An intelligent book acquisition method and system based on book classification number

Abstract

This invention discloses an intelligent book acquisition method and system based on book classification numbers, belonging to the field of book acquisition. The method includes the following steps: establishing databases of existing library book abstracts, existing book classification numbers, the general book classification number, bookseller-provided book abstracts, and bookseller-provided book classification numbers; establishing and training a machine learning model for the correspondence between existing book abstracts and existing book classification numbers; determining whether bookseller-provided book classification numbers and output book classification numbers are within the range of all book classification numbers; sorting the available book classification numbers according to their acquisition priority and calculating the book price; and further determining the books to be acquired by the acquisition personnel to finalize the acquisition list. This invention improves the accuracy, flexibility, generalization, efficiency, and quality of book acquisition, contributing to a more efficient, accurate, comprehensive, and scientific approach to book acquisition.

Description

Technical Field

[0001] The invention belongs to the field of book acquisition, and in particular relates to an intelligent book acquisition method and system based on book classification numbers. Background Art

[0002] Book acquisition refers to the process of purchasing and visiting suitable book resources for the library's collection according to the library's own collection development policy, combined with the actual library situation, acquisition principles, readers' needs and financial budget. Book acquisition is one of the primary tasks and routine basic work of library business. It is also a complex professional and technical work, which constantly puts higher requirements on the personal business ability of the interviewers. The quality of book acquisition has a decisive impact on the library's collection structure and service level as well as the readers' satisfaction. Especially in recent years, with the geometric growth of book market demand, the types and quantity of published books have continued to surge, the quality of books has also been uneven, and in the face of the diverse levels of readers and the difficulty in grasping readers' book needs, the ordering of each book requires interviewers to make correct judgments and decisions quickly; especially the library's acquisition budget is usually limited, and resources need to be reasonably allocated, requiring interviewers to more carefully evaluate the quality, demand and price of books when conducting book interviews to ensure the effective use of the budget. Therefore, book acquisition work has become more and more a tedious and difficult task with a certain degree of blindness.

[0003] Traditional book acquisition methods include subjective decision-making acquisition method and quantitative decision-making acquisition method. Subjective decision-making acquisition method is the main book acquisition method at present. It usually combines reader demand drive with subjective judgment of interviewers to conduct book acquisition. It is easily affected by factors such as uncertainty of reader demand and subjective randomness of interviewers. It has shortcomings such as low efficiency and low quality of acquisition. Quantitative decision-making acquisition method is a more advanced acquisition method. It uses an acquisition decision model constructed on the basis of comprehensive consideration of reader demand factors and multiple acquisition factors to conduct book acquisition. It is limited by the collection construction planning of each library and the readers it serves. There are shortcomings such as lack of flexibility and generalization of the acquisition model and results. In short, the current book acquisition method has problems such as long time period, high labor intensity, insufficient objective basis for decision-making, strong subjectivity in decision-making, lack of comprehensiveness, and prominent contradiction between supply and demand between collections and readers. It is difficult to achieve satisfactory book acquisition results.

[0004] Therefore, developing a new method for book acquisition and making quick and scientific decisions on purchasing books from a variety of acquisition channels is of great significance for improving the quality of collected books and meeting readers' needs. Summary of the invention

[0005] The purpose of the present invention is to provide an intelligent book acquisition method and system based on book classification numbers, which takes the book classification number as the core element, takes into account the acquisition funds of the library, and conducts book acquisition based on artificial intelligence technology, thereby solving the above-mentioned problems existing in the current book acquisition, so as to complete the book acquisition work more efficiently, more accurately, more completely and more scientifically.

[0006] According to a first aspect of the technical solution of the present invention, there is provided an intelligent book acquisition method based on book classification numbers, comprising the following steps:

[0007] Step 1: Establish a first book abstract database and a first book classification number database of the library respectively, establish a second book classification number database according to the needs of readers, and merge the first book classification number database and the second book classification number database as a third book classification number database;

[0008] Step 2: According to the book acquisition principle, all the book classification codes in the third book classification code database are sorted according to the acquisition priority level;

[0009] Step 3: using the summary information in the first book summary database as input values ​​and the classification number information in the first book classification number database as output values, establishing a machine learning model of the corresponding relationship between the two, and training it to ensure that its accuracy is not lower than the target accuracy;

[0010] Step 4: respectively establish a second book abstract database and a fourth book classification number database provided by the bookseller, input the abstract information in the second book abstract database into the machine learning model, and output the output book classification number information that matches each other;

[0011] Step 5: Determine whether the classification number information in the fourth book classification number database or the output type book classification number information is in the third book classification number database; if not, do not treat it as an interview book; if so, treat it as an interviewable book;

[0012] Step 6: sorting the book classification codes of the books that can be interviewed according to the ranking of the interview priority levels, and calculating the book prices in order and in sections;

[0013] Step 7: When the book price reaches a predetermined threshold, the calculation is stopped. At this time, the books that can be interviewed corresponding to the calculated book price are the optimized and selected books to be interviewed.

