etc. included in each formula is calculated.
- the calculation unit 38B is, for example, hypothetical logical formula HR (1) "sick (Steve)", hypothetical logical formula HR (2) “hate (Steve, Company ABC”, hypothetical logical formula HR (3) “old (Steve)”.
- the occurrence probability SP of the words "sick”, “hate”, and “old” that are included predicates is calculated with reference to, for example, the number of hits as a result of searching an article or the like on the net.
- the calculation unit 38B includes the hypothetical logic formula HR (1) "sick (Steve)", the hypothetical logic formula HR (2) “hate (Steve, CompanyABC”", and the hypothetical logic formula HR (3) “old (Steve)". , The occurrence probability SP of the word “sick”, the occurrence probability SP of the word “hate”, and the occurrence probability SP of the word “old” are output to the selection unit 38C.
- Step ST38 The selection unit 38C is based on the occurrence probability SP of the word "sick”, the occurrence probability SP of the word “hate”, and the occurrence probability SP of the word “old”, and the hypothesis logic formula HR (1) "sick (Steve)”. ) ”, Hypothesis logic formula HR (2)“ hate (Steve, Company ABC ””, Hypothesis logic formula HR (3) “old (Steve)”, the one with the highest occurrence probability SP is selected. Since the occurrence probability SP of the word "old” is the largest, the selection unit 38C selects the hypothetical logic formula HR (3) "old (Steve)".
- the selection unit 38C outputs the hypothetical logical formula HR (3) "old (Steve)" to the generation unit 36 as the output logical formula SR, that is, as the output logical formula SR "old (Steve)".
- Step ST39 In step ST38, when the output logical formula SR “old (Steve)” is output from the selection unit 38C, the generation unit 36 receives the answer sentence KB “Steve is” from the output logical formula SR “old (Steve)”. "old.” Is generated. The generation unit 36 outputs the answer sentence KB “Steve is old.” To the output unit 37.
- Step ST40 When the answer sentence KB "Steve is old.” Is output from the generation unit 36 in step ST39, the output unit 37 outputs the answer sentence KB "Steve is old.” To the user.
- the extraction unit 33 extracts a plurality of inference rules SK by searching the storage unit 34 based on the search logical formula KR.
- the processing unit 38 generates a plurality of hypothetical logical expressions HR from a plurality of inference rules SK.
- the processing unit 38 also calculates the word occurrence probability SP for the plurality of hypothetical logical formulas HR.
- the processing unit 38 further selects one hypothetical logical formula HR from the plurality of hypothetical logical formulas HR as the output logical formula SR based on the occurrence probability SPs of the plurality of hypothetical logical formulas HR.
- the answer sentence is not asked back to the user in order to obtain the answer sentence KB for the question sentence SB from the user. It becomes possible to give KB to the user.
- the information processing device 3 of the third embodiment can be used in, for example, a machine translation system or a net search system instead of the above-mentioned dialogue system.
- the inference rule SK stored in the storage unit 34 may be added and updated at any time in addition to being stored in advance.
- the storage unit 34 may be constructed in a form connected to the Internet (for example, a centralized management database system or a distributed management database system), or may be constructed in a form not connected to the Internet (for example, a stand-alone database device). You may.
- the information processing apparatus of the fourth embodiment will be described.
- the information processing device of the fourth embodiment uses "hypothesis reasoning" as in the information processing device 3 of the third embodiment.
- the information processing device of the fourth embodiment uses the simultaneous occurrence probability of a plurality of words, unlike the information processing device 3 of the third embodiment that uses the word occurrence probability SP.
- FIG. 12 is a functional block diagram of the information processing device 4 of the fourth embodiment.
- the information processing device 4 of the fourth embodiment includes an input unit 41, a conversion unit 42, an extraction unit 43, a storage unit 44, and the same as the information processing device 3 of the third embodiment.
- a generation unit 46, an output unit 47, and a processing unit 48 are included.
- the input unit 41 corresponds to the input unit 31 of the third embodiment
- the conversion unit 42 corresponds to the conversion unit 32 of the third embodiment
- the extraction unit 43 corresponds to the extraction unit 33 of the third embodiment
- the storage unit 44 corresponds to the storage unit 34 of the third embodiment
- the generation unit 46 corresponds to the generation unit 36 of the third embodiment
- the output unit 47 corresponds to the output unit 37 of the third embodiment.
- the function of the processing unit 48 is partially different from the function of the processing unit 38 of the third embodiment.
- FIG. 13 is a functional block diagram of the processing unit 48 of the fourth embodiment.
- the processing unit 48 includes a construction unit 48A, a calculation unit 48B, and a selection unit 48C.
- the construction unit 48A substitutes the solution AN obtained by collating the search logical formula KR with the inference rule SK into the plurality of inference rule SKs extracted by the extraction unit 43. Generates a plurality of hypothetical formulas HR.
- the calculation unit 48B is different from the calculation unit 48B of the third embodiment, and calculates a plurality of simultaneous occurrence probability DSPs.
- Each simultaneous occurrence probability DSP is a probability that a plurality of words included in each hypothetical formula HR occur at the same time.
- the selection unit 48C outputs the hypothesis logical formula HR having the largest simultaneous occurrence probability DSP among the plurality of hypothetical logical formula HRs based on the plurality of simultaneous occurrence probability DSPs. Select as formula SR.
- the configuration of the information processing device 4 of the fourth embodiment is the same as the configuration of the information processing device 1 of the first embodiment (shown in FIG. 2).
- FIG. 14 is a block diagram showing the operation of the information processing apparatus 4 of the fourth embodiment.
- FIG. 15 is a flowchart showing the operation of the information processing apparatus 4 of the fourth embodiment.
- Step ST41 The input unit 41 accepts the input of the question sentence SB "Steve resigns Company ABC.” From the user in the same manner as in step ST31 of the third embodiment.
- Step ST42 The input unit 41 delivers the question sentence SB "Steve resigns Company ABC.” To the conversion unit 42 in the same manner as in step ST32 of the third embodiment.
- Step ST43 The conversion unit 42 converts the question sentence SB "Steve resigns Company ABC.” To the search logical formula KR “resign (Steve, Company ABC)" under the predicate logic, as in step ST33 of the third embodiment. The conversion unit 42 outputs the search logical formula KR “resign (Steve, Company ABC)” to the extraction unit 43.
- Step ST44 The extraction unit 43 searches the storage unit 34 based on the search logical formula KR "resign (Steve, Company ABC)" in the same manner as in step ST34 of the third embodiment.
