An environmental data-oriented automated quality evaluation method and system

Abstract

The application discloses an environmental data-oriented automatic quality evaluation method and system, wherein the method aligns the source line numbers of the gold standard experiment record table and the model output table according to the literature identifier, and inserts the missing extraction placeholder record explicitly; the composite key is formed by the literature identifier and the normalized electrochemical anchor point parameter to carry out accurate matching, and meanwhile, the loose matching is carried out in the same literature group based on the relative tolerance of the anchor point parameter; for the fields of anode / cathode material, target pollutant, electrolyte, reactor configuration and operation condition, the semantic equivalence judgment is carried out by calling the term knowledge base in the field of electrochemical water treatment; the cell verification result is encoded into a visual identifier, and a process parameter quality report is generated, and the verification result can support the construction of the electrochemical water treatment process parameter library, the knowledge precipitation of the electrode material and the reaction condition, and the large language model assisted literature mining and model evaluation in the directions of electrocatalytic degradation, electro-Fenton, electric adsorption and electrodialysis.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of electrochemical water treatment and environmental information technology, and relates to a method and system for verifying the results of parameter extraction from literature related to electrochemical water treatment, electrocatalytic oxidation, electro-Fenton and electrochemical advanced reduction processes. Specifically, it relates to a method and system for verifying the structured extraction results of experimental parameters, electrode and reactor configuration, operating conditions and reaction kinetic data in electrochemical water treatment literature. Background Technology

[0002] Electrochemical water treatment is an important technological pathway for degrading recalcitrant organic pollutants and achieving advanced oxidation treatment, encompassing various forms such as anodic oxidation, electro-Fenton oxidation, electrocoagulation, electrochemical reduction, and coupled processes. In scientific research and engineering practice, a large amount of experimental data from small-scale, pilot-scale, and demonstration projects has been published in journal articles, dissertations, and technical reports. This data covers key parameters such as anode / cathode materials (e.g., BDD, DSA, Ti / IrO2, Ti / RuO2, PbO2, and carbon-based electrodes), current density, cell voltage, electrode spacing, electrode area, electrolyte composition and concentration, initial pH, temperature, target pollutant type and initial concentration, degradation rate constant k, TOC / COD removal rate, mineralization rate, specific energy consumption, reaction time, and reactor configuration (plate-and-frame, fluidized bed, etc.).

[0003] In recent years, to accelerate the construction of electrochemical water treatment process parameter databases and knowledge graphs, researchers have used large language models or specialized information extraction models to automatically extract experimental records from unstructured literature texts and write them into structured tables. However, before using the extraction results for electrode selection and comparison, process parameter regression, machine learning modeling, or inclusion in engineering databases, consistency verification with manually annotated gold standards is necessary.

[0004] Existing verification schemes mostly rely on exact matching of common strings or comparison of fixed threshold values, failing to fully consider the following characteristics of the electrochemical water treatment field:

[0005] 1. Kinetic and intensity parameters can be expressed in various ways: the same degradation rate constant may be expressed in min... -1 h -1 or s -1 The current density exists at mA / cm². 2 mA·cm -2 The writing system includes both scientific notation and fixed-point decimals.

[0006] 2. Electrode and contaminant terminology is highly specialized: for example, Ti / BDD, boron-doped diamond electrode, and boron-doped diamond electrode should be considered the same material; phenol, phenol, and PhOH should be considered the same contaminant.

[0007] 3. Multiple comparative experimental records for a single article: Comparative experiments with different current densities, different initial concentrations, different electrode ratios, or different H2O2 dosages are often missed by the model. If the missed rows are not explicitly reflected in the verification stage, the model recall rate will be overestimated.

[0008] 4. Heterogeneous descriptions of process types and reactors: Descriptions such as electrocatalytic oxidation, anodic oxidation, electrocatalytic oxidation, electro-Fenton, and electro-Fenton need to be incorporated into a unified semantic standard;

[0009] Lack of engineered output that interfaces with the process database: Only abstract classification indicators are output, which is insufficient to support electrochemical water treatment R&D personnel in verifying the quality of parameters entered into the database.

