Scientific exploration practice observation and evaluation method and system, electronic equipment and storage medium
Through a large language model, multi-dimensional evaluation of students' multi-modal observation records in scientific inquiry practice has solved the problem of human-reliance and large errors in the existing technology, and achieved a more accurate and efficient assessment of scientific observation ability, helping students improve their ability to improve their Scientific inquiry ability.
Patent Information
- Application Number
- CN202411169045.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2024-11-05
AI Technical Summary
In the prior art, students' observation and record evaluation in the practice of scientific inquiry mainly relies on teachers, resulting in high labor costs and the evaluation results are susceptible to human factors, and there are errors.
A large language model is used to analyze students' multimodal observation records, obtain key information to record, and conduct multi-dimensional evaluation based on this information, including correlation, scientificity, richness and comprehensiveness, to generate multi-dimensional evaluation information.
It reduces the labor cost of teachers, improves the accuracy of evaluation, fully grasps students' scientific inquiry performance, helps students improve their scientific observation ability, and provides direction for education.
Smart Images

Figure CN118898533A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a scientific inquiry practice observation evaluation method, system, electronic equipment and storage medium. Background Art
[0002] Scientific inquiry practice refers to the science of studying natural phenomena and laws based on experiments and observations. It includes physics, chemistry, biology, earth science and other fields. Its research results can be applied to many fields. Its development plays an important role in the progress and development of human society. It helps us better understand nature.
[0003] The existing scientific inquiry practice usually involves students making scientific observations and recording them. Scientific observation is the basis of scientific inquiry, and it provides first-hand information for scientific inquiry by directly perceiving and recording natural phenomena. In scientific experiments, students' scientific observation ability directly affects the accuracy and scientificity of observation records. Therefore, cultivating students' scientific observation ability is the key to improving their scientific inquiry ability.
[0004] However, in the prior art, teachers usually evaluate students' scientific observation activities based on their observation records to determine their scientific observation abilities. This not only increases the manpower costs of teachers, but the evaluation results may also be affected by human factors and have errors. Summary of the invention
[0005] The present invention provides a scientific inquiry practice observation evaluation method, system, electronic device and storage medium to solve the defects existing in the related technology.
[0006] The present invention provides a scientific inquiry practice observation evaluation method, comprising: Obtain multimodal observation records of students in the process of scientific inquiry practice; Based on a large language model, the multimodal observation record is parsed to obtain key information of the record, and based on the key information of the record, a multi-dimensional evaluation is performed on the multimodal observation record.
[0007] According to a scientific inquiry practice observation evaluation method provided by the present invention, based on the key information of the record, the multimodal observation record is evaluated in multiple dimensions, including: Based on the benchmark prompt information and the record key information, a combined prompt text is generated, wherein the benchmark prompt information includes instructions for generating multi-dimensional evaluation information and content requirement description information; the content requirement description information is used to describe multiple dimensions of the multi-dimensional evaluation information; The combined prompt text is input into the large language model to obtain multi-dimensional evaluation information of the multimodal observation record output by the large language model.
[0008] According to a scientific inquiry practice observation and evaluation method provided by the present invention, the multiple dimensions include the correlation between the multimodal observation records and the content of the scientific inquiry practice, the scientificity, richness and comprehensiveness of the multimodal observation records.
[0009] According to a scientific inquiry practice observation and evaluation method provided by the present invention, the benchmark prompt information also includes suggestion description information, and the suggestion description information is used to describe that the multi-dimensional evaluation information includes observation suggestions.
[0010] According to a scientific inquiry practice observation and evaluation method provided by the present invention, the recorded key information includes at least one of the key things, thing characteristics, thing connections, sequential structures, auxiliary tools and sensory methods in the multimodal observation records.
[0011] According to a scientific inquiry practice observation evaluation method provided by the present invention, the multimodal observation record is parsed based on a large language model to obtain key record information, including: Determining input text based on the multimodal observation record; The input text is input into the large language model, and the key information of the record is determined by the large language model.
