Method for designing biochemical experiment based on artificial intelligence and man-machine interaction system

Abstract

The invention belongs to the field of information data processing, and particularly relates to a method for designing a biochemical experiment based on artificial intelligence and a man-machine interaction system. The method comprises the steps of constructing a database; inputting an instruction by a user, identifying user intention, and outputting a preliminary scheme; prompting the user to feed back the initial scheme, if the user refuses to feed back or does not feed back, recording a user instruction and data, and taking the initial scheme as a final output scheme; and if the user feeds back, optimizing the scheme according to the user feedback, and outputting the scheme again. In the aspect of database construction, basic data is constructed by utilizing the basic change rule of biochemical substances, so that the data integrity and reliability are improved. The invention also designs a special algorithm to improve the accuracy of user intention recognition, and optimizes the generated experimental method based on user feedback and historical data, so as to finally obtain the experimental method with scientific process and strong individuation.

Description

Technical Field

[0001] The present invention belongs to the field of information data processing, and in particular relates to a method for designing biochemical experiments based on artificial intelligence and a human-computer interaction system. Background Art

[0002] Biological and chemical experiments are key links in promoting scientific progress, especially in emerging fields such as molecular biology and chemical biology, where the quality of experimental design is directly related to the success or failure of research. With the rapid development of science and technology, researchers need to conduct well-designed experiments to deeply explore life processes, reveal potential biological mechanisms, and discover new biomarkers and drug targets. However, the design and implementation of these experiments face many challenges.

[0003] In the experimental process, researchers often find it difficult to cope with complex operations, interdisciplinary integration, rapid iteration of technology, and diversity of terminology and standards. The lack of consistency and standardization in scientific literature makes experimental design more complicated, resulting in repeated attempts and failures becoming the norm. This situation not only wastes precious time and resources, but also hinders the progress of scientific research and increases the difficulty and cost of exploring new areas.

[0004] In recent years, with the rapid development of artificial intelligence, it is gradually penetrating into all aspects of biochemical experimental design, which can help improve efficiency, optimize experimental plans, promote innovation, and show great application potential.

[0005] Although artificial intelligence has made some progress in biochemical experimental design, it is still in its early stages of development and the technical maturity needs to be improved. In addition, artificial intelligence is highly dependent on data quality in biochemical experimental design. If there is noise, error or defect in the input data, the analysis results of artificial intelligence will also be affected, and may even lead to incorrect experimental design.

[0006] Therefore, although artificial intelligence has broad development prospects in biochemical experiment design, there are also many problems and challenges that need to be solved urgently. Summary of the invention

[0007] The purpose of the present invention is to provide a human-computer interaction system based on artificial intelligence for designing biochemical experiments.

[0008] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is: a method for designing biological and chemical experiments, comprising the following steps:

[0009] S1. Build a database; specifically include:

[0010] (1-1) Basic data collection and organization: With compounds and biomacromolecules as the core and the transformation relationship between compounds and / or biomacromolecules as the connection, a database is constructed;

[0011] (1-2) Basic data verification and supplement;

[0012] (1-3) Construct the data in the database into a knowledge graph, create nodes and edges, each node represents an entity (knowledge point), and each edge represents the relationship between nodes;

[0013] S2: User inputs a command, identifies the user's intention, and outputs a preliminary solution; specifically includes:

[0014] (2-1) Delete the emotional and personal descriptions in the instructions, perform semantic understanding on the remaining instructions after deletion, retrieve the closest matching results from the database constructed by S1, and generate an experimental process, named Plan A;

[0015] (2-2) semantic understanding of the remaining contents of the instruction after the emotional color and personal inclination are deleted, and online search is performed to retrieve the result with the highest matching degree from the search results with close time to obtain plan B; plan A and plan B are matched and compared, and the differences between the two plans are marked and analyzed to obtain the frequency of occurrence of the difference content. If the frequency of occurrence of the difference is greater than a set value, the difference content is replaced with the corresponding content in plan A; otherwise, the difference content of plan A is retained; after the comparison and replacement are completed, plan C is generated;

[0016] (2-3) Combine the database and complEx algorithm to adjust the experimental steps and parameters of Scheme C to generate Scheme D;

[0017] (2-4) Identify the descriptions with emotional colors and personal tendencies in the instructions, and understand the user's personalized needs for the solution; match and optimize solution D with the personalized needs, obtain solution E and output it;

[0018] S3, prompting the user to give feedback on Solution E. If the user refuses to give feedback or does not give feedback, the user's instructions and data are recorded, and Solution E is used as the final output solution;

[0019] If there is user feedback, optimize the solution based on the user feedback and output the solution again.

[0020] Preferably, in step (1-1), when collecting and arranging basic data, first retrieve and obtain disclosed compounds and biomacromolecules. For organic compounds, retrieve and obtain the change process of the group from the prior art. According to the directionality of the chemical reaction, infer all organic substances having the group by vector mapping, and further obtain the data information of all the change processes of the group corresponding to this type of organic substance. For inorganic compounds, retrieve and obtain the covalent bond change process. By vector mapping, obtain all inorganic compounds having the covalent bond, and further obtain the data information of all the reactions of this type corresponding to this type of compound. For biomacromolecules, respectively retrieve the physical changes, chemical changes and biological changes of biomacromolecules. After retrieving a certain type of change process of a certain type of substance, obtain the data of all possible changes of biomacromolecules disclosed in the prior art by vector mapping.

[0021] Preferably, step S1 also includes steps (1-4) of knowledge graph optimization, specifically including:

[0022] (1-4-1) Establish a relationship model:

[0023] Collect triple data containing entities and relations to form the basis of the knowledge graph and input them into the ComplEx model. The model generates complex vector representations for each entity and relation. The complex vector representations help the model capture the complex relationship characteristics between entities.

[0024] Use known relationships as positive samples, and simulate positive samples to generate non-existent relationships as negative samples to train the model and obtain a preliminary relationship model.

[0025] Verify the obtained relationship model, compare the potential relationships generated by the model simulation positive samples with the known relationships in the validation data set, and remove the relationships that are not in the validation data set;

[0026] (1-4-2) Expand the knowledge graph by filling in attributes:

[0027] Through the ComplEx embedding method, knowledge points and their associations are embedded into the same vector space. The geometric distance and direction between vectors reflect the relationship between knowledge points, so as to calculate and predict knowledge points and their potential associations.

[0028] By using a neural network method, the scores of potential knowledge points and relationships are calculated, thereby predicting knowledge points and relationships that may exist but have not been recorded;

[0029] Through online documents and other external data, use named entity recognition technology to identify entities (knowledge points) in online documents and other external data;

[0030] Use contextual embedding models to calculate the semantic similarity between entities identified in online documents and other external data and existing knowledge points in the knowledge graph;

[0031] The cosine similarity method is used to calculate the similarity between entities identified in online documents and other external data and existing knowledge points in the knowledge graph. If the similarity is higher than the set value, it will be integrated into the knowledge graph; the cosine similarity calculation formula is as follows:

[0032]

[0033] Among them: A i and B i denotes the i-th dimension of vectors A and B respectively, n is the total dimension of the vector, and w i is the weight of the i-th dimension, α ranges from 0.1 to 3.0, and Texis(A,B) is the semantic similarity calculated using the context embedding model.

