Demand prediction and intelligent allocation method and system for library collection optimization

Abstract

The invention discloses a demand prediction and intelligent allocation method and system for library collection optimization, and relates to the technical field of book management. The demand prediction and intelligent allocation method for library collection optimization comprises the following steps: S1, book borrowing evaluation; s2, evaluating user recommendation; and S3, book allocation evaluation. According to the method, whether new book inventory management is carried out or not is judged through the obtained book borrowing index of the specified book, then whether book recommendation strategy optimization is carried out or not is judged based on the obtained user borrowing index of the specified user, and finally whether book allocation optimization is carried out or not is judged based on the obtained allocation index of the specified book. The effect of improving the allocation accuracy of the new books in the library is achieved, and the problem of insufficient data during demand prediction and allocation analysis of the new books in the library in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of library management, and particularly to a demand prediction and intelligent allocation method and system for optimizing library collections. Background Art

[0002] With the rapid development of information technology and the advent of the digital age, libraries are not only repositories of knowledge and culture but also key platforms for promoting learning and social interaction. However, traditional library operation methods face many challenges, especially in resource allocation and demand management. In particular, the efficiency improvement in aspects such as library collection management, resource allocation, and user demand satisfaction. There may be situations of over-accumulation or shortage of library collection resources. Especially, the demand for some popular books is high, while it is difficult to borrow other less popular books. How to reasonably allocate library resources, avoid waste, and reduce costs is an urgent problem to be solved.

[0003] In the prior art, through the analysis of a large amount of historical borrowing data, user behavior data, book types and other data, it is possible to predict the borrowing demand and trend of books, provide accurate demand prediction, and achieve an improvement in resource allocation efficiency.

[0004] For example, a book management method and device for a community library based on big data disclosed in the invention patent announcement with the publication number of CN109785212B includes: adding tag information to each book and obtaining the reading behavior data of each book, and then judging whether to retain the book according to the reading behavior data of the book. By calculating the similarity of the retained books between adjacent community libraries, the books with a similarity greater than the second threshold are recommended to adjacent community libraries.

[0005] For example, an inventory management method and system for a library disclosed in the patent application with the publication number of CN119228271A includes: obtaining the metadata of the library, including book information and user borrowing records, constructing a knowledge graph based on these data, and using a heterogeneous graph matching algorithm to extract the user's query intention to obtain a set of target nodes. Then, based on multi-resolution wavelet transform and frequency domain analysis, the semantic representation information of the user's intention is extracted. Using the sparse Bayesian learning algorithm to predict the future borrowing demand and generate a demand prediction result. Through a multi-objective programming algorithm and a convex optimization method, combined with inventory allocation and logistics costs, an inventory allocation plan is generated, and this plan is optimized by the Lagrangian relaxation method. Finally, the inventory allocation plan is adjusted in real time in combination with the dynamic system state estimation method.

[0006] However, in the process of implementing the inventive technical solution in the embodiments of the present application, it is found that the above technologies have at least the following technical problems: In the prior art, it is difficult to predict the demand for new books that have just been put on the shelves due to the lack of historical borrowing records. In addition, for new users and users with few borrowing times, there is a lack of sufficient borrowing history, which affects the accuracy of personalized recommendations. There is also a problem of insufficient data for demand prediction and allocation analysis of new books in the library. Summary of the invention

[0007] The embodiments of the present application solve the problem of insufficient data in the prior art for demand forecasting and allocation analysis of new books in the library by providing a demand forecasting and intelligent allocation method and system for library collection optimization, thereby improving the accuracy of new book allocation in the library.

[0008] The embodiment of the present application provides a demand forecasting and intelligent allocation method for library collection optimization, comprising the following steps: S1, evaluating the book borrowing index of a specified book based on the actual number of borrowers and book forecast data obtained, judging whether to perform new book inventory management based on the book borrowing index, and the book borrowing index is used to quantitatively evaluate the borrowing capacity of the specified book in the library demand forecast; S2, if new book inventory management is performed, evaluating the user borrowing index of the specified user based on the obtained new user recommendation data, judging whether to perform book recommendation strategy optimization based on the user borrowing index, and the user borrowing index is used to quantitatively evaluate the effectiveness of recommending books to the specified user in the library demand forecast; S3, if book recommendation strategy optimization is performed, evaluating the allocation index of the specified book based on the obtained book allocation data, judging whether to perform book allocation optimization based on the allocation index, and the allocation index is used to quantitatively evaluate the efficiency of book allocation in the library intelligent allocation.

[0009] Further, the book prediction data is obtained by analyzing historical borrowing data, and the predicted borrowing quantity of new books is predicted through the historical borrowing data in the demand prediction system; the historical borrowing data includes book categories and the number of borrowing users; the actual borrowing data includes specified book categories, the actual number of borrowers, and the total number of new books; the book prediction data includes the number of books in the same category, the borrowing prediction accuracy rate, the predicted borrowing rate, and the book similarity; the new user recommendation data includes the book recommendation precision coefficient, the book recommendation coverage coefficient, the user click-through rate, and the new book recommendation coefficient; the book allocation data includes the user borrowing waiting time, the user actual borrowing time, the user borrowing return rate, and the book borrowing coefficient; the borrowing prediction accuracy rate is obtained by performing a ratio operation on the absolute value of the difference between the predicted borrowing quantity and the actual borrowing quantity and the predicted borrowing quantity; the predicted borrowing rate is obtained by performing a ratio operation on the predicted borrowing quantity and the total number of new books; the book similarity is represented by the maximum value of the similarity between the specified book and other books calculated by the cosine similarity for the specified book category and author and the book category and author; the book recommendation precision coefficient is obtained by performing a ratio operation on the number of books recommended for a specified user and the total number of recommended books; the book recommendation coverage coefficient is obtained by performing a ratio operation on the number of books recommended for a specified user and the total number of books borrowed by the specified user; the new book recommendation coefficient is obtained by performing a ratio operation on the difference between the actual maximum borrowing quantity of the specified book in the recommended books of new users and the actual maximum borrowing quantity of the specified book in the non-recommended books and the total number of books borrowed by the user; the book borrowing coefficient is obtained by performing a ratio operation on the actual borrowing quantity of the book within a preset time period and the total number of library books.

[0010] Further, the specific method for evaluating the book borrowing index of a specified book based on the obtained actual number of borrowers and book prediction data is as follows: The book borrowing deviation, book similarity, borrowing weights, and the number of books in the same category factor obtained from a preset database are processed to obtain the book borrowing index; the book borrowing deviation includes the actual number of borrowers deviation, the predicted borrowing rate deviation, and the borrowing prediction accuracy rate deviation; the actual number of borrowers deviation is used to reflect the deviation degree between the actual number of borrowers and the preset number of borrowers data; the predicted borrowing rate deviation is used to reflect the deviation degree between the predicted borrowing rate and the preset borrowing rate; the borrowing prediction accuracy rate deviation is used to reflect the deviation degree between the borrowing prediction accuracy rate and the preset borrowing prediction accuracy rate; the borrowing weights include the number of borrowers weight, the borrowing rate weight, the borrowing accuracy weight, and the similarity weight.

[0011] Further, the specific method for determining whether to conduct new book inventory management based on the book borrowing index is as follows: Determine whether the book borrowing index of the specified book is less than the preset book borrowing threshold obtained from the preset database: The preset book borrowing threshold includes a first preset book borrowing threshold and a second preset book borrowing threshold; If the book borrowing index of the specified book is less than the first preset book borrowing threshold obtained from the preset database, gradually reduce the inventory of the specified book; If the book borrowing index of the specified book is not less than the first preset book borrowing threshold obtained from the preset database and not greater than the second preset book borrowing threshold obtained from the preset database, do not conduct new book inventory management; If the book borrowing index of the specified book is greater than the second preset book borrowing threshold obtained from the preset database, gradually increase the inventory of the specified book.