[0012] According to a scientific inquiry practice observation and evaluation method provided by the present invention, the multimodal observation record includes at least one of text records, pictures, audio and video.
[0013] According to a scientific inquiry practice observation and evaluation method provided by the present invention, the multimodal observation record includes pictures and audio; The step of determining the input text based on the multimodal observation record comprises: Extracting the first text from the image based on optical character recognition technology, and performing content recognition on the image to obtain the second text; transcribing the audio into a third text; The first text, the second text and the third text are merged to obtain the input text.
[0014] The present invention also provides a scientific inquiry practice observation evaluation system, comprising: The record acquisition module is used to obtain students' multimodal observation records during the scientific inquiry practice process; The multi-dimensional evaluation module is used to parse the multi-modal observation record based on a large language model to obtain key information of the record, and to perform a multi-dimensional evaluation on the multi-modal observation record based on the key information of the record.
[0015] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the scientific inquiry practice observation and evaluation method as described in any one of the above is implemented.
[0016] The scientific inquiry practice observation and evaluation system provided by the present invention further includes at least one of an audio acquisition module, a video acquisition module, an image acquisition module and a text input module; The audio acquisition module is used to collect audio during the scientific inquiry practice process; The video acquisition module is used to collect videos during the scientific exploration practice; The image acquisition module is used to acquire images during the scientific exploration practice process; The text input module is used to collect the text input by the students during the scientific inquiry practice process.
[0017] The present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements the scientific inquiry practice observation and evaluation method as described in any one of the above.
[0018] The present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements the scientific inquiry practice observation and evaluation method as described in any one of the above.
[0019] The scientific inquiry practice observation evaluation method, system, electronic device and storage medium provided by the present invention first obtain the multimodal observation records of students in the process of scientific inquiry practice; then use a large language model to parse the multimodal observation records to obtain key information of the records, and based on the key information of the records, perform a multi-dimensional evaluation of the multimodal observation records. This method automates the evaluation of multimodal observation records through a large language model, which can not only reduce the teacher's labor cost, but also improve the accuracy of the evaluation information. Moreover, by performing a multi-dimensional evaluation of the multimodal observation records to obtain multi-dimensional evaluation information, the performance of students in the process of scientific inquiry practice can be fully grasped, and then the students' scientific observation ability can be judged, which is helpful for subsequent targeted assistance to students to improve their scientific observation ability, and can provide teachers with the education direction of students. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] In order to more clearly illustrate the technical solutions in the present invention or related technologies, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0021] Figure 1 It is one of the flow charts of the scientific inquiry practice observation and evaluation method provided by the present invention.
[0022] Figure 2 This is the second flow chart of the scientific inquiry practice observation and evaluation method provided by the present invention.
[0023] Figure 3 It is a structural schematic diagram of the scientific inquiry practice observation and evaluation system provided by the present invention.
[0024] Figure 4 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION
[0025] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0026] In the prior art, the observation records of students in the process of scientific inquiry practice are mainly evaluated by teachers, which not only increases the manpower cost of teachers, but also the evaluation results may be affected by human factors and have errors, thereby making wrong judgments about students. Based on this, an embodiment of the present invention provides a method for evaluating scientific inquiry practice observation.
[0027] Figure 1 Schematic diagram of a flow chart of a scientific inquiry practice observation and evaluation method provided in an embodiment of the present invention, such as Figure 1 As shown, the method includes: S1, obtaining multimodal observation records of students in the process of scientific inquiry practice; S2, based on a large language model, parsing the multimodal observation record to obtain key information of the record, and based on the key information of the record, performing a multi-dimensional evaluation on the multimodal observation record.
[0028] Specifically, the scientific inquiry practice observation evaluation method provided in the embodiment of the present invention is implemented by a scientific inquiry practice observation evaluation device, which can be configured in a client terminal, and the client terminal can include a computer, a personal computer (PC), a smart phone, a tablet computer, or other smart devices. For example, the device can be software installed in a computer or a personal computer, or an application (App) installed in a smart phone or a tablet computer, which is not specifically limited here.
