Computer data management system based on artificial intelligence

Abstract

The invention relates to the technical field of machine learning, in particular to a computer data management system based on artificial intelligence, which comprises an information acquisition module used for acquiring user reading information and book information; the user analysis module is used for analyzing the reading habit of the user; the time analysis module is used for analyzing the time attenuation coefficient of book reading of each user; the correlation analysis module is used for analyzing the correlation degree between the books in the book library and the reading habits of the user; the recommendation module is used for recommending books to the user according to the association degree between the books in the book library and the reading habits of the user and the obtained book reading score of the user; the optimization module is used for optimizing the book recommendation process; and the conversion analysis module is used for analyzing the conversion abnormity of the recommended book according to the reading proportion of the user to the recommended book in the monitoring period. According to the invention, the book recommendation efficiency is improved, and the user experience is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of machine learning, and in particular to a computer data management system based on artificial intelligence. Background Art

[0002] With the popularity of digital reading, people can obtain rich book resources through various online platforms. However, faced with a huge number of book choices, users often find it difficult to find books that they are really interested in. Traditional recommendation systems mainly rely on user rating data and simple collaborative filtering algorithms. This method not only has serious data sparsity problems, but also has difficulty in capturing users' personalized needs. Therefore, developing a more intelligent and efficient book recommendation system has become an urgent problem to be solved.

[0003] Chinese patent publication number CN106202184A discloses a method for personalized book recommendation for university libraries, which solves the problems of large-scale data storage and query, scalability and poor recommendation effect in the existing book recommendation algorithms of university libraries. The basic idea is as follows: first, readers and books in the library are taken as nodes to build a graph model; second, the reader's operation log file is converted into a reader-book category preference matrix, and the similarity between readers is calculated together with the reader's personal information matrix, and these operations and mined information are used as edges to build an association graph; second, the association graph is combined with spectral clustering to propose a new personalized book recommendation model, and the cluster distribution of readers is calculated; finally, when book recommendations are needed, a recommended book list is calculated in the cluster corresponding to the reader according to the collaborative filtering algorithm; it can be seen that when recommending books, this scheme only analyzes the reader's similarity, and there are problems of low book recommendation efficiency and low user satisfaction. Summary of the invention

[0004] To this end, the present invention provides a computer data management system based on artificial intelligence to overcome the problems of low book recommendation efficiency and low user satisfaction in the prior art.

[0005] To achieve the above object, the present invention provides a computer data management system based on artificial intelligence, comprising:

[0006] Information acquisition module, used to obtain user reading information and book information;

[0007] A user analysis module is used to analyze the user's reading habits based on the average sentence length, percentage of complex words, category of books read by the user, and author of books read by the user;

[0008] A time analysis module is used to analyze the time decay coefficient of each user's book reading according to the acquired user's book reading end time;

[0009] The association analysis module is used to analyze the association between the books in the library and the user's reading habits based on the analysis results of the user's reading habits and the obtained book authors, book categories, average sentence lengths, percentages of complex words and number of book browsing times in the library;

[0010] The recommendation module is used to recommend books to users based on the correlation between the books in the library and the user's reading habits and the obtained user reading book ratings;

[0011] An optimization module is used to optimize the book recommendation process according to the number of readers in common with the books in the library and the number of readers of the books;

[0012] The conversion analysis module is used to analyze the abnormality of the conversion of recommended books according to the reading ratio of recommended books by users during the monitoring period.

[0013] Furthermore, the user analysis module is provided with a readability analysis unit and a preference analysis unit. The readability analysis unit is used to analyze the complexity index of the books read by the user based on the average sentence length and complex word percentage of the books read by the user. The readability analysis unit sets the complexity index of the i-th book read by the user as Yi, and sets: Yi=0.4×αi×(YAi+YPi). The readability analysis unit sets the complexity index of the book read by the user as Yit, and sets:

[0014] Among them, Yai is the average sentence length of the i-th book read by the user, YPi is the percentage of complex words in the i-th book read by the user, αi is the time decay coefficient of the i-th book read by the user, and I is the number of books read by the user.

[0015] Furthermore, the preference analysis unit is used to analyze the user's author preference coefficient based on the obtained author of the book the user reads, and the preference analysis unit sets the user's preference coefficient for the qth author of the book to be read as Hq, and sets Hq=mq / m;

[0016] The preference analysis unit analyzes the user's category preference coefficient according to the obtained category of the book the user reads, and the preference analysis unit sets the user's preference coefficient for the rth category of the book to be read as Rr, and sets Rr=mr / m;

[0017] Among them, mq is the number of books of the qth author among the books read by the user, m is the number of books read by the user, and mr is the number of books of the rth category among the books read by the user.

[0018] Furthermore, the time analysis module analyzes the time attenuation coefficient of the user's reading of the book based on the acquired end time of the user's reading of the book, wherein:

[0019] When Ti is shorter than 30 days, the time analysis module determines that the time interval between the i-th book read by the user is short, and sets αi to a1;

[0020] When Ti is shorter than 90 days and Ti is longer than 30 days, the time analysis module determines that the time interval between the i-th book read by the user is normal, and sets αi to a2;

[0021] When Ti is longer than 90 days, the time analysis module determines that the time interval between the i-th book read by the user is normal, and sets αi to a3;

[0022] Wherein, Ti is the time interval between the end time of the user reading the i-th book and the current time, a1 is the first preset attenuation factor, a2 is the second preset attenuation factor, and a3 is the third preset attenuation factor.

