Manuscript review task distribution method and system

Abstract

The invention provides a manuscript review task allocation method and system. The method comprises the steps of receiving a target manuscript review task, and determining a research theme of the target manuscript review task; screening out manuscript reviewing personnel matched with the research subject; predicting the manuscript reviewing capability and response speed of the manuscript reviewing personnel for the target manuscript reviewing task, and calculating a corresponding manuscript reviewing suitability index according to a prediction result; selecting the manuscript reviewing personnel of which the manuscript reviewing suitability index reaches a preset condition to form a manuscript reviewing group; and fusing the manuscript reviewing opinions of the manuscript reviewing personnel corresponding to the manuscript reviewing group by using an integrated learning strategy to form a manuscript reviewing conclusion of the target manuscript reviewing task. According to the technical scheme provided by the invention, the corresponding manuscript reviewing suitability index can be calculated according to the predicted manuscript reviewing capability and response speed of the manuscript reviewing personnel for the target manuscript reviewing task, so that the manuscript reviewing personnel suitable for the target manuscript reviewing task are screened out, and the distribution efficiency of the manuscript reviewing task is improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computer technology, and in particular to a method and system for allocating review tasks. Background Art

[0002] In the field of modern academic publishing, the review process is a key link in ensuring research quality and academic integrity. With the increasing number of scientific research results, the number of review tasks is also increasing, which puts higher demands on the efficiency and quality of the review process. The traditional way of manually assigning review tasks is not only time-consuming and labor-intensive, but also easily leads to uneven review quality.

[0003] The currently commonly used manual allocation method mainly relies on the editorial department's experience and judgment, and editors usually manually select appropriate reviewers based on the subject of the manuscript and the reviewers' professional fields.

[0004] However, although this method is intuitive and easy to understand, it is inefficient when processing a large number of manuscripts, and it is difficult to ensure that the most suitable candidate can be found every time, resulting in the problem of low efficiency in the allocation of review tasks. Summary of the invention

[0005] The embodiments of the present application provide a method and system for allocating review tasks, so as to solve the problem of low efficiency in allocating review tasks in the prior art.

[0006] In a first aspect, an embodiment of the present application provides a method for allocating review tasks, including: Receive target review tasks and determine the research topics of the target review tasks; From the pre-built reviewer database, select reviewers that match the research topic; Based on the reviewer's historical review records and current available time, predict the reviewer's reviewing ability and response speed for the target review task through data analysis, and calculate the corresponding review suitability index according to the prediction result; Selecting reviewers whose review suitability index meets the predetermined conditions to form a review group, and planning a suitable review time period for each reviewer in the review group through an intelligent time management algorithm, so that each reviewer can provide review opinions for the target review task at the corresponding review time; An integrated learning strategy is used to integrate the review opinions of multiple reviewers corresponding to the review group to form a review conclusion for the target review task.

[0007] Optionally, the historical manuscript review records include at least: manuscript review time data and quality assessment data; The method of predicting the reviewing ability and response speed of the reviewer for the target reviewing task through data analysis based on the reviewer's historical reviewing records and current available time includes: Predicting the review response speed of the reviewer for the target review task based on the review time data in the reviewer's historical review records and the current available time; Predicting the reviewing ability of the reviewer for the target reviewing task based on the quality assessment data in the reviewer's historical review records; Taking the review response speed and the review ability as prediction results; The following calculation formula is used to predict the review response speed of the reviewer for the target review task: ; in, The review time data in the reviewer's historical review records; is the current available time of the reviewer in question; Represents the data at a given historical review time and current available time Under the condition of Actual completion time Expected value; It is the posterior distribution calculated based on the historical review time and the current available time; is the expected completion time of the reviewer for the target review task, that is, the predicted value of the review response speed; is the combined effect of review time and current available time; the denominator It is the time for all possible review Multiply it by its corresponding current available time The result of post-integration; The following calculation formula is used to predict the reviewer's review ability for the target review task: ; in, is the feature vector of the reviewer; is the prediction function of the random forest model; are the parameters of the random forest model; It is the review ability score for historical review tasks; is the number of samples in the training dataset; is the reviewer's review ability score for the target review task; Indicates finding a set of optimal parameters , so that the actual review ability score of all training samples With the model prediction value The sum of squared differences between them is minimal.

[0008] Optionally, predicting the review response speed of the reviewer for the target review task according to the review time data in the reviewer's historical review records and the current available time includes: Based on the review time data in the historical review records of the reviewers, a Bayesian statistical model is constructed to describe the prior distribution of the review time; Using the Bayesian update rule, the prior distribution is updated according to the latest review time data of the reviewer to obtain the posterior distribution; Based on the posterior distribution and in combination with the current available time of the reviewer, the expected completion time of the reviewer for the target review task is estimated as a predicted value of the reviewer's review response speed.

[0009] Optionally, the predicting the reviewing ability of the reviewer for the target reviewing task based on the quality assessment data in the reviewer's historical review records includes: Collecting quality assessment data from the reviewers' historical review records, including the adoption rate of the review opinions, the level of detail of the review reports, and the usefulness scores of the review opinions; Constructing a training data set based on the quality assessment data, wherein each piece of training data in the training data set contains a feature vector of the reviewer and a reviewing ability score for a historical review task, wherein the feature vector includes the reviewer's title, institution, professional field, number of past reviews, and average review time; Training a random forest model, using the feature vectors in the training data set as input and the reviewing ability scores of the corresponding historical review tasks as labels, to learn the mapping relationship between the reviewing ability and the feature vectors; The trained random forest model is used to output the reviewer's review ability score for the target review task, and used as a result of predicting the reviewer's review ability for the target review task.

[0010] Optionally, in the process of predicting the reviewing ability and response speed of the reviewer for the target reviewing task through data analysis, the method further includes: Using a deep neural network to model the behavior patterns of the reviewers, combining the review preferences, historical behavior trajectories, and external influencing factors of the reviewers, to refine the prediction of the reviewing ability and the response speed, so as to update the prediction results; A reinforcement learning algorithm is introduced to dynamically adjust the parameters of the prediction model according to the completion status of the historical review tasks by the reviewers at different time points to improve the prediction accuracy of the prediction results, wherein the adjusted parameters of the prediction model include at least the prior probability distribution parameters in the Bayesian statistical model, the hyperparameters in the random forest algorithm, the reward function and discount factor in the reinforcement learning algorithm, and the weight and bias parameters in the deep neural network.

[0011] Optionally, the step of calculating the corresponding review suitability index according to the prediction result includes: Based on the current available time of the reviewer, a time series analysis algorithm is used to predict the available time period of the reviewer in the future; Calculating the review suitability index of the reviewer according to the prediction result and the available time period; The review suitability index of the reviewer is calculated by the following formula: ; in, represents the review suitability index; represents the number of reviewers; is the reviewer's comprehensive weight factor, which is a score determined based on the reviewer's feature vector and the reviewer ability score for historical review tasks; It indicates the expected completion time of the reviewer for the target review task. is a very small positive number used to prevent division by zero errors; is with The relevant exponential factor, using the exponential function , to emphasize the importance of reviewer response time, a nonlinear coefficient determined based on the statistical distribution characteristics of reviewers' past response times; It represents the predicted reviewer's review ability score for the target review task; is with The relevant exponential factor is expressed in square roots. , used to adjust the impact of reviewer ability score; It represents the estimated value of the available time period of the reviewer for the target review task, which is a quantitative evaluation of the future available time period predicted by the time series analysis algorithm; is with The relevant exponential factor is , used to adjust the importance of the available time periods; It indicates the number of time conflicts that the reviewer will have with other review tasks in the future, which is a negative indicator; is with The relevant exponential factor is , used to adjust the impact of the number of conflicts.