[0012] Further, the specific method for evaluating the user borrowing index of the specified user based on the obtained new user recommendation data is as follows: Obtain the book recommendation precision coefficient deviation according to the relative relationship between the book recommendation precision coefficient and the preset book recommendation precision coefficient obtained from the preset database. The preset book recommendation precision coefficient includes a first preset book recommendation precision coefficient and a second preset book recommendation precision coefficient, and the book recommendation precision coefficient deviation is used to reflect the deviation degree between the book recommendation precision coefficient and the preset book recommendation precision coefficient; Obtain the book recommendation coverage coefficient deviation according to the relative relationship between the book recommendation coverage coefficient and the preset book recommendation coverage coefficient obtained from the preset database. The preset book recommendation coverage coefficient includes a first preset book recommendation coverage coefficient and a second preset book recommendation coverage coefficient, and the book recommendation coverage coefficient deviation is used to reflect the deviation degree between the book recommendation coverage coefficient and the preset book recommendation coverage coefficient; Obtain the new book recommendation coefficient deviation according to the relative relationship between the new book recommendation coefficient and the preset new book recommendation coefficient obtained from the preset database. The preset new book recommendation coefficient includes a first preset new book recommendation coefficient and a second preset new book recommendation coefficient, and the new book recommendation coefficient deviation is used to reflect the deviation degree between the new book recommendation coefficient and the preset new book recommendation coefficient; Obtain the user click-through rate deviation according to the relative relationship between the user click-through rate and the preset user click-through rate obtained from the preset database. The user click-through rate deviation is used to reflect the deviation degree between the user click-through rate and the preset user click-through rate; Process the book recommendation precision coefficient deviation, the book recommendation coverage coefficient deviation, the new book recommendation coefficient deviation, the user click-through rate deviation, and the user borrowing weight obtained from the preset database to obtain the user borrowing index; The user borrowing weight includes a recommendation precision weight, a recommendation coverage weight, a new book recommendation weight, and a user click weight.

[0013] Further, the specific method for determining whether to optimize the book recommendation strategy based on the user borrowing index is as follows: Determine whether the user borrowing index of the specified user is less than the preset user borrowing threshold obtained from the preset database. If the user borrowing index of the specified user is not less than the preset user borrowing threshold obtained from the preset database, do not optimize the book recommendation strategy. If the user borrowing index of the specified user is less than the preset user borrowing threshold obtained from the preset database, optimize the book recommendation strategy.

[0014] Further, the specific method for evaluating the allocation index of a specified book based on the obtained book allocation data is as follows: Obtain the user borrowing waiting time deviation according to the relative relationship between the user borrowing waiting time and the preset user borrowing waiting time obtained from the preset database. The user borrowing waiting time deviation is used to reflect the deviation degree between the user borrowing waiting time and the preset user borrowing waiting time. Obtain the user actual borrowing time deviation according to the relative relationship between the user actual borrowing time and the preset user book borrowing time obtained from the preset database. The user actual borrowing time deviation is used to reflect the deviation degree between the user actual borrowing time and the preset user book borrowing time. Obtain the user borrowing return rate deviation according to the relative relationship between the user borrowing return rate and the preset user borrowing return rate obtained from the preset database. The user borrowing return rate deviation is used to reflect the deviation degree between the user borrowing return rate and the preset user borrowing return rate. Obtain the book borrowing coefficient deviation according to the relative relationship between the book borrowing coefficient and the preset book borrowing coefficient obtained from the preset database. The book borrowing coefficient deviation is used to reflect the deviation degree between the book borrowing coefficient and the preset book borrowing coefficient. When the user borrowing return rate is not less than the preset user borrowing return rate and the book borrowing coefficient is not less than the preset book borrowing coefficient, process the user borrowing waiting time deviation, the user actual borrowing time deviation, the user borrowing return rate deviation, the book borrowing coefficient deviation, and the allocation weight obtained from the preset database to obtain the allocation index. Otherwise, record the allocation index as 0. The allocation weight includes the borrowing waiting weight, the borrowing time weight, the borrowing return rate weight, and the borrowing coefficient weight.

[0015] Further, the specific method for determining whether to optimize the book allocation based on the allocation index is as follows: Determine whether the allocation index of the library is less than the preset allocation threshold obtained from the preset database. If the allocation index of the library is not less than the preset allocation threshold obtained from the preset database, do not optimize the book allocation. If the allocation index of the library is less than the preset allocation threshold obtained from the preset database, optimize the book allocation. The book allocation optimization includes dynamic resource scheduling and inventory management optimization. The dynamic resource scheduling means dynamically adjusting the book location according to the historical borrowing data. The inventory management optimization means adjusting the inventory of books whose borrowing user quantity is not within the preset borrowing user quantity range.

[0016] Further, after the optimization of book allocation, it also includes optimization judgment, which is specifically as follows: judge whether the allocation index obtained after dynamic resource adjustment is less than the preset allocation threshold obtained from the preset database; if the allocation index obtained after dynamic resource adjustment is less than the preset allocation threshold obtained from the preset database, perform inventory management optimization, otherwise end the book allocation optimization; judge whether the allocation index obtained after inventory management optimization is less than the preset allocation threshold obtained from the preset database; if the allocation index obtained after inventory management optimization is less than the preset allocation threshold obtained from the preset database, give feedback, otherwise end the book allocation optimization.

[0017] The embodiment of the present application provides a demand prediction and intelligent allocation system for library collection optimization, including: a book borrowing evaluation module, a user recommendation evaluation module, and a book allocation evaluation module; wherein, the book borrowing evaluation module is used to evaluate the book borrowing index of a specified book based on the obtained actual borrowing number and book prediction data, and judge whether to perform new book inventory management based on the book borrowing index, and the book borrowing index is used to quantitatively evaluate the borrowing ability of the specified book in the library demand prediction; the user recommendation evaluation module is used to, if new book inventory management is performed, evaluate the user borrowing index of a specified user based on the obtained new user recommendation data, and judge whether to optimize the book recommendation strategy based on the user borrowing index, and the user borrowing index is used to quantitatively evaluate the effectiveness of recommending books to the specified user in the library demand prediction; the book allocation evaluation module is used to, if the book recommendation strategy is optimized, evaluate the allocation index of a specified book based on the obtained book allocation data, and judge whether to optimize the book allocation based on the allocation index, and the allocation index is used to quantitatively evaluate the efficiency of book allocation in the library intelligent allocation.

[0018] One or more technical solutions provided in the embodiment of the present application have at least the following technical effects or advantages: 1. By judging whether to perform new book inventory management based on the obtained book borrowing index of the specified book, then judging whether to optimize the book recommendation strategy based on the obtained user borrowing index of the specified user from the new user recommendation data, and finally judging whether to optimize the book allocation based on the obtained allocation index of the specified book from the book allocation data, the library collection inventory is optimized, and further the accuracy of new book allocation in the library is improved, effectively solving the problem of insufficient data in the prior art when predicting the demand and analyzing the allocation of new books in the library.