- the extraction unit 43 uses hypothetical reasoning to conclude that the atomic logical formula “resign (X, Y)” symbolizing the search logical formula KR “resign (Steve, Company ABC)” is “resulted”.
- the three inference rules SK are the inference rule SK (1) “sick (X) ⁇ resign (X, Y)” and the inference rule SK (2) “hate (X, Y)” as in step ST34 of the third embodiment.
- ⁇ resign (X, Y) and the inference rule SK (3)“ old (X) ⁇ resign (X, Y) ”.
- the extraction unit 43 includes an inference rule SK (1) “sick (X) ⁇ resign (X, Y)”, an inference rule SK (2) “hate (X, Y) ⁇ resign (X, Y)”, and an inference rule SK. (3) Output “old (X) ⁇ resign (X, Y)” to the processing unit 48.
- the construction unit 48A has the hypothetical logical formula HR (1) “sick (Steve)”, the hypothetical logical formula HR (2) “hate (Steve, Company ABC”, and the hypothetical logical formula HR (3), as in step ST36 of the third embodiment. ) "Output old (Steve) to the calculation unit 38B.
- Step ST47 The calculation unit 48B is different from step ST47 of the third embodiment, and has a hypothetical logical formula HR (1) “sick (Steve)”, a hypothetical logical formula HR (2) “hate (Steve, Company ABC”, and a hypothetical logical formula. HR (3) Calculate the simultaneous occurrence probability DSP of a plurality of words included in "old (Steve)".
- the calculation unit 48B has a hypothetical formula HR (1) "sick (Steve)", a hypothetical logical formula HR (2) “hate (Steve, CompanyABC”", and a hypothetical logical formula HR (3) “old (Steve)”.
- the simultaneous occurrence probability DSP which is the probability that each of the predicates "sick", “hate”, and “old” included in each expression of "" and the argument "Steve” of each of the above expressions occur at the same time, is calculated.
- the calculation unit 48B is, for example, a hypothetical logical formula HR (1) “sick (Steve)”, a hypothetical logical formula HR (2) “hate (Steve, Company ABC”, and a hypothetical logical formula HR (3) “old (Steve)”.
- the calculation of the simultaneous occurrence probability DSP is performed with reference to, for example, writing to the SNS (Social Networking Service) by Mr. Steve himself and writing to the SNS by a third party other than Mr. Steve himself.
- the simultaneous occurrence probability DSP of the word “sick” and the word “Steve” the simultaneous occurrence probability DSP of the word “hate” and the word “Steve”
- the simultaneous occurrence probability DSP of the word “old” and the word “Steve” the simultaneous occurrence probability DSP of the word “old” and the word “Steve”
- the simultaneous occurrence probability DSP of the word “old” and the word “Steve” is the highest.
- the calculation unit 48B includes hypothetical logical formula HR (1) "sick (Steve)", hypothetical logical formula HR (2) “hate (Steve, Company ABC”, hypothetical logical formula HR (3) “old (Steve)", and The above-mentioned three simultaneous occurrence probability DSPs are output to the selection unit 48C.
- Step ST48 As in step ST38 of the third embodiment, the selection unit 48C has the hypothetical logical formula HR (1) “sick (Steve)” and the hypothetical logical formula HR (2) based on the above-mentioned three simultaneous occurrence probability DSPs. From “hate (Steve, Company ABC”, hypothetical formula HR (3) "old (Steve)", select the one with the highest simultaneous occurrence probability DSP. As described above, the word “hate” and the word “Steve” Since the simultaneous occurrence probability DSP with the above is the largest, the selection unit 48C selects the hypothetical logical formula HR (2) “hate (Steve, Company ABC””.
- the selection unit 48C outputs the hypothetical logical formula HR (2) "hate (Steve, Company ABC") as the output logical formula SR, that is, as the output logical formula SR "hate (Steve, Company ABC"" to the generation unit 46.
- Step ST49 The generation unit 46 generates the answer sentence KB “Steve hates Company ABC.” From the output logical formula SR “hate (Steve, Company ABC”” in the same manner as in step ST39 of the third embodiment.
- the generation unit 46 generates the answer sentence KB “Steve hates Company ABC.” KB “Steve hates Company ABC.” Is output to the output unit 47.
- Step ST50 The output unit 47 outputs the answer sentence KB "Steve hates Company ABC.” To the user in the same manner as in step ST40 of the third embodiment.
- the extraction unit 43 extracts a plurality of inference rules SK by searching the storage unit 44 based on the search logical formula KR as in the third embodiment.
- the processing unit 48 generates a plurality of hypothetical formulas HR from a plurality of inference rules SK as in the third embodiment.
- the processing unit 48 also calculates the simultaneous occurrence probability DSP of a plurality of words for the plurality of hypothetical logical formulas HR, unlike the third embodiment. Further, as in the third embodiment, the processing unit 48 uses one hypothetical logical formula HR as the output logical formula SR among the plurality of hypothetical logical formulas HR based on the simultaneous occurrence probability DSP of the plurality of hypothetical logical formulas HR. select.
- the information processing device 4 of the fourth embodiment like the information processing device 3 of the third embodiment, answers the question sentence SB from the user without asking the user back in order to obtain the answer sentence KB. It is possible to give the sentence KB to the user.
- the information processing device 4 of the fourth embodiment can be used in, for example, a machine translation system or a net search system instead of the above-mentioned dialogue system.
- the inference rule SK stored in the storage unit 44 may be added and updated at any time in addition to being stored in advance.
- the storage unit 44 may be constructed in a form connected to the Internet (for example, a centralized management database system or a distributed management database system), or may be constructed in a form not connected to the Internet (for example, a stand-alone database device). You may.
- the information processing device of the fifth embodiment uses "hypothesis reasoning" as in the information processing device 3 and the like of the third embodiment.
- the information processing device of the fifth embodiment is different from the information processing device 3 and the like of the third embodiment, and does not use the occurrence probability SP and the simultaneous occurrence probability DSP.
- FIG. 16 is a functional block diagram of the information processing device 5 of the fifth embodiment.
- the information processing device 5 of the fifth embodiment includes an input unit 51, a conversion unit 52, an extraction unit 53, a storage unit 54, and the same as the information processing device 3 of the third embodiment.
- a generation unit 56, an output unit 57, and a processing unit 58 are included.
- the input unit 51 corresponds to the input unit 31 of the third embodiment
- the conversion unit 52 corresponds to the conversion unit 32 of the third embodiment
- the extraction unit 53 corresponds to the extraction unit 33 of the third embodiment
- the storage unit 54 corresponds to the storage unit 34 of the third embodiment
- the generation unit 56 corresponds to the generation unit 36 of the third embodiment
- the output unit 57 corresponds to the output unit 37 of the third embodiment.