[0012] Optionally, screening out reviewers matching the research topic from a pre-built reviewer database includes: Extract keywords and phrases from the target review task to identify the core elements of the research topic; Converting the core elements into a structured data format, and matching the core elements of the structured data format with the professional field information of the reviewers stored in the reviewer database to identify reviewers matching the research topic; Filter out the best list of reviewers based on their professional field matching, past review experience, and review quality evaluation information; Select qualified reviewers from the optimal reviewer list.

[0013] Optionally, the selecting reviewers whose review suitability index meets predetermined conditions to form a review group, and planning a suitable review period for each reviewer in the review group through an intelligent time management algorithm, includes: According to the review suitability index of the reviewers, a predetermined threshold is set, and reviewers with a suitability index higher than the predetermined threshold are screened out; Include the selected reviewers in the candidate review team list; Analyzing the current workload and current available time of each reviewer in the candidate review panel list using an intelligent time management algorithm; Based on each reviewer's current workload and current available time, combined with each reviewer's available time period in the future, an optimization algorithm is used to assign the most suitable review time period to each reviewer, to ensure that each reviewer can focus on the review task of the target review task within the specified time period, so as to improve the review efficiency and quality.

[0014] Optionally, the use of an integrated learning strategy to integrate the review opinions of multiple reviewers corresponding to the review team to form a review conclusion for the target review task includes: Collecting the review opinions provided by each reviewer in the review team and converting the review opinions into standardized review scores; Using an integrated learning framework, the standardized review score is input into an integrated model, wherein the integrated model is composed of multiple basic models, wherein the basic models at least include support vector machines, logistic regression, and neural networks, and each basic model independently evaluates the review score; Based on the evaluation results of the multiple basic models, the final review conclusion is determined by weighted voting or averaging strategy, wherein the weighted voting strategy assigns different weights according to the historical performance of each basic model to ensure the objectivity and fairness of the final review conclusion; The review conclusion for the target review task is formed by the following calculation formula: ; in, It is the review score converted from the final review conclusion. It is the standardized review score converted from the review opinions provided by each reviewer in the review team; It is The prediction function of the base model; is the number of base models; It is The weights of the base models; is the regulating factor; It is The historical loss value of the base model; is a normalization term that ensures that the sum of the weights of all base models is 1.

[0015] In a second aspect, an embodiment of the present application provides a system for allocating review tasks, including: A determination module, used to receive a target review task and determine a research topic of the target review task; A screening module is used to screen out reviewers matching the research topic from a pre-built reviewer database; A prediction module, used to predict the reviewing ability and response speed of the reviewer through data analysis based on the reviewer's historical review records and current available time, and calculate the corresponding review suitability index according to the prediction results; A processing module is used to select reviewers whose review suitability index meets the predetermined conditions to form a review group, and plan a suitable review time period for each reviewer in the review group through an intelligent time management algorithm, so that each reviewer can provide review opinions for the target review task at the corresponding review time; A generation module is used to use an integrated learning strategy to integrate the review opinions of multiple reviewers corresponding to the review group to form a review conclusion for the target review task.

[0016] In an embodiment of the present application, a target review task is received, and a research topic of the target review task is determined; reviewers matching the research topic are screened out; the reviewer's reviewing ability and response speed for the target review task are predicted, and the corresponding review suitability index is calculated based on the prediction result; reviewers whose review suitability index meets predetermined conditions are selected to form a review group; an integrated learning strategy is used to fuse the review opinions of multiple reviewers corresponding to the review group to form a review conclusion for the target review task. The technical solution provided in the present application can calculate the corresponding review suitability index based on the predicted reviewer's reviewing ability and response speed for the target review task, so as to screen out reviewers suitable for the target review task, thereby improving the allocation efficiency of the review task.

[0017] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 A flowchart of a method for allocating review tasks provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of a system for distributing manuscript review tasks provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0021] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0022] Figure 1 A flowchart of a method for assigning review tasks is provided for an embodiment of the present application, such as Figure 1 As shown, the method includes: 101. Receive the target review task and determine the research topic of the target review task; In this step, the editor or system receives an article for review and identifies the research field or specific topic to which the article belongs.

[0023] In this embodiment of the application, it is assumed that an editor receives a manuscript submitted to the journal Computer Vision and Pattern Recognition via email. The editor reads the abstract and keywords and confirms that the research belongs to the topic of "Application of Deep Learning in Image Recognition".

[0024] 102. Screen out reviewers who match the research topic from the pre-built reviewer database; In this step, the established expert database is used to find suitable reviewers based on the research field.

[0025] In an embodiment of the present application, the system automatically filters out a list of reviewers marked with relevant expertise tags such as "deep learning" and "image processing" from a database containing information on thousands of scholars.

[0026] Optionally, the step 102 of "screening out reviewers matching the research topic in a pre-constructed reviewer database" includes: extracting keywords and phrases in the target review task to determine the core elements of the research topic; converting the core elements into a structured data format, and matching the core elements in the structured data format with the professional field information of the reviewers stored in the reviewer database to identify reviewers matching the research topic; screening out an optimal list of reviewers based on the reviewers' professional field matching, past review experience and review quality evaluation information; and selecting qualified reviewers from the optimal list of reviewers.

[0027] Keywords: refers to words that can summarize the main content or research focus of the article.

[0028] Phrase: A language unit consisting of two or more words that expresses a specific meaning.

[0029] Core elements: The most critical parts of the research topic extracted through analysis of the text content.

[0030] Structured data format: Convert unstructured information into a form that is easy for computers to process.

[0031] Area of ​​expertise: Describe the specific subject area in which the reviewer has specialized during their career.

[0032] Match: An indicator that measures the relevance between the reviewer's professional background and the topic of the paper to be reviewed.

[0033] Review experience: refers to how many times the reviewer has participated in similar reviews in the past.

[0034] Review Quality Rating: A rating of the quality of feedback provided by reviewers based on historical records.

[0035] First, use natural language processing technology to automatically identify the main words that can represent the research direction from the target review task document. Determine the core elements and convert them into structured data: further refine these words to form a concise and clear research outline and encode it into a machine-readable form. Secondly, compare the previously obtained data with all the reviewer information stored in the database to find those who cover the knowledge required for the current research in their professional field. Consider each candidate's historical performance (such as the number of reviews, the rate of positive reviews received, etc.), and select the most suitable people for this review. Finally, make a final decision from the list generated in the previous step based on actual conditions such as time arrangements.

[0036] In an embodiment of the present application, it is assumed that there is a paper on "The Application of Artificial Intelligence in Medical Image Diagnosis" that needs to be reviewed. First, the system will automatically capture frequently mentioned terms in the article such as "deep learning", "image recognition", "tumor detection", etc. as keywords. Then, this information is organized into a short description: "Explore how to use advanced AI technology to improve the accuracy of medical image analysis". This description is then encoded and compared with the existing reviewer archive, and it was found that several experts specializing in health information technology development met the requirements very well. Based on their previous work experience and the evaluations given by their peers, three candidates with the best performance were selected. Finally, considering that two of them were busy with other projects, a third expert was invited to conduct this review.

[0037] This automated and refined method can effectively reduce subjective bias caused by manual intervention, and can also more quickly locate reviewers with relevant knowledge backgrounds, thus ensuring that each manuscript can receive objective, fair and professional evaluations. In addition, this method can also help promote interdisciplinary exchanges, allowing researchers from different fields to learn from each other and jointly promote scientific progress.

[0038] 103. Based on the reviewer's historical review records and current available time, predict the reviewer's reviewing ability and response speed for the target review task through data analysis, and calculate the corresponding review suitability index according to the prediction result; In this step, each potential reviewer's past performance (e.g., review quality, punctuality) and their current schedule are analyzed to assess their ability to complete the new task.

[0039] In the embodiment of the present application, each candidate reviewer is scored, for example, based on the number of their past reviews, quality feedback, average response time, and recent workload, etc. Finally, a value between 0 and 10 is generated for each person as a suitability index.