[0034] Preferably, step S1 also includes step (1-5) model training, using the knowledge graph optimized in step (1-4) to train the relationship model constructed in step (1-4), specifically including:

[0035] (1-5-1) Based on the text understanding task in the literature, the long short-term memory network is used to take the embedding vector in the knowledge graph as part of the input layer, and concatenate it with the text features as the input of the model;

[0036] (1-5-2) During the training process, the back-propagation algorithm Adam optimizer is used to adjust the weight and bias parameters of the model and set the dynamic learning rate α t In the embedded model, the learning rate α t formula:

[0037]

[0038] Where: α max and α min are the maximum and minimum learning rates, t is the current step number, T is the total number of adjustment steps, and L i is the loss of the ith validation, L min and L max are the minimum and maximum values ​​of the loss, N is the number of recent validations, and γ is the parameter that controls the polynomial decay;

[0039] (1-5-3) After training, the performance of the model is evaluated through the confusion matrix, and the calculation indicators include accuracy, recall, F1 score and AUC. For models whose results do not meet the standards, the indicator parameters of the model are analyzed, the model training steps are traced back, γ and T in the learning rate formula are adjusted, and the model training and performance evaluation steps are repeated to retrain the model to obtain a model with satisfactory results.

[0040] Preferably, step S2 further includes step (2-0), establishing a reinforcement learning method for action recognition to semantically understand the instructions, and the establishment method includes:

[0041] (2-0-1) Match the user instruction with the content in the database constructed in step S1, and retrieve the result with the highest matching degree as the original solution;

[0042] (2-0-2) Based on semantic extraction technology, the entire experimental process of the original solution is decomposed into steps or stages;

[0043] (2-0-3) Feature extraction and clustering of high-dimensional state and action data of each step or stage to form a simplified abstract space;

[0044] (2-0-4) In the abstract space formed by step (2-0-3), set a step as a sub-strategy, train the sub-strategy for each step separately, and use the DQN algorithm for training calculations;

[0045] (2-0-5) While training the sub-strategies, train the high-level strategy to select and manage the execution order of each sub-strategy, and use the policy gradient algorithm to train to obtain a more efficient execution order simulation list;

[0046] (2-0-6) Integrate sub-strategies and high-level strategies for overall testing and optimization.

[0047] Preferably, the method of adjusting the experimental steps and parameters of Scheme C to generate Scheme D in step (2-3) comprises:

[0048] (2-3-1) Obtain historical data and knowledge about the experiment to be designed from the database constructed in step S1, clean up redundant information in the data through statement comparison technology, and format the data into a unified standard;

[0049] (2-3-2) Using the ComplEx algorithm to extract the potential relationship features between the experimental objects and methods from the data obtained in step (2-3-1), and using the ComplEx training relationship data to learn the complex relationship between the experimental conditions and results; the ComplEx algorithm formula used is:

[0050] f(s,r,t)=Re(<es+ws·φ(t),er+wr·Φ(t)> )

[0051] Among them, s is the subject, r is the relationship, and t is the timestamp.

[0052] The timing ComplEx (TComplEx) scoring function is defined as follows:

[0053] f(s,r,t)=Re(<es(t),er(t)> )

[0054] is a time-dependent complex vector; <·,·,·> represents a trilinear product; Re(·) represents the real part,

[0055] The above function is defined by the following formula:

[0056] es(t)=es+ws·Φ(t); er(t)=er+wr·Φ(t)

[0057] is a static complex vector embedding; is the time weight vector, Φ(t) is the time encoding function, φ(t) = [sin(ωt), cos(ωt), sin(2ωt), cos(2ωt), ..., sin(kωt), cos(kωt)], where ω is the frequency parameter and k is the encoding dimension;

[0058] (2-3-3) Based on the prediction results of the ComplEx model, identify the experimental steps and parameters that can be improved in Plan C, combine the best practices in the database and the optimization suggestions of the model, and generate Plan D.

[0059] Preferably, in step S3, the step of optimizing the solution according to user feedback includes:

[0060] Use the large language model to analyze user feedback, identify the content that needs to be modified in Plan E, collect and record users' historical instructions and historical feedback data, and adjust Plan E in combination with the database and user historical data; analyze user historical data through the centralized function S,

[0061]

[0062] Among them, x i and x j represents different eigenvalues, c i and c j Represents feature x i and x j The corresponding context information, w ij (c i ,c j ) represents the context-aware weight,

[0063] Calculate feature x i and x j , in the given context c i and c j The attention score in is then normalized by the softmax function:

[0064]

[0065] Among them, f(c i ,c j ) is the similarity measure of the context feature vector, d k is the scaling factor.

[0066] Accordingly, a computer device / equipment / system comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of any one of the methods.

[0067] Accordingly, a computer-readable storage medium stores a computer program / instruction thereon, and when the computer program / instruction is executed by a processor, the steps of the method are implemented.

[0068] Accordingly, a computer program product comprises a computer program / instruction, which implements the steps of the method when executed by a processor.

[0069] The present invention has the following beneficial effects: The present invention provides a new human-computer interaction system for designing biochemical experiments using artificial intelligence. In database construction, basic data is constructed using the basic change laws of biochemical substances, thereby improving data integrity and reliability. Method optimization is performed in knowledge graph design to improve the density of the knowledge graph, thereby improving the efficiency and quality of human-computer interaction. The present invention also designs a special algorithm to improve the accuracy of user intent recognition, and optimizes the generated experimental method based on user feedback and historical data, so as to ultimately obtain a scientific process and highly personalized experimental method. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 It is a schematic diagram of the operation flow of the present invention;

[0071] Figure 2 A schematic diagram of the process of constructing a database for the present invention;

[0072] Figure 3 A schematic diagram of the process flow of the scheme for generating the present invention. DETAILED DESCRIPTION

[0073] The present invention provides a human-computer interaction system that can output specific biological and chemical experimental processes according to user instructions. Figure 1As shown, the specific steps include:

[0074] S1. Database construction. The purpose of this step is to build a database containing known chemicals and biomacromolecules, as well as the mutual transformation relationships between them, and further build a knowledge graph based on the database as the underlying computing data source for human-computer interaction. The process of database construction is as follows: Figure 2 As shown, the following steps are included:

[0075] (1-1) Collection and organization of basic data.

[0076] A molecular database is constructed with compounds and biomacromolecules as the core and the conversion relationship between compounds and / or biomacromolecules as the connection. Specific data sources include: collecting all publicly available and retrievable compounds and biomacromolecules from publicly available prior art (academic journals, textbooks, research papers and patents). The database can be MySQL. It should be noted that the term "all" in the present invention refers to all substances and corresponding reactions that can be obtained based on public literature, vector mapping and logical reasoning, and does not mean that all objectively existing substances and reactions can be obtained.

[0077] The following basic principles are generally recognized in the field of chemistry: (1) The essence of compound reactions is the breaking and forming of chemical bonds; (2) The transformation between organic compounds is mainly reflected in the change of radicals (functional groups); (3) The transformation between inorganic compounds is mainly reflected in the change of covalent bonds.

[0078] Based on the above principles, for organic compounds, we retrieve the change process of a certain type of group from the prior art (for example, -OH and -COOH react to form esters), and then, based on the directionality of the chemical reaction, we infer all organic substances with this group through vector mapping, and further obtain the data information of all reactions of this type corresponding to this type of organic substance. Similarly, for inorganic compounds, we retrieve the change process of a certain type of covalent bond (for example, H + and OH - Acid-base reaction occurs, and in the generated H2O, a covalent bond is formed between the H atom and the O atom), and then all inorganic compounds with the covalent bond are obtained by vector mapping, and then the data information of all such reactions corresponding to such compounds is further obtained. In this way, data on all possible reactions of all compounds disclosed in the prior art are obtained and included in the database. The chemical reaction data included in the database can optionally include: reaction ID (unique number for easy retrieval), reactants and products, reaction conditions, reaction type, reaction energetics (endothermic, exothermic, activation energy change, etc.), reaction path, and references.