[0010] Therefore, there is an urgent need for a parameter extraction result verification technology solution for the field of electrochemical water treatment that can incorporate complex situations such as missed sampling, format drift and synonyms into a unified standard, and can be deployed on computer equipment or systems. Summary of the Invention

[0011] To address the issues that arise when constructing process parameter databases for electrocatalytic oxidation, electro-Fenton, electrolysis, electrocoagulation, and advanced electrochemical reduction processes, such as differences in unit conversion of degradation kinetic parameters, inconsistencies in the dimensions of current density and cell voltage, synonymous descriptions of electrode materials and pollutants, and omissions in comparative experimental records from large language models, leading to errors in parameter entry and distorted statistical results, this invention provides an automated quality assessment method and system for environmental data. The verification results can be used to support the construction of electrochemical water treatment process parameter databases, the accumulation of knowledge on electrode materials and reaction conditions, and the auxiliary literature mining and model evaluation for large language models in electrocatalytic degradation, electro-Fenton, electroadsorption, and electrodialysis.

[0012] The objective of this invention is achieved through the following technical solution:

[0013] An automated quality assessment method for environmental data includes the following steps:

[0014] Step S1, Data Access and Electrochemical Parameter Preprocessing: Receive structured extraction verification requests for literature in the field of electrochemical water treatment, and obtain the gold standard experimental parameter table and at least one experimental parameter table output by the model corresponding to the verification request.

[0015] Step S2, Source Line Alignment: Group the gold standard experimental parameter table and the experimental parameter table output by the model according to the document identifier field, compare the number of records in each group, and when the number of records in a certain document group on the model side is less than that on the standard side, insert a placeholder row for missed sampling at the end of the corresponding document identifier group in the model output table and mark the placeholder row for missed sampling as a missed detection status.

[0016] Step S3, Third-order row alignment matching for electrochemical parameters:

[0017] For each record row in the model output table, excluding the placeholder rows that were missed, execute the following steps in sequence:

[0018] First-order precise complex bond matching: Based on the literature identifier and the standardized electrochemical anchor point parameters, precise complex bonds are constructed, and first-order precise complex bond matching is performed in the standard table;

[0019] Secondary source intra-anchor loose matching: If the exact bond is not matched within the same literature group, secondary source intra-anchor loose matching is performed within the same literature identifier group based on the relative error threshold of the electrochemical anchor parameters.

[0020] Level 3 failure explicit marking: If the match is still not found, the row is marked as an incorrect fill and will not be included in subsequent field-level comparisons;

[0021] Step S4, Heterogeneous Field Judgment in the Electrochemical Water Treatment Field: For the successfully aligned record rows, perform heterogeneous matching judgment in the electrochemical water treatment field field by field, including applying the first numerical tolerance to anchor point numerical fields, applying the second numerical tolerance to non-anchor point numerical fields, performing semantic equivalence judgment on text fields based on a preset electrochemical water treatment terminology knowledge base, and performing four-class mapping on null values ​​to obtain the correct, consistent null values, incorrect filling or missing filling status of each cell;

[0022] Step S5, Engineering Result Output: Encode the cell judgment results into color or symbol visual identifiers, generate annotation files that can be reviewed by R&D personnel, and output a quality report for the construction of the electrochemical water treatment process parameter library.

[0023] A verification device for extracting experimental parameters from electrochemical water treatment literature using the above method includes a data access module, a parameter standardization module, a line record alignment module, a domain matching module, and a result output module, wherein:

[0024] The data access module is used to receive verification requests and obtain the gold standard experimental parameter table and the experimental parameter table output by the model. The experimental parameter table is structured data of experimental records from electrochemical water treatment literature.

[0025] The parameter standardization module is used to perform unit conversion, numerical standardization and terminology field preprocessing for the electrochemical anchor point parameters in the experimental parameter table.

[0026] The row record alignment module is used to perform source intra-row cardinality alignment and three-level row alignment: first-level precise composite key matching, second-level source intra-anchor loose matching, and third-level explicit failure marking.

[0027] The domain matching module is used to call a pre-set electrochemical water treatment terminology knowledge base to perform heterogeneous matching judgment and null value four-class classification mapping in the field of electrochemical water treatment.

