A kind of auxiliary teaching method and device based on artificial intelligence
By integrating word segmentation, matching sentence templates, and multi-source databases, natural and coherent teaching content is generated, solving the efficiency and effectiveness issues of AI-assisted teaching and improving the efficiency and fun of instructional design and classroom implementation.
Patent Information
- Application Number
- CN202310468049.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-04-26
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2043-04-26
AI Technical Summary
Existing technologies are insufficient to effectively utilize artificial intelligence to assist teaching, and there is a lack of efficient instructional design and classroom implementation tools.
By acquiring user question text, performing word segmentation and part-of-speech tagging, matching sentence templates, identifying target keywords, combining multiple databases to obtain recommendation results, and integrating them to output teaching content.
It enables the natural and coherent text generation of multi-turn dialogues, assisting teachers in designing courses, improving teaching efficiency and interest, and enhancing students' comprehension.
Smart Images

Figure CN116662494B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of artificial intelligence teaching assistants, and in particular to an artificial intelligence-based auxiliary teaching method and device. Background Art
[0002] At present, artificial intelligence technology is widely used, and it can be applied to teachers' teaching and training work and daily teaching work. The generative AI teaching assistant system has a wide range of application scenarios in the vertical field of education. Based on artificial intelligence technology, it can help teachers design teaching, organize and implement classroom teaching, and improve teaching level.
[0003] Therefore, how to realize assisted teaching based on artificial intelligence technology is an issue worthy of study. Summary of the invention
[0004] In view of this, an embodiment of the present invention provides an artificial intelligence-based auxiliary teaching method and device, which is used to implement auxiliary teaching according to artificial intelligence technology.
[0005] An aspect of an embodiment of the present invention provides an artificial intelligence-based auxiliary teaching method, comprising:
[0006] Obtaining the user's question text and performing word segmentation on the question text to obtain a word segmentation result;
[0007] Searching for a sentence template matching the question text from a preset template library;
[0008] Identifying target keywords from the word segmentation results, and filling the sentence template according to the target keywords to obtain the question intention;
[0009] Obtaining recommendation results corresponding to the question intention from multiple databases respectively, and fusing the recommendation results;
[0010] The fusion result corresponding to the recommendation result is output so that the user can be taught according to the fusion result.
[0011] Optionally, the word segmentation of the question text to obtain a word segmentation result includes:
[0012] The question text is segmented according to the part of speech, and the part of speech of each segmented word is marked to obtain the segmentation result.
[0013] Optionally, searching a preset template library for a sentence template matching the question text includes:
[0014] According to the part of speech of each word segment, a sentence template matching the question text is searched from a preset template library.
[0015] Optionally, identifying a target keyword from the word segmentation result, and filling the sentence template according to the target keyword to obtain the question intention includes:
[0016] Identify the verbs and nouns in the word segmentation results as target keywords;
[0017] The sentence template is filled in according to the intention category corresponding to each verb and the intention category corresponding to each noun to obtain the question intention.
[0018] Optionally, obtaining recommendation results corresponding to the question intention from multiple databases respectively and fusing the recommendation results includes:
[0019] Obtaining a first recommendation result corresponding to the question intention from a professional production content database;
[0020] Acquire a second recommendation result corresponding to the question intention from a user-generated content database;
[0021] Acquire a third recommendation result corresponding to the question intention from an artificial intelligence generated content database;
[0022] A first connecting word is added between the first recommendation result and the second recommendation result, and a second connecting word is added between the second recommendation result and the third recommendation result to obtain a fusion result.
[0023] Optionally, the outputting a fusion result corresponding to the recommendation result includes:
[0024] The first recommendation result, the first connecting word, the second recommendation result, the second connecting word, and the third recommendation result are output in sequence.
[0025] Optionally, the method further comprises:
[0026] Convert the first recommendation result, the second recommendation result, and the third recommendation result into semantic embedding vectors respectively;
[0027] Calculating the cosine similarity between the semantic embedding vectors pairwise;
[0028] Marking the first recommendation result, the second recommendation result, and the third recommendation result with a preset label according to the cosine similarity;
[0029] The label is output together with the fusion result.