[0040] Optionally, in step 103, "the historical review records include at least: review time data, quality assessment data; the reviewing ability and response speed of the reviewer for the target review task are predicted through data analysis based on the reviewer's historical review records and current available time", includes: predicting the review response speed of the reviewer for the target review task based on the review time data in the reviewer's historical review records and current available time; predicting the review ability of the reviewer for the target review task based on the quality assessment data in the reviewer's historical review records; and taking the review response speed and the review ability as prediction results.

[0041] Review time data: records the time required for reviewers to complete previous review tasks.

[0042] Quality assessment data: Contains quality feedback from other experts or authors on the review opinions.

[0043] Current available time: refers to the amount of time that a reviewer can use for new tasks in the future.

[0044] Review response speed: the speed at which reviewers are expected to start reviewing new manuscripts after receiving them.

[0045] Reviewer Skill: The reviewer’s ability to provide high-quality feedback, estimated based on past performance.

[0046] Reviewer suitability index: A value that combines response speed and review ability to measure whether the reviewer is suitable for the current task.

[0047] First, we use the data on how long the reviewer has previously spent on similar tasks (such as the average review cycle) and combine it with his / her current schedule (i.e., available time) to estimate how long it will take for him / her to start reviewing new manuscripts. Secondly, we analyze the quality scores of the reviewer's previous reviews, such as the proportion of positive reviews, specific scores, etc., and use this as a basis to estimate the level of review opinions he / she can provide. Finally, we combine the above two factors and use a specific algorithm to generate a score value that represents the overall suitability of the reviewer.

[0048] In this embodiment of the application, suppose there is a reviewer A, whose historical review record shows that he has completed 50 article reviews in the field of machine learning in the past five years, and it took an average of 7 days per article. In addition, these reviews received an 85% favorable rating. Now there is a new paper on deep neural network optimization technology that needs to be reviewed. By checking A's schedule, it is found that he has two weeks of free time in the next month. Therefore, the following formula can be used to estimate A's adaptability to this work: Review response speed = (past average review days) / (current available weeks * 7) Assuming A can start reviewing immediately within the next two weeks, then: Review response speed = 7 / (2*7)≈0.5 (which means that the review can be started within half a week after receiving the task); The review ability score is directly based on the favorable review rate: 85%; Review suitability index = review response speed * weight + review ability * weight; If the response speed weight is set to 0.3 and the capability weight is set to 0.7, then: Review suitability index = 0.5*0.3+85*0.7=0.15+59.5=59.65; In this way, editors can not only select appropriate reviewers more scientifically, but also better plan the timeline of the entire review process. This will not only help speed up the publication of research results, but also ensure that each manuscript receives high-quality professional reviews. In addition, this method can promote fair competition, giving those reviewers with excellent performance more opportunities, thereby motivating the entire academic community to improve the review standards.

[0049] This application takes into account that in the field of academic publishing, the efficiency and quality of the review process are crucial. The following problems exist in the prior art: First, editors usually rely on subjective judgment to estimate the response speed of reviewers, lacking objective data support; second, the available time of reviewers is often not fully considered, resulting in the assigned tasks may exceed the time range of reviewers, thereby affecting the speed and quality of the review; finally, traditional prediction methods fail to make full use of the information in historical data and cannot provide accurate prediction results. Therefore, an embodiment of the present invention proposes a method based on Bayesian statistics to predict the review response speed of reviewers for target review tasks, and conducts a comprehensive evaluation in combination with the current available time to solve the above-mentioned problems and improve the efficiency and accuracy of the review process.

[0050] The options are as follows: Optionally, in step 103, “predicting the review response speed of the reviewer for the target review task based on the review time data in the reviewer's historical review records and the current available time” includes: The review response speed of the reviewer for the target review task is predicted by the following calculation formula: ;

[0051] in, The review time data in the reviewer's historical review records; is the current available time of the reviewer in question; Represents the data at a given historical review time and current available time Under the condition of Actual completion time Expected value; It is the posterior distribution calculated based on the historical review time and the current available time; is the expected completion time of the reviewer for the target review task, that is, the predicted value of the review response speed; is the combined effect of review time and current available time; the denominator It is the time for all possible review Multiply it by its corresponding current available time The result of post-integration; Review time data : A record of the time a reviewer has spent on completing similar review tasks in the past.

[0052] Current available time : The amount of time a reviewer can devote to new review tasks.

[0053] Expected Value : Given the historical review time and the current available time, the expectation of the actual completion time based on the posterior distribution.

[0054] Posterior distribution : The probability distribution after combining prior knowledge (such as the general performance of reviewers) with observed data (such as historical review time).

[0055] Review response speed : The predicted completion time of a reviewer for a specific task.

[0056] First, we obtain the time data that reviewers have needed to complete similar manuscripts in the past. Secondly, we understand the actual time that reviewers can use for reviewing in the future. We further use the Bayesian formula to combine the historical review time data with the current available time to obtain a posterior distribution that describes the possibility of reviewers completing new tasks. Then, we use the provided formula to calculate the reviewer's expected completion time for the new task. Finally, we combine the predicted response speed with other factors (such as reviewing ability) to obtain a comprehensive evaluation index.

[0057] In the present embodiment, it is assumed that a reviewer B has completed 100 biomedical research papers in the past five years, and each review takes an average of 10 days. Now there is a new biomedical research article to be reviewed. Checking B's schedule, it is found that he has 20 days in the next month to review.

[0058] Historical review time data Days, current available time sky.

[0059] Compute the posterior distribution: , since this is a simplified example, we can assume that the result of the integral part is a constant, so that Simplified to proportional relationship.

[0060] Calculate the review response speed according to the formula: days, which means that after taking into account the current available time, reviewer B is expected to complete the review task in approximately 10 days.

[0061] This method can more accurately predict the response speed of reviewers, thereby helping editors to arrange review tasks reasonably and avoid delays due to insufficient reviewer time. At the same time, by combining historical data with actual conditions, the reliability of predictions is improved, which helps to improve the efficiency of the entire review process. In addition, this method can also serve as a basis for further developing more complex models to more comprehensively evaluate the overall performance of reviewers.

[0062] Optionally, the step 103 of "predicting the review response speed of the reviewer for the target review task based on the review time data in the reviewer's historical review records and the current available time" includes: constructing a Bayesian statistical model based on the review time data in the reviewer's historical review records to describe the prior distribution of the review time; using the Bayesian update rule to update the prior distribution based on the latest review time data obtained from the reviewer to obtain the posterior distribution; based on the posterior distribution and in combination with the reviewer's current available time, estimating the reviewer's expected completion time for the target review task as a predicted value of the reviewer's review response speed.

[0063] Prior distribution: The probability distribution of possible values ​​of a parameter based on existing knowledge before obtaining new observations.

[0064] Posterior distribution: The probability distribution after combining prior information with new observation data, reflecting the updated knowledge state.

[0065] Bayesian Update Rule: A mathematical formula for calculating the posterior distribution from a prior distribution and new data.

[0066] Expected Completion Time: The time the model predicts a reviewer will take to complete a specific review task.

[0067] Review response speed: that is, the expected completion time, which represents the speed from when the reviewer starts reviewing to when he submits his review opinions.

[0068] First, we use the time data in the reviewers’ historical review records to build an initial probability distribution model describing the review time. Secondly, when we obtain the latest review time data, we use Bayes’ theorem to update the prior distribution and form a posterior distribution that reflects the latest situation. Finally, we use the posterior distribution to estimate the expected time for reviewers to complete new tasks based on their current available time.

[0069] Assume that there is a normal distribution as the prior distribution, where the mean and variance It is based on historical data. When there is new review time data When it occurs, the following Bayesian update rule can be used: Updated mean ; Updated variance ; in, is the variance of the new data, assumed to be known or can be estimated from the data.

[0070] In this example, it is assumed that a reviewer C has completed 50 paper reviews in the field of machine learning in the past, with an average time of 8 days and a standard deviation of 2 days. This means that the prior distribution can be represented by a normal distribution, with a mean of ,variance .