[0079] For biological macromolecules, polysaccharides (starch, cellulose, etc.), proteins and nucleic acids (DNA and RNA) are used as the core to retrieve and summarize the physical changes, chemical changes and biological changes that may occur in these biological macromolecules. After retrieving a certain type of change process of a certain type of substance, the data information of all such changes of this type of biological macromolecule is obtained through vector mapping. In this way, the data of all possible changes of biological macromolecules that have been disclosed in the prior art are obtained and included in the database. The biological reaction data included in the database can optionally include: reaction ID, enzyme / catalyst, substrate and product, reaction conditions, metabolic pathway, organism information, and references.

[0080] (1-2) Data cleaning: basic data verification and supplementation.

[0081] In this step, the data collected in step (1-1) is verified, increased or decreased to avoid further amplification of errors in the basic data during the subsequent construction of the knowledge graph, which would cause obvious errors in the calculation basis of human-computer interaction.

[0082] (1-2-1) Use existing technology to verify, add missing data, and delete data that is not available with existing technology.

[0083] The changes / reactions that can theoretically occur in compounds and biomacromolecules are affected by many other factors and may not necessarily occur in practice. At the same time, in addition to the data obtained in step (1-1), there may be individual special reactions. Therefore, in order to improve the accuracy and completeness of the data in the database, natural language processing tools such as SciBERT are used to extract information on chemical reactions and biological reactions from existing scientific literature, patents, and databases (such as PubChem, ChEBI, KEGG, Reactome, etc.). The information from public sources is compared with the information collected in step (1-1), and the reaction data that is not disclosed in the prior art is deleted. At the same time, the reaction data missing from the database established in step (1-1) is also supplemented from the prior art.

[0084] (1-2-2) Use existing technologies to verify, delete, and merge duplicate data.

[0085] Then, combined with the public information on the Internet and the public information of existing technologies, the data in the database is compared one by one, and duplicate data is deleted and merged. The operations of deleting and merging duplicate data include:

[0086] (1-2-2-1) For the same reaction of the same substance, there may be two or more groups of the same data because the descriptions in different literatures are slightly different, or the data are searched and collected in the chemical compound and biological macromolecule categories at the same time. For multiple groups of data, only the group with the most public information is retained, and the other data information is deleted. For example, chitosan may be classified as an organic compound and a biological macromolecule during classification and search, and the reaction data are obtained separately, resulting in data duplication.

[0087] (1-2-2-2) Searches were conducted and data was collected for different names and expressions of the same substance. For example, chitobiose and (1,4)-2-amino-2-deoxy-β-D-glucose. For this type of duplicate data, if common knowledge (such as textbooks and reference books) clearly states that they are the same substance, they will be directly merged and deleted in the database. The specific method is: retain the terms and expressions that appear more frequently in the prior art, and use the different terms for the same substance as notes for the data. If there is no common knowledge that clearly states that different terms are the same substance, all these data will be retained.

[0088] When performing steps (1-2-2-1) and (1-2-2-2) to delete and merge redundant data, multiple groups of identical data are compared between groups, and the difference information between the groups of data is extracted, and the difference information is used as the annotation information of the retained data.

[0089] (1-3) Build a knowledge graph.

[0090] In order to improve the efficiency and quality of human-computer interaction, the database obtained in step (1-2) is constructed into a knowledge graph. The main advantages include: the knowledge graph organizes information in the form of entities and relationships, making it easier for computers to understand and process complex knowledge structures; based on the graph structure, the system can perform logical reasoning, discover implicit relationships and knowledge, and complete more complex problems in human-computer interaction; by constructing a user knowledge graph, the system can provide a more personalized interactive experience, which is conducive to meeting different user needs.

[0091] You can choose to use the Neo4j tool to construct the reaction data in the database into a knowledge graph, which includes creating nodes and edges and adding attribute information to each node and edge. Each node represents an entity (compound or biomacromolecule), and each edge represents the relationship between nodes (chemical or biological reactions), such as the conversion relationship from reactants to products or the sequence of experimental steps. Other information and annotation information about each node in the database (including different names of the same substance, molecular structures, reaction conditions, operating parameters, etc.) are used as attribute information of nodes and edges. Then use a graph database (such as Neo4j) to store and manage the knowledge graph for subsequent calls and queries.

[0092] (1-4) Knowledge graph optimization.

[0093] Although the prior art has disclosed a large amount of chemical reaction and biological reaction knowledge, the knowledge graph directly constructed with this knowledge has too sparse data and cannot meet the requirements of high-quality and efficient human-computer interaction. In order to solve this problem, the present invention also includes preferred solution steps (1-4) to optimize the knowledge graph. Specifically, it includes: combining relationship prediction and attribute filling, through deep learning and embedding technology, introducing generative adversarial networks to generate entity relationships, multi-source data fusion, adaptive weighting mechanism and interactive learning, so as to effectively solve the sparsity problem of the knowledge graph and improve the dynamic adaptability and self-optimization ability of the knowledge graph.

[0094] Embedding model is a technology that maps discrete objects (such as words, entities or relationships) to continuous vector space. In the knowledge graph, the embedding model is used to represent the entities and relationships in the graph as vectors, capturing the semantic information between entities and relationships. The ComplEx algorithm is an algorithm specifically used for knowledge graph embedding. It uses complex vectors to represent entities and relationships and can better handle asymmetric relationships. The optimization method of the knowledge graph specifically includes the following steps:

[0095] (1-4-1) Establishment of the relationship model

[0096] The complEx algorithm captures the relationships and attributes between entities by training a recursive neural network. The trained model can capture entities (compounds and biomacromolecules) and identify potential relationships (reactions) between knowledge points (entities), and can predict relationships that have not appeared before based on this. The specific steps are as follows:

[0097] (1-4-1-1) Collect triples of entities and relations (e.g. entity 1, relation, entity 2), which form the basis of the knowledge graph. Input these data into the ComplEx model, and the model will generate complex vector representations for each entity and relation. These vector representations help the model capture the complex relationship features between entities.

[0098] (1-4-1-2) Using the entity and relationship vectors from step (1-4-1-1), use known relationships as positive samples, and train the model by simulating positive samples to generate non-existent relationships as negative samples. Obtain a preliminary relationship model.

[0099] (1-4-1-3) Verify the obtained relationship model. Introduce additional verification data sets, which come from public databases (such as PubChem, ChEBI, KEGG, etc.) or literature. Use natural language processing tools such as SciBERT to compare the potential relationships generated by the model's simulated positive samples with the known relationships in the verification data set, and remove relationships that are not in the verification data set. SciBERT evaluates the authenticity of relationships by using the cosine similarity formula to calculate the similarity between two semantic vectors and output a normalized value ranging from 0 to 1.

[0100] (1-4-1-4) The quality of the model is evaluated by the scoring function. The principle is to calculate the strength of the relationship between entities and relationships based on the complex vectors of entities and relationships. The scoring function is as follows: score(h,r,t)=Re(<h,r,t> ). Where h is the complex vector representation of entities and relations, r is the relation vector, t is the conjugate complex vector of the tail entity vector, and Re represents the real part operation. The scoring function is normalized using the Softmax function so that the sum of all output probabilities is 1. Models whose normalized scoring function is greater than the set value, such as greater than 0.85, are judged as reasonable relationship prediction models.

[0101] (1-4-2) Expanding the knowledge graph by filling in attributes

[0102] The purpose of this step is to fill in the missing knowledge points of entities in the knowledge graph and the correlation between them through reasoning and prediction, so as to make the knowledge graph denser and improve the efficiency and quality of human-computer interaction. Specific methods include:

[0103] (1-4-2-1) Through the ComplEx embedding method, knowledge points and their associations are embedded into the same vector space. The geometric distance and direction between vectors reflect the relationship between knowledge points, thereby calculating and predicting knowledge points and their potential associations.