[0030] Optionally, the method further comprises:
[0031] Acquire the teaching dialogue generation type, teaching dialogue source content and additional conditions selected by the user for the question text, and generate a prompt described in natural language;
[0032] Calling the artificial intelligence generated content database service interface, taking the prompt as input, and obtaining the corresponding target text;
[0033] The question text and the target text are combined into a final result for output, and the target text in the final result is marked.
[0034] Optionally, the teaching dialogue generation type selected by the user for the question text includes at least one of teaching dialogue continuation, teaching dialogue supplement, teaching dialogue rewriting, teaching dialogue expansion, teaching dialogue abbreviation or teaching dialogue overview.
[0035] Another aspect of the embodiment of the present invention further provides an auxiliary teaching device based on artificial intelligence, comprising:
[0036] A question text segmentation unit is used to obtain the user's question text and segment the question text to obtain a segmentation result;
[0037] A sentence template search unit, used to search for a sentence template matching the question text from a preset template library;
[0038] A question intention determination unit, used to identify target keywords from the word segmentation results, and fill the sentence template according to the target keywords to obtain the question intention;
[0039] A recommendation result fusion unit, used to obtain recommendation results corresponding to the question intention from multiple databases respectively, and fuse the recommendation results;
[0040] The fusion result output unit is used to output the fusion result corresponding to the recommendation result so that the user can be taught according to the fusion result.
[0041] Another aspect of an embodiment of the present invention further provides an electronic device, including a processor and a memory;
[0042] The memory is used to store programs;
[0043] The processor executes the program to implement any of the above methods.
[0044] Another aspect of the embodiments of the present invention further provides a computer-readable storage medium, wherein the storage medium stores a program, and the program is executed by a processor to implement any of the above methods.
[0045] The embodiment of the present invention also discloses a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. A processor of a computer device can read the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the above method.
[0046] The present invention can identify target keywords based on the user's question text and fill in sentence templates based on the target keywords, and then obtain recommendation results corresponding to the sentence templates from multiple databases and merge and output them; since the recommendation results are highly correlated with the target keywords, the recommendation results of the present invention can help users who are teachers to design teaching, organize and implement classroom teaching, improve teaching level, etc. Based on the present invention, at least the following beneficial effects can be achieved:
[0047] 1) It can be used as a teaching aid to automatically generate natural and coherent texts through multiple rounds of dialogue to empower classroom teaching.
[0048] 2) Provide creative ideas for teachers' course design, assist in retrieving and organizing literature, and generate complete course materials such as syllabus, lesson plans and reading materials.
[0049] 3) Assist teachers in participating in teaching research and lesson preparation, conduct knowledge search, and automatically generate general and normalized teaching content, helping teachers save initial brainstorming time and improve lesson preparation efficiency.
[0050] 4) As a classroom assistant, it provides ideas for classroom activities, increases the fun and richness of the classroom, helps students understand complex content and concepts, and becomes an AI teaching assistant, after-school tutor, or debate partner for teachers.
[0051] 5) Provide rich explanations, methods, information, and ideas to help students understand knowledge more deeply and creatively, and improve their learning efficiency and interest. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0053] Figure 1 A flowchart of an artificial intelligence-based auxiliary teaching method provided by an embodiment of the present invention;
[0054] Figure 2 An example diagram of a teaching dialogue content generation part provided by an embodiment of the present invention;
[0055] Figure 3 A structural block diagram of an artificial intelligence-based auxiliary teaching device provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0056] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0057] Reference Figure 1 The embodiment of the present invention provides an artificial intelligence-based auxiliary teaching method, which specifically includes the following steps:
[0058] S100: Obtain a user's question text and perform word segmentation on the question text to obtain a word segmentation result.
[0059] Specifically, the question text is segmented according to the part of speech, and the part of speech of each segmented word is marked to obtain the segmentation result. The part of speech of the question text may include prepositions, verbs, punctuation marks, nouns, coordinating conjunctions, adverbs or approximate numbers, etc., as well as other parts of speech.
[0060] S110: searching a preset template library for a sentence template that matches the question text.
[0061] Specifically, according to the part of speech of the word segmentation marked on the question text in step S100, a sentence template matching the question text can be searched from a preset template library.
[0062] S120: Identify target keywords from the word segmentation results, and fill in the sentence template according to the target keywords to obtain the question intention.