[0028] The result output module is used to encode cell status as visual identifiers, generate annotation files and quality reports for the construction of an electrochemical water treatment process parameter library.

[0029] A verification system for the extraction results of experimental parameters from electrochemical water treatment literature includes the aforementioned verification device, literature data interface, model service interface, and display terminal, wherein:

[0030] The literature data interface is used to obtain structured tables to be verified from literature databases or experimental data management platforms;

[0031] The model service interface is used to receive the experimental parameter table output by the large language model or information extraction model;

[0032] The display terminal is used to display the labeled verification result documents and quality reports.

[0033] An electronic device includes one or more processors and a memory; the memory is used to store one or more programs; when the one or more programs are executed by one or more processors, the one or more processors implement the above-described automated quality assessment method for environmental data.

[0034] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described automated quality assessment method for environmental data.

[0035] Compared with existing general text comparison schemes, the present invention has the following advantages:

[0036] 1. Significantly reduces missed detections and misjudgments in the extraction and verification of electrochemical water treatment parameters;

[0037] 2. Standardize the format and unit differences of key parameters such as processing rate constant, current density, removal rate, and specific energy consumption;

[0038] 3. Utilize the electrochemical water treatment terminology knowledge base to correctly identify BDD / DSA electrode materials, target pollutants, and synonymous expressions for processes such as electro-Fenton / electrocatalytic oxidation;

[0039] 4. Supports manual review at the cell level and comparison of multiple models;

[0040] 5. The solution is implemented in the form of a device or system, meeting the requirements of the Patent Law regarding technical solutions. Attached Figure Description

[0041] Figure 1Flowchart of the method for verifying the results of parameter extraction from electrochemical water treatment literature;

[0042] Figure 2 This is a diagram illustrating three-level line alignment matching.

[0043] Figure 3 This is a schematic diagram of the verification device.

[0044] Figure 4 This is a diagram illustrating the deployment of the verification system. Detailed Implementation

[0045] The technical solution of the present invention will be further described below with reference to the accompanying drawings, but it is not limited thereto. Any modifications or equivalent substitutions to the technical solution of the present invention that do not depart from the spirit and scope of the technical solution of the present invention should be covered within the protection scope of the present invention.

[0046] This invention provides an automated quality assessment method for environmental data: First, the gold standard experimental record table and the model output table are aligned in terms of source row numbers based on literature identifiers, and placeholder records for missed samples are explicitly inserted. Next, precise matching is performed using a composite bond formed by the literature identifiers and standardized electrochemical anchor parameters (such as degradation rate constant k, current density, TOC / COD removal rate, etc.), while a more relaxed matching is performed within the same literature group based on the relative tolerance of the anchor parameters. Then, semantic equivalence determination is performed by calling a terminology knowledge base in the field of electrochemical water treatment for fields such as anode / cathode materials, target pollutants, electrolytes, reactor configuration, and operating conditions. Finally, the cell verification results are encoded into visual identifiers, and a process parameter quality report is generated. This invention deploys the verification process in a verification device or system containing a processor, memory, and data communication interface, which can interface with literature databases, experimental data management platforms (LIMS), and model inference services. This significantly reduces missed detections and misjudgments in the extraction and evaluation of electrochemical water treatment parameters, providing support for engineering applications such as process parameter library construction, electrode selection and comparison, literature mining, and model evaluation. The specific steps are as follows:

[0047] Step S1, Data Access and Electrochemical Parameter Preprocessing: Receive structured extraction verification requests for literature in the field of electrochemical water treatment, and obtain the gold standard experimental parameter table and at least one model output experimental parameter table corresponding to the verification request; wherein, the gold standard experimental parameter table and the model output experimental parameter table are both derived from experimental records of literature related to electrocatalytic oxidation, electro-Fenton, electrolysis, electrocoagulation, electrodialysis, or electrochemical advanced reduction, and at least include a literature identification field, an electrochemical anchor parameter field, and a process attribute field; the electrochemical anchor parameter field includes at least one of degradation rate constant, current density, cell voltage, charge, specific energy consumption, pollutant removal rate, mineralization rate, or TOC / COD removal rate; the process attribute field includes at least one of anode material, cathode material, anode area, cathode area, target pollutant, initial pollutant concentration, electrolyte type, electrolyte concentration, pH, temperature, electrode spacing, effective volume, reaction time, or reactor type.