[0019] 2. The actual borrowing number deviation is obtained from the actual borrowing number and the preset borrowing number data. Then, the predicted borrowing rate deviation is obtained based on the predicted borrowing rate and the preset borrowing rate. Next, the borrowing prediction accuracy deviation is obtained according to the borrowing prediction accuracy and the preset borrowing prediction accuracy. Finally, the book borrowing deviation, book similarity, and borrowing weight are processed to obtain the book borrowing index, thereby quantitatively evaluating the borrowing ability of the specified book in the library demand prediction, and then realizing the adjustment of the inventory of newly introduced library collections.

[0020] 3. The user borrowing waiting time deviation is obtained from the user borrowing waiting time and the preset user borrowing waiting time. Then, the actual user borrowing time deviation is obtained based on the actual user borrowing time and the preset user book borrowing time. Finally, the user borrowing waiting time deviation, the actual user borrowing time deviation, the user borrowing return rate, the book borrowing coefficient, the preset book borrowing coefficient, the preset user borrowing return rate, and the allocation weight are processed to obtain the allocation index, thereby quantitatively evaluating the efficiency of book allocation in the library intelligent allocation, and then realizing the improvement of the library collection intelligent allocation efficiency. Brief Description of the Drawings

[0021] Figure 1 It is a flowchart of the demand prediction and intelligent allocation method for library collection optimization provided by the embodiment of the present application; Figure 2 It is a schematic structural diagram of the demand prediction and intelligent allocation system for library collection optimization provided by the embodiment of the present application. Detailed Embodiment

[0022] The embodiment of the present application provides a demand prediction and intelligent allocation method and system for library collection optimization, which solves the problem of insufficient data in the prior art when predicting the demand and analyzing the allocation of new books in the library. The book borrowing index of the specified book is obtained through evaluation based on the actual borrowing number and book prediction data obtained. Whether to manage the new book inventory is judged based on the book borrowing index. Then, the user borrowing index of the specified user is obtained through evaluation based on the new user recommendation data obtained. Whether to optimize the book recommendation strategy is judged based on the user borrowing index. Finally, the allocation index of the specified book is obtained through evaluation based on the book allocation data obtained. Whether to optimize the book allocation is judged based on the allocation index, realizing the improvement of the accuracy of the new book allocation in the library.

[0023] The technical solution in the embodiment of the present application is to solve the problem of insufficient data in the above-mentioned demand prediction and allocation analysis of new books in the library. The general idea is as follows: Determine whether to conduct new book inventory management by specifying the book borrowing index of a book, then obtain the user borrowing index of a specified user based on the new user recommendation data to determine whether to optimize the book recommendation strategy, and finally obtain the allocation index of a specified book based on the book allocation data to determine whether to optimize the book allocation, achieving the effect of improving the accuracy of new book allocation in the library.

[0024] To better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings of the specification and specific embodiments.

[0025] As Figure 1 shown, it is a flowchart of a demand prediction and intelligent allocation method for library collection optimization provided by an embodiment of the present application. The method includes the following steps: S1, book borrowing assessment: evaluate according to the obtained actual borrowing number and book prediction data to obtain the book borrowing index of a specified book, and determine whether to conduct new book inventory management based on the book borrowing index. The book borrowing index is used to quantitatively evaluate the borrowing ability of a specified book in the library demand prediction; S2, user recommendation assessment: if new book inventory management is carried out, evaluate according to the obtained new user recommendation data to obtain the user borrowing index of a specified user, and determine whether to optimize the book recommendation strategy based on the user borrowing index. The user borrowing index is used to quantitatively evaluate the effectiveness of recommending books to a specified user in the library demand prediction; S3, book allocation assessment: if the book recommendation strategy is optimized, evaluate according to the obtained book allocation data to obtain the allocation index of a specified book, and determine whether to optimize the book allocation based on the allocation index. The allocation index is used to quantitatively evaluate the efficiency of book allocation in the library intelligent allocation.

[0026] Among them, the book prediction data is obtained by analyzing the predicted borrowing quantity of new books and the actual borrowing data. The predicted borrowing quantity of new books is predicted through the historical borrowing data in the demand prediction system. The historical borrowing data includes book categories and the number of borrowing users. The actual borrowing data includes specified book categories, the actual number of borrowers, and the total number of new books. The book prediction data includes the quantity of similar books, the borrowing prediction accuracy rate, the predicted borrowing rate, and the book similarity. The new user recommendation data includes the book recommendation precision coefficient, the book recommendation coverage coefficient, the user click-through rate, and the new book recommendation coefficient. The book allocation data includes the user borrowing waiting duration, the user actual borrowing duration, the user borrowing return rate, and the book borrowing coefficient. The historical borrowing data and the actual borrowing data are obtained through the library management system. The borrowing prediction accuracy rate is obtained by performing a ratio operation on the absolute value of the difference between the predicted borrowing quantity and the actual borrowing quantity and the predicted borrowing quantity. The predicted borrowing rate is obtained by performing a ratio operation on the predicted borrowing quantity and the total number of new books. The book similarity is represented by the maximum value of the similarity between the specified book and other books calculated by the cosine similarity for the specified book categories and authors and the book categories and authors. The quantity of similar books is obtained by counting the number of books with a specified book similarity greater than the preset similarity. The book recommendation precision coefficient is obtained by performing a ratio operation on the number of recommended books borrowed by a specified user and the total number of recommended books. The book recommendation coverage coefficient is obtained by performing a ratio operation on the number of recommended books borrowed by a specified user and the total number of books borrowed by the specified user. The user click-through rate is obtained by counting the number of times a specified user clicks on a book in the book recommendation system. The new book recommendation coefficient is obtained by performing a ratio operation on the difference between the actual maximum borrowing quantity of a specified book in the recommended books for new users and the actual maximum borrowing quantity of the specified book in the non-recommended books and the total number of books borrowed by the user. The user borrowing waiting duration is obtained by recording the time difference from when the user requests to borrow a book to when the borrowing is successful. The user actual borrowing duration is obtained by recording the time difference between the return time and the borrowing time of the book. The user borrowing return rate is obtained by performing a ratio operation on the number of books returned on time within a preset time period in the library and the total number of borrowings. The book borrowing coefficient is obtained by performing a ratio operation on the actual borrowing quantity of books within a preset time period and the total number of library books. The user borrowing return rate and the book borrowing coefficient are data comprehensively obtained by the library based on the borrowing data of all users.

[0027] In this embodiment, the designated books represent newly introduced books in the library; the designated users represent users without historical borrowing data in the library; demand forecasting is carried out through an existing demand forecasting system to predict borrowing demand and assist the library in inventory management. For example, in the demand forecasting system, an AutoRegressive Integrated Moving Average (ARIMA) model is used to predict the predicted borrowing quantity of new books based on the quantity of books of the same category as the new books and the number of borrowing users of books of the same category; books are recommended to the designated users through an existing book recommendation system to improve the borrowing experience of the designated users.

[0028] Through the above steps, the borrowing efficiency of the library is improved, and further, the accuracy of new book allocation in the library is improved.