[0023] Furthermore, the association analysis module is provided with a first analysis unit, a second analysis unit and an association degree analysis unit, wherein the first analysis unit is used to analyze the complexity index of the books by using the average sentence length and the percentage of complex words of the acquired books, and the readability analysis unit sets the complexity index of the jth book in the library as Dj, and sets: Dj=0.4×(DAj+DPj);

[0024] The first analysis unit analyzes the book complexity recommendation coefficient according to the analysis result of the complexity index of the j-th book, and the first analysis unit sets the complexity recommendation coefficient of the j-th book in the library as Fj, and sets Fj=s in(|Yit-Dj| / Yit);

[0025] Among them, DAj is the average sentence length of the j-th book in the library, and DPj is the percentage of complex words in the j-th book in the library;

[0026] The second analysis unit is used to compare the acquired browsing times of books in the library with the preset browsing times c0, and analyze the popularity coefficient of the books according to the comparison result, wherein:

[0027] When cj≤c0, the second analysis unit sets the popularity recommendation coefficient of the j-th book in the library to Lj1, and sets Lj1=0;

[0028] When cj>c0, the second analysis unit sets the popularity recommendation coefficient of the jth book in the library to Lj2, setting Lj2=lg[1+(cj-c0) / cj)];

[0029] Among them, cj is the number of views of the jth book in the library.

[0030] Further, the correlation analysis unit analyzes the correlation between the book and the user's reading habits according to the book author, book category, book complexity recommendation coefficient and book popularity recommendation coefficient in the book library, and the correlation analysis unit sets the correlation coefficient between the jth book in the book library and the user's reading habits as Gj, and sets Gj=w1×Hq+w2×Rr+w3×Ljx+w4×Fj;

[0031] Among them, when the author of the j-th book in the library is not the q-th author of the book read by the user, the value of Hq is 0, when the category of the j-th book in the library is not the r-th category of the book read by the user, the value of Rr is 0, Ljx is the analysis result of the book popularity coefficient, x = 1,2; w1 is the author weight, w2 is the category weight, w3 is the complexity weight, and w4 is the popularity weight.

[0032] Furthermore, the recommendation module analyzes the recommendation coefficient of each book in the library according to the obtained book ratings and the analysis results of the association between the book and the user's reading habits, wherein:

[0033] When PFj≤PF0, the recommendation module determines that the rating of the j-th book in the library is abnormal, and sets the recommendation coefficient of the j-th book in the library to Uj1, setting Uj1=Gj×{1-β×exp[-(PF0-PFj) / PF0]};

[0034] When PFj>PF0, the recommendation module determines that the score of the j-th book in the library is normal, and sets the recommendation coefficient of the j-th book in the library to Uj2, setting Uj2=Gj;

[0035] The recommendation module sorts the recommendation coefficients of the books in the library from large to small, and selects the books with the top N recommendation coefficients as recommended books to recommend books to the user;

[0036] Among them, PF0 is the preset rating threshold, and PFj is the rating of the jth book in the library.

[0037] Furthermore, the optimization module is provided with a first optimization unit and a second optimization unit, wherein the first optimization unit is used to compare the number of common readers nj of the j-th book in the acquired library and the book read by the user with the preset number of common readers n0, and optimize the book recommendation process once according to the comparison result, wherein:

[0038] When nj≤n0, the first optimization unit determines that the number of common readers of the j-th book in the library and the book read by the user is normal, and no optimization is performed;

[0039] When nj>n0, the first optimization unit determines that the number of common readers of the jth book in the library and the book read by the user is abnormal, and optimizes the book recommendation process once, setting the optimized correlation coefficient to Gj', and setting Gj'=Gj×{1+ln[(nj-n0) / (nj+n0)+1]×cos{arctan[(nj-n0) / (nj+n0)+1]}}.

[0040] Furthermore, the second optimization unit is used to compare the number of readers sj of the jth book in the obtained library with the preset number of readers s0, and perform secondary optimization on the book recommendation process according to the comparison result, wherein:

[0041] When sj≤s0, the second optimization unit determines that the number of readers of the jth book in the library is normal and does not perform optimization;

[0042] When sj>s0, the second optimization unit determines that the number of readers of the jth book in the library is abnormal, and performs secondary optimization on the book recommendation process, and sets the optimized preset number of common readers to n0'.