[0063] Specifically, the verbs and nouns in the word segmentation results can be identified as target keywords; each verb and noun can be preset with a corresponding intent category, and different verbs or nouns can correspond to different intent categories. Then, the sentence template is filled according to the intent category corresponding to each verb and the intent category corresponding to each noun to obtain the question intent.
[0064] S130: Obtain recommendation results corresponding to the question intention from multiple databases respectively, and merge the recommendation results.
[0065] Specifically, the database may be a database including teaching contents, such as a PGC (Professional Generated Content) database, a UGC (User Generated Content) database, and an AIGC (AI generated content) database.
[0066] Taking the above three databases as an example, the PGC recommendation results, UGC recommendation results and AIGC recommendation results obtained therefrom are respectively used as the first recommendation result, the second recommendation result, and the third recommendation result, and then the three recommendation results are fused to obtain the fusion result.
[0067] In order to ensure the consistency of the fusion result, the fusion process may include: adding a first connecting word between the first recommendation result and the second recommendation result, and adding a second connecting word between the second recommendation result and the third recommendation result to obtain the fusion result.
[0068] S140: Outputting the fusion result corresponding to the recommendation result, so that the user can be taught according to the fusion result.
[0069] Specifically, the first recommendation result, the first connecting word, the second recommendation result, the second connecting word and the third recommendation result after merging in the above step 130 may be output in sequence.
[0070] Furthermore, in order to allow the user to understand the recommendation level of each recommendation result, the embodiment of the present invention may also include the following process:
[0071] S1. Convert the first recommendation result, the second recommendation result, and the third recommendation result into semantic embedding vectors respectively.
[0072] Specifically, the first recommendation result, the second recommendation result, and the third recommendation result may be semantically encoded respectively and converted into computable semantic embedding vectors.
[0073] S2. Calculate the cosine similarity between the semantic embedding vectors pairwise.
[0074] S3. Marking the first recommendation result, the second recommendation result, and the third recommendation result with preset labels according to the cosine similarity.
[0075] Specifically, the preset labels may include "highly recommended", "generally recommended" or "for reference only", etc., as well as other optional labels.
[0076] S4. Output the label together with the fusion result.
[0077] Specifically, outputting the labels and fusion result intentions can allow users to clearly understand the relevance of each recommended result to the question text, and flexibly adjust teaching strategies based on this to better realize AI-assisted teaching.
[0078] Furthermore, the embodiment of the present invention may also include a process of generating teaching dialogue content according to the question text, which is specifically as follows:
[0079] S1. Obtain the teaching dialogue generation type, teaching dialogue source content and additional conditions selected by the user for the question text, and generate prompts described in natural language.
[0080] Specifically, the teaching dialogue generation type selected by the user for the question text may include at least one of teaching dialogue continuation, teaching dialogue supplement, teaching dialogue rewriting, teaching dialogue expansion, teaching dialogue abbreviation or teaching dialogue overview.
[0081] S2. Call the artificial intelligence generated content database service interface, take the prompt as input, and obtain the corresponding target text.
[0082] S3: merging the question text and the target text into a final result for output, and marking the target text in the final result.
[0083] Next, the application process of the present invention will be described with specific examples.
[0084] Specifically, the embodiment of the present invention may include two parts: generating teaching recommendation content and generating teaching dialogue content.
[0085] First, the generation of teaching recommendation content is explained, which may include the following:
[0086] 1.1 Question intention recognition.
[0087] (1) Question text segmentation: The question text initiated by the user is segmented to obtain the segmentation results. For example, when the user enters the question text "When explaining the elementary school mathematics class "Fractions and Decimals", how to guide students to analyze problems and recommend some corpora", the following results can be obtained after segmentation: When explaining the elementary school mathematics class "Fractions and Decimals", how to guide students to analyze problems and recommend some corpora, where \ represents the word segmentation separator.