[0071] Recently, C completed 3 more reviews, taking 7 days, 9 days, and 6 days respectively. These new data can be added to the model: the average value of the new data , assuming the standard deviation of the new data is Now (for simplicity) we apply the Bayesian update rule: Updated mean ; Updated variance ; If reviewer C is available for 20 days in the next month, he can be expected to complete the new review task in about 7.5 days, which is reasonable considering his current available time.

[0072] By introducing the Bayesian statistical model and continuously updating it, the estimation of the reviewer's response speed can be dynamically adjusted, making the prediction closer to the actual situation. This not only helps editors make more reasonable time arrangements, but also ensures that reviewers complete their review work in their best condition, thereby improving the efficiency and quality of the entire review process. In addition, this data-driven approach can better cope with changes in reviewer workload and promote the smooth progress of academic exchanges.

[0073] Optionally, "the calculation of the corresponding review suitability index based on the prediction results" in step 103 includes: using a time series analysis algorithm to predict the available time period of the reviewer in the future based on the current available time of the reviewer; and calculating the review suitability index of the reviewer based on the prediction results and the available time period.

[0074] Time series analysis: A statistical method used to analyze data points arranged in time sequence to discover trends, cyclical or seasonal patterns.

[0075] Available time period: The specific time interval that a reviewer can use to conduct reviews in the future, based on the prediction results.

[0076] Review suitability index: An indicator calculated by combining review response speed and future available time period to measure the suitability of the reviewer for the current review task.

[0077] First, we use a time series analysis algorithm to predict the available time period: by performing a time series analysis on the reviewers’ historical work schedule data, we can identify their daily work patterns and potential time patterns, and thus predict the future available time period. Secondly, we combine the predicted available time period with the length of time required for the review task to evaluate whether the time period is sufficient to complete the review task, and calculate an index reflecting the suitability of the review accordingly.

[0078] In the present embodiment, suppose there is a reviewer D, whose history shows that he has two days a week on average completely free in the past two years, which can be used for reviewing. By using ARIMA (autoregressive integrated moving average model) to perform time series analysis on D's work schedule data, the distribution of his available time every day in the next month can be predicted.

[0079] Assume that the predicted value of the number of available hours per day for reviewer D in the next month has been obtained through the ARIMA model. For example: Week 1: 4 hours on Monday, 6 hours on Tuesday, 3 hours on Wednesday... Week 2: 2 hours on Monday, 5 hours on Tuesday, 7 hours on Wednesday... This data can help determine which days are most appropriate for reviewing manuscripts.

[0080] If the target review task is expected to take 10 hours to complete, then we can use the above prediction results to evaluate which days are most suitable for reviewing. For example: If Monday and Tuesday of the first week are chosen (4+6=10 hours), the review suitability index is high because it meets the task requirements and has strong continuity. If Tuesday and Wednesday of the second week are chosen (5+7=12 hours), although the total time exceeds the required time, it also shows that reviewer D has enough buffer space during this period to deal with possible delays or other unexpected events.

[0081] A simple review suitability index formula can be defined as follows: Suitability index = predicted available time × required time × consecutive days coefficient Suitability index = required time × predicted available time × consecutive days coefficient, where the consecutive days coefficient can be set according to actual needs, such as giving a higher weight to two consecutive days.

[0082] Assuming the coefficient for two consecutive days is 1.2, the suitability index for the first week is: Suitability Index 1 = 1010 × 1.2 = 1.2 Suitability Index 1 = 1010 × 1.2 = 1.2, and the suitability index for the second week is: Suitability Index 2 = 1210 × 1.2 = 1.44 Suitability Index 2 = 1012 × 1.2 = 1.44. From this example, it can be seen that Tuesday and Wednesday of the second week are more suitable for arranging this review task.

[0083] By introducing time series analysis to predict the future available time periods of reviewers and calculating the review suitability index based on this, editors can arrange review tasks more accurately. This method not only takes into account the reviewer's current available time, but also foresees future schedules, which helps avoid task delays due to time conflicts. In addition, it can help editors optimize resource allocation and ensure that each reviewer can complete the review work in the best condition, thereby improving the efficiency and quality of the entire review process. Ultimately, this data-driven approach promotes the smooth conduct of academic exchanges and accelerates the publication of scientific research results.

[0084] 104. Selecting reviewers whose review suitability index meets predetermined conditions to form a review group, and planning a suitable review time period for each reviewer in the review group through an intelligent time management algorithm, so that each reviewer can provide review opinions for the target review task at the corresponding review time; In this step, appropriate reviewers are selected to form a team based on the suitability index obtained in the previous step, and specific review time windows are allocated to them taking into account their personal schedules.

[0085] In the present embodiment, 5 reviewers with scores of more than 7 points are selected to form a review group. Time optimization software is used to check their available time periods, and then the manuscript is sent out, while specifying the expected feedback date.

[0086] Optionally, the step 104 of "selecting reviewers whose review suitability index meets predetermined conditions to form a review group, and planning appropriate review time periods for each reviewer in the review group through an intelligent time management algorithm" includes: setting a predetermined threshold value according to the reviewer's review suitability index, and screening out reviewers whose index is higher than the predetermined threshold value; incorporating the screened reviewers into a candidate review group list; using an intelligent time management algorithm to analyze the current workload and current available time of each reviewer in the candidate review group list; based on the current workload and current available time of each reviewer, combined with the available time period of each reviewer in the future, allocating the most suitable review time period to each reviewer through an optimization algorithm, so as to ensure that each reviewer can focus on the review task of the target review task within the specified time period, so as to improve the review efficiency and quality.

[0087] Predetermined threshold: A minimum standard value set to screen out qualified reviewers.

[0088] Candidate review panel list: A list of potential reviewers formed after preliminary screening.

[0089] Smart Time Management Algorithm: An automated algorithm that analyzes and plans reviewers’ time allocation to ensure that review tasks are optimally performed.

[0090] Current workload: The number and complexity of all tasks the reviewer is currently working on.

[0091] Current available time: The amount of time the reviewer currently has available for new tasks.

[0092] Optimization algorithm: A mathematical method used to find the optimal solution or a near-optimal solution, which is used here to determine the most appropriate review time for each reviewer.

[0093] First, a threshold is set based on the review suitability index, and reviewers with a threshold above this threshold are selected as candidates. Second, the selected reviewers are included in a temporary list for further evaluation. The intelligent time management algorithm is further used to conduct a detailed analysis of the work situation and time resources of each reviewer on the candidate list. Finally, based on information such as workload, available time, and available time period, the optimization algorithm is used to arrange the best review time window for each reviewer.

[0094] In the embodiment of the present application, it is assumed that there are four reviewers E, F, G, and H, whose review suitability indexes are 1.5, 1.8, 2.0, and 1.7 respectively. The predetermined threshold is set to 1.6.

[0095] E(1.5)<predetermined threshold (1.6), not selected.

[0096] F(1.8)>predetermined threshold(1.6), selected.

[0097] G(2.0)>predetermined threshold(1.6), selected.

[0098] H(1.7)>predetermined threshold(1.6), selected.

[0099] The final list of candidate reviewers includes three reviewers: F, G, and H.

[0100] Using intelligent time management algorithms, you can obtain the following information: F: The current workload is low, and there will be an average of 5 hours of free time per day in the next week.

[0101] G: The current workload is high, but there is more free time on weekends.

[0102] H: The current workload is moderate, and I have a fixed 3 hours of free time every night.

[0103] You need to allocate review time for a paper that is expected to take 20 hours to review. Assume that you use an optimization algorithm based on linear programming to allocate time: For F, you can arrange for him to devote 5 hours every day during this week, for a total of 25 hours.

[0104] For G, you can arrange for him to spend 10 hours on each of the two days of the weekend, for a total of 20 hours.

[0105] For H, you can arrange for him to devote 3 hours every night in the next week, a total of 21 hours.