[0043] Furthermore, the conversion analysis module compares the user's reading ratio of the recommended books with the preset reading ratio b0, and analyzes the abnormality of the recommended book reading ratio according to the comparison result, wherein:

[0044] When bj≤b0, the conversion analysis module determines that the reading ratio of the jth book in the library as a recommended book is abnormal;

[0045] When bj>b0, the conversion analysis module determines that the reading ratio of the jth book in the library as a recommended book is normal;

[0046] The conversion analysis module updates the recommendation process of books in the next monitoring period according to the analysis results of the abnormal reading ratio of recommended books in the monitoring period, wherein:

[0047] When Ej / ETj≤γ, the conversion analysis module determines that the conversion of the jth book in the library of the current monitoring period as a recommended book is normal;

[0048] When Ej / ETj>γ, the conversion analysis module determines that the conversion of the jth book in the library of the current monitoring period as a recommended book is abnormal, and updates the book recommendation process of the next monitoring period, and sets the updated preset adjustment coefficient to β', setting β'=β×cos{sin[(Ej / ETj-γ)×(π / 2) / (Ej / ETj+γ)]};

[0049] Among them, Ej is the number of abnormal reading ratios when the j-th book in the library during the monitoring period is used as a recommended book, ETj is the number of times the j-th book in the library during the monitoring period is used as a recommended book, and γ is the preset abnormal ratio.

[0050] Compared with the prior art, the beneficial effects of the present invention are that the user analysis module quantifies the user's preference for text complexity by calculating the complexity index of the user's reading books, thereby improving the accuracy of the user's reading habit analysis, and improves the accuracy of the analysis of the association between the books in the library and the user's reading habits by calculating the author preference coefficient and the category preference coefficient. The time analysis module reduces the influence of the time factor on the analysis of the user's reading habits by setting the first preset attenuation factor, the second preset attenuation factor and the third preset attenuation factor, thereby improving the accuracy of the analysis of the user's reading habits. The association analysis module calculates the complexity recommendation coefficient of the books in the library to improve the accuracy of the analysis of the association between the books in the library and the user's reading habits, and sets a preset number of browsing times to improve the analysis of the popularity coefficient of the books. The accuracy of the analysis of the correlation between the books in the library and the user's reading habits is improved. The accuracy of the analysis of the correlation between the books in the library and the user's reading habits is improved by setting various weights. The recommendation module improves the accuracy of the analysis of the book recommendation coefficient by setting a preset scoring threshold and a preset adjustment coefficient. The optimization module adds the preferences of readers of the same type to analyze the book recommendation coefficient by setting a preset number of common readers to improve the accuracy of the analysis of the book recommendation coefficient. The accuracy of the book recommendation analysis is improved by setting a preset number of readers, thereby improving the accuracy of the book recommendation analysis to users. The conversion analysis module improves the accuracy of the abnormality analysis of the conversion of recommended books by setting a preset abnormality ratio and a preset reading ratio, thereby improving the accuracy of the book recommendation analysis, and ultimately improves the efficiency of book recommendations, while also improving user satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 Schematic diagram of the structure of the computer data management system based on artificial intelligence in this embodiment.

[0052] Figure 2 Schematic diagram of the structure of the user analysis module of this embodiment.

[0053] Figure 3 Schematic diagram of the structure of the association analysis module in this embodiment.

[0054] Figure 4 This is a schematic diagram of the structure of the optimization module in this embodiment. DETAILED DESCRIPTION

[0055] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0056] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0057] It should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0058] See also Figure 1 As shown, it is a schematic diagram of the structure of a computer data management system based on artificial intelligence in this embodiment, and the system includes:

[0059] The information acquisition module is used to obtain user reading information and book information. The user reading information includes the category of books read by the user, the author of the book read by the user, the number of books read by the user, the average sentence length of the books read by the user, the percentage of complex words and the end time of the book reading by the user. The book information includes the category of books in the library, the author of the book, the book rating, the average sentence length of the book, the percentage of complex words in the book, the number of readers of the book and the number of book browsing times; the books read by the user are the books that the user has finished reading. The categories of books read by the user include literary novels, science fiction novels, suspense and mystery, history, biography, popular science, philosophy, psychology, economic management, etc. This embodiment does not specifically limit the setting of book categories. Those skilled in the art can set them freely as long as the setting requirements of the book categories are met; the average sentence length of the books read by the user is the average length of the sentences in a single sentence. The average number of words in the sentence is calculated by dividing the total number of words in the book read by the user by the number of sentences in the book read by the user. The percentage of complex words is the percentage of words with three or more characters, which is calculated by the ratio of the number of words with three or more characters in the book read by the user to the total number of words. The calculation method of the average sentence length and the percentage of complex words in the book are the same as the above method, which will not be repeated in this embodiment. The end time of the user reading the book is the time when the user finishes reading the book, and the book rating is based on a ten-point system. The book library in this embodiment is the books that the user has not read on the book reading platform. In this embodiment, the method for obtaining the user reading information and the book information is not specifically limited, and the technicians in this field can freely set it as long as the requirements for obtaining the user reading information and the book information are met. Among them, the average sentence length and the percentage of complex words in the book can be obtained through the jieba library of the Python language. The category of the book read by the user, the author of the book read by the user, the number of books read by the user, the rating of the book read by the user, the end time of the book read by the user and the book information can be obtained by importing the background information of the book reading platform;

[0060] A user analysis module, which obtains the average sentence length, percentage of complex words, category and author of books read by users and analyzes the reading habits of users. The user analysis module is connected to the information acquisition module;

[0061] A time analysis module, used to analyze the time attenuation coefficient of each user's reading of the book according to the acquired end time of the user's reading of the book, and the time analysis module is connected to the user analysis module;