[0104] (1-4-2-2) Using the neural network method, the scores of potential knowledge points and relationships are calculated to predict knowledge points and relationships that may exist but have not been recorded. This step uses, for example, cosine similarity to calculate the score and evaluate the strength of the potential relationship between knowledge points. Relationships with cosine similarity higher than a set value, such as higher than 0.75, are selected as potential relationships that can be supplemented.

[0105] (1-4-2-3) To enrich the knowledge graph, named entity recognition technology is used to identify entities in the text through online literature and other external data. Using tools such as BioBERT, the identified entities are divided into different types of knowledge points such as inorganic matter, organic matter, protein, enzyme, fat, cellulose, etc. Accurate entity classification helps to extract and verify more precise relationships later.

[0106] (1-4-2-4) Through SpaCy's dependency parsing module, we can determine which knowledge points are subjects, objects, or modifiers, as well as the logical or functional relationships between them. These relationships help to more fully understand and represent the interactions between knowledge points.

[0107] (1-4-2-5) Use a context-related model (such as BERT) to perform a similarity calculation method. Use the cosine similarity method to match the knowledge points identified from the text with the existing knowledge points in the knowledge graph. If the similarity is higher than a set value, such as 0.75, it will be incorporated into the knowledge graph. The preferred solution is to improve the cosine similarity calculation formula by introducing weights, such as introducing w i As the weight of the i-th dimension, it can better reflect the importance of different features and avoid the excessive influence of some unimportant features on the similarity; combined with context-related models, such as introducing the balance coefficient α, it can better capture complex semantic information, especially in sentences or short texts. This method can improve the perception of semantic similarity; using a combination of multiple word vector models, such as introducing BERT (A, B), using BERT to calculate the semantic similarity score of the context, capturing semantic information at different levels, and thus more comprehensively evaluating the cosine similarity.

[0108] The improved cosine similarity calculation formula is as follows:

[0109]

[0110] Among them: A i and B i w represents the i-th dimension of vector A and B respectively, and n is the total dimension of the vector. i is the weight of the i-th dimension, which can be set according to the number of knowledge points, reflecting the importance of this dimension to the similarity. α is an adjustment coefficient used to balance the influence between traditional cosine similarity and contextual similarity, with a value range of 0.1-3.0, preferably 1.2-2.0. Texis(A,B) is a semantic similarity score calculated using contextual embedding (such as BERT), which is used to enhance the ability to capture complex semantic relationships.

[0111] (1-4-2-6) Use tools such as Neo4j to manage and store knowledge graph data, and structure the knowledge graph data to ensure that it can be quickly called and queried in human-computer interaction. Track the changes and update history of the knowledge graph through version control and logging for auditing and tracking. Repeat the above process regularly to continuously extract and update the knowledge graph from new text data to maintain its dynamic and timeliness, thereby continuously enhancing the comprehensiveness and accuracy of the knowledge graph.

[0112] (1-5) Model training. Use the (preferably optimized) complEx algorithm and knowledge graph to train the artificial intelligence model so that it can accurately understand and apply the information in the knowledge base to improve the intelligence level of the system. The training methods include:

[0113] (1-5-1) Based on the text understanding task in the literature, a long short-term memory network (LSTM) is used to take the embedding vector in the knowledge graph as part of the input layer and concatenate it with the text features as the input of the model. This combination can effectively integrate text information and background knowledge in the knowledge graph and enhance the semantic understanding ability of the model.

[0114] (1-5-2) During the training process, the back propagation algorithm Adam optimizer is used to adjust the weight and bias parameters of the model. In order to improve the quality of the model, the dynamic learning rate α t Embedded in the model, it is used to help the model more accurately capture the complex patterns between entities and relationships, thereby building a high-quality knowledge graph. Learning rate α t Adaptive cosine annealing formula:

[0115]

[0116] Where: α t is the learning rate. max and α min are the maximum and minimum learning rates (α max Preferably 0.001, α min Preferably 0.00001). t is the current step number, and T is the total adjustment step number. i is the loss (or other performance indicator) of the ith validation, L min and L max are the minimum and maximum values ​​of the loss, respectively. N is the number of recent validations, which can be set as needed, preferably 100-150. γ is a parameter that controls polynomial decay (preferably 0.85). Continuously iterate until the loss function change rate is less than 5%, then stop iterating.

[0117] The advantages of the new training mechanism are as follows: (a) When the loss is large, the learning rate is increased to speed up the convergence; when the loss is small, the learning rate is reduced to stabilize the training; (b) Polynomial decay is introduced to normalize the verification loss and weighted average it, and its influence is controlled by the parameter γ (γ ranges between 0.5-3.0, preferably 1.9). This can make the learning rate dynamically adapt to different training stages and further optimize the training process.

[0118] (1-5-3) After training, the performance of the model is evaluated through the confusion matrix, and the calculated indicators include accuracy, recall, F1 score and AUC. These indicators provide the performance of the model in different aspects and help identify the quality of the model. Set the values ​​of accuracy, recall, F1 score and AUC, for example: set the accuracy to be higher than 80%, the recall and F1 score to be higher than 70%, and the AUC to be higher than 70%. For models whose results do not meet the standards, according to the analysis results, by analyzing the above four indicator parameters of the model (accuracy, recall, F1 score and AUC), trace back to the model training steps, adjust γ and T in the dynamic learning rate formula, and repeat the model training and performance evaluation steps to retrain the model. Repeat to obtain a model with satisfactory results.

[0119] S2, input instructions, intention recognition and preliminary solution generation.

[0120] The user inputs the instruction of designing a certain experimental process to the system. After receiving the instruction, the system recognizes the intention of the user's instruction. The process is as follows Figure 3 As shown, specifically including:

[0121] (2-1) Delete the descriptions with emotional and personal tendencies in the instructions. These descriptions are specifically reflected in the adjectives and adverbs in the instructions. Perform semantic understanding on the remaining instructions after deletion, retrieve the closest matching results from the database, and preliminarily generate the experimental process, which is named Plan A.

[0122] (2-2) After deleting the emotional color and personal inclination, the remaining content of the instruction is semantically understood, and an online search is performed. The closest matching results are retrieved from the research results with similar time (the time limit for online search can be set in advance or manually set by the user as needed) to obtain Plan B. Plan A and Plan B are matched and compared, and the differences between the two plans are marked. The online search results are analyzed to obtain the frequency of the difference content appearing in the online search documents (for example, 10 documents are searched, and the difference content appears in 6 documents at the same time, then the frequency is 60%). If the frequency of the difference is greater than the set value, the difference content in the online search results is judged to be correct, and the content is replaced with the corresponding content in Plan A; otherwise, the corresponding content of Plan A is retained. After the comparison and replacement are completed, Plan C is generated.

[0123] Since the states (such as experimental conditions, reactant states) and actions (such as experimental steps, parameter adjustments) involved in biochemical experiments are very large and complex, directly identifying the overall instructions requires exploration and decision-making in a huge possibility space, resulting in high computational complexity and slow convergence. Therefore, the preferred scheme includes: before performing step (2-1) and / or step (2-2) to identify user instructions, semantic understanding and recognition, it also includes step (2-0) to establish a reinforcement learning method for action recognition to optimize the processing steps of complex tasks and improve semantic and intent recognition capabilities. Specifically including:

[0124] (2-0-1) First, simply match the user instructions with the content in the database, and retrieve the result with the highest matching degree as the original solution.

[0125] (2-0-2) Based on semantic extraction technology, the system breaks down the entire experimental process of the original plan into several steps or stages, such as sample preparation, reaction condition setting, experimental operation, data recording, etc.