[0106] In order to ensure the quality and efficiency of the review, priority may be given to reviewers who can devote more time continuously. For example, G can be selected to complete the review in a concentrated manner on the weekend, or F can be asked to complete it in a dispersed manner during the weekdays. If three people are required to participate together, they can be reasonably allocated according to their best time.

[0107] By introducing the review suitability index and intelligent time management algorithm, this solution can form a review team in a more scientific and systematic manner and reasonably arrange the review time for each member. This not only helps to ensure that the review task can be completed with high quality within the specified time, but also avoids review quality problems caused by excessive load. In addition, this method improves the transparency and fairness of the entire review process, promotes the effective allocation of review resources, and thus speeds up the publication of academic results. At the same time, it also reduces the workload of editors, allowing them to focus on other key links, further improving the overall operational efficiency of publishing institutions.

[0108] 105. Use an integrated learning strategy to integrate the review opinions of multiple reviewers corresponding to the review group to form a review conclusion for the target review task.

[0109] In this step, the comments of all reviewers are summarized and analyzed using machine learning methods to provide comprehensive evaluations and suggestions.

[0110] In the embodiment of this application, natural language processing technology is used to analyze the positive and negative feedback in each review report; a weighted voting mechanism or other aggregation model is used to decide whether to accept the manuscript for re-review after revision or directly reject it. The final report not only includes the overall decision, but also provides specific improvement directions for the author's reference.

[0111] Optionally, in step 105, "the use of an integrated learning strategy to integrate the review opinions of multiple reviewers corresponding to the review group to form a review conclusion for the target review task" includes: collecting the review opinions provided by each reviewer in the review group, and converting the review opinions into standardized review scores; using an integrated learning framework to input the standardized review scores into an integrated model, wherein the integrated model is composed of multiple basic models, wherein the basic models include at least support vector machines, logistic regression, and neural networks, and each basic model independently evaluates the review scores; based on the evaluation results of multiple basic models, determining the final review conclusion through weighted voting or averaging strategy, wherein the weighted voting strategy assigns different weights according to the historical performance of each basic model to ensure the objectivity and fairness of the final review conclusion.

[0112] Ensemble learning: A machine learning method that improves overall prediction performance by combining the prediction results of multiple models (base models).

[0113] Standardized review scoring: Convert textual comments provided by reviewers into numerical scores in a unified format for easy machine processing.

[0114] Support Vector Machine (SVM): A supervised learning model used for classification and regression analysis.

[0115] Logistic Regression: A generalized linear model used for binary or multivariate classification problems.

[0116] Neural Network: A computational model that simulates the structure of the human brain and is suitable for complex pattern recognition and nonlinear relationship modeling.

[0117] Weighted voting: Give different weights to each model based on its historical performance, and then aggregate these weighted results to reach the final decision.

[0118] Averaging strategy: directly take the average of all model evaluation results as the final conclusion.

[0119] First, we obtain the review opinions from each reviewer and convert them into a scoring form under a unified standard. Secondly, we design an integrated learning system that includes multiple types of basic models (such as SVM, logistic regression, neural network, etc.). Then, we let each basic model evaluate the standardized review scores separately. Finally, we use weighted voting or averaging strategies to integrate the evaluation results of each basic model to obtain the final review conclusion.

[0120] In the present embodiment, it is assumed that there is a review team consisting of three reviewers A, B, and C. They each provide a review opinion on a certain paper. These opinions are converted into a standard score of 0 to 10: A gives a score of 8.5, B gives a score of 7.0, and C gives a score of 9.0. Next, these scores are input into an integrated learning framework that contains three basic models: SVM, logistic regression, and neural network. Assume that the performance of these three models before is as follows: The historical accuracy of SVM is 80%, and the weight is set to 0.3; The historical accuracy of logistic regression is 75%, and the weight is set to 0.25; The historical accuracy of the neural network is 85%, and the weight is set to 0.45; The SVM evaluation result for the score was 8.0; The logistic regression evaluation result for the score was 7.5; The neural network's evaluation result for the score was 8.8; In order to reach the final review conclusion, a weighted voting strategy can be used to combine the three evaluation results: Final score = (8.0×0.3) + (7.5×0.25) + (8.8×0.45) Final score = (8.0×0.3) + (7.5×0.25) + (8.8×0.45) Final score = 2.4 + 1.875 + 3.96 Final score = 2.4 + 1.875 + 3.96 Final score = 8.235 Final score = 8.235.

[0121] If a simple averaging strategy is adopted, the calculation can be done directly: Final score = 8.0 + 7.5 + 8.83 Final score = 38.0 + 7.5 + 8.8 Final score = 8.1 Final score = 8.1.

[0122] In this example, both the weighted voting and the average strategy obtained relatively high final scores, indicating that the quality of this paper is good. However, the weighted voting strategy takes into account the historical performance of each model, so its results are more valuable for reference.

[0123] By using an ensemble learning strategy to integrate the opinions of multiple reviewers, this method can reduce the subjective bias brought by a single reviewer and improve the consistency and reliability of the review conclusions. At the same time, due to the integration of multiple types of machine learning models, different types of data features can be better captured, thereby improving the robustness and generalization ability of the entire system. In addition, the weighted voting strategy makes those models that have performed well in history have greater influence, further ensuring the objectivity and fairness of the final review conclusions. This not only helps to improve the efficiency of review, but also enhances the transparency of academic review and promotes the high-quality publication of scientific research results.

[0124] This application takes into account that in the field of academic publishing, the review process is a key link in ensuring the quality of research. However, the following problems exist in the prior art: first, the opinions of a single reviewer may have subjective biases, affecting the objectivity of the final conclusion; second, the opinions of different reviewers may be inconsistent, and it is difficult for editors to make a comprehensive judgment; finally, with the development of data science, how to use machine learning technology to improve the consistency and accuracy of review conclusions has become an urgent problem to be solved. Therefore, an embodiment of the present invention proposes a method based on an integrated learning strategy to integrate the opinions of multiple reviewers, and form a comprehensive review conclusion for the target review task through a specific weighted formula to solve the above-mentioned problems and improve the objectivity and fairness of the review process.

[0125] Optionally, in step 105, "using an integrated learning strategy to integrate the review opinions of multiple reviewers corresponding to the review group to form a review conclusion for the target review task" includes: The review conclusion for the target review task is formed through the following calculation formula: ; in, It is the review score converted from the final review conclusion. It is the standardized review score converted from the review opinions provided by each reviewer in the review team; It is The prediction function of the base model; is the number of base models; It is The weights of the base models; is the regulating factor; It is The historical loss value of the base model; is a normalization term that ensures that the sum of the weights of all base models is 1.

[0126] First, we obtain the review opinions from each reviewer and convert them into a scoring form under a unified standard. Secondly, we select various types of machine learning models as basic models, such as support vector machines, logistic regression, neural networks, etc. We further use the given formula to calculate the weight of each basic model, and the weight depends on the historical performance of the model (i.e., historical loss value). Finally, we multiply the prediction results of each basic model by the corresponding weight, and then sum them up to get the final review score.

[0127] Suppose there is a review team consisting of three reviewers A, B, and C. They each provide a review opinion on a paper. Convert these opinions into a standard score of 0 to 10: A gives a score of 8.5, B gives a score of 7.0, and C gives a score of 9.0.

[0128] Next, these scores are input into an integrated learning framework, which contains three basic models: SVM, logistic regression, and neural network. Assume that the previous performance of these three models is as follows: the historical loss value of SVM is 0.3, the historical loss value of logistic regression is 0.4, and the historical loss value of neural network is 0.2, and the adjustment factor λ=1λ=1 is set.

[0129] Calculate the weights for each base model: ; It is assumed that the basic models evaluated the standardized review scores as follows: the SVM evaluation result of the score was 8.0, the logistic regression evaluation result of the score was 7.5, and the neural network evaluation result of the score was 8.8.