[0062] A correlation analysis module, used to analyze the correlation between the books in the library and the user's reading habits based on the analysis results of the user's reading habits and the obtained book authors, book categories, average sentence lengths, percentages of complex words and number of book browsing in the library, and the correlation analysis module is connected to the time analysis module;

[0063] A recommendation module, used to recommend books to users based on the correlation between books in the library and the user's reading habits and the obtained user reading book ratings, and the recommendation module is connected to the correlation analysis module;

[0064] The optimization module is used to optimize the book recommendation process according to the obtained number of common readers of the books in the library and the books read by the user and the number of readers of the books, and the optimization module is connected to the recommendation module; the common readers in this embodiment are readers who have read two or more of the books in the library and the books read by the user at the same time; in this embodiment, no specific limitation is made on the method for obtaining the number of common readers of the books in the library and the books read by the user, and those skilled in the art can freely set it, and it only needs to meet the requirements for obtaining the number of common readers of the books in the library and the books read by the user, wherein the number of common readers of the books in the library and the books read by the user can be obtained by importing the background data of the book reading platform;

[0065] The conversion analysis module is used to analyze the abnormality of the conversion of recommended books according to the reading ratio of recommended books by users within the monitoring period. The conversion analysis module is connected to the optimization module. In this embodiment, the reading ratio of recommended books is the ratio of the number of words in the recommended books that have been read by users to the number of words in the recommended books. In this embodiment, the method for obtaining the reading ratio of recommended books is not specifically limited. Those skilled in the art can set it freely, and it is only necessary to meet the requirements for obtaining the reading ratio of recommended books. Among them, the reading ratio of recommended books can be obtained by importing the background data of the book reading platform. In this embodiment, there is no specific limitation on the setting of the monitoring period. Those skilled in the art can set it freely, and it is only necessary to meet the setting requirements of the monitoring period. Among them, the monitoring period can be set to 30 days, 50 days, etc.

[0066] See also Figure 2 As shown, it is a schematic diagram of the structure of the user analysis module of this embodiment, including:

[0067] A readability analysis unit, used to analyze the complexity index of the books read by the user based on the average sentence length and percentage of complex words obtained from the books read by the user;

[0068] The preference analysis unit is used to analyze the author preference coefficient of the user according to the obtained authors of the books read by the user, and is also used to analyze the category preference coefficient of the user according to the obtained categories of the books read by the user. The preference analysis unit is connected to the readability analysis unit.

[0069] See also Figure 3 As shown, it is a schematic diagram of the structure of the association analysis module of this embodiment, including:

[0070] The first analysis unit is used to analyze the book complexity recommendation coefficient according to the average sentence length and complex word percentage of the acquired books;

[0071] A second analysis unit, used for analyzing the popularity coefficient of the books according to the acquired browsing times of the books in the library, the second analysis unit is connected to the first analysis unit;

[0072] A correlation analysis unit is used to analyze the correlation between the books in the library and the user's reading habits according to the book authors, book categories, book complexity recommendation coefficients and book popularity recommendation coefficients in the library. The correlation analysis unit is connected to the first analysis unit and the second analysis unit.

[0073] See also Figure 4 As shown, it is a schematic diagram of the structure of the optimization module of this embodiment, including:

[0074] The first optimization unit is used to optimize the book recommendation process according to the number of common readers of the books in the library and the books read by the user;

[0075] The second optimization unit is used to perform secondary optimization on the book recommendation process according to the number of readers of the books in the library, and the second optimization unit is connected to the first optimization unit.

[0076] Specifically, the artificial intelligence-based computer data management system described in this embodiment is applied to the processing of book reading platform data. It analyzes user preferences and adjusts the analysis process in combination with the time attenuation coefficient. It calculates the correlation between books in the library and books read by users in combination with user preferences, and recommends books to users, thereby improving the efficiency of book recommendations and user satisfaction.

[0077] Specifically, the readability analysis unit analyzes the complexity index of the books read by the user based on the average sentence length and percentage of complex words obtained from the books read by the user. The readability analysis unit sets the complexity index of the i-th book read by the user to Yi, and sets: Yi=0.4×αi×(YAi+YPi). The readability analysis unit sets the complexity index of the book read by the user to Yit, and sets:

[0078] Among them, Yai is the average sentence length of the i-th book read by the user, YPi is the percentage of complex words in the i-th book read by the user, αi is the time decay coefficient of the i-th book read by the user, and I is the number of books read by the user.

[0079] Specifically, the readability analysis unit quantifies the user's preference for text complexity by calculating the complexity index of the user's book reading, thereby improving the accuracy of the user's reading habit analysis, thereby improving the accuracy of the correlation analysis between the books in the library and the user's reading habits, and ultimately improving the efficiency of book recommendations while improving user satisfaction.

[0080] Specifically, the preference analysis unit analyzes the user's author preference coefficient based on the obtained author of the book the user reads, and the preference analysis unit sets the user's preference coefficient for the qth author of the book to be read as Hq, and sets Hq=mq / m;

[0081] The preference analysis unit analyzes the user's category preference coefficient according to the obtained category of the book the user reads, and the preference analysis unit sets the user's preference coefficient for the rth category of the book to be read as Rr, and sets Rr=mr / m;

[0082] Among them, mq is the number of books of the qth author among the books read by the user, m is the number of books read by the user, and mr is the number of books of the rth category among the books read by the user.