[0088] (2) Part-of-speech tagging: Tag each word in the word segmentation result one by one. For example, after tagging the above word segmentation result, the following result can be obtained:
[0089] In (P)\explain (VV)\"(PU)\Fractions(NN)\and(CC)\Decimals(NN)\" (PU)\primary school(NN)\math(NN)\class(NN)\,\(PU)\how(AD)\guide(VV)\students(NN)\analyze(VV)\problems(NN)\,\(PU)\and(CC)\recommend(VV)\some(CD)\corpora(NN), among which the corresponding relationships among various parts of speech are as follows: P represents preposition, VV represents verb, PU represents punctuation, NN represents noun, CC represents coordinate conjunction, AD represents adverb, and CD represents approximate number.
[0090] (3) Target keyword identification: Based on the word segmentation and part-of-speech tagging results, target keywords such as nouns and verbs are identified, extracted, and classified. For example, after performing target keyword identification on the above word segmentation and part-of-speech tagging results, the following results are obtained:
[0091] Recommended scope: Primary school (NN) mathematics (NN).
[0092] Recommended topics: Fractions (NN) Decimals (NN).
[0093] Recommended requirements: guidance (VV) analysis (VV) problem (NN).
[0094] Recommended object: Corpus (NN).
[0095] (4) Sentence template matching: Based on the preset question template library, sentence templates that match the user's question text are retrieved. For example, for the above question, the rule corresponding to the user's question "When explaining the elementary school mathematics class "Fractions and Decimals", how to guide students to analyze problems and recommend some corpus" is matched from the question template library. The sentence template is as follows:
[0096] [Topic].*? {recommend|give|given|[verb]}[Topic].*? {corpus|discourse|dialogue|[noun]}.*? [Topic].
[0097] where .*? represents any character, [] represents entity type or part of speech, {} represents keywords, and | represents or.
[0098] The user's question text is matched to the sentence template, and then the operation execution instruction corresponding to the sentence template can be executed to achieve "recommended corpus".
[0099] (5) Semantic slot filling: According to the matching results of the sentence template, the semantic slot corresponding to the intent is filled in according to the pre-defined intent category and the target keyword recognition results. For example, based on the above target keyword recognition results, different components such as subject (mathematics), stage (primary school), topic (fraction, decimal), and demand (guidance, analysis, problem) are identified, and the corresponding values are filled into the intent semantic slot, and finally the corpus recommendation behavior instruction with the constraint condition of "primary school mathematics on fraction, decimal topic and guidance analysis problem demand" is obtained.
[0100] 1.2 Multi-source data recommendation.
[0101] 1.2.1PGC recommendation.
[0102] Based on the question intention recognition results, the question requirements in the intention are searched for keyword exact matching and fuzzy matching in the PGC database to obtain PGC recommendation results, where the PGC database may include but is not limited to the "Compulsory Education Curriculum Plan and Curriculum Standards" and other professional institutions or user-written / generated content. For example, for the above example, the PGC recommendation results of "being able to perform simple decimal and fraction arithmetic operations and mixed operations, understanding the consistency of operations, and developing calculation ability and reasoning awareness" can be matched, where "decimals" and "fractions" are consistent with the recommended topics of the user's question, and "reasoning" and "analysis" are related.
[0103] 1.2.2 UGC recommendation.
[0104] Based on the question intention recognition results, the question requirements in the intention are matched with the text in the UGC database to obtain the UGC recommendation results, where the UGC database may include but is not limited to the dialogue between teachers and students in the classroom teaching field. For example, for the above example, according to the corpus recommendation behavior instruction with the constraint condition of "primary school mathematics on fractions, decimals and guiding analysis of problem requirements", the corpus content and discourse text that meet the corresponding theme can be retrieved from the UGC database.
[0105] 1.2.3AIGC recommended.
[0106] First, we call ChatGPT and other third-party service interfaces, take the user's question text as input, and get the output AIGC, i.e., the original answer. Then, we use the topic model to model AIGC and get the subject word set. We then merge it with the target keyword recognition result set of 1.1(3) to get the expanded subject word set. Finally, we divide AIGC into sentences and get the AIGC result set. We then determine whether each sentence covers any word in the subject word set. If so, we keep it. Otherwise, we remove it from the result set.
[0107] 1.3 Multi-source data fusion.
[0108] 1.3.1 Semantic relevance evaluation.
[0109] The semantic relevance of the recommendation results generated in 1.2 is evaluated, including the following:
[0110] (1) Semantically encode the PGC, UGC, and AIGC recommendation results generated in 1.2 and convert them into computable semantic embedding vectors E PGC ,E UGC ,E AIGC .