[0126] (2-0-3) Feature extraction and clustering of high-dimensional state and action data of each step to form a simplified abstract space. The high-dimensional state refers to the overall system state described by multiple variables or features in each step of the experimental process, which is the comprehensive information of the experiment at a certain moment; these variables or features may include temperature, pressure, chemical concentration, time point, equipment parameters, etc. Action data refers to the specific operations or decisions taken according to the current state during the experiment; these actions can be adjusting temperature, adjusting pressure, adding reagents, changing equipment settings, etc.

[0127] (2-0-3-1) Feature extraction steps, specifically including:

[0128] Data collection. Collect all data and parameters that occur during the experiment. This data may include temperature, time, reactant concentration, stirring speed, instrument settings, etc.

[0129] Data preprocessing: Use filters or smoothing algorithms to remove noise from the data; fill in or delete missing data to ensure data integrity; standardize data of different dimensions to the same range (such as 0 to 1) to facilitate comparison and analysis.

[0130] Feature extraction method. Calculate statistics such as mean, variance, standard deviation, maximum, minimum, etc.; extract the frequency components of the signal through Fourier transform; use PCA (principal component analysis) to project high-dimensional data into low-dimensional space.

[0131] (2-0-3-2) Clustering methods, specifically including:

[0132] Select a clustering algorithm. Select the K-means clustering algorithm based on the data characteristics.

[0133] Set parameters. Set the number of clusters K.

[0134] Clustering result analysis: Analyze the center point of each cluster to represent the typical characteristics of this type of data; evaluate the similarity and consistency of the data within the cluster.

[0135] Form an abstract space. Map the cluster labels in the clustering results to the original data to form a simplified abstract space; train sub-strategies within each cluster and use the DQN algorithm for reinforcement learning.

[0136] (2-0-4) In the abstract space formed by step (2-0-3), set a step as a sub-strategy, train the sub-strategy for each step, and use the DQN (Deep Q-Network) algorithm for training calculations. Specifically, it includes:

[0137] Each step in the human-computer interaction process is defined as a reinforcement learning environment. The state is the data extracted from the features, the action is the response that the system can take, and the reward function is defined to evaluate the effect of each response. The reward function can be based on indicators such as user satisfaction and task completion. The network structure of DQN includes an input layer, several hidden layers, and an output layer. The input layer receives state data, and the output layer provides the Q value of each possible response. The hidden layer uses a nonlinear activation function to capture complex state-action relationships.

[0138] Training process: Randomly initialize the weight parameters of the DQN network, set the exploration rate and discount factor, and control the probability of exploring new actions and the weight of future rewards respectively. After each human-computer interaction, store the state, action, reward, and next state in the experience replay buffer, regularly extract small batches of samples from the buffer, and use these samples to train the DQN network. Use the target network to stabilize the training process. The target network has the same structure as the main network, but the parameters are updated more slowly to prevent shocks during training.

[0139] (2-0-5) While training the sub-strategies, train a high-level strategy to select and manage the execution order of these sub-strategies, and use the policy gradient algorithm to train to obtain a more efficient execution order simulation list. The method of training the high-level strategy includes the following steps:

[0140] (2-0-5-1) Data Collection and Strategy Execution

[0141] In the actual human-computer interaction process, the system continuously records detailed information about each interaction. This information includes the current state (such as environmental characteristics, user input, etc.), the selected sub-strategy (such as specific operation steps or decisions), the execution results (such as success or failure, task completion), and the feedback obtained (such as user satisfaction ratings, system logs). These data can provide real interaction scenarios and results, thereby helping high-level strategies learn how to select the optimal sub-strategy in different situations. In order to improve the representativeness and coverage of the data, data should be collected under as diverse environmental and task conditions as possible. The collected data should include not only normal operating scenarios, but also edge cases and abnormal situations to ensure that the strategy can handle various possible interaction scenarios.

[0142] (2-0-5-2) Strategy Gradient Calculation and Parameter Update

[0143] Using the REINFORCE algorithm in the policy gradient method, the effect of the current policy is evaluated by calculating the gradient of the policy. After each interaction, the gradient of the policy is calculated based on the sub-policy selected by the policy and the reward obtained. The gradient reflects the direction and size of the expected reward of selecting a sub-policy.

[0144] Use the gradient ascent method to update the weight parameters of the high-level policy network: define the loss function as the negative cumulative reward (or use other suitable reward functions), and then calculate the gradient through the gradient descent / ascent method. Adjust the weight parameters and discount factors of the policy network so that the policy is more likely to choose a sub-policy with higher rewards in future interactions. Among them, the weight parameter refers to the learning rate, which determines the step size of each update; the discount factor is the weight that affects future rewards.

[0145] Since there are few real data on human-computer interaction and there are a lot of low-quality data, it is not enough to meet the artificial intelligence learning samples. In order to solve this problem, the preferred solution is: introduce virtual environment simulation, use the data enhancement method in DeepMind technology, generate a large amount of high-quality interaction data, and help train the human-computer interaction system. However, virtual environment simulation technology may not fully reflect the reality due to the low-fidelity virtual environment. If you want to create or maintain a high-fidelity environment. It is necessary to consume a lot of manpower and material resources. Therefore, the preferred solution is: introduce an intelligent agent system (such as AnyLogic tool) to simulate user behavior, so as to overcome these shortcomings in virtual environment simulation and improve human-computer interaction capabilities.

[0146] The intelligent agent system generates rich and diverse interaction data by simulating the behavior and decision-making process of real users to improve the realism and data quality of the virtual environment. The intelligent agent system can adaptively adjust its behavior and simulate the interaction methods of different user groups and situations. The working process of the intelligent agent system is as follows:

[0147] A. Intelligent Agent System Training

[0148] Collect actual user interaction data, perform data labeling and cleaning, and ensure data quality and accuracy. Divide the data into training sets, validation sets, and test sets for training and evaluating intelligent agent models. Use supervised learning methods to train intelligent agent models and simulate user decision-making and behavior processes. Improve the generalization ability and performance of the model through cross-validation and hyperparameter optimization.

[0149] B. Virtual environment integration

[0150] Embed the trained intelligent agent system into the virtual environment to replace or supplement user behavior. The intelligent agent adaptively adjusts its behavior based on the state and feedback in the virtual environment to generate high-quality interaction data. Introduce multiple intelligent agents to simulate the interaction methods of different user groups and situations, and increase the diversity and coverage of data.

[0151] C. Data Generation and Strategy Training

[0152] The intelligent agent automatically performs interactive tasks in the virtual environment and records the state, action and feedback data during the interaction. Through the behavior simulation of the agent system, a large amount of high-quality and representative interaction data is generated. The reinforcement learning algorithm DQN is used to train strategies in the virtual environment, and the data generated by the intelligent agent is used to accelerate the training process.

[0153] (2-0-6) Integrate all abstract sub-strategies and high-level strategies together for overall testing and optimization.

[0154] (2-0-6-1) Feature extraction and conversion

[0155] Detailed feature extraction is performed for each stage of the experimental process, including key features such as experimental conditions, user behavior and feedback. These features can represent important information in the entire experimental process, such as temperature, time, operation steps, etc. Then the high-dimensional raw data is simplified and transformed to form an easy-to-process state representation. This process can use dimensionality reduction techniques such as PCA (principal component analysis) to ensure that the extracted state features are representative and concise.

[0156] (2-0-6-2) Parallel training of sub-strategies and high-level strategies

[0157] While training sub-strategies, high-level strategies are used for scheduling and management. High-level strategies select appropriate sub-strategies to execute based on the current state and context information. Sub-strategies are trained using the DQN algorithm, and each sub-strategy independently optimizes its operation steps. High-level strategies are trained using the policy gradient algorithm to optimize the execution order of sub-strategies to ensure the overall efficiency of the system.