- the function of the processing unit 58 is partially different from the function of the processing unit 38 of the third embodiment.
- FIG. 17 is a functional block diagram of the processing unit 58 of the fifth embodiment.
- the processing unit 58 includes a first construction unit 58A, an inquiry unit 58B, and a second construction unit 58C.
- first construction unit 58A corresponds to the "first construction unit”
- the "inquiry unit 58B” corresponds to the “inquiry unit”
- the "second construction unit 58C” corresponds to the "second construction unit”. Corresponds to "Construction Department”.
- the first construction unit 58A substitutes one solution AN obtained by collating the search logical formula KR with the plurality of inference rules SK into the plurality of inference rules SK extracted by the extraction unit 53, thereby causing a plurality of inference rules SK. Generate the hypothetical formula HR.
- the inquiry unit 58B queries the storage unit 54 for other solution ANs for the plurality of hypothetical formulas HR.
- the second construction unit 58C constructs the output logical formula SR from the search logical formula KR, the hypothetical logical formula HR from which the other solution AN could be obtained, and the other solution AN.
- the configuration of the information processing device 5 of the fifth embodiment is the same as the configuration of the information processing device 1 of the first embodiment (shown in FIG. 2).
- FIG. 18 is a block diagram showing the operation of the information processing apparatus 5 of the fifth embodiment.
- FIG. 19 is a flowchart showing the operation of the information processing apparatus 5 of the fifth embodiment.
- Step ST51 The input unit 51 accepts the input of the question sentence SB "Who resigns Company ABC?" From the user.
- Step ST52 The input unit 51 delivers the question sentence SB "Who resigns Company ABC?" To the conversion unit 52.
- Step ST53 The conversion unit 52 converts the question sentence SB "Who resigns Company ABC?" To the search logical formula KR “resign (X, Company ABC)" under the predicate logic. The conversion unit 52 outputs the search logical formula KR “resign (X, Company ABC)" to the extraction unit 53.
- Step ST54 The extraction unit 53 searches the storage unit 54 based on the search logical formula KR "resign (X, CompanyABC)".
- the extraction unit 53 uses three inference rules SK in which the atomic logical formula “resign (X, Y)” symbolizing the search logical formula KR “resign (Steve, Company ABC)” is the “result part”. Extract.
- the three inference rules SK are the inference rule SK (1) “sick (X) ⁇ resign (X, Y)” and the inference rule SK (2) “hate (X, Y)” as in step ST34 of the third embodiment. ⁇ resign (X, Y) ”and the inference rule SK (3)“ old (X) ⁇ resign (X, Y) ”.
- the extraction unit 53 includes an inference rule SK (1) “sick (X) ⁇ resign (X, Y)”, an inference rule SK (2) “hate (X, Y) ⁇ resign (X, Y)”, and an inference rule SK. (3) Output “old (X) ⁇ resign (X, Y)” to the processing unit 58.
- Step ST55 In the processing unit 58, the first construction unit 58A has the search logical expression KR “resign (X, CompanyABC”” and the inference rule SK (1) “sick (X)” as in step ST35 of the third embodiment.
- ⁇ resign (X, Y) consequence
- the first construction unit 58A has hypothetical formula HR (1) "sick (X)”, hypothetical logical formula HR (2) “hate (X, CompanyABC”, hypothetical logical formula HR (3) "old ( X) ”is constructed.
- Step ST57 The query unit 58B uses the hypothetical formula HR (1) “sick (X)”, the hypothetical formula HR (2) “hate (X, CompanyABC”, and the hypothetical formula HR (3) “old (X)”.
- the storage unit 54 is inquired about the existence or nonexistence of the specific contents of the variable "X" of the three formulas and the contents.
- the storage unit 54 has only the knowledge about the hypothetical logical formula HR (3) "old (X)", and on the other hand, the knowledge about the hypothetical logical formula HR (1) "sick (X)” and the hypothesis. It is assumed that the person does not have knowledge about the logical formula HR (2) "hate (X, Company ABC”.
- the second construction unit 58C constructs the output logical formula SR (1) “resign (Steve, Company ABC)” and the output logical formula SR (2) “old (Steve)”.
- the second construction unit 58C outputs the output logical formula SR (1) "resign (Steve, CompanyABC)" and the output logical formula SR (2) “old (Steve)" to the generation unit 56.
- Step ST59 The generation unit 56 receives the answer sentence KB (1) "Steve resigns” from the output logical formula SR (1) “resign (Steve, Company ABC)” and the output logical formula SR (2) “old (Steve)". "Company ABC.” And the answer sentence KB (2) “Steve is old.” Are generated.
- the generation unit 56 outputs the answer sentence KB (1) “Steve resigns Company ABC.” And the answer sentence KB (2) “Steve is old.” To the output unit 57.
- the generation unit 56 receives the answer sentence KB (1) from the output logical formula SR (1) “resign (Steve, Company ABC)” which is the answer corresponding to the search logical formula KR “resign (X, Company ABC)” which is the question. ) After generating to "Steve resigns Company ABC.”, Generate from the output logical formula SR (2) "old (Steve)" to the answer sentence KB (2) "Steve is old.”.
- Step ST60 The output unit 57 outputs the answer sentence KB (1) "Steve resigns Company ABC.” And the answer sentence KB (2) "Steve is old.” To the user.
- the extraction unit 53 searches the storage unit 54 based on the search logical formula KR in the same manner as the information processing device 3 of the third embodiment, thereby performing a plurality of inference rules. Extract SK.
- the processing unit 58 generates a plurality of hypothetical logical expressions HR from a plurality of inference rules SK as in the third embodiment.
- the processing unit 58 also inquires the storage unit 54 of the solution AN, which is a specific content, for the variable of the hypothetical logical formula HR.
- the processing unit 58 further constructs an output logical formula SR from the search logical formula KR, the hypothetical logical formula HR, and the solution AN.
- the information processing device 5 of the fifth embodiment reaches the output logic expression SR from the question sentence SB in order to obtain the answer sentence KB for the question sentence SB from the user, similarly to the information processing device 3 and the like of the third embodiment. It is possible to give the answer sentence KB to the user without asking the user for the information necessary for the purpose.
- the information processing device 5 of the fifth embodiment can be used in, for example, a machine translation system or a net search system instead of the above-mentioned dialogue system.
- the inference rule SK stored in the storage unit 54 may be added and updated at any time in addition to being stored in advance.