[0029] First, step S1 is performed to obtain multimodal observation records of students in the process of scientific inquiry practice. The content of scientific inquiry practice may include observing and recording the growth status, physical reaction process or chemical reaction process of animals and plants throughout their life cycle. Animals may include earthworms, silkworms, cicadas, cicadas, fish, etc., and plants may include sunflowers, cabbage, carrots, green radish, garlic sprouts, bean sprouts, flowers, etc.
[0030] Multimodal observation records refer to different types of observation records generated during the scientific inquiry practice process, such as pictures, text records, recorded audio, video, etc. collected at different key nodes in the scientific inquiry practice process. That is, multimodal observation records can include at least one of text records, pictures, audio and video.
[0031] The pictures may include visualization information of plants and animals at different key nodes, and visualization information of physical and chemical reaction processes at different key nodes. Here, the visualization information may include environmental information, state information, etc.
[0032] The text record may include text description information of the observed phenomena at different key nodes.
[0033] The audio may include descriptive information of the observed phenomenon input by the student's voice.
[0034] The video can include visual information with audio output at different key points.
[0035] The multimodal observation records can be imported into the client terminal after being acquired by the user, or can be directly acquired through the acquisition module configured on the client terminal. For example, the client terminal can be configured with at least one of an audio acquisition module, a video acquisition module, an image acquisition module, and a text input module, which are respectively used to acquire audio, video, and images during the scientific inquiry practice process, and text input by students during the scientific inquiry practice process.
[0036] Then, step S2 is executed, and a large language model can be used to perform multi-dimensional evaluation on the multimodal observation records. Here, a large language model (LLM) is referred to as a large model, which refers to a natural language processing (NLP) model with a large parameter scale. The number of model parameters and / or the complexity of the model structure exceeds a preset threshold. The model processes large-scale text data during training and has the ability to understand and generate natural language. For example, a large language model can include the Spark large model, etc.
[0037] Furthermore, the large-scale language model here can be a large-scale language model in a general field, or a large-scale language model in the field of scientific inquiry practice observation obtained by fine-tuning scientific inquiry practice observation data. For example, the large-scale language model can be fine-tuned on the basis of a pre-trained large-scale language model using pre-set prompt information, multimodal observation record samples, and multi-dimensional evaluation information of the multimodal observation record samples. The embodiment of the present invention does not specifically limit this.
[0038] Here, the multimodal observation records can be input into a large language model, and the multimodal observation records can be parsed using the large language model to obtain the key information of the records. The parsing process can include converting the multimodal observation records into text information, then semantically understanding and analyzing the text information, and combining it with the content of scientific inquiry practice to obtain the key information of the records.
[0039] The key information of the record refers to the core information that is crucial to the evaluation of the multimodal observation record, and may include at least one of the key things, thing features, thing connections, sequential structures, auxiliary tools, and sensory methods in the multimodal observation record.
[0040] Key things are entities or phenomena that require special attention at different stages of the observation process. For example, they can be newly produced chemicals, environmental conditions (such as temperature, pressure, light), or certain phenomena (such as chemical reactions, physical changes).
[0041] Clarifying key things helps observers focus their attention, reduce interference from irrelevant information, and ensure the effectiveness and pertinence of observation.
[0042] Characteristics of things refer to the attributes or characteristics involved in describing key things, such as shape, size, color, texture, smell, sound, temperature, etc. Detailed records of these characteristics help to establish a comprehensive understanding of scientific inquiry practices and provide a basis for subsequent analysis and comparison. For example, when observing a flower, you may record the color, shape, number of its petals, and the structure of its stamens.
[0043] Things are connected to each other. This includes causal relationships, correlations, spatial relationships, etc. Identifying and describing these connections helps to reveal the laws behind observed phenomena and promote the development of scientific theories. For example, in ecology, observing the food chain relationship between different species is an important aspect of exploring the interactions and dependencies between them.