[0158] (2-0-6-3) Feedback Mechanism and Dynamic Adjustment

[0159] Through the real-time feedback mechanism, the parameters of sub-strategies and high-level strategies are dynamically adjusted. For example, if a sub-strategy does not work well in a specific situation, the high-level strategy will adjust its execution order or select a different sub-strategy based on the feedback. This feedback mechanism ensures the collaboration between sub-strategies and high-level strategies, improving the adaptability and flexibility of the system. At the same time, the strategy performance is regularly evaluated and fine-tuned with a small amount of real data to further optimize the strategy.

[0160] (2-3) Combining the database and the optimized complEx algorithm, the system automatically adjusts the experimental steps and parameters of Scheme C to optimize the experimental results and generates Scheme D. Scheme D includes detailed information such as specific operation steps, required equipment and reagents, reaction conditions, etc. Each step lists the required materials, operation methods, and expected results in detail so that users can perform the experiment according to the instructions.

[0161] The specific operation process of this step includes:

[0162] (2-3-1) Obtain historical data and knowledge about experimental objects, experimental methods, reaction conditions, etc. from the database constructed in step S1. Then, through statement comparison technology, clean up redundant information in the data and format the data into a unified standard.

[0163] (2-3-2) Use the ComplEx algorithm to extract the potential relationship features between the experimental objects and methods from the data obtained in step (2-3-1), including extracting the associations between various experimental parameters (such as temperature, time, concentration, etc.) and the experimental results, and then use ComplEx to train the relationship data to learn the complex relationship between experimental conditions and results.

[0164] During the experiment, there is a strict sequence between the experimental steps, but the existing ComplEx algorithm is static and cannot accurately reflect this feature. In order for the ComplEx algorithm to better capture the temporal relationship and more accurately extract the potential relationship characteristics between the experimental objects and methods, the following new algorithm is preferably used:

[0165] f(s,r,t)=Re(<es+ws·φ(t),er+wr·φ(t)> )

[0166] Among them, s is the subject, r is the relation, and t is the timestamp.

[0167] The timing ComplEx (TComplEx) scoring function is defined as follows:

[0168] f(s,r,t)=Re(<es(t),er(t)> )

[0169] in: is a time-dependent complex vector; <·,·,·> represents a trilinear product; Re(·) means taking the real part.

[0170] Calculate f(s,r,t) for each experimental step. Set a threshold value θ. If f(s,r,t)≥θ, then it is judged that there is a significant relationship between the step and the designed experiment, and it belongs to the step in the experiment to be designed; if f(s,r,t)<θ, then it is judged that the step is not strongly related to the designed experiment, and the step is deleted. Calculate and compare the values ​​of f(s1,r,t) and f(s2,r,t) of any two experimental steps s1 and s2. The one with a larger value belongs to the previous step. For example, f(s1,r,t)>f(s2,r,t), then it is judged that step s1 comes before s2.

[0171] The above function can be defined using the following formula:

[0172] es(t)=es+ws·φ(t); er(t)=er+wr·φ(t)

[0173] in: is a static complex vector embedding; is the time weight vector. φ(t) is the time encoding function, φ(t) = [sin(ωt), cos(ωt), sin(2ωt), cos(2ωt), ..., sin(kωt), cos(kωt)] where ω is the frequency parameter and k is the encoding dimension.

[0174] The advantages of the above formula include: the introduction of time information through φ(t) enables the model to capture time-related patterns; the use of trigonometric functions can naturally represent periodic relationships and expand the application field; by adjusting ω and k, the accuracy and complexity of time encoding can be controlled; the time weight vectors ws and wr can be interpreted as the degree to which entities and relationships change over time.

[0175] (2-3-3) Based on the prediction and calculation results of the ComplEx model, determine whether each experimental step has a significant relationship with the experiment to be designed and whether the order of the steps is correct, so as to modify Scheme C and generate Scheme D. Scheme D includes detailed information such as the optimized operation steps, required equipment and reagents, reaction conditions, etc. Scheme D is output as a detailed experimental manual containing all necessary information for reference by experimenters.

[0176] (2-4) Identify the descriptions with emotional colors and personal tendencies in the instructions, and understand the user's personalized needs for the solution. Match Solution D with the user's intention (personalized needs) and perform targeted optimization to obtain Solution E.

[0177] For example, the user instruction is "generate an XXX method that can quickly complete the reaction", and the additional description is "only qualitative results are required, and the accuracy requirements are not high." Extract personal preference keywords, "fast", "qualitative results", and "low accuracy requirements." Based on the keywords, search the database for additional operations that can increase the reaction rate and add them to Scheme D (such as adding catalysts, increasing reaction temperature / pressure, etc.), and delete the relevant operations in Scheme D for improving accuracy (for example, if the attribute information of the database records that adding a substance dropwise can reduce impurity formation and improve product accuracy, then delete this step accordingly). If there is no preference description in the user instruction, the default selection and optimization is the scheme with the highest match and the most complete details.

[0178] The specific operations of user intent matching include:

[0179] (2-4-1) Extraction of trending keywords

[0180] User instructions are parsed through natural language processing technology to extract keywords representing experimental needs and preferences. For example, if the user instruction is "generate a method that can quickly complete the reaction, only qualitative results are required, and the accuracy requirements are not high", then "fast", "qualitative results" and "low accuracy requirements" are extracted as keywords. The natural language model BERT and the sentiment analysis tool VADER are used to identify and classify these keywords to ensure an accurate understanding of user needs. The extracted keywords are preferably classified into different experimental needs and preference categories. The classification criteria can be set according to actual needs based on various aspects of the experiment, such as reaction speed, result type, accuracy requirements, etc. By classifying keywords, matching and retrieval can be performed more efficiently in the experimental database. For example, "fast" may correspond to the addition of accelerators or increasing the temperature in the experimental steps, and "qualitative results" correspond to the type of experimental results and analysis methods.

[0181] (2-4-2) Matching and Optimization

[0182] Based on the extracted keywords, the operations and parameters that match the user's needs are searched in the database built in step S1. The keywords are matched with the operations and parameters in the database to find the experimental steps and conditions that best suit the user's needs. For example, the keyword "fast" will trigger a database search for all operations related to the designed experiment that can accelerate the reaction, such as adding catalysts or increasing the reaction temperature. Based on the search results, the experimental plan is optimized, and operation steps are added or deleted to meet user needs. For example, if the user emphasizes a fast reaction, a catalyst step can be added to the experimental plan to increase the reaction temperature or pressure; if the user emphasizes qualitative results, the experimental steps can be simplified to reduce unnecessary precision operations.

[0183] The preferred solution is: when extracting the preferred keywords in (2-4-1), the complEx algorithm is optimized to better identify and understand user preferences and tendencies, thereby optimizing the solution. The complEx algorithm optimization method includes: introducing a self-attention layer in the middle layer of the ComplEx model. The specific method is to introduce the classic self-attention mechanism model Transformer, automatically pay attention to important features and relationships, capture long-distance dependencies in the human-computer interaction process, and enable the model to perform better when processing complex user instructions and emotional understanding. Through adaptive feature capture, the model can dynamically adjust feature capture according to changes in input data, and enhance the emotional understanding and preference response capabilities of user needs. The complEx algorithm optimization method specifically includes:

[0184] In the self-attention calculation process, for each embedding vector x i , calculate its value with other vectors x j The dot product of ij Next, these weights are normalized using the Softmax function to ensure that their sum is 1, resulting in normalized attention weights. These attention weights are then used to perform a weighted summation of the embedding vectors to generate a new feature representation z i In addition, through the multi-head self-attention mechanism, multiple attention weights are calculated in parallel, and the outputs of each head are spliced ​​together to form the final feature representation, thereby enhancing the model's ability to capture features.