[0083] Specifically, the preference analysis unit improves the accuracy of the user reading habit analysis by calculating the author preference coefficient and the category preference coefficient, thereby improving the accuracy of the correlation analysis between the books in the library and the user's reading habits, ultimately improving the efficiency of book recommendations and user satisfaction; the number of books by the qth author and the number of books in the rth category among the books read by the user in this embodiment can be obtained by importing the background data of the book reading platform.

[0084] Specifically, the time analysis module analyzes the time attenuation coefficient of the user's reading of the book based on the acquired end time of the user's reading of the book, wherein:

[0085] When Ti is shorter than 30 days, the time analysis module determines that the time interval between the i-th book read by the user is short, and sets αi to a1;

[0086] When Ti is shorter than 90 days and Ti is longer than 30 days, the time analysis module determines that the time interval between the i-th book read by the user is normal, and sets αi to a2;

[0087] When Ti is longer than 90 days, the time analysis module determines that the time interval between the i-th book read by the user is normal, and sets αi to a3;

[0088] Wherein, Ti is the time interval between the end time of the user reading the i-th book and the current time, a1 is the first preset attenuation factor, a2 is the second preset attenuation factor, a3 is the third preset attenuation factor, and 1>a1>a2>a3>0.6.

[0089] Specifically, Ti described in this embodiment can be calculated and obtained through the datetime module of Python. In this embodiment, the method for obtaining Ti is not specifically limited, and those skilled in the art can set it freely, and it is only necessary to meet the requirements for obtaining Ti; the time analysis module reduces the influence of time factors on the analysis of user reading habits by setting the first preset attenuation factor, the second preset attenuation factor and the third preset attenuation factor, thereby improving the accuracy of the analysis of user reading habits, thereby improving the accuracy of the analysis of the correlation between books in the library and user reading habits, and ultimately improving the efficiency of book recommendations, while also improving user satisfaction; in this embodiment, the setting of the first preset attenuation factor, the second preset attenuation factor and the third preset attenuation factor is not specifically limited, and those skilled in the art can set it freely, and it is only necessary to meet the setting requirements of the first preset attenuation factor, the second preset attenuation factor and the third preset attenuation factor, wherein the optimal value of a1 is 1, the optimal value of a2 is 0.9, and the optimal value of a3 is 0.7.

[0090] Specifically, the first analysis unit analyzes the complexity index of the books by using the average sentence length and complex word percentage of the acquired books, and the readability analysis unit sets the complexity index of the jth book in the library as Dj, and sets: Dj=0.4×(DAj+DPj);

[0091] The first analysis unit analyzes the book complexity recommendation coefficient according to the analysis result of the complexity index of the j-th book, and the first analysis unit sets the complexity recommendation coefficient of the j-th book in the library as Fj, and sets Fj=s in(|Yit-Dj| / Yit);

[0092] Among them, DAj is the average sentence length of the j-th book in the library, and DPj is the percentage of complex words in the j-th book in the library.

[0093] Specifically, the first analysis unit calculates the complexity recommendation coefficient of the books in the library to improve the accuracy of the correlation analysis between the books in the library and the user's reading habits, thereby improving the accuracy of the book recommendation analysis to the user, ultimately improving the efficiency of book recommendations and improving user satisfaction.

[0094] Specifically, the second analysis unit compares the acquired browsing times of books in the library with the preset browsing times c0, and analyzes the popularity coefficient of the books according to the comparison result, wherein:

[0095] When cj≤c0, the second analysis unit sets the popularity recommendation coefficient of the j-th book in the library to Lj1, and sets Lj1=0;

[0096] When cj>c0, the second analysis unit sets the popularity recommendation coefficient of the jth book in the library to Lj2, setting Lj2=lg[1+(cj-c0) / cj)];

[0097] Among them, cj is the number of views of the jth book in the library.

[0098] Specifically, the second analysis unit improves the accuracy of the book popularity coefficient analysis by setting a preset number of browsing times, thereby improving the accuracy of the correlation analysis between the books in the library and the user's reading habits, thereby improving the accuracy of the book recommendation analysis to the user, and ultimately improving the efficiency of book recommendations while improving user satisfaction. In this embodiment, there is no specific limitation on the value of the preset number of browsing times, and technical personnel in this field can set it freely, as long as the setting requirements of the preset number of browsing times are met. Among them, the best value of c0 is 5000.

[0099] Specifically, the correlation analysis unit analyzes the correlation between books and user reading habits according to the book authors, book categories, book complexity recommendation coefficients and book popularity recommendation coefficients in the library. The correlation analysis unit sets the correlation coefficient between the jth book in the library and the user reading habits as Gj, and sets Gj=w1×Hq+w2×Rr+w3×Ljx+w4×Fj;

[0100] Among them, when the author of the j-th book in the library is not the q-th author of the book read by the user, the value of Hq is 0, when the category of the j-th book in the library is not the r-th category of the book read by the user, the value of Rr is 0, Ljx is the analysis result of the book popularity coefficient, x=1,2; w1 is the author weight, w2 is the category weight, w3 is the complexity weight, w4 is the popularity weight, w1+w2+w3+w4=1, w1>w2>w4>w3.