[0044] Sequential structure refers to the order in which key things occur during the observation process or the order in which the observation steps are arranged. Following a certain sequential structure for observation can ensure the comprehensiveness and systematicness of the observation and avoid missing important information. For example, when conducting a chemical experiment, it is necessary to add reagents, control conditions, and record reaction phenomena in a timely manner according to specific steps to ensure the accuracy and repeatability of the experimental results.
[0045] Auxiliary tools are an indispensable part of scientific observation. They can enhance students' perception and improve the accuracy and efficiency of observation. Common auxiliary tools include microscopes, telescopes, thermometers, pressure gauges, spectrometers, cameras, etc. These tools can magnify the observed object, measure physical quantities, record the change process, etc., providing strong support for scientific observation.
[0046] Sensory modalities refer to the way students receive and perceive information about an object through different sensory organs. Humans mainly obtain information through five senses: vision, hearing, smell, taste, and touch. In scientific observation, vision is the most commonly used sensory modality because it can intuitively show the form and changes of the object being observed. However, in some cases, other sensory modalities are equally important. For example, when tasting food, the combination of taste and smell can provide a more comprehensive sensory experience; when touching an object, touch can perceive its characteristics such as texture and temperature.
[0047] The key things, characteristics of things, connections between things, sequential structures, auxiliary tools and sensory methods in scientific observation practice together constitute the basic framework and elements of scientific observation. By comprehensively applying these elements, observers can fully and accurately describe the observed phenomena and provide strong support for scientific research.
[0048] Furthermore, with the help of the generation capability of large language models, we can analyze the key information of the records and conduct multi-dimensional evaluation of the multimodal observation records to obtain multi-dimensional evaluation information of the multimodal observation records. Multi-dimensional evaluation information refers to evaluation information of multiple dimensions, and each dimension can include the correlation between the multimodal observation records and the content of scientific inquiry practice, the scientificity, richness and comprehensiveness of the multimodal observation records, etc.
[0049] Correlation refers to the connection between multimodal observation records and the content of scientific inquiry practice. In scientific inquiry practice, multimodal observation records need to be highly relevant to the content of scientific inquiry practice in order to more accurately reflect the true state of the research object, thereby helping researchers discover new scientific phenomena, verify scientific hypotheses or construct scientific theories.
[0050] Scientificity refers to whether the recording process of multimodal observation records follows scientific methods and principles. Only scientificity can ensure the accuracy and reliability of multimodal observation records. This requires that multimodal observation records have clear observation purposes and hypotheses to ensure that the content of the records is consistent with the goals of scientific inquiry. Moreover, standardized methods and tools need to be used in the recording process to reduce errors and biases. For example, in video recording, high-definition cameras and stable shooting environments should be used to ensure the clarity and stability of the video images. At the same time, students are required to have relevant knowledge and skills to accurately identify and record the key features of the observed objects.
[0051] Richness refers to the amount of information and diversity of types that multimodal observation records can provide. Compared with single-modal records, multimodal observation records can capture multiple types of data at the same time, such as text descriptions, image displays, video records, audio analysis, etc. These different types of data complement each other and together constitute a comprehensive and multi-angle description of the research object. For example, in the process of observing animals and plants, students can describe the morphological characteristics of organisms in words, use images to show their appearance, then record their behavioral habits through videos, and finally judge their vocal characteristics through audio analysis. This rich information enables researchers to have a more comprehensive understanding of the research object and discover more valuable scientific problems.
[0052] Comprehensiveness refers to the breadth of coverage and depth of multimodal observation records. Comprehensiveness requires that multimodal observation records not only cover all aspects and levels of the research object, but also deeply explore its internal connections and laws. In scientific inquiry practice, students need to pay attention to multiple aspects of the research object, such as morphology, structure, function, behavior, etc., and try to reveal the interactions and relationships between them. Through in-depth analysis and exploration of multimodal observation records, some deep-level laws and mechanisms hidden behind surface phenomena can be discovered.