- the storage unit 54 may be constructed in a form connected to the Internet (for example, a centralized management database system or a distributed management database system), or may be constructed in a form not connected to the Internet (for example, a stand-alone database device). You may.
- the information processing device of the sixth embodiment uses "hypothesis reasoning" as in the information processing device 3 and the like of the third embodiment.
- the information processing device of the sixth embodiment uses “vectorization" unlike the information processing device 3 and the like of the third embodiment.
- FIG. 20 is a functional block diagram of the information processing device 6 of the sixth embodiment.
- the information processing device 6 of the sixth embodiment includes an input unit 61, a conversion unit 62, an extraction unit 63, a storage unit 64, and the same as the information processing device 3 of the third embodiment.
- a generation unit 66, an output unit 67, and a processing unit 68 are included.
- the input unit 61 corresponds to the input unit 31 of the third embodiment
- the conversion unit 62 corresponds to the conversion unit 32 of the third embodiment
- the extraction unit 63 corresponds to the extraction unit 33 of the third embodiment
- the storage unit 64 corresponds to the storage unit 34 of the third embodiment
- the generation unit 66 corresponds to the generation unit 36 of the third embodiment
- the output unit 67 corresponds to the output unit 37 of the third embodiment.
- the function of the processing unit 68 is partially different from the function of the processing unit 38 of the third embodiment.
- FIG. 21 is a functional block diagram of the processing unit 68 of the sixth embodiment.
- the processing unit 68 includes a first construction unit 68A, a disposal unit 68B, a vectorization unit 68C, an acquisition unit 68D, a selection unit 68E, a second construction unit 68F, and the like. Has.
- first construction unit 68A corresponds to the "first construction unit”
- vehicleization unit 68C corresponds to the “vectorization unit”
- acquisition unit 68D corresponds to the "acquisition unit”.
- second construction unit 68F corresponds to the "second construction unit”.
- the first construction unit 68A substitutes one solution AN obtained by collating the search logical formula KR with the plurality of inference rules SK into the plurality of inference rules SK extracted by the extraction unit 63, thereby causing a plurality of inference rules SK. Construct the hypothetical formula HR.
- the sorting unit 68B selects one hypothetical logical formula HR from among the plurality of hypothetical logical formulas HR based on the amount of deficiency of the plurality of hypothetical logical formulas HR, and discards the other hypothetical logical formula HR.
- the vectorization unit 68C vectorizes one hypothetical formula HR.
- the acquisition unit 68D acquires a plurality of associative logical formulas RR in which the distance between the vectors is within a predetermined distance range based on one vectorized hypothetical logical formula HR.
- the selection unit 68E selects one associative logical formula RR from among a plurality of associative logical formulas RR, for example, under a unique condition (for example, whether or not the proper nouns match).
- the second construction unit 68F constructs the output logical formula SR by substituting another solution AN obtained by collating the search logical formula KR with the associative logical formula RR into one associative logical formula RR.
- the configuration of the information processing device 6 of the sixth embodiment is the same as the configuration of the information processing device 1 of the first embodiment (shown in FIG. 2).
- FIG. 22 is a block diagram showing the operation of the information processing apparatus 6 of the sixth embodiment.
- 23 and 24 are flowcharts showing the operation of the information processing apparatus 6 of the sixth embodiment.
- Step ST61 The input unit 61 accepts the input of the question sentence SB "Who resigns Company ABC?" From the user.
- Step ST62 The input unit 61 delivers the question text SB "Who resigns Company ABC?" To the conversion unit 62.
- Step ST63 The conversion unit 62 converts the question sentence SB "Who resigns Company ABC?" To the search logical formula KR “resign (X, Company ABC)" under the predicate logic. The conversion unit 62 outputs the search logical formula KR “resign (X, Company ABC)” to the extraction unit 63.
- Step ST64 The extraction unit 63 searches the storage unit 54 based on the search logical formula KR "resign (X, CompanyABC)".
- the extraction unit 63 uses hypothetical reasoning to change the atomic logic formula “resign (X, Y)” symbolizing the search logic formula KR “resign (Steve, Company ABC)” to “resign (X, Y)”.
- the three inference rule SKs which are the "results", are extracted.
- the three inference rules SK are the inference rule SK (1) “sick (X) ⁇ resign (X, Y)” and the inference rule SK (2) “hate (X, Y)” as in step ST54 of the fifth embodiment. ⁇ resign (X, Y) ”and the inference rule SK (3)“ old (X) ⁇ resign (X, Y) ”.
- the extraction unit 63 includes an inference rule SK (1) “sick (X) ⁇ resign (X, Y)”, an inference rule SK (2) “hate (X, Y) ⁇ resign (X, Y)”, and an inference rule SK. (3) Output “old (X) ⁇ resign (X, Y)” to the processing unit 68.
- Step ST65 In the processing unit 68, the first construction unit 68A has the search logical formula KR “resign (X, Company ABC”” and the inference rule SK (1) “sick (X)” as in step ST55 of the fifth embodiment.
- ⁇ resign (X, Y) consequence
- Step ST66 The first construction unit 68A also has the inference rule SK (1), the assumption part of the inference rule SK (1) “sick (X) ⁇ resign (X, Y)”, the inference rule SK (similar to the step ST56 of the fifth embodiment). 2)
- the first construction unit 68A has the hypothetical formula HR (1) "sick (X)", the hypothetical logical formula HR (2) "hate (X, CompanyABC”", and the hypothetical logical formula HR (3) "old ( X) ”is constructed.
- Step ST67 Hypothetical logical formula HR (1) "sick (X)", hypothetical logical formula HR (2) “hate (X, CompanyABC”, hypothetical logical formula HR (3) “old (X)”
- the “deficiency amount” refers to the number of words (including variables) contained in each of the hypothetical formulas HR (1), (2), and (3) with respect to the number of words. The ratio of the number of variables.
- the discarding unit 68B calculates that the missing amount of the hypothetical logical formula HR (1) “sick (X)” is “50%”.
- the missing amount of the hypothetical logical formula HR (2) “hate (X, Company ABC” is “33%”, and the hypothetical logical formula HR (3) “old (X)”.
- the amount of deficiency is calculated to be "50%”.
- Step ST68 The disposal unit 68B further includes hypothetical formula HR (1) “sick (X)”, hypothetical formula HR (2) “hate (X, Company ABC”, hypothetical formula HR (3) “old (X)”. ) ”Is discarded.
- the discarding unit 68B selects the hypothetical logical formula HR (2) “hate (X, Company ABC”” having a missing amount “33%” which is less than the threshold value “50%”, that is, outputs it to the vectorization unit 68C in the subsequent stage.