[0014] Furthermore, in step 1, the first book abstract database is an existing book abstract database.

[0015] Furthermore, in step 1, the first book classification number database is an existing book classification number database.

[0016] Furthermore, in step 1, the second book classification number database is a newly added book classification number database.

[0017] Furthermore, in step 1, the third book classification number database is a general book classification number database.

[0018] Furthermore, in step 3, the target accuracy of the machine learning model is not less than 0.9.

[0019] Furthermore, in step 3, the machine learning model is a deep reinforcement learning model.

[0020] Furthermore, in step 7, the predetermined threshold of the fee is 1.0 to 1.5 times the book acquisition funds that the library intends to invest.

[0021] Furthermore, after step 7, the method further includes: book interviewers intervene to further determine the books to be interviewed and finally determine the books to be interviewed.

[0022] According to a second aspect of the technical solution of the present invention, an intelligent book acquisition system based on book classification numbers is provided, the system comprising: a processor and a memory for storing executable instructions; wherein the processor is configured to execute the executable instructions to execute the intelligent book acquisition method based on book classification numbers as described in any of the above aspects.

[0023] According to a third aspect of the technical solution of the present invention, there is provided a computer-readable storage medium, wherein a computer program is stored thereon, and when the computer program is executed by a processor, the intelligent book acquisition method based on book classification number as described in any of the above aspects is implemented.

[0024] Beneficial effects of the present invention:

[0025] 1. The technical solution of this application adopts a machine learning model for book interviews, so that efficient and accurate book interviews can be conducted based on book summary data and book classification number data, according to changes in reader demand and library acquisition funds, etc., providing new ideas and methods for book interviews, significantly reducing the cost of book interviews, greatly improving the accuracy, flexibility, generalization, efficiency and quality of book interviews, and being able to meet readers' changing needs faster and better.

[0026] 2. In the technical solution of this application, artificial intelligence book interviews can also be combined with interviewer book interviews, which can solve the problems of high cost, low efficiency, long cycle, many errors, low consistency and difficulty in meeting the ever-changing needs of readers in the current book interview methods. It can reduce the labor intensity of book interviewers, reduce the waste of manpower, material and financial resources, reduce costs, save time, and help to quickly realize the library's personalized and characteristic book interviews, improve readers' satisfaction, lay the foundation for future library book interviews, and point out a new direction.

[0027] 3. The technical solution of this application takes the book classification number as the core element. Since the book classification number is a specific and clear identification symbol assigned to the book according to the content and subject attributes of the book in accordance with certain classification tools (such as the "Chinese Library Classification"), it helps to achieve scientific classification and orderly management of books. It is a subject classification identification with strong scientificity recognized in the field of books. Therefore, combining the book classification number with the machine learning model makes the book acquisition process more efficient and accurate, and the book acquisition results are more complete and scientific. It is a new model for book acquisition that integrates scientific, standardized and efficient methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0029] Figure 1 The present invention is a flow chart of an intelligent book acquisition method based on book classification numbers. DETAILED DESCRIPTION

[0030] The present invention is described in detail below in conjunction with the embodiments. It should be noted that the following embodiments are only used to further illustrate the present invention and cannot be understood as limiting the scope of protection of the present invention. Those skilled in the art may make non-essential improvements and adjustments based on the contents of the present invention.

[0031] This embodiment provides an intelligent book acquisition method based on book classification numbers. Figure 1 As shown, the following steps are included:

[0032] Step 101: Establish a first book abstract database and a first book classification number database of the library respectively, establish a second book classification number database according to readers' needs, and merge the first book classification number database and the second book classification number database as a third book classification number database.

[0033] In a preferred embodiment, the first book abstract database is an existing book abstract database.

[0034] In a preferred embodiment, the first book classification number database is an existing book classification number database.

[0035] In a preferred embodiment, the second book classification number database is a newly added book classification number database.

[0036] In a preferred embodiment, the third book classification number database is a general book classification number database.

[0037] Here, the reader needs of the library include the needs for books with existing book classification numbers in the library and the needs for new books that the library does not have book classification numbers for yet. Therefore, according to the unified standard "Chinese Library Classification", an existing book classification number database including the existing book classification numbers of the library is established, and a new book classification number database of the library's new books that are about to be added to meet the new needs of readers is also established. The establishment method is the same as that of the library's existing book classification number database, and the establishment method of this database is well known in the art.

[0038] Step 102: According to the book acquisition principle, all the book classification codes in the third book classification code database are sorted according to the acquisition priority level.