[0053] The scientific inquiry practice observation evaluation method provided in the embodiment of the present invention first obtains the multimodal observation records of students in the process of scientific inquiry practice; then uses a large language model to parse the multimodal observation records to obtain key information of the records, and based on the key information of the records, performs a multi-dimensional evaluation of the multimodal observation records. This method automates the evaluation of multimodal observation records through a large language model, which can not only reduce the manpower cost of teachers, but also improve the accuracy of evaluation information. Moreover, by performing a multi-dimensional evaluation of multimodal observation records to obtain multi-dimensional evaluation information, it is possible to fully grasp the performance of students in the process of scientific inquiry practice, and then judge the students' scientific observation ability, which is helpful for subsequent targeted assistance to students to improve their scientific observation ability, and can provide teachers with the education direction of students.
[0054] On the basis of the above embodiment, based on the key information of the record, the multimodal observation record is evaluated in multiple dimensions, including: Based on the benchmark prompt information and the record key information, a combined prompt text is generated, wherein the benchmark prompt information includes instructions for generating multi-dimensional evaluation information and content requirement description information; the content requirement description information is used to describe multiple dimensions of the multi-dimensional evaluation information; The combined prompt text is input into the large language model to obtain multi-dimensional evaluation information of the multimodal observation record output by the large language model.
[0055] Specifically, when performing multi-dimensional evaluation on multimodal observation records, the baseline prompt information and the record key information may be used to generate a combined prompt text, which may be a text obtained by splicing the baseline prompt information and the record key information.
[0056] The benchmark prompt information may be a prompt for a large language model for generating multi-dimensional evaluation information. The benchmark prompt information is used to specify the generation requirements of the multi-dimensional evaluation information at the input stage of the large language model, that is, the benchmark prompt information is used to instruct the large language model to generate multi-dimensional evaluation information. Therefore, the benchmark prompt information may include instructions for instructing the large language model to generate multi-dimensional evaluation information. The instructions here are in the form of natural language text, for example, "Please generate multi-dimensional evaluation information for evaluating the above multimodal observation records", or "As an elementary school science teacher, please comment on the students' scientific observations based on the uploaded input text content, based on the following four dimensions, with a word count of 100-150 words."
[0057] At the same time, the benchmark prompt information also includes content requirement description information, which is used to describe multiple dimensions of multi-dimensional evaluation information, for example, it can be "1. Relevance: Please check whether the content uploaded by students is related to the observation activities, whether it records the observed phenomena and characteristics, etc. If there are no obvious keywords in the content, you need to read and understand the text carefully to determine whether the students are doing observation activities. 2. Scientificity: Please evaluate whether the students' descriptions are in line with daily life and whether there are obvious scientific problems. Pay attention to check whether the description content is consistent with objective facts and avoid obvious misstatements. For example, do not accept obviously wrong statements such as "I am 4 meters tall" or "The earth is the center of the universe." 3. Richness: Encourage students to record diaries through a combination of pictures and texts to more vividly show their observations. 4. Comprehensiveness: Please pay attention to evaluating whether students' observations are multi-dimensional, multi-angle and planned. For lower grade students, we understand that they may not be able to be very comprehensive, so the requirements can be appropriately relaxed during the evaluation. However, students are still encouraged to observe and record as comprehensively as possible. Please comment on students' multimodal observation records based on the above four dimensions."
[0058] In addition, the content requirement description information can also be used to describe the format and style of multi-dimensional evaluation information. For example, it can be "Pay attention to a few points: first, please change the tone to the tone of speaking to students instead of speaking to me, second, do not divide the points, use a smooth paragraph to comment, and finally, pay attention to the length, you do not need to introduce the dimensions of the comment to students, just use a short paragraph (about 100 words) to tell them what is a good observation activity."
[0059] In addition, the content requirement description information can also be used to describe the output template of the multi-dimensional evaluation information. For example, it can be "Please follow the following template when outputting the final comments: xxx from x class of x year, <comment content>, looking forward to your next work!".
[0060] Thereafter, the combined prompt text is input into a large language model, which analyzes the multimodal observations and records the evaluation information of each dimension mentioned in the baseline prompt information.
