CN118097078A - Application scene analysis system of virtual library - Google Patents
Application scene analysis system of virtual library Download PDFInfo
- Publication number
- CN118097078A CN118097078A CN202410288311.6A CN202410288311A CN118097078A CN 118097078 A CN118097078 A CN 118097078A CN 202410288311 A CN202410288311 A CN 202410288311A CN 118097078 A CN118097078 A CN 118097078A
- Authority
- CN
- China
- Prior art keywords
- browsing
- resource data
- user
- intersection
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 31
- 230000006399 behavior Effects 0.000 claims abstract description 48
- 238000012163 sequencing technique Methods 0.000 claims abstract description 15
- 238000005457 optimization Methods 0.000 claims abstract description 8
- 238000003062 neural network model Methods 0.000 claims abstract description 7
- 238000012216 screening Methods 0.000 claims abstract description 6
- 238000000034 method Methods 0.000 claims description 29
- 230000000694 effects Effects 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 4
- 230000003247 decreasing effect Effects 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims 3
- 238000005065 mining Methods 0.000 abstract description 3
- 230000003190 augmentative effect Effects 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to the technical field of virtual augmented reality, in particular to an application scene analysis system of a virtual library; comprising the following steps: the system comprises a scene analysis module, a database and a scene optimization module; analyzing the browsing behaviors of the user to judge the interest recommendation index of the user and deeply mining the overlapping degree between the resource data to judge the association relation between the resource data in each relevant field of the user so as to obtain preference parameters; the requirements and interests of the user can be more accurately understood, and support is provided for personalized and accurate resource recommendation; screening out resource data related to the search word in an autonomous search mode through a neural network model and a TF-I DF algorithm, and calculating a recommended value, so that the resource data with higher relativity can be rapidly presented when a search result is displayed, and the search time of a user is saved; through overlapping degree sequencing and long-chain recommendation, the sequencing and recommendation sequence of the resource data is optimized, so that a user can find interested resources more easily, and the efficiency of information acquisition is improved.
Description
Technical Field
The invention relates to the technical field of virtual augmented reality, in particular to an application scene analysis system of a virtual library.
Background
The metauniverse is a virtual space generated by a computer, integrates a plurality of information technologies such as digital image processing, computer graphics, artificial intelligence, multimedia technology and the like, and is becoming a novel internet form along with the rapid development of virtual reality technology and the internet; wherein the virtual (metauniverse) library is an important direction of metauniverse application, is a digitized library, is presented in a virtual reality form, and can experience various information services provided by the reality library;
the number of resources in the virtual library is large, the content is wide in relation range, and a user needs to carry out a large amount of screening and filtering to find out the required resources; therefore, the current virtual library has the problem that a user cannot quickly find required resource data in the virtual application scene, so that bad experience is brought to the user.
Disclosure of Invention
The invention aims to provide an application scene analysis system of a virtual library, which aims to solve the problems in the background technology.
The aim of the invention can be achieved by the following technical scheme: an application scenario analysis system of a virtual library, comprising: the system comprises a database, a scene analysis module and a scene optimization module;
the database is used for storing resource data of the virtual library and browsing behavior information of the user;
The scene analysis module analyzes the learning state of the user through the browsing behavior of the user and judges whether to generate an interest recommendation strategy; meanwhile, carrying out overlapping analysis according to the browsing behaviors of the user so as to refine preference parameters of the user;
the scene optimization module optimizes the searching mode according to the preference parameters, and specifically comprises the following steps:
Setting two search modes in the virtual library, and screening out resource data related to the search word by using a neural network model to be recorded as related resource data if the user selects an autonomous search mode; the method comprises the steps of calling keywords of related resource data, comparing the keywords with preference parameters to obtain the number of the related keywords in the related resource data, marking the number as U1, calculating weights of the related keywords in the related resource data by using a TF-IDF algorithm, and summing the weights to obtain a total weight value which is marked as U2; substituting the number U1 of the related keywords and the total weight U2 into a formula Ua=a1×U1+a2×U2 to calculate a recommended value Ua, wherein a1 and a2 are respectively set proportionality coefficients, and sequencing related resource data according to the corresponding recommended values from large to small and displaying the related resource data as a search result;
if the user selects the recommended auxiliary mode:
K1: the resource data of the last browsing action of the user is called as initial resources;
k2: the method comprises the steps of calling resource data in the same field as an initial resource and overlapping degree between each resource data and the initial resource, and sequencing the resource data according to the sequence from high overlapping degree to low overlapping degree to be used as recommendation sequencing;
K3: repeating steps K1-K2 until the user turns off the recommended auxiliary mode.