[0029] Furthermore, the specific method for evaluating the book borrowing index of the designated books based on the obtained actual borrowing number and book prediction data is as follows: the actual borrowing number deviation is obtained according to the relative relationship between the actual borrowing number and the preset borrowing number data obtained from the preset database (i.e., ), and the preset borrowing number data includes the preset minimum borrowing number and the preset maximum borrowing number; the predicted borrowing rate deviation is obtained according to the relative relationship between the predicted borrowing rate and the preset borrowing rate obtained from the preset database (i.e., ); the borrowing prediction accuracy deviation is obtained according to the relative relationship between the borrowing prediction accuracy and the preset borrowing prediction accuracy obtained from the preset database (i.e., ); the book borrowing deviation, book similarity, borrowing weights, and the number factor of books of the same category obtained from the preset database are processed to obtain the book borrowing index; the book borrowing deviation includes the actual borrowing number deviation, the predicted borrowing rate deviation, and the borrowing prediction accuracy deviation; the actual borrowing number deviation is used to reflect the deviation degree between the actual borrowing number and the preset borrowing number data; the predicted borrowing rate deviation is used to reflect the deviation degree between the predicted borrowing rate and the preset borrowing rate; the borrowing prediction accuracy deviation is used to reflect the deviation degree between the borrowing prediction accuracy and the preset borrowing prediction accuracy; the borrowing weights include the borrowing number weight, the borrowing rate weight, the borrowing accuracy weight, and the similarity weight.

[0030] Among them, the specific method for obtaining the book borrowing index is: ; ; ; ; In the formula, i represents the serial number of the type of the designated books, , N represents the total number of types of the designated books, represents the book borrowing index of the i-th specified book, XT represents the factor of the number of books in the same category, represents the deviation of the actual borrowing population of the i-th specified book, represents the deviation of the predicted borrowing rate of the i-th specified book, represents the deviation of the borrowing prediction accuracy rate of the i-th specified book, represents the book similarity of the i-th specified book, represents the preset borrowing rate, represents the preset borrowing prediction accuracy rate, represents the actual borrowing population of the i-th specified book, represents the preset minimum borrowing population, represents the preset maximum borrowing population, represents the predicted borrowing rate of the i-th specified book, represents the borrowing prediction accuracy rate of the i-th specified book, represents the weight of the borrowing population, represents the weight of the borrowing rate, represents the weight of the borrowing accuracy, represents the weight of the similarity.

[0031] In this embodiment, the preset borrowing rate is represented by the average value of the predicted borrowing rates within the historical time period; the preset borrowing prediction accuracy rate is represented by the average value of the borrowing prediction accuracy rates within the historical time period; the preset minimum borrowing population is represented by the minimum value of the actual borrowing populations within the historical time period; the preset maximum borrowing population is represented by the maximum value of the actual borrowing populations within the historical time period.

[0032] The borrowing weights are obtained from the preset database. The weight of the borrowing population represents the influence degree of the actual borrowing population on the book borrowing index. The weight of the borrowing rate represents the influence degree of the predicted borrowing rate on the book borrowing index. The weight of the borrowing accuracy represents the influence degree of the borrowing prediction accuracy rate on the book borrowing index. The weight of the similarity represents the influence degree of the book similarity on the book borrowing index; the sum of the four is 1. For example, the actual borrowing population and the preset book borrowing index form a mapping set, and the real-time actual borrowing population is input into the mapping set to obtain the corresponding weight of the borrowing population; the predicted borrowing rate and the preset book borrowing index form a mapping set, and the real-time predicted borrowing rate is input into the mapping set to obtain the corresponding weight of the borrowing rate; the borrowing prediction accuracy rate and the preset book borrowing index form a mapping set, and the real-time borrowing prediction accuracy rate is input into the mapping set to obtain the corresponding weight of the borrowing accuracy; the book similarity and the preset book borrowing index form a mapping set, and the real-time book similarity is input into the mapping set to obtain the corresponding weight of the borrowing accuracy; the mapping relationship therein can be one-to-one or many-to-one.

[0033] The quantity factor of similar books is obtained from the database. The quantity factor of similar books represents the degree of influence of the quantity of similar books on the book borrowing index. For example, a mapping set of the quantity of similar books and their corresponding influence factors is constructed based on the relationship between the historical quantity of similar books and book prediction data (such as borrowing prediction accuracy, predicted borrowing rate), and the real-time quantity of similar books is input into the mapping set to obtain the corresponding quantity factor of similar books.

[0034] In this algorithm, the book borrowing index is obtained by processing multiple independent variables (actual borrowing number, borrowing prediction accuracy, predicted borrowing rate, and book similarity). There are mutual influence relationships among these independent variables; the higher the predicted borrowing rate, the more likely it is to cause an increase in the actual borrowing number; the higher the book similarity, the more likely it is to cause an increase in the predicted borrowing rate and the actual borrowing number; the lower the borrowing prediction accuracy, the more likely it is to cause a shortage of the inventory of the specified book and a decrease in the predicted borrowing rate, which may in turn lead to a decrease in the actual borrowing rate and the actual borrowing number of the specified book; for a specified book with a higher predicted borrowing rate, the actual borrowing number is usually also larger. Therefore, the predicted borrowing rate is one of the key factors affecting the actual borrowing number.

[0035] In this algorithm, the book borrowing index is positively correlated with the actual borrowing number, borrowing prediction accuracy, predicted borrowing rate, and book similarity.

[0036] Taking the quantity factor of similar books as 1, and the weights of the borrowing number, borrowing rate, borrowing accuracy, and similarity as 0.25, 0.2, 0.3, and 0.25 respectively, the change statistical table of the book borrowing index is shown in Table 1: Table 1 Change Statistical Table of Book Borrowing Index From the first and second groups of data in the table, it can be seen that the book borrowing index increases with the increase of the deviation of the actual borrowing number; from the second and third groups of data, it can be seen that the book borrowing index increases with the increase of the deviation of the predicted borrowing rate; from the third and fourth groups, it can be seen that the book borrowing index increases with the increase of the deviation of the borrowing prediction accuracy; from the fourth and fifth groups of data, it can be seen that the book borrowing index increases with the increase of the book similarity.

[0037] Through the above steps, the borrowing ability of the specified book in the library demand prediction is quantitatively evaluated, and then the adjustment of the inventory of newly introduced library collections is realized.

[0038] Further, the specific method for determining whether to conduct new book inventory management based on the book borrowing index is as follows: Determine whether the book borrowing index of the specified book is less than the preset book borrowing threshold obtained from the preset database. The preset book borrowing threshold includes the first preset book borrowing threshold and the second preset book borrowing threshold. If the book borrowing index of the specified book is less than the first preset book borrowing threshold obtained from the preset database, gradually reduce the inventory of the specified book. If the book borrowing index of the specified book is not less than the first preset book borrowing threshold obtained from the preset database and not greater than the second preset book borrowing threshold obtained from the preset database, do not conduct new book inventory management. If the book borrowing index of the specified book is greater than the second preset book borrowing threshold obtained from the preset database, gradually increase the inventory of the specified book.

[0039] In this embodiment, the first preset book borrowing threshold is represented by the difference between the average value of the qualified book borrowing indices within the historical time period and three times the variance. The second preset book borrowing threshold is represented by the sum of the average value of the qualified book borrowing indices within the historical time period and three times the variance.

[0040] Gradually reduce the inventory of the specified book: Each time, reduce 10% of the original inventory of the specified book. Gradually increase the inventory of the specified book: Each time, increase 10% of the original inventory of the specified book.

[0041] Through the above steps, the borrowing ability of the specified book in the library demand prediction is quantitatively evaluated, and then the adjustment of the inventory of the newly introduced library collections is realized.