[0101] Specifically, the association analysis unit sets various weights to improve the accuracy of the analysis of the association between books in the library and user reading habits, thereby improving the accuracy of book recommendation analysis to users, ultimately improving the efficiency of book recommendations and user satisfaction. In this embodiment, there is no specific limitation on the setting of each weight, and technical personnel in this field can set it freely as long as the setting requirements of each weight are met. Among them, the best value of w1 is 0.38, the best value of w2 is 0.3, the best value of w3 is 0.15, and the best value of w4 is 0.17.

[0102] Specifically, the recommendation module analyzes the recommendation coefficient of each book in the library according to the obtained book ratings and the analysis results of the correlation between the book and the user's reading habits, where:

[0103] When PFj≤PF0, the recommendation module determines that the rating of the j-th book in the library is abnormal, and sets the recommendation coefficient of the j-th book in the library to Uj1, setting Uj1=Gj×{1-β×exp[-(PF0-PFj) / PF0]};

[0104] When PFj>PF0, the recommendation module determines that the score of the j-th book in the library is normal, and sets the recommendation coefficient of the j-th book in the library to Uj2, setting Uj2=Gj;

[0105] The recommendation module sorts the recommendation coefficients of the books in the library from large to small, and selects the books with the top N recommendation coefficients as recommended books to recommend books to the user;

[0106] Among them, PF0 is the preset rating threshold, PFj is the rating of the jth book in the library, β is the preset adjustment coefficient, 0.2<β<0.36, and N is the preset number of recommendations.

[0107] Specifically, the recommendation module improves the accuracy of book recommendation coefficient analysis by setting a preset scoring threshold and a preset adjustment coefficient, thereby improving the accuracy of book recommendation analysis to users, ultimately improving the efficiency of book recommendation, and at the same time improving user satisfaction; in this embodiment, no specific limitation is made to the setting of the preset scoring threshold and the preset adjustment coefficient, and those skilled in the art can set them freely, as long as the setting requirements of the preset scoring threshold and the preset adjustment coefficient are met, among which the optimal value of PF0 is 6 points, the optimal value of β is 0.23, and the optimal value of N is 3; in this embodiment, no specific limitation is made to the sorting method, and those skilled in the art can set it freely, such as sorting by bubble sorting.

[0108] Specifically, the first optimization unit compares the number of common readers nj of the j-th book in the obtained library and the book read by the user with the preset number of common readers n0, and optimizes the book recommendation process according to the comparison result, wherein:

[0109] When nj≤n0, the first optimization unit determines that the number of common readers of the j-th book in the library and the book read by the user is normal, and no optimization is performed;

[0110] When nj>n0, the first optimization unit determines that the number of common readers of the jth book in the library and the book read by the user is abnormal, and optimizes the book recommendation process once, setting the optimized correlation coefficient to Gj', and setting Gj'=Gj×{1+ln[(nj-n0) / (nj+n0)+1]×cos{arctan[(nj-n0) / (nj+n0)+1]}}.

[0111] Specifically, the first optimization unit sets a preset number of common readers to add preferences of readers of the same type to jointly analyze the book recommendation coefficient, so as to improve the accuracy of the book recommendation coefficient analysis, thereby improving the accuracy of the book recommendation analysis to users, and ultimately improving the efficiency of book recommendations while improving user satisfaction. In this embodiment, the setting of the preset number of common readers is not specifically limited, and those skilled in the art can set it freely, as long as the setting requirements of the preset number of common readers are met, where the optimal value of n0 is 100.

[0112] Specifically, the second optimization unit compares the number of readers sj of the jth book in the library with the preset number of readers s0, and performs secondary optimization on the book recommendation process according to the comparison result, wherein:

[0113] When sj≤s0, the second optimization unit determines that the number of readers of the jth book in the library is normal and does not perform optimization;

[0114] When sj>s0, the second optimization unit determines that the number of readers of the jth book in the library is abnormal, and performs secondary optimization on the book recommendation process, setting the optimized preset number of common readers to n0', and setting n0'=n0×{1+exp{arctan[(sj-s0)×(π / 4) / (sj+s0)]}}.

[0115] Specifically, the second optimization unit improves the accuracy of book recommendation analysis by setting a preset number of readers, thereby improving the accuracy of book recommendation analysis to users, ultimately improving the efficiency of book recommendations, and at the same time improving user satisfaction; in this embodiment, there is no specific limitation on the setting of the preset number of readers, and technical personnel in this field can set it freely, as long as the setting requirements of the preset number of readers are met, where the optimal value of s0 is 2000.