[0130] Calculate the final review score: ; By using an ensemble learning strategy combined with a specific weighting formula to integrate the opinions of multiple reviewers, this method can reduce the subjective bias brought by a single reviewer and improve the consistency and reliability of the review conclusions. At the same time, since multiple types of machine learning models are integrated and the weight distribution takes into account the historical performance of each model, it can better capture different types of data features, thereby improving the robustness and generalization ability of the entire system. In addition, this weighting method makes those models that have performed well in history have greater influence, further ensuring the objectivity and fairness of the final review conclusions. This not only helps to improve the efficiency of review, but also enhances the transparency of academic review and promotes the high-quality publication of scientific research results.

[0131] This application takes into account that in the field of academic publishing, the review process is a key link in ensuring the quality of research. However, there are the following problems in the prior art: first, editors usually evaluate the ability of reviewers based on limited information (such as personal experience and simple statistics), which lacks systematicity and objectivity; second, traditional evaluation methods fail to make full use of the rich information in historical data, resulting in inaccurate evaluation of reviewers' abilities; finally, with the development of machine learning technology, how to use advanced algorithms to improve the accuracy of reviewer ability prediction has become an urgent problem to be solved. Therefore, an embodiment of the present invention proposes a method based on a random forest model to predict the reviewer's review ability for a target review task, so as to solve the above-mentioned problems and improve the efficiency and quality of the review process.

[0132] Optionally, the predicting the reviewing ability of the reviewer for the target reviewing task based on the quality assessment data in the reviewer's historical review records includes: The reviewer's reviewing ability for the target reviewing task is predicted by the following calculation formula: ; in, is the feature vector of the reviewer; is the prediction function of the random forest model; are the parameters of the random forest model; It is the review ability score for historical review tasks; is the number of samples in the training dataset; is the reviewer's review ability score for the target review task; Indicates finding a set of optimal parameters , so that the actual review ability score of all training samples With the model prediction value The sum of squared differences between them is minimal.

[0133] First, relevant features are extracted from the reviewer's historical review records to form a feature vector The random forest algorithm is then used as the prediction model, which can handle a large number of features and provide stable prediction results. Secondly, the data set in the historical review records is used to train the model by minimizing the sum of the squared differences between the actual score and the predicted score to find the optimal parameters. Finally, the relevant features of the target review task are input into the trained model to obtain the reviewer's predicted review ability score for the task. .

[0134] Suppose there is a historical review record dataset containing 100 reviewers. Each record includes the following features: professional background (numerical representation), number of previous reviews, average score, review completion time, and article type preference. These features form a feature vector In addition, each record also includes an actual reviewer ability score . 100 samples were extracted from the historical data, each containing 5 features and the corresponding actual review ability score. For example: Sample 1: [1,20,9.0,7,8.5]->8.8 Sample 2: [2,15,8.5,5,7.0]->7.5 ... Sample 100: [3, 30, 9.5, 6, 9.0] -> 9.2 Use the random forest algorithm to train the model. Assume that after training, the model finds the optimal parameters .

[0135] Now we need to predict the review ability of a new reviewer A for a certain paper. The feature vector of reviewer A is: professional background: 2, number of previous reviews: 18, average score: 8.7, review completion time: 6, article type preference: 8.0.

[0136] These features are fed into the trained random forest model: , suppose the model outputs a predicted score of 8.6.

[0137] By using the random forest model to predict the reviewer's review ability, this method can more systematically utilize the information in historical data and improve the accuracy and reliability of the prediction. At the same time, the random forest model has strong robustness and generalization ability, and can handle high-dimensional and complex data structures. In addition, this data-driven approach reduces the impact of subjective judgment and enhances the transparency and fairness of the review process. This not only helps editors make more reasonable reviewer selections, but also promotes the high-quality publication of scientific research results and enhances the trust of the entire academic community.

[0138] Optionally, the method of predicting the reviewer's reviewing ability for the target review task based on the quality assessment data in the reviewer's historical review records includes: collecting the quality assessment data in the reviewer's historical review records, including the adoption rate of the review opinions, the detail level of the review report, and the usefulness score of the review opinions; constructing a training data set based on the quality assessment data, wherein each training data in the training data set contains a feature vector of the reviewer and a reviewing ability score for historical review tasks, the feature vector including the reviewer's title, institution, professional field, number of past reviews, and average review time; training a random forest model, using the feature vector in the training data set as input and the corresponding reviewing ability score of the historical review tasks as labels to learn the mapping relationship between reviewing ability and feature vectors; using the trained random forest model to output the reviewer's reviewing ability score for the target review task, and use it as a result of predicting the reviewer's reviewing ability for the target review task.

[0139] Quality assessment data: refers to a data set that evaluates reviewers’ past performance.

[0140] Acceptance rate: the proportion of review comments accepted and revised by the authors.

[0141] Level of detail: A measure of the thoroughness of the review report.

[0142] Usefulness rating: A rating given to the value of the review comments based on feedback from peers or other readers.

[0143] Feature vector: A set of numerical attributes that describe the characteristics of an object (here, the reviewers).

[0144] Random Forest Model: An ensemble learning method that uses multiple decision trees to vote to determine the final classification result.

[0145] First, the quality assessment data of each reviewer, such as the rate of opinion adoption, is extracted from the historical records. Secondly, the collected information is organized into a form that includes reviewer characteristics and their corresponding ability scores. The random forest algorithm is further used to train a model that can reflect the relationship between the two, using the feature vector as the input variable and the historical ability score as the output variable. Finally, the trained model is used to predict the ability scores of potential reviewers under the newly submitted target review task.

[0146] In the example of this application, suppose a journal needs to find a suitable reviewer for a paper in the field of machine learning. First, the system will automatically screen out all registered reviewers with relevant background knowledge and collect their historical review records; then, based on these records, the feature values ​​of each candidate are calculated, such as the average time required for each review, the number of articles published in the professional field, etc.; then, the previously established random forest model is used to predict the scores of all candidates; finally, the experts with the highest scores are recommended as the best candidates for editors to refer to and select.

[0147] By implementing this plan, we can significantly improve the objectivity and fairness of the review process, reduce the deviation caused by human factors, and effectively shorten the review cycle and speed up the dissemination of scientific research results. In addition, it can help young scholars get growth opportunities faster and promote the healthy development of the academic community.

[0148] Optionally, in the process of predicting the reviewer's reviewing ability and response speed for the target review task through data analysis, it also includes: using a deep neural network to model the acquired behavior pattern of the reviewer, combining the acquired reviewer's review preferences, historical behavior trajectory and external influencing factors, to refine the prediction of the reviewing ability and the response speed to update the prediction result; introducing a reinforcement learning algorithm to dynamically adjust the parameters of the prediction model according to the reviewer's completion of historical review tasks at different time points to improve the prediction accuracy of the prediction result, wherein the adjusted parameters of the prediction model include at least the prior probability distribution parameters in the Bayesian statistical model, the hyperparameters in the random forest algorithm, the reward function and discount factor in the reinforcement learning algorithm, and the weight and bias term parameters in the deep neural network.

[0149] Deep neural network: An artificial intelligence algorithm that mimics the way the human brain processes information and can be used to identify complex patterns in data.

[0150] Review preference: refers to the selection tendency shown by reviewers when accepting or rejecting review invitations.

[0151] Historical behavior trajectory: including but not limited to behavioral data accumulated over a long period of time, such as review time, frequency, and field preferences.

[0152] External factors: Such as holidays, meeting schedules, etc., which may affect the reviewer's available time.

[0153] Bayesian statistical model: a statistical method that uses prior probability distribution to update posterior probability.

[0154] Hyperparameters: Parameters in a machine learning model that need to be manually set, such as the number of trees in a random forest.

[0155] Reward function: A function that defines the goal in reinforcement learning and is used to guide the learning direction of the algorithm.

[0156] Discount factor: A weight that measures the importance of future rewards and is used in reinforcement learning to calculate cumulative rewards.