[0042] Further, the specific method for obtaining the user borrowing index of the specified user through evaluation based on the obtained new user recommendation data is as follows: Obtain the deviation of the book recommendation accuracy coefficient (i.e., ) according to the relative relationship between the book recommendation accuracy coefficient and the preset book recommendation accuracy coefficient obtained from the preset database. The preset book recommendation accuracy coefficient includes the first preset book recommendation accuracy coefficient and the second preset book recommendation accuracy coefficient. The deviation of the book recommendation accuracy coefficient is used to reflect the deviation degree between the book recommendation accuracy coefficient and the preset book recommendation accuracy coefficient. Obtain the deviation of the book recommendation coverage coefficient (i.e., ) according to the relative relationship between the book recommendation coverage coefficient and the preset book recommendation coverage coefficient obtained from the preset database. The preset book recommendation coverage coefficient includes the first preset book recommendation coverage coefficient and the second preset book recommendation coverage coefficient. The deviation of the book recommendation coverage coefficient is used to reflect the deviation degree between the book recommendation coverage coefficient and the preset book recommendation coverage coefficient. Obtain the deviation of the new book recommendation coefficient (i.e., ), the preset new book recommendation coefficient includes a first preset new book recommendation coefficient and a second preset new book recommendation coefficient, and the new book recommendation coefficient deviation is used to reflect the deviation degree between the new book recommendation coefficient and the preset new book recommendation coefficient; the user click-through rate deviation is obtained according to the relative relationship between the user click-through rate and the preset user click-through rate obtained from the preset database (i.e., ), and the user click-through rate deviation is used to reflect the deviation degree between the user click-through rate and the preset user click-through rate; the book recommendation accuracy coefficient deviation, the book recommendation coverage coefficient deviation, the new book recommendation coefficient deviation, the user click-through rate deviation and the user borrowing weight obtained from the preset database are processed to obtain the user borrowing index; the user borrowing weight includes a recommendation accuracy weight, a recommendation coverage weight, a new book recommendation weight and a user click weight.

[0043] Among them, the specific method for obtaining the user borrowing index is: ; ; ; ; In the formula, m represents the number of the specified user, , M represents the total number of specified users, represents the user borrowing index of the m-th specified user, represents the book recommendation accuracy coefficient deviation of the m-th specified user, represents the book recommendation coverage coefficient deviation of the m-th specified user, represents the user click-through rate of the m-th specified user, represents the new book recommendation coefficient deviation of the m-th specified user, represents the preset user click-through rate, represents the book recommendation accuracy coefficient of the m-th specified user, represents the first preset book recommendation accuracy coefficient, represents the second preset book recommendation accuracy coefficient, represents the book recommendation coverage coefficient of the m-th specified user, represents the first preset book recommendation coverage coefficient, represents the second preset book recommendation coverage coefficient, represents the new book recommendation coefficient of the m-th specified user, represents the first preset new book recommendation coefficient, represents the second preset new book recommendation coefficient, represents the recommendation accuracy weight, represents the recommendation coverage weight, represents the user click weight, represents the new book recommendation weight.

[0044] In this embodiment, the preset user click-through rate is represented by the average value of the user click-through rate within a historical time period; the first preset book recommendation accuracy coefficient is represented by the difference between the average value of the book recommendation accuracy coefficient within a historical time period and three times the variance, and the second preset book recommendation accuracy coefficient is represented by the sum of the average value of the book recommendation accuracy coefficient within a historical time period and three times the variance; the first preset book recommendation coverage coefficient is represented by the difference between the average value of the book recommendation coverage coefficient within a historical time period and three times the variance, and the second preset book recommendation coverage coefficient is represented by the sum of the average value of the book recommendation coverage coefficient within a historical time period and three times the variance; the first preset new book recommendation coefficient is represented by the difference between the average value of the new book recommendation coefficient within a historical time period and three times the variance, and the second preset new book recommendation coefficient is represented by the sum of the average value of the new book recommendation coefficient within a historical time period and three times the variance.

[0045] The user borrowing weight is obtained from a preset database. The recommendation accuracy weight represents the influence degree of the book recommendation accuracy coefficient on the user borrowing index, the recommendation coverage weight represents the influence degree of the book recommendation coverage coefficient on the user borrowing index, the new book recommendation weight represents the influence degree of the new book recommendation coefficient on the user borrowing index, and the user click weight represents the influence degree of the user click-through rate on the user borrowing index; the sum of the four is 1. For example, the book recommendation accuracy coefficient forms a mapping set with the preset user borrowing index, and the real-time book recommendation accuracy coefficient is input into the mapping set to obtain the corresponding recommendation accuracy weight; the book recommendation coverage coefficient forms a mapping set with the preset user borrowing index, and the real-time book recommendation coverage coefficient is input into the mapping set to obtain the corresponding recommendation coverage weight; the new book recommendation coefficient forms a mapping set with the preset user borrowing index, and the real-time new book recommendation coefficient is input into the mapping set to obtain the corresponding new book recommendation weight; the user click-through rate forms a mapping set with the preset user borrowing index, and the real-time user click-through rate is input into the mapping set to obtain the corresponding user click weight; the mapping relationship therein can be a one-to-one or many-to-one relationship.

[0046] In this algorithm, the user borrowing index is obtained by processing multiple independent variables (book recommendation accuracy coefficient, book recommendation coverage coefficient, user click-through rate, and new book recommendation coefficient). There are mutual influence relationships among these independent variables; the higher the book recommendation accuracy coefficient, the more likely the specified user is to be interested in the recommended books, thereby increasing the corresponding user click-through rate and book recommendation coverage coefficient; if the book recommendation system more accurately meets the user's needs, the corresponding user click-through rate is higher, so the number of successfully recommended books will increase, thereby increasing the book recommendation accuracy coefficient and book recommendation coverage coefficient of the corresponding specified user; the higher the new book recommendation coefficient, it may lead to an increase in the book recommendation accuracy coefficient, book recommendation coverage coefficient, and user click-through rate of the corresponding specified user.

[0047] In this algorithm, the user borrowing index is positively correlated with the book recommendation accuracy coefficient, the book recommendation coverage coefficient, the user click-through rate, and the new book recommendation coefficient.

[0048] Through the above steps, the effectiveness of recommending books to a specified user in library demand prediction is quantitatively evaluated, and thus the accuracy of book recommendation in the library is improved.

[0049] Furthermore, the specific method for determining whether to optimize the book recommendation strategy based on the user borrowing index is as follows: Determine whether the user borrowing index of the specified user is less than the preset user borrowing threshold obtained from the preset database: If the user borrowing index of the specified user is not less than the preset user borrowing threshold obtained from the preset database, then do not optimize the book recommendation strategy; If the user borrowing index of the specified user is less than the preset user borrowing threshold obtained from the preset database, then optimize the book recommendation strategy; Optimizing the book recommendation strategy means adjusting the recommendation strategy in real time according to the basic information and user behavior of new users.

[0050] In this embodiment, the preset user borrowing threshold is represented by the average value of the qualified user borrowing indices within the historical time period.

[0051] Use the basic information of new users (such as age, subject, interests, etc.) for initial book recommendations: For example, a new user registers for a library account and fills in basic information indicating that they are a college student majoring in computer science. The book recommendation system can recommend some popular books and classic textbooks in computer science, such as "Introduction to Algorithms", "Data Structures and Algorithm Analysis", etc. based on the basic information; It can also recommend according to market hotspots, such as new works by Nobel laureates in literature.

[0052] Adjust the book recommendation strategy according to the user click record and search record. For example, a new user starts borrowing books on machine learning. The book recommendation system can identify the user's interest in machine learning and recommend more books related to machine learning, such as "Python Machine Learning", etc.

[0053] Through the above steps, the accuracy of book recommendation for new users in the library is improved.