- the disposal unit 68B has a hypothetical logical formula HR (1) "sick (X)” and a hypothetical logical formula HR (3) “old (X)” having a missing amount "50%” which is equal to or higher than the threshold value "50%”. That is, it is not output to the vectorization unit 68C in the subsequent stage.
- Step ST69 The vectorization unit 68C vectorizes the hypothetical logical formula HR (2) “hate (X, Company ABC”.
- vectorization vectorization at 0/1
- word2vec vectorization at 0/1
- the hypothetical logical formula HR (2) "hate (X, Company ABC” is a word in natural language, that is, a variable. It is desirable to clearly distinguish it from words that do not contain. In order to make this distinction, for example, assuming that the length of the word after vectorization is N, N 9s (in the case of decimal numbers), N. Substitution by F (in the case of hexadecimal numbers) can be adopted.
- Step ST70 The acquisition unit 68D is associated with the vectorized hypothetical formula HR (2) "hate (X, CompanyABC”", and the associative formula RR in which the distance between the vectors is within a predetermined distance range.
- the associative formula RR associated with the hypothetical formula HR (2) “hate (X, Company ABC”" is acquired from the storage unit 64.
- the acquisition unit 68D is pre-vectorized by using "word2vec" in the storage unit 64, that is, among a plurality of associative formulas RR for which associative memory learning is performed, the vectorized hypothetical logic.
- Equation HR (2) Acquires an associative formula RR in which the distance between the vectorized one associative formula RR and “hate (X, Company ABC” is within a predetermined distance range.
- the acquisition unit 68D is, for example, an associative logical expression RR (1) "dislike (Steve, CompanyABC)", an associative logical expression RR (2) “hate (Tom, CompanyDEF)", and an associative logical expression RR (3) “love (Steve)”. , CompanyDEF) ”.
- Step ST71 The selection unit 68E has an associative formula RR (1) "dislike (Steve, CompanyABC)", an associative formula RR (2) “hate (Tom, CompanyDEF)", and an associative formula RR (3) “love ( Select one of "Steve, Company DEF)".
- the selection unit 68E is from the viewpoint of whether or not the proper nouns (for example, personal name and company name) match in relation to the hypothetical logical formula HR (2) "hate (X, CompanyABC)".
- Associative formula RR (1) Select "dislike (Steve, Company ABC)".
- the logical formula SR (1) "resign (Steve, Company ABC" is constructed.
- Step ST74 The second construction unit 68F uses the associative formula RR (1) “dislike (Steve, Company ABC)” word “dislike” and the hypothetical formula HR (2) “hate (X, Company ABC)” word. Replace with “hate”. As a result, the second construction unit 68F constructs the output logical formula SR, that is, the output logical formula SR (2) “hate (Steve, Company ABC)”.
- the second construction unit 68F outputs the output logical formula SR (1) "resign (Steve, Company ABC)" and the output logical formula SR (2) “hate (Steve, Company ABC)" to the generation unit 66.
- Step ST75 The generation unit 66 receives the answer sentences KB (1) "Steve” from the output logical formula SR (1) “resign (Steve, CompanyABC)” and the output logical formula SR (2) “hate (Steve, CompanyABC)". Generate “resigns Company ABC.” And answer text KB (2) “Steve hates Company ABC.”.
- the generation unit 66 outputs the answer sentence KB (1) "Steve resigns Company ABC.” And the answer sentence KB (2) "Steve hates Company ABC.” To the output unit 67.
- the output logical formula SR (1) “resign (Steve, CompanyABC)” is the specific content of the search logical formula KR “resign (X, CompanyABC)", that is, the corresponding answer.
- the output logical formula SR (2) “hate (Steve, Company ABC)” is a supplement to the output logical formula SR (1) "resign (Steve, Company ABC)". Therefore, the generation unit 66 first generates the answer sentence KB (1) "Steve resigns Company ABC.”, And then generates the answer sentence KB (2) "Steve hates Company ABC.”.
- Step ST76 The output unit 67 outputs the answer sentence KB (1) "Steve resigns Company ABC.” And the answer sentence KB (2) "Steve hates Company ABC.” To the user.
- the extraction unit 63 searches the storage unit 64 based on the search logical formula KR as in the information processing device 3 of the third embodiment, thereby performing a plurality of inference rules. Extract SK.
- the processing unit 68 also vectorizes one hypothetical formula HR whose missing amount does not exceed the threshold among the plurality of hypothetical formula HR constructed from the plurality of inference rules SK.
- the processing unit 68 further acquires the associative logical formula RR associated with the vectorized one hypothetical logical formula HR.
- the information processing device 6 of the sixth embodiment reaches the output logic expression SR from the question sentence SB in order to obtain the answer sentence KB for the question sentence SB from the user, similarly to the information processing device 3 and the like of the third embodiment. It is possible to give the answer sentence KB to the user without asking the user for the information necessary for the purpose.
- the information processing apparatus 6 of the sixth embodiment can be used in, for example, a machine translation system or a net search system instead of the above-mentioned dialogue system.
- the inference rule SK stored in the storage unit 64 may be added and updated at any time in addition to being stored in advance.
- the storage unit 64 may be constructed in a form connected to the Internet (for example, a centralized management database system or a distributed management database system), or may be constructed in a form not connected to the Internet (for example, a stand-alone database device). You may.
- 1 information processing device 11 input unit, 12 conversion unit, 13 extraction unit, 14 storage unit, 15 construction unit, 16 generation unit, 17 output unit, 2 information processing device, 21 input unit, 22 conversion unit, 23 extraction unit, 24 storage unit, 25 construction unit, 26 generation unit, 27 output unit, 3 information processing device, 31 input unit, 32 conversion unit, 33 extraction unit, 34 storage unit, 36 generation unit, 37 output unit, 38 processing unit, 38A Construction unit, 38B calculation unit, 38C selection unit, 4 information processing device, 41 input unit, 42 conversion unit, 43 extraction unit, 44 storage unit, 46 generation unit, 47 output unit, 48 processing unit, 48A construction unit, 48B calculation Unit, 48C selection unit, 5 information processing device, 51 input unit, 52 conversion unit, 53 extraction unit, 54 storage unit, 56 generation unit, 57 output unit, 58 processing unit, 58A first construction unit, 58B inquiry unit, 58C 2nd construction unit, 6 information processing device, 61 input unit, 62 conversion unit, 63 extraction unit, 64 storage unit, 66 generation unit, 67 output unit, 68 processing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Databases & Information Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Machine Translation (AREA)
Abstract
Description
実施形態1の情報処理装置1について説明する。
〈実施形態1の構成〉
図1は、実施形態1の情報処理装置1の機能ブロック図である。
実施形態1の情報処理装置1の動作について説明する。
検索論理式KR:resign(Steve, CompanyABC)
推論規則SK:resign(X, Y)→get(X, money)
解AN:X=Steve
出力論理式SR:get(Steve, money)
回答文章KB:Steve gets money.