[0116] Specifically, the conversion analysis module compares the user's reading ratio of the recommended books with the preset reading ratio b0, and analyzes the abnormality of the recommended book reading ratio according to the comparison result, wherein:

[0117] When bj≤b0, the conversion analysis module determines that the reading ratio of the j-th book in the library as a recommended book is abnormal, and bj is the reading ratio of the j-th book in the library as a recommended book;

[0118] When bj>b0, the conversion analysis module determines that the reading ratio of the jth book in the library as a recommended book is normal;

[0119] The conversion analysis module updates the recommendation process of books in the next monitoring period according to the analysis results of the abnormal reading ratio of recommended books in the monitoring period, wherein:

[0120] When Ej / ETj≤γ, the conversion analysis module determines that the conversion of the jth book in the library of the current monitoring period as a recommended book is normal;

[0121] When Ej / ETj>γ, the conversion analysis module determines that the conversion of the jth book in the library of the current monitoring period as a recommended book is abnormal, and updates the book recommendation process of the next monitoring period, and sets the updated preset adjustment coefficient to β', setting β'=β×cos{sin[(Ej / ETj-γ)×(π / 2) / (Ej / ETj+γ)]};

[0122] Among them, Ej is the number of abnormal reading ratios when the j-th book in the library during the monitoring period is used as a recommended book, ETj is the number of times the j-th book in the library during the monitoring period is used as a recommended book, and γ is the preset abnormal ratio.

[0123] Specifically, the conversion analysis module improves the accuracy of the abnormality analysis of the conversion of recommended books by setting a preset abnormality ratio and a preset reading ratio, thereby improving the accuracy of the book recommendation analysis, and ultimately improving the efficiency of book recommendations while improving user satisfaction. In this embodiment, there is no specific limitation on the setting of the preset abnormality ratio and the preset reading ratio, and technical personnel in this field can set them freely as long as the setting requirements of the preset abnormality ratio and the preset reading ratio are met. Among them, the optimal value of b0 is 0.45, and the optimal value of γ is 0.3.

[0124] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A computer data management system based on artificial intelligence, It is characterized in that include, Information acquisition module, used to obtain user reading information and book information; A user analysis module is used to analyze the user's reading habits based on the average sentence length, percentage of complex words, category of books read by the user, and author of books read by the user; A time analysis module is used to analyze the time decay coefficient of each user's book reading according to the acquired user's book reading end time; The association analysis module is used to analyze the association between the books in the library and the user's reading habits based on the analysis results of the user's reading habits and the obtained book authors, book categories, average sentence lengths, percentages of complex words and number of book browsing in the library; The recommendation module is used to recommend books to users based on the correlation between the books in the library and the user's reading habits and the obtained user reading book ratings; An optimization module is used to optimize the book recommendation process according to the number of readers in common with the books in the library and the number of readers of the books; The conversion analysis module is used to analyze the abnormality of the conversion of recommended books according to the reading ratio of recommended books by users during the monitoring period.

2. The computer data management system based on artificial intelligence according to claim 1, It is characterized in that The user analysis module is provided with a readability analysis unit and a preference analysis unit. The readability analysis unit is used to analyze the complexity index of the books read by the user based on the average sentence length and complex word percentage of the books read by the user. The readability analysis unit sets the complexity index of the i-th book read by the user as Yi, and sets: Yi=0.4×αi×(YAi+YPi). The readability analysis unit sets the complexity index of the book read by the user as Yit, and sets: Among them, Yai is the average sentence length of the i-th book read by the user, YPi is the percentage of complex words in the i-th book read by the user, αi is the time decay coefficient of the i-th book read by the user, and I is the number of books read by the user.

3. The computer data management system based on artificial intelligence according to claim 2, It is characterized in that The preference analysis unit is used to analyze the author preference coefficient of the user according to the obtained author of the book read by the user, and the preference analysis unit sets the user's preference coefficient for the qth author of the book read as Hq, and sets Hq=mq / m; The preference analysis unit analyzes the user's category preference coefficient according to the obtained category of the book the user reads, and the preference analysis unit sets the user's preference coefficient for the rth category of the book to be read as Rr, and sets Rr=mr / m; Among them, mq is the number of books of the qth author among the books read by the user, m is the number of books read by the user, and mr is the number of books of the rth category among the books read by the user.

4. The computer data management system based on artificial intelligence according to claim 3, It is characterized in that The time analysis module analyzes the time attenuation coefficient of the user's reading of the book based on the acquired end time of the user's reading of the book, wherein: When Ti is shorter than 30 days, the time analysis module determines that the time interval between the i-th book read by the user is short, and sets αi to a1; When Ti is shorter than 90 days and Ti is longer than 30 days, the time analysis module determines that the time interval between the i-th book read by the user is normal, and sets αi to a2; When Ti is longer than 90 days, the time analysis module determines that the time interval between the i-th book read by the user is normal, and sets αi to a3; Wherein, Ti is the time interval between the end time of the user reading the i-th book and the current time, a1 is the first preset attenuation factor, a2 is the second preset attenuation factor, and a3 is the third preset attenuation factor.