[0054] Further, the specific method for obtaining the allocation index of a specified book by evaluating the obtained book allocation data is as follows: Obtain the user borrowing waiting time deviation (i.e., JYP) based on the relative relationship between the user borrowing waiting time and the preset user borrowing waiting time obtained from the preset database. The user borrowing waiting time deviation is used to reflect the deviation degree between the user borrowing waiting time and the preset user borrowing waiting time; obtain the user actual borrowing time deviation (i.e., JTP) based on the relative relationship between the user actual borrowing time and the preset user book borrowing time obtained from the preset database. The user actual borrowing time deviation is used to reflect the deviation degree between the user actual borrowing time and the preset user book borrowing time; obtain the user borrowing return rate deviation (i.e., ), the user borrowing return rate deviation is used to reflect the deviation degree between the user borrowing return rate and the preset user borrowing return rate; obtain the book borrowing coefficient deviation (i.e., ), the book borrowing coefficient deviation is used to reflect the deviation degree between the book borrowing coefficient and the preset book borrowing coefficient; when the user borrowing return rate is not less than the preset user borrowing return rate and the book borrowing coefficient is not less than the preset book borrowing coefficient, process the user borrowing waiting time deviation, the user actual borrowing time deviation, the user borrowing return rate deviation, the book borrowing coefficient deviation, and the allocation weight obtained from the preset database to obtain the allocation index, otherwise record the allocation index as 0; the allocation weight includes the borrowing waiting weight, the borrowing time weight, the borrowing return rate weight, and the borrowing coefficient weight; the specific method for obtaining the allocation index is: ; ; ; In the formula, i represents the number of the specified book, , N represents the total number of types of the specified book, m represents the number of the specified user, , M represents the total number of the specified users, TP represents the allocation index of the library, JYP represents the user borrowing waiting time deviation, JTP represents the user actual borrowing time deviation, JGH represents the user borrowing return rate, JYX represents the book borrowing coefficient, represents the preset user borrowing return rate, represents the preset book borrowing coefficient, represents the user borrowing waiting time of the m-th specified user borrowing the i-th specified book, represents the user actual borrowing time of the m-th specified user borrowing the i-th specified book, represents the preset user borrowing waiting time, Indicates the preset borrowing time of user books. Indicates the borrowing waiting weight, Indicates the borrowing time weight, represents the borrowing return rate weight, Represents the borrowing coefficient weight.

[0055] In this embodiment, the preset user borrowing return rate is represented by the average value of the user borrowing return rate in the historical time period; the preset book borrowing coefficient is represented by the average value of the book borrowing coefficient in the historical time period; the preset user borrowing return rate and the preset book borrowing coefficient are data comprehensively obtained by the library based on the borrowing data of all users.

[0056] The allocation weight is obtained from a preset database, the borrowing waiting weight indicates the influence of the user's borrowing waiting time on the allocation index, the borrowing time weight indicates the influence of the user's actual borrowing time on the allocation index, the borrowing return rate weight indicates the influence of the user's borrowing return rate on the allocation index, and the borrowing coefficient weight indicates the influence of the book borrowing coefficient on the allocation index; the sum of the four is 1, for example, the user's borrowing waiting time and the preset allocation index form a mapping set, and the real-time user's borrowing waiting time is input into the mapping set to obtain the corresponding borrowing waiting weight; the user's actual borrowing time and the preset allocation index form a mapping set, and the real-time user's actual borrowing time is input into the mapping set to obtain the corresponding borrowing time weight; the user's borrowing return rate and the preset allocation index form a mapping set, and the real-time user's borrowing return rate is input into the mapping set to obtain the corresponding borrowing return rate weight; the book borrowing coefficient and the preset allocation index form a mapping set, and the real-time book borrowing coefficient is input into the mapping set to obtain the corresponding borrowing coefficient weight; the mapping relationship can be one-to-one or many-to-one.

[0057] The allocation index in this algorithm involves processing multiple independent variables (user borrowing waiting time, user actual book borrowing time, user borrowing return rate and book borrowing coefficient), and there is a mutual influence relationship between these independent variables; if the borrowing waiting time is longer, the borrowing willingness of any user in the library may decrease, causing them to give up borrowing books, which in turn leads to a decrease in the book borrowing coefficient; a higher borrowing return rate may lead to an increase in the user's borrowing waiting time; a higher book borrowing coefficient means that the library resource utilization rate is higher and the borrowing index is higher. A lower book borrowing coefficient may mean that a large number of books are ignored or not borrowed in time, thus affecting the book borrowing index, and a shorter actual book borrowing time may lead to a shorter user borrowing waiting time.

[0058] In this algorithm, when the user's borrowing waiting time is not greater than the preset user borrowing waiting time, the smaller the user's borrowing waiting time, the larger the allocation index; when the user's actual book borrowing time is not greater than the preset user book borrowing time, the smaller the user's actual book borrowing time, the larger the allocation index; the allocation index is positively correlated with the user borrowing return rate and the book borrowing coefficient.

[0059] Taking the preset user borrowing return rate as 0.3, the preset book borrowing coefficient as 0.3, and the borrowing waiting weight, borrowing duration weight, borrowing return rate weight, and borrowing coefficient weight as 0.25, 0.2, 0.25, and 0.3 respectively, the change statistical table of the allocation index is shown in Table 2: Table 2 Change Statistical Table of Allocation Index From the first and second groups of data in the table, it can be seen that the allocation index increases with the increase of the deviation of the user's borrowing waiting time; from the second and third groups of data, it can be seen that the allocation index increases with the increase of the deviation of the user's actual borrowing time; from the third and fourth groups, it can be seen that the allocation index increases with the increase of the user borrowing return rate; from the fourth and fifth groups of data, it can be seen that the allocation index increases with the increase of the book borrowing coefficient.

[0060] Through the above steps, the efficiency of book allocation in the intelligent allocation of the library is quantitatively evaluated, and then the improvement of the intelligent allocation efficiency of the library collection is realized.

[0061] Furthermore, the specific method for judging whether to optimize book allocation based on the allocation index is as follows: judge whether the allocation index of the library is less than the preset allocation threshold obtained from the preset database: if the allocation index of the library is not less than the preset allocation threshold obtained from the preset database, then no book allocation optimization is performed; if the allocation index of the library is less than the preset allocation threshold obtained from the preset database, then book allocation optimization is performed; book allocation optimization includes dynamic resource scheduling and inventory management optimization; dynamic resource scheduling means dynamically adjusting the book location according to historical borrowing data; inventory management optimization means adjusting the inventory of books whose borrowing user numbers are not within the preset borrowing user number range.

[0062] It should be added that after the optimization of book allocation, it also includes optimization judgment, specifically as follows: judging whether the allocation index obtained after dynamic resource adjustment is less than the preset allocation threshold obtained from the preset database; if the allocation index obtained after dynamic resource adjustment is less than the preset allocation threshold obtained from the preset database, inventory management optimization is carried out, otherwise the book allocation optimization ends; judging whether the allocation index obtained after inventory management optimization is less than the preset allocation threshold obtained from the preset database; if the allocation index obtained after inventory management optimization is less than the preset allocation threshold obtained from the preset database, feedback is given to the preset personnel, otherwise the book allocation optimization ends.

[0063] In this embodiment, the preset allocation threshold is represented by the average value of the qualified allocation indexes within the historical time period.

[0064] Dynamic resource adjustment: The demand prediction system is used to predict the borrowing trend based on historical borrowing data. For example, during the exam season, it is predicted that the borrowing volume of professional books will increase significantly. Therefore, professional books can be allocated to popular areas (such as the study rooms or reading rooms in the library) in advance to meet the needs of readers.