[0039] Here, the book acquisition principle itself includes corresponding acquisition priority levels for all book classification numbers.

[0040] Step 103: Using the summary information in the first book summary database as input values ​​and the classification number information in the first book classification number database as output values, a machine learning model of the corresponding relationship between the two is established, and training is performed to ensure that its accuracy is not lower than the target accuracy.

[0041] In a preferred embodiment, the target accuracy of the machine learning model is not less than 0.9.

[0042] In a preferred embodiment, the machine learning model is a deep reinforcement learning model.

[0043] Here, before step 3, the step further includes: performing text vectorization processing on the summary information in the first book summary database and the classification number information in the first book classification number database. Therefore, the summary information in the first book summary database is vectorized summary information; and the classification number information in the first book classification number database is vectorized classification number information.

[0044] Step 104: Establish a second book abstract database and a fourth book classification number database provided by booksellers respectively, input the abstract information in the second book abstract database into the machine learning model, and output the output book classification number information that matches each other.

[0045] Step 105: Determine whether the classification number information in the fourth book classification number database or the output type book classification number information is in the third book classification number database; if not, do not treat it as an interview book; if so, treat it as an interviewable book.

[0046] Step 106: sort the book classification codes of the books that can be interviewed according to the ranking of the interview priority levels, and calculate the book prices in sequence and in sections.

[0047] Step 107: When the book price reaches a predetermined threshold, the calculation is stopped. At this time, the books that can be interviewed and corresponding to the calculated book price are the optimized and screened books to be interviewed.

[0048] In a preferred embodiment, the predetermined fee threshold is 1.0 to 1.5 times the book acquisition funds that the library intends to invest.

[0049] In a preferred embodiment, the step 7 may also include: interviewers intervene to further determine and decide on the books to be interviewed, determine the final books to be interviewed, and complete the book interview work.

[0050] Here, the interviewer can make further judgments and decisions based on the above optimized and screened books to be interviewed. This step is what is currently done by manual interviews and is well known to relevant personnel in the field. However, all the books to be interviewed by manual interviews are directly from booksellers, which have the aforementioned problems such as long time period, high labor intensity, insufficient objective basis for decision-making, strong subjectivity in decision-making, lack of comprehensiveness, and more prominent contradictions between supply and demand between collections and readers. One of the innovations of the technical solution of the present invention is to use artificial intelligence for optimized screening. After such processing, the workload and existing problems of subsequent manual interviews are greatly reduced, which not only plays the advantages of artificial intelligence optimized screening, but also plays the professional advantages of interviewers based on the individual needs of the library.

[0051] The technical solution of the present invention also provides an intelligent book acquisition system based on book classification numbers, the system comprising: a processor and a memory for storing executable instructions; wherein the processor is configured to execute the executable instructions to execute the intelligent book acquisition method based on book classification numbers as described above.

[0052] The technical solution of the present invention further provides a computer-readable storage medium, wherein a computer program is stored thereon, and when the computer program is executed by a processor, the intelligent book acquisition method based on book classification numbers as described above is implemented.

[0053] The book acquisition method provided by the present invention is described in detail below through a specific example.

[0054] Example

[0055] The data of the database established in this embodiment are derived from the data of books purchased by the Beijing University of Science and Technology Library in recent years, the data in the "Chinese Library Classification" and the book data provided by booksellers.

[0056] Establish the existing book abstract database of University of Science and Technology Beijing Library, the existing book classification number database of University of Science and Technology Beijing Library and the newly added book classification number database of University of Science and Technology Beijing Library respectively; merge the existing book classification number database of University of Science and Technology Beijing Library and the newly added book classification number database of University of Science and Technology Beijing Library to establish the total book classification number database of University of Science and Technology Beijing Library;

[0057] According to the book acquisition principles of the Beijing University of Science and Technology Library, all book classification numbers in the Beijing University of Science and Technology Library's general book classification number database are sorted by acquisition priority;

[0058] Performing text vectorization processing on the summary information in the existing book summary database and the book classification number information in the existing book classification number database, so that the summary information in the existing book summary database is vectorized summary information, and the book classification number information in the existing book classification number database is vectorized book classification number information;

[0059] Taking the summary information in the existing book summary database as input value and the book classification number information in the existing book classification number database as output value, a deep reinforcement learning model of the corresponding relationship between the two is established;

[0060] Train the deep reinforcement learning model to a precision of at least 0.9;

[0061] Establish a database of book summaries provided by booksellers and a database of book classification codes provided by booksellers respectively;

[0062] Input the summary information in the book summary database provided by the bookseller into the deep reinforcement learning model, and output the matching output book classification number information;

[0063] Determine whether the book classification number information in the book classification number database or the output book classification number information provided by the bookseller is within the range of all book classification numbers in the total book classification number database of the University of Science and Technology Beijing Library; if not, do not consider the book as one that the University of Science and Technology Beijing Library can acquire; if so, consider the book as one that the University of Science and Technology Beijing Library can acquire;

[0064] According to the requirements of the access priority level, the book classification codes of the books that can be accessed by the Beijing University of Science and Technology Library are sorted, and the book prices are calculated in order;

[0065] When the book price reaches 1.2 times the book acquisition funds that Beijing University of Science and Technology Library intends to invest, the book price calculation will stop. At this time, the books that can be acquired corresponding to the calculated book price are the optimized and selected books that Beijing University of Science and Technology Library intends to acquire.