実施形態1の情報処理装置1によれば、変換部12は、入力部11から入力された質問文章SB「Steve resigns CompanyABC.」を検索論理式KR「resign(Steve, CompanyABC)」に変換する。抽出部13は、検索論理式KR「resign(Steve, CompanyABC)」に基づき、推論規則SKが蓄積されている蓄積部14を検索することにより、検索論理式KR「resign(Steve, CompanyABC)」に関連する推論規則SK「resign(X, Y)→get(X, money)」を抽出し、推論規則SK「resign(X, Y)→get(X, money)」及び解AN「X=Steve」を構築部15に受け渡す。構築部15は、推論規則SK「resign(X, Y)→get(X, money)」及び解AN「X=Steve」から出力論理式SR「get(Steve, money)」を構築する。生成部16は、出力論理式SR「get(Steve, money)」から回答文章KB「Steve gets money.」を生成する。出力部17は、回答文章KB「Steve gets money.」をユーザへ出力する。
実施形態1の情報処理装置1は、上記した対話システムに代えて、例えば、機械翻訳システム、ネット検索システムでも用いることが可能である。
変換部12及び生成部16での、自然言語及び述語論理間での変換及び生成には、以下の文献に記載された技術を用いることができる。
Hinaut, X. et al., “Cortico-Striatal Response Selection in Sentence Production: Insights from neural network simulation with Reservoir Computing.”, Brain and Language, vol. 150, Nov. 2015, pp. 54-68.
US Patent 8180758, “Data management system utilizing predicate logic”, Amazon Technologies, Inc.
実施形態2の情報処理装置について説明する。実施形態1では質問文章SBが、「肯定文」であることと相違して、実施形態2では、質問文章SBが、「疑問文」である。
図5は、実施形態2の情報処理装置2の機能ブロック図である。
〈実施形態2の動作〉
実施形態2の情報処理装置2の動作について説明する。
検索論理式KR:resign(X, CompanyABC)
解AN:X=Steve
出力論理式SR:resign(Steve, CompanyABC)
回答文章KB:Steve resigns CompanyABC.
実施形態2の情報処理装置2によれば、質問文章SBが「疑問文」であっても、質問文章SBが「肯定文」である実施形態1の情報処理装置1と同様に、ユーザからの質問文章SBについて回答文章KBを得るべく、ユーザに問い返すということを行うことなく、回答文章KBをユーザに与えることが可能となる。
上記したステップST24で、抽出部23は、検索論理式KR「resign(X, CompanyABC)」に関連する具体的な内容の解AN(例えば、上記した解AN(X=Steve))に代えて、具体的な内容の解ANが存在しない旨を示す解AN「X=0」を抽出する。抽出部23は、検索論理式KR「resign(X, CompanyABC)」、及び、解AN「X=0」を、構築部25を経由して、生成部26へ出力する。
実施形態2の情報処理装置2は、上記した対話システムに代えて、例えば、機械翻訳システム、ネット検索システムでも用いることが可能である。
実施形態3の情報処理装置について説明する。実施形態3の情報処理装置では、実施形態1の情報処理装置1及び実施形態2の情報処理装置2と相違して、「仮説推論」を行う。
〈実施形態3の構成〉
図8は、実施形態3の情報処理装置3の機能ブロック図である。
実施形態3の情報処理装置3の動作について説明する。
検索論理式KR:resign(Steve, CompanyABC)
推論規則SK(1):sick(X)→resign(X, Y)
推論規則SK(2):hate(X, Y)→resign(X, Y)
推論規則SK(3):old(X)→resign(X, Y)
解AN(1):X=Steve
解AN(2):Y=CompanyABC
仮説論理式HR(1):sick(Steve)
仮説論理式HR(2):hate(Steve, CompanyABC)
仮説論理式HR(3):old(Steve)
出力論理式SR:old(Steve)
回答文章KB:Steve is old.
実施形態3の情報処理装置3によれば、抽出部33が、検索論理式KRに基づき、蓄積部34を検索することにより、複数の推論規則SKを抽出する。
実施形態3の情報処理装置3は、上記した対話システムに代えて、例えば、機械翻訳システム、ネット検索システムでも用いることが可能である。
実施形態4の情報処理装置について説明する。実施形態4の情報処理装置は、実施形態3の情報処理装置3と同様に、「仮説推論」を用いる。他方で、実施形態4の情報処理装置は、単語の生起確率SPを用いる実施形態3の情報処理装置3と相違して、複数の単語の同時生起確率を用いる。
図12は、実施形態4の情報処理装置4の機能ブロック図である。
図14は、実施形態4の情報処理装置4の動作を示すブロック図である。図15は、実施形態4の情報処理装置4の動作を示すフローチャートである。
検索論理式KR:resign(Steve, CompanyABC)
推論規則SK(1):sick(X)→resign(X, Y)
推論規則SK(2):hate(X, Y)→resign(X, Y)
推論規則SK(3):old(X)→resign(X, Y)
解AN(1):X=Steve
解AN(2):Y=CompanyABC
仮説論理式HR(1):sick(Steve)
仮説論理式HR(2):hate(Steve, CompanyABC)
仮説論理式HR(3):old(Steve)
出力論理式SR:hate(Steve, CompanyABC)
回答文章KB:Steve hates CompanyABC.
実施形態4の情報処理装置4によれば、抽出部43が、実施形態3と同様に、検索論理式KRに基づき、蓄積部44を検索することにより、複数の推論規則SKを抽出する。
実施形態4の情報処理装置4は、上記した対話システムに代えて、例えば、機械翻訳システム、ネット検索システムでも用いることが可能である。
実施形態5の情報処理装置について説明する。
図16は、実施形態5の情報処理装置5の機能ブロック図である。
図18は、実施形態5の情報処理装置5の動作を示すブロック図である。図19は、実施形態5の情報処理装置5の動作を示すフローチャートである。
検索論理式KR:resign(X, CompanyABC)
推論規則SK(1):sick(X)→resign(X, Y)
推論規則SK(2):hate(X, Y)→resign(X, Y)
推論規則SK(3):old(X)→resign(X, Y)
解AN(1):Y=CompanyABC
仮説論理式HR(1):sick(X)
仮説論理式HR(2):hate(X, CompanyABC)
仮説論理式HR(3):old(X)
解AN(2):X=Steve
出力論理式SR(1):resign(Steve, CompanyABC)
出力論理式SR(2):old(Steve)
回答文章KB(1):Steve resigns CompanyABC.