[0065] Inventory management optimization: The preset borrowing user quantity range is represented by the difference and sum of the average value of the borrowing user quantities within the historical time period and three times the variance. The preset minimum borrowing user quantity is represented by the difference between the average value of the borrowing user quantities within the historical time period and three times the variance. The preset maximum borrowing user quantity range is represented by the sum of the average value of the borrowing user quantities within the historical time period and three times the variance; when the borrowing user quantity of any book is less than the preset minimum borrowing user quantity, the inventory of the corresponding book is reduced; when the borrowing user quantity of any book is greater than the preset maximum borrowing user quantity, the inventory of the corresponding book is increased.

[0066] Through the above steps, the improvement of the intelligent allocation efficiency of the library collection is achieved.

[0067] Such as Figure 2As shown in the figure, it is a schematic structural diagram of a demand prediction and intelligent allocation system for library collection optimization provided by an embodiment of the present application. The demand prediction and intelligent allocation system for library collection optimization provided by an embodiment of the present application includes: a book borrowing evaluation module, a user recommendation evaluation module, and a book allocation evaluation module; among them, the book borrowing evaluation module is used to evaluate based on the obtained actual borrowing number and book prediction data to obtain the book borrowing index of a specified book, and determine whether to conduct new book inventory management based on the book borrowing index. The book borrowing index is used to quantitatively evaluate the borrowing ability of a specified book in the library demand prediction; the user recommendation evaluation module is used to, if new book inventory management is carried out, evaluate based on the obtained new user recommendation data to obtain the user borrowing index of a specified user, and determine whether to optimize the book recommendation strategy based on the user borrowing index. The user borrowing index is used to quantitatively evaluate the effectiveness of recommending books to a specified user in the library demand prediction; the book allocation evaluation module is used to, if the book recommendation strategy is optimized, evaluate based on the obtained book allocation data to obtain the allocation index of a specified book, and determine whether to optimize the book allocation based on the allocation index. The allocation index is used to quantitatively evaluate the efficiency of book allocation in the library intelligent allocation.

[0068] In this embodiment, the book borrowing evaluation module uses the demand prediction system to predict the borrowing demand based on historical borrowing data, obtains the book borrowing index of a specified book, and determines whether it is necessary to conduct new book inventory management for the specified book. For example, the library may decide to replenish the inventory of this book; through the user recommendation evaluation module, by analyzing the borrowing preferences and borrowing frequencies of users, the book recommendation system can more accurately recommend books that users may be interested in, thereby improving the borrowing experience of users; through the evaluation of the book allocation efficiency by the book allocation evaluation module, the library can dynamically adjust the location or quantity of books to ensure the maximum utilization of resources, reduce the waiting time of readers, and improve the overall operation efficiency of the library; by using quantitative indicators to evaluate and optimize the resource allocation and user experience of the library, the efficient management of the library collection is realized. In summary, the embodiment of the present application determines whether to conduct new book inventory management by obtaining the book borrowing index of a specified book, then determines whether to optimize the book recommendation strategy based on the user borrowing index of a specified user obtained from the obtained new user recommendation data, and finally determines whether to optimize the book allocation based on the allocation index of a specified book obtained from the obtained book allocation data, thereby optimizing the library collection inventory, and further improving the accuracy of the new book allocation in the library, effectively solving the problem of insufficient data in the prior art when predicting the demand and analyzing the allocation of new books in the library.

[0069] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) that contain computer-usable program code.

[0070] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0071] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing devices to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0072] These computer program instructions can also be loaded onto a computer or other programmable data processing devices, such that a series of operation steps are executed on the computer or other programmable devices to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable devices provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0073] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concepts. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications that fall within the scope of the present invention.

[0074] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A demand forecasting and intelligent allocation method for library collection optimization, characterized in that: The following steps are involved: S1, evaluating the book borrowing index of a specified book based on the actual number of borrowers and book forecast data, and determining whether to perform new book inventory management based on the book borrowing index, wherein the book borrowing index is used to quantitatively evaluate the borrowing capacity of a specified book in the library demand forecast; S2, after the new book inventory management is performed, the user borrowing index of the specified user is obtained by evaluating the obtained new user recommendation data, and whether to optimize the book recommendation strategy is determined based on the user borrowing index. The user borrowing index is used to quantitatively evaluate the effectiveness of recommending books to the specified user in the library demand forecast; S3: After optimizing the book recommendation strategy, the allocation index of the specified book is obtained through evaluation based on the acquired book allocation data, and whether to optimize the book allocation is determined based on the allocation index. The allocation index is used to quantitatively evaluate the efficiency of book allocation in intelligent library allocation.

2. The demand forecasting and intelligent allocation method for library collection optimization as claimed in claim 1, characterized in that: The book prediction data is obtained by analyzing the historical borrowing data, and the predicted number of new book borrowings is obtained by predicting the historical borrowing data in the demand prediction system; The historical borrowing data includes book categories and the number of borrowing users; The actual borrowing data includes the designated book category, the actual number of borrowers and the total number of new books; The book prediction data includes the number of similar books, borrowing prediction accuracy, predicted borrowing rate and book similarity; The new user recommendation data includes book recommendation accuracy coefficient, book recommendation coverage coefficient, user click rate and new book recommendation coefficient; The book allocation data includes the user's borrowing waiting time, the user's actual borrowing time, the user's borrowing return rate and the book borrowing coefficient; The borrowing prediction accuracy is obtained by calculating the ratio of the absolute value of the difference between the predicted borrowing quantity and the actual borrowing quantity to the predicted borrowing quantity; The predicted borrowing rate is obtained by calculating the ratio of the predicted borrowing quantity to the total number of new books; The book similarity is represented by the maximum value of the cosine similarity between the specified book and the author and the similarity between the specified book and other books calculated by the book category and the author; The book recommendation accuracy coefficient is obtained by calculating the ratio of the number of recommended books borrowed by the specified user to the total number of recommended books; The book recommendation coverage coefficient is obtained by calculating the ratio of the number of recommended books borrowed by the designated user to the total number of books borrowed by the designated user; The new book recommendation coefficient is obtained by performing a ratio operation on the difference between the actual maximum borrowing quantity of the specified book in the recommended books of the new user and the actual maximum borrowing quantity of the specified book in the non-recommended books and the total number of books borrowed by the user; The book borrowing coefficient is obtained by recording the actual number of books borrowed within a preset time period and performing a ratio calculation with the total number of books in the library.

3. The demand forecasting and intelligent allocation method for library collection optimization as claimed in claim 1, characterized in that: The specific method of evaluating and obtaining the book borrowing index of a specified book based on the actual number of borrowers and book forecast data is as follows: The book borrowing deviation, book similarity, borrowing weight and quantity factor of similar books of the specified book are processed to obtain the book borrowing index; The book borrowing deviation includes the actual number of borrowers deviation, the predicted borrowing rate deviation and the borrowing prediction accuracy deviation; The actual number of borrowers deviation is used to reflect the degree of deviation between the actual number of borrowers and the preset number of borrowers data; The predicted borrowing rate deviation is used to reflect the deviation degree between the predicted borrowing rate and the preset borrowing rate; The borrowing prediction accuracy deviation is used to reflect the degree of deviation between the borrowing prediction accuracy and the preset borrowing prediction accuracy; The borrowing weights include borrower number weight, borrowing rate weight, borrowing accuracy weight and similarity weight.