[0185] The feature representation generated by the self-attention mechanism is used as the input of the ComplEx model to continue triple scoring and prediction. The ComplEx model uses these new feature representations to score triples and uses the cross entropy loss function to optimize the model. According to the specific needs of human-computer interaction tasks, the core parameters (such as the number of heads, hidden layer dimensions, learning rate, etc.) are tuned to improve the performance and robustness of the model, thereby achieving more accurate emotion prediction and more efficient interaction.

[0186] S3. User feedback and solution optimization.

[0187] According to the result of step S2, the experimental process scheme E is output, and a prompt message is given to prompt the user to review the scheme E and provide feedback on the scheme E. The feedback form can be to select a fixed option, input feedback, or a combination of the two. The feedback content includes: whether there are obvious logical loopholes in the scheme, what is the optimization tendency of the scheme (whether to obtain a simplified version or a detailed version, or other personalized needs).

[0188] If the user refuses to give feedback or there is no feedback, the user instructions and related data are recorded, and plan E is used as the final output plan.

[0189] If the user provides feedback, the system uses a large language model to analyze the user feedback, identify the parts of Plan E that the user is dissatisfied with and need to be modified, and collect and record the user's historical instructions and historical feedback data, and adjust Plan E in combination with the database and user historical data (if any).

[0190] The specific adjustment process includes: for the identified content that needs to be modified, supplementary search is performed in the database, and supplementary search and comparison are performed in combination with online data on the Internet, and the content with the highest frequency of occurrence in the database and online data search results is extracted. If there are multiple contents with high frequency of occurrence and the frequency difference is not significant, the content closest to the time of the search is selected to replace this part of the content of Plan E, and Plan F is obtained.

[0191] The preferred solution is: based on Solution E or F, according to the optimization tendency of user feedback and the tendency data recorded in the historical data (the tendency of this user feedback is given priority, if the user has no clear feedback, refer to the tendency in the historical data), optimize Solution E or F in the manner of steps (2-4) and output Solution G.

[0192] Among them, user historical data and past tendencies are analyzed using the user history weighted function algorithm, specifically including: the user history weighted function algorithm analyzes user preferences by weighted processing of user historical interaction data. First, the system collects user interaction records, cleans and extracts features from the data, and extracts key features such as user instructions and operation frequency. Next, according to the timestamp and feedback nature of the data, weights are set for each feature point, and high weights are given to recent and positive feedback data. Then, these weights are used to perform a weighted average on the features to generate the user's personalized feature vector, and a new weighted function S is designed to integrate the weighted features into a user preference score. The formula for the weighted function S is as follows:

[0193]

[0194] Among them, x i and x j Represents different feature values ​​(user personalized tendency keywords), uses the BERT embedding model to convert the extracted user personalized tendency keywords into vectors, and then enters the formula here. i and c j Represents feature x i and x j The corresponding context information. j (c i ,c j ) represents the context-aware weight, which is dynamically calculated based on the context information.

[0195]

[0196] γ is a positive constant, ranging from 1 to 100, preferably 30 to 45. i -x j ∥ 2 is the eigenvector x i and x j The square of the Euclidean distance between them.

[0197] Use the attention mechanism to calculate the context-aware weights. Specifically, calculate the feature x i and x j , in the given context c i and c j The attention score in is then normalized by the softmax function:

[0198]

[0199] f(c i , c j )=c i ·c j

[0200] Among them, f(c i ,c j ) is the dot product or other similarity measure of the context feature vector, d k is the scaling factor (the dimension of the input data processed by each head in the multi-head attention mechanism). For example: a Transformer model with a hidden layer dimension of 512 and a multi-head attention mechanism with 8 heads, then dk = 512 / 8 = 64.

[0201] Split scheme E or F into several sub-parts (including experimental parameter settings, operation steps, and equipment selection), and use the centralized weight function S to calculate and score each sub-part. If the centralized weight function S score calculated for a sub-part is lower than the set value, it is judged that there is a problem with the sub-part. A low score indicates that the content it represents (including unreasonable parameter settings, insufficient experimental conditions, unreasonable experimental steps, etc.) is unreasonable, and adjustments should be made accordingly. For example, if the temperature setting of an experimental scheme is low at 20°C, the temperature parameters should be adjusted accordingly, and different temperature values ​​should be tried within other appropriate ranges and the S value should be recalculated. Adjust the calculated S value to a value higher than the set value, and output the optimized scheme G. Generally speaking, for exploratory experiments, the set value of S is 0.65, and for standard experiments or verification experiments, the set value of S is 0.75.

[0202] The weighting function S has the following advantages: (a) Traditional weighting functions usually use linear weighting methods to linearly combine features. However, user behavior and preferences often exhibit complex, nonlinear relationships. Therefore, the present invention introduces nonlinear combinations (such as polynomials) to help the weighting function better capture these complex relationships; (b) Traditional models may ignore the impact of contextual information on user behavior. The present invention dynamically adjusts weights based on the user's contextual information to help the model adapt to different user states more flexibly; (c) The weighting function S also introduces an attention mechanism (dot product attention) to dynamically adjust the similarity between contextual features and target features, so that it can automatically focus on important features and relationships to improve the accuracy of understanding complex user instructions and emotions, and predict user needs and intentions in future interactions.

[0203] The system outputs solution F or G and submits it to the user for review again. If the user passes the review, the operation is completed; if the review fails, step S3 is repeated to continue to modify and optimize the solution based on the user's feedback until the user passes the review. At the same time, each user's feedback information is recorded. The user loop feedback and optimization mechanism can continuously improve the accuracy and practicality of the output solution, and gradually form a system that can meet the personalized needs of users.

[0204] The embodiments described above are only descriptions of the preferred modes of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations, modifications, and substitutions made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for designing biological and chemical experiments, characterized in that: The steps include: S1. Build a database; specifically include: (1-1) Basic data collection and organization: With compounds and biomacromolecules as the core and the transformation relationship between compounds and / or biomacromolecules as the connection, a database is constructed; (1-2) Data cleaning: basic data verification and supplementation; (1-3) Construct the data in the database into a knowledge graph, create nodes and edges, each node represents an entity, and each edge represents the relationship between nodes; S2: User inputs a command, identifies the user's intention, and outputs a preliminary solution; specifically includes: (2-1) Delete the emotional and personal descriptions in the instructions, perform semantic understanding on the remaining instructions after deletion, retrieve the closest matching results from the database constructed by S1, and generate an experimental process, named Plan A; (2-2) semantically understand the remaining content of the instruction after removing the emotional color and personal inclination, conduct online search, retrieve the result with the highest matching degree from the search results, and obtain solution B; Plan A and Plan B are matched and compared, and the differences between the two plans are marked and analyzed to obtain the frequency of occurrence of the difference content. If the frequency of occurrence of the difference content is greater than a set value, the difference content is replaced with the corresponding content in Plan A; otherwise, the difference content of Plan A is retained; after the comparison and replacement are completed, Plan C is generated; (2-3) Combining the database constructed in step S1 with the complEx algorithm, adjusting the experimental steps and parameters of scheme C to generate scheme D; (2-4) Identify the descriptions with emotional colors and personal tendencies in the instructions, and understand the user's personalized needs for the solution; match and optimize solution D with the personalized needs, obtain solution E and output it; S3, prompting the user to give feedback on Solution E. If the user refuses to give feedback or does not give feedback, the user's instructions and data are recorded, and Solution E is used as the final output solution; If there is user feedback, optimize the solution based on the user feedback and output the solution again.