[0066] The interview staff of the Beijing University of Science and Technology Library intervened to further determine the books to be interviewed and finally determine the books to be interviewed.

[0067] The technical solution of the present invention is used to conduct book interviews, and the book interview work can be completed efficiently in a time that is at least 3 / 4 to 1 / 2 shorter than that of current manual interviews. The accuracy of the interviewed books is increased by more than 10% compared with that of current manual interviews, and the library's collection structure can be better enriched to improve its perfection by no less than 5%. The scientific rationality of the interviewed books is at least 10% to 20% higher than that of current manual interviews.

[0068] In summary, the technical solution of the present invention first establishes a database of the library's existing book summaries, existing book classification numbers, total book classification numbers, book summaries provided by booksellers, and book classification numbers provided by booksellers; establishes and trains a machine learning model for the correspondence between existing book summaries and existing book classification numbers; determines whether the book classification numbers provided by booksellers and the output book classification numbers are within the range of all book classification numbers; sorts the classification numbers of interviewable books according to the interview priority level, and calculates the book price; further determines the books to be interviewed based on the book price, and determines the final interviewed books. The technical solution of the present invention can improve the accuracy, flexibility, generalization, efficiency and quality of book interviews, and help to complete book interviews more efficiently, accurately, completely and scientifically.

[0069] The embodiments of the present invention are described above, but the present invention is not limited to the above-mentioned specific implementation modes. The above-mentioned specific implementation modes are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which are all within the protection of the present invention.

Claims

1. An intelligent book acquisition method based on book classification number, characterized in that: The steps include: Step 1: Establish a first book abstract database and a first book classification number database of the library respectively, establish a second book classification number database according to the needs of readers, and merge the first book classification number database and the second book classification number database as a third book classification number database; Step 2: According to the book acquisition principle, all the book classification codes in the third book classification code database are sorted according to the acquisition priority level; Step 3: using the summary information in the first book summary database as input values ​​and the classification number information in the first book classification number database as output values, establishing a machine learning model of the corresponding relationship between the two, and training it to ensure that its accuracy is not lower than the target accuracy; Step 4: respectively establish a second book abstract database and a fourth book classification number database provided by the bookseller, input the abstract information in the second book abstract database into the machine learning model, and output the output book classification number information that matches each other; Step 5: Determine whether the classification number information in the fourth book classification number database or the output type book classification number information is in the third book classification number database; if not, do not treat it as an interview book; if so, treat it as an interviewable book; Step 6: sorting the book classification codes of the books that can be interviewed according to the ranking of the interview priority levels, and calculating the book prices in order and in sections; Step 7: When the book price reaches a predetermined threshold, the calculation is stopped. At this time, the books that can be interviewed corresponding to the calculated book price are the optimized and selected books to be interviewed.

2. The intelligent book acquisition method according to claim 1, characterized in that: In the step 1, the first book abstract database is an existing book abstract database, and the first book classification number database is an existing book classification number database.

3. The intelligent book acquisition method according to claim 1, characterized in that: In the step 1, the second book classification number database is a newly added book classification number database.

4. The intelligent book acquisition method according to claim 1, characterized in that: In the step 1, the third book classification number database is a general book classification number database.

5. The intelligent book acquisition method according to claim 1, characterized in that: In step 3, the target accuracy of the machine learning model is not less than 0.

9.

6. The intelligent book acquisition method according to claim 5, characterized in that: In step 3, the machine learning model is a deep reinforcement learning model.

7. The intelligent book acquisition method according to claim 1, characterized in that: In step 7, the predetermined threshold of the fee is 1.0 to 1.5 times the book acquisition funds that the library intends to invest.

8. The intelligent book acquisition method according to claim 1, characterized in that: The step 7 further includes: book interviewers intervene to further determine the books to be interviewed and finally determine the books to be interviewed.

9. An intelligent book acquisition system based on book classification numbers, the system comprising: A processor and a memory for storing executable instructions; characterized in that the processor is configured to execute the executable instructions to execute the intelligent book acquisition method based on book classification numbers according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the intelligent book acquisition method based on book classification numbers according to any one of claims 1 to 8 is implemented.

Images