[0061] The embodiment of the present invention generates a combined prompt text as an input of a large language model through benchmark prompt information and record key information, thereby realizing multi-dimensional evaluation information generation based on the large language model, and thus realizing automatic multi-dimensional evaluation information generation by calling on the excellent semantic understanding and analysis capabilities of the large language model.
[0062] Based on the above embodiment, the benchmark prompt information also includes suggestion description information, and the suggestion description information is used to describe that the multi-dimensional evaluation information includes observation suggestions.
[0063] Specifically, the benchmark prompt information includes not only instructions for generating multi-dimensional evaluation information and content requirement description information, but also suggestion description information. The suggestion description information is used to describe that the multi-dimensional evaluation information includes observation suggestions, for example, "Please conduct a multi-dimensional evaluation of the students' multimodal observation records, give positive encouragement and constructive suggestions, and help them further improve their observation and recording abilities."
[0064] Observation suggestions can be constructive advice and encouraging information used to guide students in their subsequent scientific inquiry practices. For example, they can be “You can take more clear pictures”, “To better study animal habits, you can wait for a long time to record and upload relevant information about the animals”, etc.
[0065] In the embodiment of the present invention, observation suggestions are automatically given through a large language model, which helps to improve students' scientific observation ability.
[0066] On the basis of the above embodiment, the multimodal observation record is parsed based on a large language model to obtain key information of the record, including: Determining input text based on the multimodal observation record; The input text is input into the large language model, and the key information of the record is determined by the large language model.
[0067] Specifically, in the process of determining the key information of the record, the multimodal observation record can be preprocessed first, and the multimodal observation record can be converted into text form to obtain the input text. Then, the input text is input into the large language model, and the powerful text parsing function of the large language model is used to determine the key information of the record, so that the obtained key information of the record can be made more accurate, thereby ensuring the accuracy of the multi-dimensional evaluation information finally obtained.
[0068] Based on the above embodiment, the multimodal observation record includes pictures and audio; The step of determining the input text based on the multimodal observation record comprises: Extracting the first text from the image based on optical character recognition technology, and performing content recognition on the image to obtain the second text; transcribing the audio into a third text; The first text, the second text and the third text are merged to obtain the input text.
[0069] Specifically, the multimodal observation record may include pictures and audio.
[0070] When determining the input text, for observation records other than text records, they can be converted into text types first, and then all texts can be merged to obtain the input text.
[0071] For a picture, the optical character recognition (OCR) technology can be used to extract the text in the picture as the first text, and then the content of the picture is recognized to obtain the second text of the picture content.
[0072] For audio, audio transcription technology can be used to transcribe the audio into a third text.
[0073] Afterwards, the texts are merged to obtain the input text.
[0074] If the multimodal observation record also includes video, then for the video, the audio in the video can be separated from the silent video, and the silent video can be divided into multiple video frames. Then, the audio can be transcribed into text using audio transcription technology, and the text in each video frame can be extracted using OCR technology. The content of each video frame can be recognized to obtain the corresponding text.
[0075] Afterwards, all the obtained texts are fused to obtain the input text.
[0076] In the embodiment of the present invention, each type of observation record is processed separately and finally fused to obtain the input text, which can reduce the data processing amount of the large language model.
[0077] like Figure 2 FIG. 1 is a flow chart of a scientific inquiry practice observation and evaluation method provided in an embodiment of the present invention, the method comprising: Obtain multimodal observation records of students during scientific inquiry practice.
[0078] Preprocess the multimodal observation records. For example, for images, extract the first text through OCR technology, and obtain the second text by performing content recognition on the image; for audio, obtain the third text of the audio through voice writing technology.
[0079] The text record in the multimodal observation record is fused with the first text, the second text and the third text to obtain an input text.
[0080] The input text is input into the large language model, and the input text is parsed by the large language model to obtain the key information of the record. The key information of the record includes the key things, thing features, thing connections, sequential structures, auxiliary tools and sensory methods in the multimodal observation record.