4. The demand forecasting and intelligent allocation method for library collection optimization as claimed in claim 1, characterized in that: The specific method for determining whether to perform new book inventory management based on the book borrowing index is as follows: Determine whether the book borrowing index of the specified book is less than the preset book borrowing threshold obtained from the preset database: The preset book borrowing threshold includes a first preset book borrowing threshold and a second preset book borrowing threshold; If the book borrowing index of the specified book is less than the first preset book borrowing threshold obtained from the preset database, the inventory of the specified book is reduced step by step; If the book borrowing index of the specified book is not less than the first preset book borrowing threshold obtained from the preset database, and is not greater than the second preset book borrowing threshold obtained from the preset database, new book inventory management is not performed; If the book borrowing index of the specified book is greater than a second preset book borrowing threshold obtained from the preset database, the inventory of the specified book is increased step by step.

5. The demand forecasting and intelligent allocation method for library collection optimization as claimed in claim 1, characterized in that: The specific method of evaluating and obtaining the user borrowing index of a specified user based on the acquired new user recommendation data is as follows: Obtaining a book recommendation precision coefficient deviation according to a relative relationship between the book recommendation precision coefficient and a preset book recommendation precision coefficient obtained from a preset database, wherein the preset book recommendation precision coefficient includes a first preset book recommendation precision coefficient and a second preset book recommendation precision coefficient, and the book recommendation precision coefficient deviation is used to reflect the degree of deviation between the book recommendation precision coefficient and the preset book recommendation precision coefficient; Obtaining a book recommendation coverage coefficient deviation according to a relative relationship between the book recommendation coverage coefficient and a preset book recommendation coverage coefficient obtained from a preset database, wherein the preset book recommendation coverage coefficient includes a first preset book recommendation coverage coefficient and a second preset book recommendation coverage coefficient, and the book recommendation coverage coefficient deviation is used to reflect the degree of deviation between the book recommendation coverage coefficient and the preset book recommendation coverage coefficient; Obtaining a new book recommendation coefficient deviation according to a relative relationship between the new book recommendation coefficient and a preset new book recommendation coefficient obtained from a preset database, wherein the preset new book recommendation coefficient includes a first preset new book recommendation coefficient and a second preset new book recommendation coefficient, and the new book recommendation coefficient deviation is used to reflect the degree of deviation between the new book recommendation coefficient and the preset new book recommendation coefficient; Obtaining a user click rate deviation according to a relative relationship between the user click rate and a preset user click rate obtained from a preset database, wherein the user click rate deviation is used to reflect the degree of deviation between the user click rate and the preset user click rate; The book recommendation precision coefficient deviation, book recommendation coverage coefficient deviation, new book recommendation coefficient deviation, user click rate deviation and user borrowing weight obtained from the preset database are processed to obtain the user borrowing index; The user borrowing weight includes recommendation accuracy weight, recommendation coverage weight, new book recommendation weight and user click weight.

6. The demand forecasting and intelligent allocation method for library collection optimization as claimed in claim 1, characterized in that: The specific method for determining whether to optimize the book recommendation strategy based on the user borrowing index is as follows: Determine whether the user borrowing index of the specified user is less than the preset user borrowing threshold obtained from the preset database: If the user borrowing index of the specified user is not less than the preset user borrowing threshold obtained from the preset database, the book recommendation strategy will not be optimized; If the user borrowing index of the specified user is less than a preset user borrowing threshold obtained from a preset database, the book recommendation strategy is optimized.

7. The demand forecasting and intelligent allocation method for library collection optimization as claimed in claim 1, characterized in that: The specific method of evaluating the allocation index of a specified book based on the acquired book allocation data is as follows: Obtaining a user borrowing waiting time deviation according to a relative relationship between the user borrowing waiting time and a preset user borrowing waiting time obtained from a preset database, wherein the user borrowing waiting time deviation is used to reflect the degree of deviation between the user borrowing waiting time and the preset user borrowing waiting time; The actual borrowing time deviation of the user is obtained according to the relative relationship between the actual borrowing time of the user and the preset borrowing time of the user book obtained from the preset database, and the actual borrowing time deviation of the user is used to reflect the deviation degree between the actual borrowing time of the user and the preset borrowing time of the user book; Obtaining a user borrowing return rate deviation according to a relative relationship between the user borrowing return rate and a preset user borrowing return rate obtained from a preset database, wherein the user borrowing return rate deviation is used to reflect the degree of deviation between the user borrowing return rate and the preset user borrowing return rate; Obtaining a book borrowing coefficient deviation according to a relative relationship between the book borrowing coefficient and a preset book borrowing coefficient obtained from a preset database, wherein the book borrowing coefficient deviation is used to reflect the degree of deviation between the book borrowing coefficient and the preset book borrowing coefficient; When the user borrowing return rate is not less than the preset user borrowing return rate, and the book borrowing coefficient is not less than the preset book borrowing coefficient, the user borrowing waiting time deviation, the user actual borrowing time deviation, the user borrowing return rate deviation, the book borrowing coefficient deviation and the allocation weight obtained from the preset database are processed to obtain the allocation index, otherwise the allocation index is recorded as 0; The allocation weights include borrowing waiting weight, borrowing duration weight, borrowing return rate weight and borrowing coefficient weight.

8. The demand forecasting and intelligent allocation method for library collection optimization as claimed in claim 1, characterized in that: The specific method for determining whether to perform book allocation optimization based on the allocation index is as follows: Determine whether the library's allocation index is less than the preset allocation threshold obtained from the preset database: If the allocation index of the library is not less than the preset allocation threshold obtained from the preset database, no book allocation optimization is performed; If the allocation index of the library is less than the preset allocation threshold obtained from the preset database, book allocation optimization is performed; The book allocation optimization includes dynamic resource scheduling and inventory management optimization; The dynamic resource scheduling means dynamically adjusting the book location according to historical borrowing data; The inventory management optimization means adjusting the inventory of books whose number of borrowers is not within a preset range of the number of borrowers.

9. The demand forecasting and intelligent allocation method for library collection optimization as claimed in claim 8, characterized in that: The book allocation optimization also includes optimization judgment, which is as follows: Determine whether the allocation index obtained after the dynamic resource adjustment is less than a preset allocation threshold obtained from a preset database; If the allocation index obtained after dynamic resource adjustment is less than the preset allocation threshold obtained from the preset database, then the inventory management optimization is performed, otherwise the book allocation optimization is terminated; Determine whether the allocation index obtained after the inventory management optimization is less than a preset allocation threshold obtained from a preset database; If the allocation index obtained after the inventory management optimization is less than the preset allocation threshold obtained from the preset database, feedback is given, otherwise the book allocation optimization is terminated.

10. The demand forecasting and intelligent allocation system for library collection optimization is characterized by: include: Book borrowing evaluation module, user recommendation evaluation module and book allocation evaluation module; The book borrowing evaluation module is used to evaluate the book borrowing index of a specified book based on the actual number of borrowers and book forecast data, and determine whether to perform new book inventory management based on the book borrowing index. The book borrowing index is used to quantitatively evaluate the borrowing capacity of a specified book in the library demand forecast; The user recommendation evaluation module is used to evaluate the user borrowing index of the specified user according to the acquired new user recommendation data after the new book inventory management is carried out, and judge whether to optimize the book recommendation strategy based on the user borrowing index. The user borrowing index is used to quantitatively evaluate the effectiveness of recommending books to the specified user in the library demand forecast; The book allocation evaluation module is used to evaluate the allocation index of a specified book based on the acquired book allocation data after optimizing the book recommendation strategy, and to determine whether to optimize the book allocation based on the allocation index. The allocation index is used to quantitatively evaluate the efficiency of book allocation in intelligent library allocation.

Images