[0048] Step S2, Source Line Alignment: Group the gold standard experimental parameter table and the model output experimental parameter table according to the document identification field (such as DOI, CNKI article title ID or internal document number), compare the number of records in each group, and when the number of records in a certain document group on the model side is less than that on the standard side, insert a placeholder row for missed detection at the end of the corresponding document identification group in the model output table, and mark the placeholder row for missed detection as a missed detection, so that the missed detection behavior of the model is included in the subsequent statistics, and avoids the recall rate being overestimated.

[0049] Step S3, Level 3 row alignment matching for electrochemical parameters: Perform the following steps sequentially on each record row in the model output table, excluding any missing placeholder rows:

[0050] First-order precise complex bond matching: Based on the literature identifier and the standardized electrochemical anchor parameters (such as the converted degradation rate constant k, current density, TOC / COD removal rate) to form precise complex bonds, first-order precise complex bond matching is performed in the standard table. Among them, the standardized electrochemical anchor parameters need to convert the degradation rate constant, current density or removal rate into a uniform unit and format it as a fixed-point decimal or a uniform significant number format.

[0051] Secondary source anchor loose matching: If an exact bond is not matched within the same document group, secondary source anchor loose matching is performed within the same document identifier group based on the relative error threshold (e.g., 1%) of the electrochemical anchor parameters (degradation rate constant k, current density, or TOC / COD removal rate) to tolerate scientific notation, significant figures, and slight experimental fluctuations. When multiple standard rows that meet the secondary source anchor loose matching conditions exist within the same document identifier group, a one-to-one optimal allocation algorithm is used to determine the matching relationship.

[0052] Level 3 failure explicit marking: If the match is still not found, the row is marked as an incorrect fill and will not be included in subsequent field-level comparisons;

[0053] Step S4: Heterogeneous Field Determination in the Electrochemical Water Treatment Field: For successfully aligned record rows, perform heterogeneous matching determination in the electrochemical water treatment field field field by field. This includes applying a first numerical tolerance (i.e., a stricter tolerance) to anchor point numerical fields (k, current density, removal rate, etc.), applying a second numerical tolerance (i.e., a looser tolerance) to non-anchor point numerical fields (pH, temperature, electrode spacing, specific energy consumption, etc.), and performing semantic equivalence determination on text fields (anodine / cathode materials, pollutants, electrolytes, reactor types, and process types, etc.) based on a pre-set electrochemical water treatment terminology knowledge base. Null values ​​are then mapped using a four-category method to obtain the correct, consistent, incorrectly filled, or missing values ​​for each cell. The first numerical tolerance is 1%, and the second numerical tolerance is 5%. The electrochemical water treatment terminology knowledge base includes at least electrode material equivalence classes, pollutant name equivalence classes, electrolyte equivalence classes, and reactor type equivalence classes. The electrode material equivalence classes cover boron-doped diamond electrodes, Ti / BDD electrodes, Ti / IrO2 electrodes, Ti / RuO2 electrodes, and PbO2 electrodes. Synonyms for electrode, DSA electrode, and three-dimensional particle electrode.

[0054] Step S5, Engineering Result Output: Encode the cell judgment results with color or symbol identifiers, generate annotation files that can be reviewed by R&D personnel, and output a quality report for the construction of the electrochemical water treatment process parameter library, including field-level accuracy, recall, and missed detection statistics grouped by anode material, target pollutant, reactor type, or process type.

[0055] Step S6, Device and System Deployment: Modularize the above functions into a data access module, a parameter standardization module, a line record alignment module, a domain matching module, and a result output module, and deploy them in a verification device, verification system, or electronic device including a processor and memory, and obtain the data to be verified through the literature data interface and the model service interface.

[0056] The above method also includes: unifying the dimensions of electrochemical parameters in the experimental parameter table before verification, and converting current density, pollutant concentration, rate constant, removal rate and specific energy consumption into preset benchmark units.