[0157] First, we use deep neural networks to learn the behavior patterns of reviewers, inputting multi-dimensional information such as personal preferences and historical behavior trajectories. Secondly, we combine the above model with data that takes external factors into account to obtain more accurate predictions about reviewer ability and response speed. We further use reinforcement learning technology to continuously optimize model parameters based on actual feedback to ensure that the model evolves over time. Specifically, we update key parameter values ​​such as the prior probability in the Bayesian model, the hyperparameters of the random forest, and the weights of the deep learning network regularly or on demand.

[0158] In an embodiment of the present application, it is assumed that a journal is looking for a suitable reviewer to review a recent research article in the field of artificial intelligence. First, the system collects historical behavior data of all potential reviewers in the field, such as what topics they have reviewed in the past, the average time required to complete a review, etc. Then, a trained deep neural network model is used to analyze this group of data to obtain each candidate's estimated review quality score for the current specific article and an estimate of the speed at which they may respond. At the same time, considering that a large international AI conference is about to be held recently, this may make some experts very busy, so the priority of those known participants is appropriately lowered in the final recommendation list. Finally, as more actual review result data is accumulated, the system will also automatically tune the various algorithm parameters used internally, thereby continuously improving the accuracy of predictions for future tasks.

[0159] By introducing advanced deep learning and reinforcement learning technologies, this solution can not only reflect the real working status of reviewers to a greater extent, but also flexibly respond to changes in demand in different situations, providing editors with more scientific and reasonable decision-making support tools. This will help further improve the transparency and fairness of academic exchanges, and also promote the rapid dissemination of high-quality research results.

[0160] This application takes into account that in the academic publishing process, the selection of suitable reviewers must not only consider their professional ability and historical performance, but also the reviewer's response speed and available time period. Existing methods often only focus on one aspect, such as determining reviewers based solely on review ability or response time. This may result in the selected reviewers not being suitable for the current task requirements in actual operations, such as delaying the review due to being busy. Therefore, this application proposes a method that comprehensively considers multiple factors to calculate the review suitability index, aiming to improve the overall efficiency and quality of the review process by more comprehensively evaluating the match of reviewers.

[0161] Optionally, calculating the review suitability index of the reviewer according to the prediction result and the available time period includes: The review suitability index of the reviewer is calculated by the following formula: ; in, represents the review suitability index; represents the number of reviewers; is the reviewer's comprehensive weight factor, which is a score determined based on the reviewer's feature vector and the reviewer ability score for historical review tasks; It indicates the expected completion time of the reviewer for the target review task. is a very small positive number used to prevent division by zero errors; is with The relevant exponential factor, using the exponential function , to emphasize the importance of reviewer response time, a nonlinear coefficient determined based on the statistical distribution characteristics of reviewers' past response times; It represents the predicted reviewer's review ability score for the target review task; is with The relevant exponential factor is expressed in square roots. , used to adjust the impact of reviewer ability score; It represents the estimated value of the available time period of the reviewer for the target review task, which is a quantitative evaluation of the future available time period predicted by the time series analysis algorithm; is with The relevant exponential factor is , used to adjust the importance of the available time periods; It indicates the number of time conflicts that the reviewer will have with other review tasks in the future, which is a negative indicator; is with The relevant exponential factor is , used to adjust the impact of the number of conflicts.

[0162] Comprehensive weight factor : A weighted value determined based on the reviewer’s feature vector and their historical review task scores.

[0163] Predicting review response time : The time required for the reviewer to complete the target review task based on the model prediction.

[0164] Exponential Factor : Coefficient used to adjust the influence of each factor on the final suitability index.

[0165] Reviewer Ability Rating : A quantitative indicator that reflects the reviewer's expertise level for a specific task.

[0166] Available time period evaluation value : The amount of time available for reviewers in the future, obtained through methods such as time series analysis.

[0167] Number of time conflicts : The number of other review tasks that the reviewer has accepted, which is considered as a negative factor.

[0168] First, collect relevant information of all potential reviewers, including but not limited to personal background, historical review records, etc., and clean and format them. Then set corresponding weight factors according to the specific situation of each reviewer. , Exponential Factor and very small positive numbers Then substitute the above parameters into the provided calculation formula to calculate the review suitability index of each candidate. Finally, the candidate list is arranged from high to low according to the suitability index obtained, and the most suitable candidates are selected for invitation.

[0169] In the present embodiment, suppose a journal needs to find a suitable reviewer for an article on machine learning. The system first identifies five experts with relevant professional knowledge as candidates. For one of the experts (set as the first), the known data is as follows: Comprehensive weight factor , predicted review response time T_1 = 7 days, exponential factor Reviewer ability score C_1=4.6 (full score 5), index factor , available time period evaluation value A_1 = 10 hours / week, exponential factor , the number of time conflicts K_1=2, the exponential factor , the smallest positive number .

[0170] Use the given formula to calculate the review suitability index of the expert : ; After substituting the specific values, we get: ; This process was repeated to calculate the suitability indexes of the remaining four experts, and the final selection was made based on this.

[0171] By adopting this multi-dimensional evaluation system, it can more accurately reflect the adaptability of each reviewer under specific conditions, thereby helping editors quickly find the most suitable reviewer. This not only improves the fairness and transparency of the review process, but also ensures the quality and timeliness of paper reviews, thereby promoting the effective dissemination and development of scientific research results.

[0172] Figure 2 A schematic diagram of a system for distributing manuscript review tasks is provided for an embodiment of the present application, such as Figure 2 As shown, the device comprises: A determination module 21 is used to receive a target review task and determine a research topic of the target review task; A screening module 22 is used to screen out reviewers matching the research topic from a pre-built reviewer database; Prediction module 23, used to predict the reviewing ability and response speed of the reviewer through data analysis based on the reviewer's historical review records and current available time, and calculate the corresponding review suitability index according to the prediction result; Processing module 24, for selecting reviewers whose review suitability index meets the predetermined conditions to form a review group, and planning a suitable review time period for each reviewer in the review group through an intelligent time management algorithm, so that each reviewer can provide review opinions for the target review task at the corresponding review time; The generation module 25 is used to use an integrated learning strategy to integrate the review opinions of multiple reviewers corresponding to the review group to form a review conclusion for the target review task.

[0173] Figure 2 The system for distributing manuscript review tasks can be executed Figure 1 The implementation principle and technical effect of the method for allocating manuscript review tasks described in the embodiment are not described in detail. The specific way in which each module and unit performs operations in the manuscript review task allocation system in the above embodiment has been described in detail in the embodiment of the method, and will not be described in detail here.

[0174] In one possible design, Figure 2 A review task distribution system of the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32; The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0175] The processing component 32 is used to: receive a target review task and determine a research topic for the target review task; screen out reviewers who match the research topic; predict the reviewing ability and response speed of the reviewers for the target review task, and calculate the corresponding review suitability index based on the prediction result; select reviewers whose review suitability index meets predetermined conditions to form a review team; use an integrated learning strategy to integrate the review opinions of multiple reviewers corresponding to the review team to form a review conclusion for the target review task.

[0176] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 A method for assigning review tasks in the illustrated embodiment.

[0177] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0178] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0179] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0180] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for assigning manuscript review tasks, characterized in that: include: Receive target review tasks and determine the research topics of the target review tasks; From the pre-built reviewer database, select reviewers that match the research topic; Based on the reviewer's historical review records and current available time, predict the reviewer's reviewing ability and response speed for the target review task through data analysis, and calculate the corresponding review suitability index according to the prediction result; Selecting reviewers whose review suitability index meets the predetermined conditions to form a review group, and planning a suitable review time period for each reviewer in the review group through an intelligent time management algorithm, so that each reviewer can provide review opinions for the target review task at the corresponding review time; An integrated learning strategy is used to integrate the review opinions of multiple reviewers corresponding to the review group to form a review conclusion for the target review task.