[0081] Based on the benchmark prompt information and recording key information, a combined prompt text is generated.
[0082] A large language model uses combined prompt texts to automatically generate multi-dimensional evaluation information of multimodal observation records.
[0083] like Figure 3 As shown, based on the above embodiments, a scientific inquiry practice observation and evaluation system is provided in an embodiment of the present invention, including: The record acquisition module 31 is used to acquire the multimodal observation records of students in the process of scientific inquiry practice; The multi-dimensional evaluation module 32 is used to parse the multi-modal observation record based on a large language model to obtain key information of the record, and perform a multi-dimensional evaluation on the multi-modal observation record based on the key information of the record.
[0084] Based on the above embodiment, the multi-dimensional evaluation module is specifically used for: Based on the benchmark prompt information and the record key information, a combined prompt text is generated, wherein the benchmark prompt information includes instructions for generating multi-dimensional evaluation information and content requirement description information; the content requirement description information is used to describe multiple dimensions of the multi-dimensional evaluation information; The combined prompt text is input into the large language model to obtain multi-dimensional evaluation information of the multimodal observation record output by the large language model.
[0085] Based on the above embodiment, the multiple dimensions include the correlation between the multimodal observation records and the content of the scientific inquiry practice, and the scientificity, richness and comprehensiveness of the multimodal observation records.
[0086] Based on the above embodiment, the benchmark prompt information also includes suggestion description information, and the suggestion description information is used to describe that the multi-dimensional evaluation information includes observation suggestions.
[0087] Based on the above embodiment, the recorded key information includes at least one of the key things, thing features, thing connections, sequential structures, auxiliary tools and sensory methods in the multimodal observation record.
[0088] Based on the above embodiment, the multi-dimensional evaluation module is specifically used for: Determining input text based on the multimodal observation record; The input text is input into the large language model, and the key information of the record is determined by the large language model.
[0089] Based on the above embodiment, the multimodal observation record includes at least one of text records, pictures, audio and video.
[0090] Based on the above embodiment, the multimodal observation record includes pictures and audio; The multi-dimensional evaluation module is also specifically used for: Extracting the first text from the image based on optical character recognition technology, and performing content recognition on the image to obtain the second text; transcribing the audio into a third text; The first text, the second text and the third text are merged to obtain the input text.
[0091] Specifically, the functions of each module in the scientific inquiry practice observation and evaluation system provided in the embodiment of the present invention correspond one-to-one to the operating procedures of each step in the above-mentioned method embodiment, and the effects achieved are also consistent. Please refer to the above-mentioned embodiment for details, and no further details will be given in the embodiment of the present invention.
[0092] Figure 4 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 4 As shown, the electronic device may include: a processor 410, a communication interface 420, a memory 430 and a communication bus 440, wherein the processor 410, the communication interface 420 and the memory 430 communicate with each other through the communication bus 440. The processor 410 may call the logic instructions in the memory 430 to execute the scientific inquiry practice observation and evaluation methods provided in the above embodiments.
[0093] In addition, the logic instructions in the above-mentioned memory 430 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the relevant technology or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable 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 method 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 and other media that can store program codes.
[0094] Based on the above embodiment, the electronic device further includes at least one of an audio acquisition module, a video acquisition module, an image acquisition module and a text input module; The audio acquisition module is used to collect audio during the scientific inquiry practice process; The video acquisition module is used to collect videos during the scientific exploration practice; The image acquisition module is used to acquire images during the scientific exploration practice process; The text input module is used to collect the text input by the students during the scientific inquiry practice process.
[0095] Specifically, by integrating at least one of an audio acquisition module, a video acquisition module, an image acquisition module and a text input module into an electronic device, the electronic device can have the functions of acquiring audio, video, images and text, thereby reducing the hardware cost required for multi-dimensional evaluation of students' multimodal observation records.
[0096] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the scientific inquiry practice observation and evaluation methods provided by the above-mentioned methods.
[0097] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the scientific inquiry practice observation and evaluation methods provided by the above-mentioned methods.