2. The method according to claim 1, characterized in that: In step (1-1), when collecting and organizing basic data, first search and collect published compounds and biomacromolecules; For organic compounds, the change processes of the groups are retrieved and collected from the prior art. According to the directionality of the chemical reaction, all the organic substances with the group are inferred through vector mapping, and then the data information of all the change processes of the group corresponding to this type of organic substances is further obtained; for inorganic compounds, the change processes of covalent bonds are retrieved and collected. Through vector mapping, all the inorganic compounds with the covalent bonds are obtained, and then the data information of all the reactions of this type corresponding to this type of compounds is further obtained; for biological macromolecules, the physical changes, chemical changes and biological changes of the biological macromolecules are retrieved respectively. After retrieving a certain type of change process of a certain type of substance, the data of all possible changes of the biological macromolecules that have been disclosed in the prior art are obtained through vector mapping.

3. The method according to claim 1, characterized in that: Step S1 also includes steps (1-4) of knowledge graph optimization, which specifically include the following steps: (1-4-1) Establish a relationship model: Collect triple data containing entities and relations to form the basis of the knowledge graph and input them into the ComplEx model. The model generates complex vector representations for each entity and relation. The complex vector representations help the model capture the complex relationship characteristics between entities. Use known relationships as positive samples, and simulate positive samples to generate non-existent relationships as negative samples to train the model and obtain a preliminary relationship model. Verify the obtained relationship model, compare the potential relationship generated by the model simulation positive sample with the known relationship in the verification data set, and remove the relationship that is not in the verification data set; the verification data set comes from a public database or public literature; (1-4-2) Expand the knowledge graph by filling in attributes: Through the ComplEx embedding method, knowledge points and their associations are embedded into the same vector space. The geometric distance and direction between vectors reflect the relationship between knowledge points, so as to calculate and predict knowledge points and their potential associations. By using a neural network method, the scores of potential knowledge points and relationships are calculated, thereby predicting knowledge points and relationships that may exist but have not been recorded; Through online documents and other external data, use named entity recognition technology to identify knowledge points in online documents and other external data; Use contextual embedding models to calculate the semantic similarity between knowledge points identified in online literature and other external data and existing knowledge points in the knowledge graph; The cosine similarity method is used to calculate the similarity between the knowledge points identified in online documents and other external data and the existing knowledge points in the knowledge graph. If the similarity is higher than the set value, it will be integrated into the knowledge graph; the cosine similarity calculation formula is as follows: Among them: A i and B i denotes the i-th dimension of vectors A and B respectively, n is the total dimension of the vector, and w i is the weight of the i-th dimension, α ranges from 0.1 to 3.0, and Texis(A,B) is the semantic similarity calculated using the context embedding model.

4. The method according to claim 3, characterized in that: Step S1 also includes step (1-5) model training, which uses the knowledge graph optimized in step (1-4) to train the relationship model constructed in step (1-4), specifically including the following steps: (1-5-1) Use the long short-term memory network to take the embedding vector in the knowledge graph as part of the input layer, concatenate it with the text features, and use it as the input of the model; (1-5-2) During the training process, the back-propagation algorithm Adam optimizer is used to adjust the weight and bias parameters of the model and set the dynamic learning rate α t In the embedded model, the learning rate α t formula: Where: α max and α min are the maximum and minimum learning rates, t is the current step number, T is the total number of adjustment steps, and L i is the loss of the ith validation, L min and L max are the minimum and maximum values ​​of the loss, N is the number of recent validations, and γ is the parameter that controls the polynomial decay; (1-5-3) After training, the performance of the model is evaluated through the confusion matrix, and the calculated indicators include accuracy, recall, F1 score and AUC; for models whose indicator calculation results do not reach the set values, by analyzing the indicators, backtracking to the model training steps, adjusting γ and T in the learning rate formula, repeating the model training and performance evaluation steps to retrain the model and obtain a model with satisfactory results.

5. The method according to claim 1, characterized in that: Step S2 also includes step (2-0), establishing a reinforcement learning method for action recognition to semantically understand the instructions, and the establishment method includes: (2-0-1) Match the user instruction with the content in the database constructed in step S1, and retrieve the result with the highest matching degree as the original solution; (2-0-2) Based on semantic extraction technology, the entire experimental process of the original solution is decomposed into steps or stages; (2-0-3) Feature extraction and clustering of high-dimensional state and action data of each step or stage to form a simplified abstract space; (2-0-4) In the abstract space formed by step (2-0-3), set a step as a sub-strategy, train the sub-strategy for each step separately, and use the DQN algorithm for training calculations; (2-0-5) While training the sub-strategies, train the high-level strategy to select and manage the execution order of each sub-strategy, and use the policy gradient algorithm to train to obtain a more efficient execution order simulation list; (2-0-6) Integrate sub-strategies and high-level strategies for overall testing and optimization.

6. The method according to claim 1, characterized in that: Step (2-3) The method of adjusting the experimental steps and parameters of Scheme C to generate Scheme D includes: (2-3-1) Obtain historical data and knowledge about the experiment to be designed from the database constructed in step S1, clean up redundant information in the data through statement comparison technology, and format the data into a unified standard; (2-3-2) Using the ComplEx algorithm to extract the potential relationship features between the experimental objects and methods from the data obtained in step (2-3-1), and using the ComplEx training relationship data to learn the complex relationship between the experimental conditions and results; the ComplEx algorithm formula used is: f(s,r,t)=Re(<es+ws·φ(t),er+wr·φ(t)> ) Among them, s is the subject, r is the relationship, and t is the timestamp. The timing ComplEx (TComplEx) scoring function is defined as follows: f(s,r,t)=Re(<es(t),er(t)> ) is a time-dependent complex vector; <·,·,·> represents a trilinear product; Re(·) represents the real part, The above function is defined by the following formula: es(t)=es+ws·φ(t); er(t)=er+wr·φ(t) is a static complex vector embedding; is the time weight vector, φ(t) is the time encoding function, φ(t) = [sin(ωt), cos(ωt), sin(2ωt), cos(2ωt), ..., sin(kωt), cos(kωt)], where ω is the frequency parameter and k is the encoding dimension; For each experimental step, calculate f(s,r,t) separately; (2-3-3) According to the calculation result of step (2-3-2), set the threshold θ. If f(s,r,t)≥θ, then it is judged that there is a significant relationship between the step and the designed experiment and that the step belongs to the step in the experiment to be designed. If f(s,r,t)<θ, then delete the corresponding step. Compare the sizes of f(s1,r,t) and f(s2,r,t) of any two experimental steps s1 and s2. If f(s1,r,t)>f(s2,r,t), then it is judged that step s1 is before s2. Adjust the order of the experimental steps accordingly and generate plan D.

7. The method according to claim 1, characterized in that: In step S3, the steps of optimizing the solution according to user feedback include: Use the large language model to analyze user feedback, identify the content that needs to be modified in Plan E, collect and record users' historical instructions and historical feedback data, and adjust Plan E in combination with the database and user historical data; analyze user historical data through the centralized function S, Among them, x i and x j represents different eigenvalues, c i and c j Represents feature x i and x j The corresponding context information, w ij (c i ,c j ) represents the context-aware weight, Calculate feature x i and x j , in the given context c i and c j The attention score in is then normalized by the softmax function: f(c i ,c j )=c i ·c j Among them, f(c i ,c j ) is the similarity measure of the context feature vector, d k is the scaling factor; The scheme to be optimized is divided into several sub-parts, and each sub-part is calculated and scored using the centralized weight function S. If the score of the centralized weight function S calculated for a sub-part is lower than the set value, it is judged that there is a problem with the sub-part and optimization is required.

8. A computer device / equipment / system comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images