[0111] (2) The cosine similarity between the semantic embedding vectors generated in step (1) above is calculated to obtain a semantic relevance score ranging from 0 to 1, which is used to quantify the degree of semantic matching between different recommendation results.
[0112] 1.3.2 Generation of fusion results.
[0113] First, the AIGC recommendation result is output as the previous result.
[0114] Then, determine whether there is a matching result in 1.2.1. If so, output the PGC recommendation result as the mid-order result, and insert a connecting statement between the pre-order result and the mid-order result, such as "The following are the relevant curriculum standards retrieved from the system database:".
[0115] Then, continue to determine whether there is a matching result in 1.2.2. If so, output the UGC recommendation result as the post-order result, and determine whether there is an in-order result. If so, insert a connecting discourse between the in-order result and the post-order result, such as "The following are the relevant corpus and teaching discourse retrieved from the system database according to the above curriculum standards:"; if not, insert a connecting discourse between the pre-order result and the post-order result, such as "The following are the relevant corpus and teaching discourse retrieved from the system database:".
[0116] Finally, determine whether the semantic relevance score calculated in 1.3.1 is greater than the predetermined threshold. If so, continue to determine which recommended evaluation result interval it belongs to, and output the label of the corresponding result interval as the recommended result, such as "highly recommended" or "generally recommended", etc. Otherwise, it will be marked as "for reference only".
[0117] Next, the teaching dialogue content generation part is explained, which may include the following:
[0118] Reference Figure 2 The embodiment of the present invention provides an example diagram of a teaching dialogue content generation part, which can achieve the following six functions:
[0119] (1) Continuation of teaching dialogue: Given a previous teaching dialogue, expand on the original teaching dialogue content to generate a continuation of the teaching dialogue.
[0120] (2) Teaching dialogue completion: A coherent teaching dialogue is divided into three parts in order: the beginning, the middle, and the end. Given the contents of the beginning and the end of the teaching dialogue, the content of the middle part of the teaching dialogue is completed according to the logic of the teaching dialogue.
[0121] (3) Teaching dialogue rewriting: Given a teaching dialogue, modify and rewrite the old teaching dialogue content in different expressions according to the original semantics to generate new teaching dialogue content.
[0122] (4) Teaching dialogue expansion: Given a teaching dialogue, expand its expression breadth and depth based on the original semantics to generate new teaching dialogue content.
[0123] (5) Teaching dialogue abbreviation: Given a teaching dialogue, some lengthy expressions and non-core content are appropriately deleted, while retaining its key semantic information to generate new teaching dialogue content.
[0124] (6) Teaching dialogue summary: Given a teaching dialogue, summarize its content, retain the semantic information of the key parts of the teaching dialogue, and generate a condensed summary.
[0125] The specific process of generating teaching dialogue content is as follows:
[0126] (1) The user selects the type of content to be generated, i.e., the continuation, supplement, rewrite, expansion, abbreviation or summary of the above-mentioned teaching dialogue.
[0127] (2) The user selects teaching dialogue source content from the corpus, which may include the following:
[0128] ① For "Teaching Dialogue Completion", the user must first select at least one round of teaching discourse content, and finally select at least one round of teaching discourse content from the subsequent rounds of the selected round. The two selection round intervals must be separated by at least one round.
[0129] ②For other generation types, users need to select at least one round of teaching discourse content.
[0130] (3) The user selects additional conditions, for example:
[0131] ① Fill in the “Teaching Topic”, such as “Tangram”.
[0132] ②Fill in “Subject Literacy”, such as “Analysis” and “Induction”.
[0133] (4) Generate prompts described in natural language based on the generation type selected by the user, the source content of the teaching dialogue, and the additional conditions filled in.
[0134] (5) Call the AIGC service interface, use the generated prompt as an input parameter, and obtain the generated target text.
[0135] (6) The question text and the target text are combined into the final result for output, and the generated target text is highlighted.