5. The computer data management system based on artificial intelligence according to claim 4, It is characterized in that The association analysis module includes a first analysis unit, a second analysis unit and a correlation analysis unit. The first analysis unit is used to analyze the complexity index of the books by using the average sentence length and the percentage of complex words of the acquired books. The readability analysis unit sets the complexity index of the jth book in the library as Dj, and sets: Dj=0.4×(DAj+DPj); The first analysis unit analyzes the book complexity recommendation coefficient according to the analysis result of the complexity index of the j-th book, and the first analysis unit sets the complexity recommendation coefficient of the j-th book in the library as Fj, and sets Fj=sin(|Yit-Dj| / Yit); Among them, DAj is the average sentence length of the j-th book in the library, and DPj is the percentage of complex words in the j-th book in the library; The second analysis unit is used to compare the acquired browsing times of books in the library with the preset browsing times c0, and analyze the popularity coefficient of the books according to the comparison result, wherein: When cj≤c0, the second analysis unit sets the popularity recommendation coefficient of the j-th book in the library to Lj1, and sets Lj1=0; When cj>c0, the second analysis unit sets the popularity recommendation coefficient of the jth book in the library to Lj2, setting Lj2=lg[1+(cj-c0) / cj)]; Among them, cj is the number of views of the jth book in the library.

6. The computer data management system based on artificial intelligence according to claim 5, It is characterized in that The correlation analysis unit analyzes the correlation between the book and the user's reading habits according to the book author, book category, book complexity recommendation coefficient and book popularity recommendation coefficient in the book library. The correlation analysis unit sets the correlation coefficient between the jth book in the book library and the user's reading habits as Gj, and sets Gj=w1×Hq+w2×Rr+w3×Ljx+w4×Fj; Among them, when the author of the j-th book in the library is not the q-th author of the book read by the user, the value of Hq is 0, when the category of the j-th book in the library is not the r-th category of the book read by the user, the value of Rr is 0, Ljx is the analysis result of the book popularity coefficient, x = 1,2; w1 is the author weight, w2 is the category weight, w3 is the complexity weight, and w4 is the popularity weight.

7. The computer data management system based on artificial intelligence according to claim 6, It is characterized in that The recommendation module analyzes the recommendation coefficient of each book in the library according to the obtained book rating and the analysis result of the correlation between the book and the user's reading habits, where: When PFj≤PF0, the recommendation module determines that the rating of the j-th book in the library is abnormal, and sets the recommendation coefficient of the j-th book in the library to Uj1, setting Uj1=Gj×{1-β×exp[-(PF0-PFj) / PF0]}; When PFj>PF0, the recommendation module determines that the score of the j-th book in the library is normal, and sets the recommendation coefficient of the j-th book in the library to Uj2, setting Uj2=Gj; The recommendation module sorts the recommendation coefficients of the books in the library from large to small, and selects the books with the top N recommendation coefficients as recommended books to recommend books to the user; Among them, PF0 is the preset rating threshold, and PFj is the rating of the jth book in the library.

8. The computer data management system based on artificial intelligence according to claim 7, It is characterized in that The optimization module is provided with a first optimization unit and a second optimization unit. The first optimization unit is used to compare the number of common readers nj of the j-th book in the acquired library and the book read by the user with the preset number of common readers n0, and optimize the book recommendation process once according to the comparison result, wherein: When nj≤n0, the first optimization unit determines that the number of common readers of the j-th book in the library and the book read by the user is normal, and no optimization is performed; When nj>n0, the first optimization unit determines that the number of common readers of the jth book in the library and the book read by the user is abnormal, and optimizes the book recommendation process once, setting the optimized correlation coefficient to Gj', and setting Gj'=Gj×{1+ln[(nj-n0) / (nj+n0)+1]×cos{arctan[(nj-n0) / (nj+n0)+1]}}.

9. The computer data management system based on artificial intelligence according to claim 8, It is characterized in that The second optimization unit is used to compare the number of readers sj of the jth book in the obtained library with the preset number of readers s0, and perform secondary optimization on the book recommendation process according to the comparison result, wherein: When sj≤s0, the second optimization unit determines that the number of readers of the jth book in the library is normal and does not perform optimization; When sj>s0, the second optimization unit determines that the number of readers of the jth book in the library is abnormal, and performs secondary optimization on the book recommendation process, and sets the optimized preset number of common readers to n0'.

10. The computer data management system based on artificial intelligence according to claim 7, It is characterized in that The conversion analysis module compares the user's reading ratio of the recommended books with the preset reading ratio b0, and analyzes the abnormality of the recommended book reading ratio according to the comparison result, wherein: When bj≤b0, the conversion analysis module determines that the reading ratio of the jth book in the library as a recommended book is abnormal; When bj>b0, the conversion analysis module determines that the reading ratio of the jth book in the library as a recommended book is normal; The conversion analysis module updates the recommendation process of books in the next monitoring period according to the analysis results of the abnormality of the recommended book reading ratio in the monitoring period, wherein: When Ej / ETj≤γ, the conversion analysis module determines that the conversion of the jth book in the library of the current monitoring period as a recommended book is normal; When Ej / ETj>γ, the conversion analysis module determines that the conversion of the jth book in the library of the current monitoring period as a recommended book is abnormal, and updates the book recommendation process of the next monitoring period, and sets the updated preset adjustment coefficient to β', setting β'=β×cos{sin[(Ej / ETj-γ)×(π / 2) / (Ej / ETj+γ)]}; Among them, Ej is the number of abnormal reading ratios when the j-th book in the library during the monitoring period is used as a recommended book, ETj is the number of times the j-th book in the library during the monitoring period is used as a recommended book, and γ is the preset abnormal ratio.

Images