[0057] This invention also provides a verification device for extracting experimental parameters from electrochemical water treatment literature. The device includes a data access module, a parameter standardization module, a line record alignment module, a domain matching module, and a result output module, wherein:

[0058] The data access module is used to receive verification requests and obtain the gold standard experimental parameter table and the experimental parameter table output by the model. The experimental parameter table is structured data of experimental records from electrochemical water treatment literature.

[0059] The parameter standardization module is used to perform unit conversion, numerical standardization and terminology field preprocessing for the electrochemical anchor point parameters in the experimental parameter table.

[0060] The row record alignment module is used to perform source intra-row cardinality alignment and three-level row alignment: first-level precise composite key matching, second-level source intra-anchor loose matching, and third-level explicit failure marking.

[0061] The domain matching module is used to call a pre-set electrochemical water treatment terminology knowledge base to perform heterogeneous matching judgment and null value four-class classification mapping in the field of electrochemical water treatment.

[0062] The result output module is used to encode cell status as visual identifiers, generate annotation files and quality reports for the construction of an electrochemical water treatment process parameter library.

[0063] This invention also provides a verification system for extracting experimental parameters from electrochemical water treatment literature. The system includes the aforementioned verification device, a literature data interface, a model service interface, and a display terminal, wherein:

[0064] The literature data interface is used to obtain structured tables to be verified from literature databases or experimental data management platforms;

[0065] The model service interface is used to receive the experimental parameter table output by the large language model or information extraction model;

[0066] The display terminal is used to display the labeled verification result documents and quality reports.

[0067] The present invention also provides an electronic device comprising one or more processors and a memory; the memory is used to store one or more programs; when one or more programs are executed by one or more processors, the one or more processors implement the above-described automated quality assessment method for environmental data.

[0068] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described automated quality assessment method for environmental data.

[0069] Example 1:

[0070] like Figure 1As shown, this embodiment uses parameter extraction and verification from literature on the electrocatalytic oxidation degradation of organic pollutants as an example. An electrochemical water treatment R&D team constructed a gold standard experimental parameter table from 50 articles and verified the extraction results using three large language models. The data table fields are shown in Table 1:

[0071] Table 1

[0072]

[0073] Step S1: Data Access and Preprocessing. The data access module of the verification device loads the gold standard Excel file and the model output Excel file from the experimental data management platform. The parameter standardization module parses the column names, excluding non-verification fields such as notes and chart numbers; it loads the electrochemical water treatment terminology knowledge base, for example:

[0074] Electrode materials: Ti / BDD, BDD, boron-doped diamond, boron-doped diamond electrode, boron-doped diamond electrode, Ti / IrO2, Ti / RuO2, DSA.

[0075] Pollutants: phenol; tetracycline; TC.

[0076] Process types: electrocatalytic oxidation, anodic oxidation, electrocatalytic oxidation; electro-Fenton, electro-Fenton, EF.

[0077] Reactors: plate-and-frame electrolytic cell, fluidized bed reactor, electroadsorption cell.

[0078] At the same time, the units for fields such as current density, rate constant, and specific energy consumption are standardized, for example, mA / cm². 2 Convert to mA·cm -2 , will h -1 Convert to min -1 , will kWh / m 3 Converted to kWh·m -3 .

[0079] Step S2: Source row cardinality alignment. The row record alignment module groups by the source field. If a DOI has only 2 experimental records in the model table, but 3 in the gold standard table (e.g., comparing 10, 20, and 30 mA / cm² in the same literature), ... 2If there are three sets of current density experiments, then insert a placeholder row for missed sampling at the end of the DOI group. Except for the source field, all other fields are empty and marked as missed (e.g., filled in purple) so that the missed sampling can be included in the statistics.

[0080] Step S3: Three-level line alignment matching. For example... Figure 2 As shown, for non-placeholder rows in the model table, a first-level match is performed: the DOI is combined with the normalized rate_constant (e.g., 0.0123 formatted as 0.012300) to form a composite key, which is then searched in the standard table index. If this fails, a second-level match is performed within the same DOI group: pressing |k m - k s | / |k s | < 1% match, for example, 1.23e-2 and 0.0123 are considered a match; secondary matching can also be performed on the current_density field using the same relative error threshold. If it still fails, it is marked as incorrect filling and will not participate in the field-level comparison.