回答文章KB(2):Steve is old.
実施形態5の情報処理装置5によれば、抽出部53が、実施形態3の情報処理装置3等と同様に、検索論理式KRに基づき、蓄積部54を検索することにより、複数の推論規則SKを抽出する。
実施形態5の情報処理装置5は、上記した対話システムに代えて、例えば、機械翻訳システム、ネット検索システムでも用いることが可能である。
実施形態6の情報処理装置について説明する。
図20は、実施形態6の情報処理装置6の機能ブロック図である。
図22は、実施形態6の情報処理装置6の動作を示すブロック図である。図23及び図24は、実施形態6の情報処理装置6の動作を示すフローチャートである。
検索論理式KR:resign(X, CompanyABC)
推論規則SK(1):sick(X)→resign(X, Y)
推論規則SK(2):hate(X, Y)→resign(X, Y)
推論規則SK(3):old(X)→resign(X, Y)
解AN(1):Y=CompanyABC
仮説論理式HR(1):sick(X)
仮説論理式HR(2):hate(X, CompanyABC)
仮説論理式HR(3):old(X)
連想論理式RR(1):dislike(Steve, CompanyABC)
連想論理式RR(2):hate(Tom, CompanyDEF)
連想論理式RR(3):love(Steve, CompanyDEF)
解AN(2):X=Steve
出力論理式SR(1):resign(Steve, CompanyABC)
出力論理式SR(2):hate(Steve, CompanyABC)
回答文章KB(1):Steve resigns CompanyABC.
回答文章KB(2):Steve hates CompanyABC.
実施形態6の情報処理装置6によれば、抽出部63が、実施形態3の情報処理装置3等と同様に、検索論理式KRに基づき、蓄積部64を検索することにより、複数の推論規則SKを抽出する。
実施形態6の情報処理装置6は、上記した対話システムに代えて、例えば、機械翻訳システム、ネット検索システムでも用いることが可能である。
Claims (8)
- 質問の文章が述語論理で記述された検索論理式に関連する推論規則に基づき出力論理式を構築する構築ユニットと、
前記出力論理式から回答の文章を生成する生成ユニットと、
を含むことを特徴とする情報処理装置。 - 質問の文章を取得する取得部と、
前記取得された質問の文章を、述語論理で記述された検索論理式に変換する変換部と、
前記検索論理式に関連する推論規則を抽出する抽出部と、
前記推論規則に、前記検索論理式と前記推論規則との照合により得られる解を代入することにより、出力論理式を構築する構築部と、
前記出力論理式から、回答の文章を生成する生成部と、
を含むことを特徴とする情報処理装置。 - 質問の文章を取得する取得部と、
前記取得された質問の文章を、述語論理で記述された検索論理式に変換する変換部と、
前記検索論理式を用いて、前記検索論理式に関連する解を抽出する抽出部と、
前記推論規則に前記解を代入することにより、出力論理式を構築する構築部と、
前記出力論理式から、回答の文章を生成する生成部と、
を含むことを特徴とする情報処理装置。 - 質問の文章を取得する取得部と、
前記取得された質問の文章を、述語論理で記述された検索論理式に変換する変換部と、
前記検索論理式を用いて、前記検索論理式に関連する複数の推論規則を抽出する抽出部と、
前記複数の推論規則に、前記検索論理式と前記複数の推論規則との照合により得られる解を代入することによって、複数の仮説論理式を構築する構築部と、
各生起確率が前記各推論規則に含まれる単語が生起する確率である複数の生起確率を算出する算出部と、
前記複数の生起確率に基づき、前記複数の仮説論理式のうち、最も大きい生起確率を有する仮説論理式を、出力論理式として選択する選択部と、
前記出力論理式から、回答の文章を生成する生成部と、
を含むことを特徴とする情報処理装置。 - 質問の文章を取得する取得部と、
前記取得された質問の文章を、述語論理で記述された検索論理式に変換する変換部と、
前記検索論理式を用いて、前記検索論理式に関連する複数の推論規則を抽出する抽出部と、
前記複数の推論規則に、前記検索論理式と前記複数の推論規則との照合により得られる解を代入することによって、複数の仮説論理式を構築する構築部と、
各生起確率が前記各推論規則に含まれる複数の単語が同時に生起する確率である複数の同時生起確率を算出する算出部と、
前記複数の同時生起確率に基づき、前記複数の仮説論理式のうち、最も大きい同時生起確率を有する仮説論理式を、出力論理式として選択する選択部と、
前記出力論理式から、回答の文章を生成する生成部と、
を含むことを特徴とする情報処理装置。 - 質問の文章を取得する取得部と、
前記取得された質問の文章を、述語論理で記述された検索論理式に変換する変換部と、
前記検索論理式を用いて、前記検索論理式に関連する複数の推論規則を抽出する抽出部と、
前記複数の推論規則に、前記検索論理式と前記複数の推論規則との照合により得られる一の解を代入することによって、複数の仮説論理式を構築する第1の構築部と、
前記複数の仮説論理式について他の解を問い合わせる問合せ部と、
前記検索論理式及び前記他の解を得られた仮説論理式に、前記他の解を代入することにより、出力論理式を構築する第2の構築部と、
前記出力論理式から、回答の文章を生成する生成部と、
を含むことを特徴とする情報処理装置。 - 質問の文章を取得する第1の取得部と、
前記取得された質問の文章を、述語論理で記述された検索論理式に変換する変換部と、
前記検索論理式を用いて、前記検索論理式に関連する複数の推論規則を抽出する抽出部と、
前記複数の推論規則に、前記の検索論理式と前記複数の推論規則との照合により得られる一の解を代入することにより、複数の仮説論理式を構築する第1の構築部と、
前記複数の仮説論理式のうちの一の仮説論理式をベクトル化するベクトル化部と、
前記ベクトル化された一の仮説出力論理式から連想される連想論理式を取得する第2の取得部と、
前記連想論理式に、前記検索論理式と前記連想論理式との照合により得られる他の解を代入することにより、出力論理式を構築する第2の構築部と、
前記出力論理式から、回答の文章を生成する生成部と、
を含むことを特徴とする情報処理装置。 - 構築ユニットが、質問の文章が述語論理で記述された検索論理式に関連する推論規則に基づき出力論理式を構築し、
生成ユニットが、前記出力論理式から回答の文章を生成する、
ことを特徴とする情報処理方法。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2021/001143 WO2022153470A1 (ja) | 2021-01-15 | 2021-01-15 | 情報処理装置、及び情報処理方法 |
US18/269,572 US20240062084A1 (en) | 2021-01-15 | 2021-01-15 | Information processing apparatus and information processing method |
JP2021542352A JP7019875B1 (ja) | 2021-01-15 | 2021-01-15 | 情報処理装置、及び情報処理方法 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2021/001143 WO2022153470A1 (ja) | 2021-01-15 | 2021-01-15 | 情報処理装置、及び情報処理方法 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022153470A1 true WO2022153470A1 (ja) | 2022-07-21 |
Family
ID=80929761
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2021/001143 WO2022153470A1 (ja) | 2021-01-15 | 2021-01-15 | 情報処理装置、及び情報処理方法 |
Country Status (3)
Country | Link |
---|---|
US (1) | US20240062084A1 (ja) |
JP (1) | JP7019875B1 (ja) |
WO (1) | WO2022153470A1 (ja) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH09330306A (ja) * | 1996-06-13 | 1997-12-22 | Adoin Kenkyusho:Kk | 記号的処理を可能にする非記号処理装置と、情報処理システム及びその自動コーダ |
JP2018147042A (ja) * | 2017-03-01 | 2018-09-20 | 日本電気株式会社 | 推論結果可視化装置、推論結果可視化方法、及びプログラム |
WO2019058479A1 (ja) * | 2017-09-21 | 2019-03-28 | 日本電気株式会社 | 知識取得装置、知識取得方法、及び、記録媒体 |
WO2020044415A1 (ja) * | 2018-08-27 | 2020-03-05 | 日本電気株式会社 | 仮説推論装置、仮説推論方法、及びコンピュータ読み取り可能な記録媒体 |
CN111667072A (zh) * | 2020-05-15 | 2020-09-15 | 中国电子科技集团公司电子科学研究院 | 一种情报使用价值评估方法 |
-
2021