[0136] Reference Figure 3 , an embodiment of the present invention provides an auxiliary teaching device based on artificial intelligence, comprising:
[0137] A question text segmentation unit is used to obtain the user's question text and segment the question text to obtain a segmentation result;
[0138] A sentence template search unit, used to search for a sentence template matching the question text from a preset template library;
[0139] A question intention determination unit, used to identify target keywords from the word segmentation results, and fill the sentence template according to the target keywords to obtain the question intention;
[0140] A recommendation result fusion unit, used to obtain recommendation results corresponding to the question intention from multiple databases respectively, and fuse the recommendation results;
[0141] The fusion result output unit is used to output the fusion result corresponding to the recommendation result so that the user can be taught according to the fusion result.
[0142] The embodiment of the present invention also discloses a computer program product or a computer program, wherein the computer program product or the computer program includes a computer instruction, and the computer instruction is stored in a computer-readable storage medium. A processor of a computer device can read the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes Figure 1 The method shown.
[0143] In some selectable embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the present invention is provided by way of example, for the purpose of providing a more comprehensive understanding of technology. The disclosed method is not limited to the operation and logic flow presented herein. Selectable embodiments are expected, wherein the order of various operations is changed and the sub-operation of a part for which is described as a larger operation is performed independently.
[0144] In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise specified, one or more of the functions and / or features described may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in separate physical devices or software modules. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the present invention. More specifically, in view of the properties, functions, and internal relationships of the various functional modules in the device disclosed herein, the actual implementation of the module will be understood within the conventional skills of the engineer. Therefore, those skilled in the art can implement the present invention set forth in the claims without excessive experimentation using ordinary techniques. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, which is determined by the full scope of the appended claims and their equivalents.
[0145] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.
[0146] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.
[0147] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.
[0148] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0149] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.
[0150] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.
[0151] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the described embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of the present invention.
Claims
1. An artificial intelligence-based auxiliary teaching method, characterized in that: include: Obtaining the user's question text and performing word segmentation on the question text to obtain a word segmentation result; Searching for a sentence template matching the question text from a preset template library; Identifying target keywords from the word segmentation results, and filling the sentence template according to the target keywords to obtain the question intention; Obtaining recommendation results corresponding to the question intention from multiple databases respectively, and fusing the recommendation results; The obtaining the recommendation results corresponding to the question intention from multiple databases respectively and fusing the recommendation results includes: Obtaining a first recommendation result corresponding to the question intention from a professional production content database; Acquire a second recommendation result corresponding to the question intention from a user-generated content database; Acquire a third recommendation result corresponding to the question intention from an artificial intelligence generated content database; The step of obtaining the third recommendation result includes: Input the question text into ChatGPT to obtain the original answer; use the topic model to model the original answer to obtain a subject word set; merge the subject word set with the target keyword to obtain an expanded subject word set; divide the original answer by sentence to obtain an original answer result set; determine whether the original answer result set covers any word in the expanded subject word set, and if so, retain it as the third recommendation result; Adding a first connecting word between the first recommendation result and the second recommendation result, and adding a second connecting word between the second recommendation result and the third recommendation result, to obtain a fusion result; The steps of obtaining the fusion result include: Output the third recommendation result as the preceding result; determine whether the first recommendation result is obtained, if so, output the first recommendation result as the in-order result, and insert a linking discourse between the preceding result and the in-order result; determine whether the second recommendation result is obtained, if so, output the second recommendation result as the following result, and determine whether the in-order result exists, if so, insert a linking discourse between the in-order result and the following result, and if not, insert a linking discourse between the preceding result and the following result; determine whether the semantic relevance scores of the first recommendation result, the second recommendation result, and the third recommendation result are greater than a predetermined threshold, if so, continue to determine the recommended evaluation result intervals to which the first recommendation result, the second recommendation result, and the third recommendation result belong, and output the labels of the corresponding recommended evaluation result intervals as the recommendation results, otherwise, mark the first recommendation result, the second recommendation result, and the third recommendation result as for reference only; The fusion result corresponding to the recommendation result is output so that the user can be taught according to the fusion result.
2. The artificial intelligence-based auxiliary teaching method according to claim 1, characterized in that: The word segmentation of the question text to obtain the word segmentation result includes: The question text is segmented according to the part of speech, and the part of speech of each segmented word is marked to obtain the segmentation result.
3. The artificial intelligence-based auxiliary teaching method according to claim 2, characterized in that: The step of searching a preset template library for a sentence template matching the question text includes: According to the part of speech of each word segment, a sentence template matching the question text is searched from a preset template library.