[0081] Step S4: Electrochemical Domain Field Heterogeneity Determination. The domain matching module performs field-by-field determination on the aligned rows. For example, the anode_material field "Ti / BDD" is determined to be semantically equivalent to the standard "boron-doped diamond electrode" using the terminology knowledge base; the process_type field "electrocatalytic oxidation" is determined to be semantically equivalent to the standard "electrocatalytic oxidation"; the current_density field is determined to be correct within a 5% tolerance; and null values ​​in target_pollutant are determined to be missing from the standard non-null values.

[0082] Steps S5 to S7: Filter invalid rows, generate an Excel or HTML annotation file with color annotations, and output a quality report, including field-level recall, precision, and missed detection statistics grouped by anode_material, target_pollutant, process_type, and reactor_type.

[0083] Example 2:

[0084] like Figure 3 As shown, the verification device includes a processor, a memory, and a data communication interface interconnected via a data bus or network. The memory stores a program implementing the method of Example 1 and a knowledge base of electrochemical water treatment terminology; when the processor executes the program, the functional modules work together to complete the verification task. The device can be a workstation, server, or embedded industrial computer, deployed in an electrochemical water treatment laboratory or an environmental protection company's R&D data platform.

[0085] Example 3:

[0086] like Figure 4 As shown, the verification system includes the verification device of Example 2, a literature data interface, a model service interface, and a display terminal. The literature data interface exports records from CNKI, Web of Science, or obtains gold standard tables from the internal LIMS system; the model service interface receives structured tables returned by the model extraction API; and the display terminal allows process engineers to review the annotation results and decide whether to add them to the electrochemical water treatment process parameter database.

[0087] Example 4:

[0088] An electronic device includes a processor and a memory, wherein a computer program is stored in the memory, and when the program is executed by the processor, it implements the method of Embodiment 1.

[0089] Example 5:

[0090] A computer program stored in a computer-readable storage medium (such as a USB flash drive, ROM, RAM, or solid-state drive) is executed by a processor to implement the method of Embodiment 1.

Claims

1. An automated quality assessment method for environmental data, characterized in that... The method includes the following steps: Step S1, Data Access and Electrochemical Parameter Preprocessing: Receive structured extraction verification requests for literature in the field of electrochemical water treatment, and obtain the gold standard experimental parameter table and at least one experimental parameter table output by the model corresponding to the verification request. Step S2, Source Line Alignment: Group the gold standard experimental parameter table and the experimental parameter table output by the model according to the document identifier field, compare the number of records in each group, and when the number of records in a certain document group on the model side is less than that on the standard side, insert a placeholder row for missed sampling at the end of the corresponding document identifier group in the model output table and mark the placeholder row for missed sampling as a missed detection status. Step S3, Third-order row alignment matching for electrochemical parameters: For each record row in the model output table, excluding the placeholder rows that were missed, execute the following steps in sequence: First-order precise complex bond matching: Based on the literature identifier and the standardized electrochemical anchor point parameters, precise complex bonds are constructed, and first-order precise complex bond matching is performed in the standard table; Secondary source intra-anchor loose matching: If the exact bond is not matched within the same literature group, secondary source intra-anchor loose matching is performed within the same literature identifier group based on the relative error threshold of the electrochemical anchor parameters. Level 3 failure explicit marking: If the match is still not found, the row is marked as an incorrect fill and will not be included in subsequent field-level comparisons; Step S4, Heterogeneous Field Judgment in the Electrochemical Water Treatment Field: For the successfully aligned record rows, perform heterogeneous matching judgment in the electrochemical water treatment field field by field, including applying the first numerical tolerance to anchor point numerical fields, applying the second numerical tolerance to non-anchor point numerical fields, performing semantic equivalence judgment on text fields based on a preset electrochemical water treatment terminology knowledge base, and performing four-class mapping on null values ​​to obtain the correct, consistent null values, incorrect filling or missing filling status of each cell; Step S5, Engineering Result Output: Encode the cell judgment results into color or symbol visual identifiers, generate annotation files that can be reviewed by R&D personnel, and output a quality report for the construction of the electrochemical water treatment process parameter library.