- 2021-01-15 WO PCT/JP2021/001143 patent/WO2022153470A1/ja active Application Filing
- 2021-01-15 US US18/269,572 patent/US20240062084A1/en active Pending
- 2021-01-15 JP JP2021542352A patent/JP7019875B1/ja active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH09330306A (ja) * | 1996-06-13 | 1997-12-22 | Adoin Kenkyusho:Kk | 記号的処理を可能にする非記号処理装置と、情報処理システム及びその自動コーダ |
JP2018147042A (ja) * | 2017-03-01 | 2018-09-20 | 日本電気株式会社 | 推論結果可視化装置、推論結果可視化方法、及びプログラム |
WO2019058479A1 (ja) * | 2017-09-21 | 2019-03-28 | 日本電気株式会社 | 知識取得装置、知識取得方法、及び、記録媒体 |
WO2020044415A1 (ja) * | 2018-08-27 | 2020-03-05 | 日本電気株式会社 | 仮説推論装置、仮説推論方法、及びコンピュータ読み取り可能な記録媒体 |
CN111667072A (zh) * | 2020-05-15 | 2020-09-15 | 中国电子科技集团公司电子科学研究院 | 一种情报使用价值评估方法 |
Non-Patent Citations (2)
Title |
---|
ADAM MARBLESTONE; GREG WAYNE; KONRAD KORDING: "Towards an integration of deep learning and neuroscience", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, vol. 34, 13 June 2016 (2016-06-13), 201 Olin Library Cornell University Ithaca, NY 14853 , pages 714 - 719, XP081590532, DOI: 10.3389/fncom.2016.00094 * |
ISHIKAWA, TSUTOMU: "Generating the Well-Formed Formulas from Natural Language Sentences and the Inference Scheme", JOURNAL OF JAPANESE SOCIETY FOR ARTIFICIAL INTELLIGENCE, vol. 22, no. 5, 1 September 2007 (2007-09-01), pages 613 - 620 * |
Also Published As
Publication number | Publication date |
---|---|
US20240062084A1 (en) | 2024-02-22 |
JP7019875B1 (ja) | 2022-02-15 |
JPWO2022153470A1 (ja) | 2022-07-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US5802504A (en) | Text preparing system using knowledge base and method therefor | |
Roos et al. | Narratives in economics | |
Alani et al. | Managing reference: Ensuring referential integrity of ontologies for the semantic web | |
Ma | Evocation: analyzing and propagating a semantic link based on free word association | |
Kadhim et al. | A multi-intelligent agent architecture for knowledge extraction: novel approaches for automatic production rules extraction | |
Pethő | WHAT IS POLYSEMY?-A SURVEY OF CURRENT RESEARCH AND RESULTS¹ | |
Garg et al. | Hindi EmotionNet: A scalable emotion lexicon for sentiment classification of Hindi text | |
Zeng et al. | Automated interactive domain-specific conversational agents that understand human dialogs | |
van Loon | Three families of automated text analysis | |
WO2022153470A1 (ja) | 情報処理装置、及び情報処理方法 | |
Galitsky et al. | Chatbots for CRM and dialogue management | |
Fu et al. | Multiple interaction attention model for open-world knowledge graph completion | |
Oduro-Afriyie et al. | Knowledge graph enabled open-domain conversational question answering | |
Jain et al. | Revisiting document-level relation extraction with context-guided link prediction | |
Atkinson-Abutridy et al. | A semantically guided and domain-independent evolutionary model for knowledge discovery from texts | |
Asenbrener Katic et al. | Comparison of two versions of formalization method for text expressed knowledge | |
Kavitha et al. | A chatbot system for education nlp using deep learning | |
Guo et al. | Question answering system based on ontology and semantic web | |
Broughton | The respective roles of intellectual creativity and automation in representing diversity: human and machine generated bias | |
Danenas et al. | Enhancing the extraction of SBVR business vocabularies and business rules from UML use case diagrams with natural language processing | |
Duží | Knowing-that’vs.‘Knowing-wh | |
Harmon et al. | Revisiting computational models of creative storytelling based on imaginative recall | |
O’Shea | Natural language scripting within conversational agent design | |
Liu et al. | Generating Natural Language Responses in Robot-Mediated Referential Communication Tasks to Simulate Theory of Mind | |
Prasla et al. | Question answering system using artificial neural network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
ENP | Entry into the national phase |
Ref document number: 2021542352 Country of ref document: JP Kind code of ref document: A |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21919363 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 18269572 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 21919363 Country of ref document: EP Kind code of ref document: A1 |