4. The artificial intelligence-based auxiliary teaching method according to claim 2, characterized in that: The step of identifying target keywords from the word segmentation results and filling the sentence template according to the target keywords to obtain the question intention includes: Identify the verbs and nouns in the word segmentation results as target keywords; The sentence template is filled in according to the intention category corresponding to each verb and the intention category corresponding to each noun to obtain the question intention.
5. The artificial intelligence-based auxiliary teaching method according to claim 1, characterized in that: The outputting the fusion result corresponding to the recommendation result includes: The first recommendation result, the first connecting word, the second recommendation result, the second connecting word, and the third recommendation result are output in sequence.
6. The artificial intelligence-based auxiliary teaching method according to claim 1, characterized in that: The method further comprises: Convert the first recommendation result, the second recommendation result, and the third recommendation result into semantic embedding vectors respectively; Calculating the cosine similarity between the semantic embedding vectors pairwise; Marking the first recommendation result, the second recommendation result, and the third recommendation result with a preset label according to the cosine similarity; The label is output together with the fusion result.
7. The artificial intelligence-based auxiliary teaching method according to claim 1, characterized in that: The method further comprises: Acquire the teaching dialogue generation type, teaching dialogue source content and additional conditions selected by the user for the question text, and generate a prompt described in natural language; Calling the artificial intelligence generated content database service interface, taking the prompt as input, and obtaining the corresponding target text; The question text and the target text are combined into a final result for output, and the target text in the final result is marked.
8. The artificial intelligence-based auxiliary teaching method according to claim 7, characterized in that: The teaching dialogue generation type selected by the user for the question text includes at least one of teaching dialogue continuation, teaching dialogue supplement, teaching dialogue rewriting, teaching dialogue expansion, teaching dialogue abbreviation or teaching dialogue summary.
9. An artificial intelligence-based auxiliary teaching device, characterized in that: include: A question text segmentation unit is used to obtain the user's question text and segment the question text to obtain a segmentation result; A sentence template search unit, used to search for a sentence template matching the question text from a preset template library; A question intention determination unit, used to identify target keywords from the word segmentation results, and fill the sentence template according to the target keywords to obtain the question intention; A recommendation result fusion unit, used to obtain recommendation results corresponding to the question intention from multiple databases respectively, and fuse the recommendation results; The obtaining the recommendation results corresponding to the question intention from multiple databases respectively and fusing the recommendation results includes: Obtaining a first recommendation result corresponding to the question intention from a professional production content database; Acquire a second recommendation result corresponding to the question intention from a user-generated content database; Acquire a third recommendation result corresponding to the question intention from an artificial intelligence generated content database; The step of obtaining the third recommendation result includes: Input the question text into ChatGPT to obtain the original answer; use the topic model to model the original answer to obtain a subject word set; merge the subject word set with the target keyword to obtain an expanded subject word set; divide the original answer by sentence to obtain an original answer result set; determine whether the original answer result set covers any word in the expanded subject word set, and if so, retain it as the third recommendation result; Adding a first connecting word between the first recommendation result and the second recommendation result, and adding a second connecting word between the second recommendation result and the third recommendation result, to obtain a fusion result; The steps of obtaining the fusion result include: Output the third recommendation result as the preceding result; determine whether the first recommendation result is obtained, if so, output the first recommendation result as the in-order result, and insert a linking discourse between the preceding result and the in-order result; determine whether the second recommendation result is obtained, if so, output the second recommendation result as the following result, and determine whether the in-order result exists, if so, insert a linking discourse between the in-order result and the following result, and if not, insert a linking discourse between the preceding result and the following result; determine whether the semantic relevance scores of the first recommendation result, the second recommendation result, and the third recommendation result are greater than a predetermined threshold, if so, continue to determine the recommended evaluation result intervals to which the first recommendation result, the second recommendation result, and the third recommendation result belong, and output the labels of the corresponding recommended evaluation result intervals as the recommendation results, otherwise, mark the first recommendation result, the second recommendation result, and the third recommendation result as for reference only; The fusion result output unit is used to output the fusion result corresponding to the recommendation result so that the user can be taught according to the fusion result.