2. The automated quality assessment method for environmental data according to claim 1, characterized in that... In step S1, both the gold standard experimental parameter table and the experimental parameter table output by the model are derived from experimental records in literature related to electrocatalytic oxidation, electro-Fenton, electrolysis, electrocoagulation, electrodialysis, or electrochemical advanced reduction, and at least include a literature identification field, an electrochemical anchor parameter field, and a process attribute field. The electrochemical anchor parameter field includes at least one of the following: degradation rate constant, current density, cell voltage, specific energy consumption, pollutant removal rate, mineralization rate, or TOC / COD removal rate. The process attribute field includes at least one of the following: anode material, cathode material, anode area, cathode area, target pollutant, initial pollutant concentration, electrolyte type, electrolyte concentration, pH, temperature, electrode spacing, effective volume, reaction time, or reactor type.

3. The automated quality assessment method for environmental data according to claim 1, characterized in that... In step S3, the standardized electrochemical anchor point parameters need to be converted from degradation rate constant, current density, or removal rate to a uniform unit and formatted as a fixed-point decimal or a uniform significant number format; when there are multiple standard rows that meet the loose matching conditions of anchor points within the same document identifier group, a one-to-one optimal allocation algorithm is used to determine the matching relationship.

4. The automated quality assessment method for environmental data according to claim 1, characterized in that... In step S4, the first numerical tolerance is 1%, and the second numerical tolerance is 5%. The electrochemical water treatment terminology knowledge base includes at least electrode material equivalence classes, pollutant name equivalence classes, electrolyte equivalence classes, and reactor type equivalence classes. The electrode material equivalence classes cover synonymous expressions for boron-doped diamond electrodes, Ti / BDD electrodes, Ti / IrO2 electrodes, Ti / RuO2 electrodes, PbO2 electrodes, DSA electrodes, and three-dimensional particle electrodes.

5. The automated quality assessment method for environmental data according to claim 1, characterized in that... In step S5, the quality report includes field-level accuracy, recall, and missed detection statistics grouped by anode material, target contaminant, reactor type, or process type.

6. The automated quality assessment method for environmental data according to claim 1, characterized in that... The method further includes: before verification, performing dimensional unification of electrochemical parameters in the experimental parameter table, converting current density, pollutant concentration, rate constant, removal rate and specific energy consumption into preset benchmark units.

7. A verification device for extracting experimental parameters from electrochemical water treatment literature using the method described in any one of claims 1-6, characterized in that... The device includes a data access module, a parameter standardization module, a row record alignment module, a domain matching module, and a result output module, wherein: The data access module is used to receive verification requests and obtain the gold standard experimental parameter table and the experimental parameter table output by the model. The experimental parameter table is structured data of experimental records from electrochemical water treatment literature. The parameter standardization module is used to perform unit conversion, numerical standardization and terminology field preprocessing for the electrochemical anchor point parameters in the experimental parameter table. The row record alignment module is used to perform source intra-row cardinality alignment and three-level row alignment: first-level precise composite key matching, second-level source intra-anchor loose matching, and third-level explicit failure marking. The domain matching module is used to call a pre-set electrochemical water treatment terminology knowledge base to perform heterogeneous matching judgment and null value four-class classification mapping in the field of electrochemical water treatment. The result output module is used to encode cell status as visual identifiers, generate annotation files and quality reports for the construction of an electrochemical water treatment process parameter library.

8. A verification system for extracting experimental parameters from electrochemical water treatment literature, characterized in that... The system includes the verification device as described in claim 7, a document data interface, a model service interface, and a display terminal, wherein: The literature data interface is used to obtain structured tables to be verified from literature databases or experimental data management platforms; The model service interface is used to receive the experimental parameter table output by the large language model or information extraction model; The display terminal is used to display the labeled verification result documents and quality reports.

9. An electronic device comprising one or more processors and a memory; said memory being used to store one or more programs; characterized in that... When one or more programs are executed by one or more processors, the one or more processors implement the automated quality assessment method for environmental data as described in any one of claims 1-6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that... When the computer program is executed by a processor, it implements the automated quality assessment method for environmental data as described in any one of claims 1-6.

Images