[0098] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.
[0099] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiment.
[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A scientific inquiry practice observation and evaluation method, characterized in that: include: Obtain multimodal observation records of students in the process of scientific inquiry practice; Based on a large language model, the multimodal observation record is parsed to obtain key information of the record, and based on the key information of the record, a multi-dimensional evaluation is performed on the multimodal observation record.
2. The scientific inquiry practice observation and evaluation method according to claim 1 is characterized in that: Based on the key information of the record, a multi-dimensional evaluation is performed on the multimodal observation record, including: Based on the benchmark prompt information and the record key information, a combined prompt text is generated, wherein the benchmark prompt information includes instructions for generating multi-dimensional evaluation information and content requirement description information; the content requirement description information is used to describe multiple dimensions of the multi-dimensional evaluation information; The combined prompt text is input into the large language model to obtain multi-dimensional evaluation information of the multimodal observation record output by the large language model.
3. The scientific inquiry practice observation and evaluation method according to claim 2 is characterized in that: The multiple dimensions include the relevance between the multimodal observation records and the content of the scientific inquiry practice, and the scientificity, richness, and comprehensiveness of the multimodal observation records.
4. The scientific inquiry practice observation and evaluation method according to claim 2 is characterized in that: The benchmark prompt information also includes suggestion description information, and the suggestion description information is used to describe that the multi-dimensional evaluation information includes observation suggestions.
5. The scientific inquiry practice observation and evaluation method according to claim 1 is characterized in that: The recorded key information includes at least one of the key things, thing features, thing connections, sequential structures, auxiliary tools and sensory methods in the multimodal observation record.
6. The scientific inquiry practice observation and evaluation method according to any one of claims 1 to 5, characterized in that: The multimodal observation record is parsed based on the large language model to obtain key information of the record, including: Determining input text based on the multimodal observation record; The input text is input into the large language model, and the key information of the record is determined by the large language model.
7. The scientific inquiry practice observation and evaluation method according to any one of claims 1 to 5, characterized in that: The multimodal observation record includes at least one of a text record, a picture, an audio, and a video.
8. The scientific inquiry practice observation and evaluation method according to claim 7 is characterized in that: The multimodal observation record includes pictures and audio; The step of determining the input text based on the multimodal observation record comprises: Extracting the first text from the image based on optical character recognition technology, and performing content recognition on the image to obtain the second text; transcribing the audio into a third text; The first text, the second text and the third text are merged to obtain the input text.
9. A scientific inquiry practice observation and evaluation system, characterized in that: include: The record acquisition module is used to obtain students' multimodal observation records during the scientific inquiry practice process; The multi-dimensional evaluation module is used to parse the multi-modal observation record based on a large language model to obtain key information of the record, and to perform a multi-dimensional evaluation on the multi-modal observation record based on the key information of the record.
10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, it implements the scientific inquiry practice observation and evaluation method as described in any one of claims 1-8.
11. The electronic device according to claim 10, characterized in that: Also includes at least one of an audio acquisition module, a video acquisition module, an image acquisition module and a text input module; The audio acquisition module is used to collect audio during the scientific inquiry practice process; The video acquisition module is used to collect videos during the scientific exploration practice; The image acquisition module is used to acquire images during the scientific exploration practice process; The text input module is used to collect the text input by the students during the scientific inquiry practice process.
12. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the scientific inquiry practice observation and evaluation method as described in any one of claims 1 to 8 is implemented.
13. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the scientific inquiry practice observation and evaluation method as described in any one of claims 1 to 8 is implemented.
Citation Information
Patent Citations
Multi-modal conference evaluation system and method based on natural language understanding
CN114299564A
College student party member continuous education tracking evaluation system and method based on multi-modal data analysis
CN116777690A
Multimodal customer service method and system based on voice and image recognition
CN117010907A
Learner input degree analysis method and system based on multi-modal large language model
CN118037103A
Situation awareness data importance scoring method and device, equipment and medium
CN118037121A