2. The method according to claim 1, characterized in that The historical review records include at least: review time data and quality assessment data; The method of predicting the reviewing ability and response speed of the reviewer for the target reviewing task through data analysis based on the reviewer's historical reviewing records and current available time includes: Predicting the review response speed of the reviewer for the target review task based on the review time data in the reviewer's historical review records and the current available time; Predicting the reviewing ability of the reviewer for the target reviewing task based on the quality assessment data in the reviewer's historical review records; Taking the review response speed and the review ability as prediction results; The following calculation formula is used to predict the review response speed of the reviewer for the target review task: ； in, The review time data in the reviewer's historical review records; is the current available time of the reviewer in question; Represents the data at a given historical review time and current available time Under the condition of Actual completion time Expected value; It is the posterior distribution calculated based on the historical review time and the current available time; is the expected completion time of the reviewer for the target review task, that is, the predicted value of the review response speed; is the combined effect of review time and current available time; the denominator It is the time for all possible review Multiply it by its corresponding current available time The result of post-integration; The following calculation formula is used to predict the reviewer's review ability for the target review task: ； in, is the feature vector of the reviewer; is the prediction function of the random forest model; are the parameters of the random forest model; It is the review ability score for historical review tasks; is the number of samples in the training dataset; is the reviewer's review ability score for the target review task; Indicates finding a set of optimal parameters , so that the actual review ability score of all training samples With the model prediction value The sum of squared differences between them is minimal.

3. The method according to claim 2, characterized in that The predicting, based on the review time data in the reviewer's historical review records and the current available time, of the reviewer's review response speed for the target review task includes: Based on the review time data in the historical review records of the reviewers, a Bayesian statistical model is constructed to describe the prior distribution of the review time; Using the Bayesian update rule, the prior distribution is updated according to the latest review time data of the reviewer to obtain the posterior distribution; Based on the posterior distribution and in combination with the current available time of the reviewer, the expected completion time of the reviewer for the target review task is estimated as a predicted value of the reviewer's review response speed.

4. The method according to claim 2, characterized in that: The predicting the reviewing ability of the reviewer for the target reviewing task based on the quality assessment data in the reviewer's historical review records includes: Collecting quality assessment data from the reviewers' historical review records, including the adoption rate of the review opinions, the level of detail of the review reports, and the usefulness scores of the review opinions; Constructing a training data set based on the quality assessment data, wherein each piece of training data in the training data set contains a feature vector of the reviewer and a reviewing ability score for a historical review task, wherein the feature vector includes the reviewer's title, institution, professional field, number of past reviews, and average review time; Training a random forest model, using the feature vectors in the training data set as input and the reviewing ability scores of the corresponding historical review tasks as labels, to learn the mapping relationship between the reviewing ability and the feature vectors; The trained random forest model is used to output the reviewer's review ability score for the target review task, and used as a result of predicting the reviewer's review ability for the target review task.

5. The method according to claim 2, characterized in that: In the process of predicting the reviewer's reviewing ability and response speed for the target reviewing task through data analysis, it also includes: Using a deep neural network to model the behavior patterns of the reviewers, combining the review preferences, historical behavior trajectories, and external influencing factors of the reviewers, to refine the prediction of the reviewing ability and the response speed, so as to update the prediction results; A reinforcement learning algorithm is introduced to dynamically adjust the parameters of the prediction model according to the completion status of the historical review tasks by the reviewers at different time points to improve the prediction accuracy of the prediction results, wherein the adjusted parameters of the prediction model include at least the prior probability distribution parameters in the Bayesian statistical model, the hyperparameters in the random forest algorithm, the reward function and discount factor in the reinforcement learning algorithm, and the weight and bias parameters in the deep neural network.

6. The method according to claim 5, characterized in that The calculation of the corresponding review suitability index according to the prediction results includes: Based on the current available time of the reviewer, a time series analysis algorithm is used to predict the available time period of the reviewer in the future; Calculating the review suitability index of the reviewer according to the prediction result and the available time period; The review suitability index of the reviewer is calculated by the following formula: ； in, represents the review suitability index; represents the number of reviewers; is the reviewer's comprehensive weight factor, which is a score determined based on the reviewer's feature vector and the reviewer ability score for historical review tasks; It indicates the expected completion time of the reviewer for the target review task. is a very small positive number used to prevent division by zero errors; is with The relevant exponential factor, using the exponential function , to emphasize the importance of reviewer response time, a nonlinear coefficient determined based on the statistical distribution characteristics of reviewers' past response times; It represents the predicted reviewer's review ability score for the target review task; is with The relevant exponential factor is expressed in square roots. , used to adjust the impact of reviewer ability score; It represents the estimated value of the available time period of the reviewer for the target review task, which is a quantitative evaluation of the future available time period predicted by the time series analysis algorithm; is with The relevant exponential factor is , used to adjust the importance of the available time periods; It indicates the number of time conflicts that the reviewer will have with other review tasks in the future, which is a negative indicator; is with The relevant exponential factor is , used to adjust the impact of the number of conflicts.

7. The method according to claim 1, characterized in that The reviewers matching the research topic are screened out from the pre-built reviewer database, including: Extract keywords and phrases from the target review task to identify the core elements of the research topic; Converting the core elements into a structured data format, and matching the core elements of the structured data format with the professional field information of the reviewers stored in the reviewer database to identify reviewers matching the research topic; Filter out the best list of reviewers based on their professional field matching, past review experience, and review quality evaluation information; Select qualified reviewers from the optimal reviewer list.

8. The method according to claim 1, characterized in that The step of selecting reviewers whose review suitability index meets predetermined conditions to form a review group, and planning a suitable review time period for each reviewer in the review group through an intelligent time management algorithm includes: According to the review suitability index of the reviewers, a predetermined threshold is set, and reviewers with a suitability index higher than the predetermined threshold are screened out; Include the selected reviewers in the candidate review team list; Analyzing the current workload and current available time of each reviewer in the candidate review panel list using an intelligent time management algorithm; Based on each reviewer's current workload and current available time, combined with each reviewer's available time period in the future, an optimization algorithm is used to assign the most suitable review time period to each reviewer, to ensure that each reviewer can focus on the review task of the target review task within the specified time period, so as to improve the review efficiency and quality.

9. The method according to claim 8, characterized in that The use of an integrated learning strategy to integrate the review opinions of multiple reviewers corresponding to the review group to form a review conclusion for the target review task includes: Collecting the review opinions provided by each reviewer in the review team and converting the review opinions into standardized review scores; Using an integrated learning framework, the standardized review score is input into an integrated model, wherein the integrated model is composed of multiple basic models, wherein the basic models at least include support vector machines, logistic regression, and neural networks, and each basic model independently evaluates the review score; Based on the evaluation results of the multiple basic models, the final review conclusion is determined by weighted voting or averaging strategy, wherein the weighted voting strategy assigns different weights according to the historical performance of each basic model to ensure the objectivity and fairness of the final review conclusion; The review conclusion for the target review task is formed by the following calculation formula: ； in, It is the review score converted from the final review conclusion. It is the standardized review score converted from the review opinions provided by each reviewer in the review team; It is The prediction function of the base model; is the number of base models; It is The weights of the base models; is the regulating factor; It is The historical loss value of the base model; is a normalization term that ensures that the sum of the weights of all base models is 1.

10. A system for distributing manuscript review tasks, characterized in that: include: A determination module, used to receive a target review task and determine a research topic of the target review task; A screening module is used to screen out reviewers matching the research topic from a pre-built reviewer database; A prediction module, used to predict the reviewing ability and response speed of the reviewer through data analysis based on the reviewer's historical review records and current available time, and calculate the corresponding review suitability index according to the prediction results; A processing module is used to select reviewers whose review suitability index meets the predetermined conditions to form a review group, and plan a suitable review time period for each reviewer in the review group through an intelligent time management algorithm, so that each reviewer can provide review opinions for the target review task at the corresponding review time; A generation module is used to use an integrated learning strategy to integrate the review opinions of multiple reviewers corresponding to the review group to form a review conclusion for the target review task.

Images