Preferably, the specific process of analyzing the learning state of the user is:
Retrieving user browsing behavior information, wherein the browsing behavior information comprises fields and keywords corresponding to resource data of each browsing behavior, and browsing starting time and browsing ending time; a time cross axis is established, and each resource data is marked on the time cross axis according to the corresponding browsing start time and browsing end time to obtain a browsing behavior track graph of the user;
Calculating the interval duration of two adjacent browsing behaviors; comparing and analyzing the interval duration with a set interval so as to divide the interval duration into a high interval, a medium interval and a low interval; counting the number of high intervals, medium intervals and low intervals respectively, and marking the numbers as A1, A2 and A3 respectively; when A1 is more than or equal to A2+A3, generating a low-frequency stable learning state; when A2 is more than or equal to A1+A3, generating an intermediate frequency stable learning state; when A3 is more than or equal to A1+A2, generating a high-frequency stable learning state; otherwise, a general learning state is generated.
Preferably, the specific process of judging whether to generate the interest recommendation strategy is as follows:
Tracking the browsing behavior and the corresponding browsing time length to establish a graph of the browsing behavior changing along with time, and carrying out graph analysis on the graph to obtain a browsing activity value which is recorded as Hb;
Setting a learning frequency value corresponding to each learning state as Bo, substituting the learning frequency value and a browsing activity value Hb into a set formula BH=ln (b3×e Bo+b4×eHb +1) to calculate to obtain a user interest recommendation index BH, wherein b3 and b4 are set proportionality coefficients respectively, and e is a natural constant; and comparing and analyzing the interest recommendation index with a set recommendation threshold, and generating an interest recommendation strategy when the interest recommendation index of the user is larger than the set recommendation threshold.
Preferably, the specific process of parsing the graph is as follows:
Numbering the browsing behaviors according to the sequence of the corresponding browsing time to obtain the corresponding browsing times of each browsing behavior, and calculating the time difference between the browsing start time and the browsing end time corresponding to the browsing behaviors to obtain the browsing duration, so that the browsing duration corresponding to each browsing time can be obtained;
Constructing a two-dimensional rectangular coordinate system by taking the marks as abscissa and the browsing time length as ordinate, inputting the browsing time length into the coordinate system according to the corresponding marks, and marking the position longer than the position in the coordinate system during browsing as a browsing point; sequentially connecting browsing points by adopting a smooth curve to obtain a line graph of the browsing duration changing along with time; a curve tangent is made at the browsing points, and the slope of the tangent is calculated to obtain the slope of each browsing point to be marked as Xj; wherein j=1, 2,3 … … J, the value of J is a positive integer, J represents the total number of browsing actions, and J represents the label of any one of the browsing actions; summing the slopes smaller than zero, taking the absolute value, marking the browsing monotonic decreasing degree as H1, and summing the slopes larger than zero to obtain the browsing monotonic increasing degree as H2;
substituting the slope Xj, the browsing monotonic decrease degree H1 and the browsing monotonic increase degree H2 into a set formula And calculating to obtain a browsing activity value Hb, wherein b1 and b2 are set proportionality coefficients respectively, and e is a natural constant.
Preferably, the specific process of performing overlap analysis to refine the preference parameters of the user according to the browsing behavior of the user is as follows:
Extracting keywords of each resource data in each related field, counting the number of intersection keywords and the number of union keywords of any two resource data, and marking the intersection keywords and the number of union keywords as T1 and T2 respectively;
Performing deep analysis on the intersection weight of the keywords between any two resource data to obtain an intersection degree value which is marked as Qf;
substituting the intersection keyword number T1, the union keyword number T2 and the intersection degree value Qf into a set formula Calculating to obtain the overlapping degree TQ of any two resource data, wherein f4 and f5 are respectively set proportionality coefficients; comparing and analyzing the overlapping degree with a set overlapping threshold value, and recording the two resource data as related resources when the overlapping degree is larger than the set overlapping threshold value; extracting common keywords of all the associated resources as associated keywords of associated resource data; the associated keywords of each relevant field can be obtained and recorded as preference parameters of the user.
Preferably, the specific process of performing the depth analysis on the intersection weight of the keywords between any two resource data is as follows:
Calculating the weight of the key words corresponding to the resource data in the resource data by using a TF-IDF algorithm, thereby obtaining the weight corresponding to the key words of the resource data; summing weights corresponding to a pair of intersection keywords to obtain intersection weights, so that intersection weights corresponding to any two intersection keywords of the resource data can be obtained, and comparing and analyzing the intersection weights with a set weight interval to divide the intersection weights into primary intersection weights, secondary intersection weights and tertiary intersection weights; counting the number of the first-level intersection weight, the second-level intersection weight and the third-level intersection weight respectively, and marking the number as Q1, Q2 and Q3 respectively; summing the first-level intersection weight, the second-level intersection weight and the third-level intersection weight to obtain a first-level intersection value, a second-level intersection value and a third-level intersection value, and marking the first-level intersection value, the second-level intersection value and the third-level intersection value as Q4, Q5 and Q6 respectively;
Substituting the first-level intersection weight Q1, the second-level intersection weight Q2, the number Q3 of the third-level intersection weights, the first-level intersection value Q4, the second-level intersection value Q5 and the third-level intersection value Q6 into a set formula And calculating to obtain intersection degree values Qf of any two resource data, wherein f1, f2 and f3 are respectively set proportionality coefficients, e is a natural constant, and f1 is more than f2 is more than f3 is more than 0.
The invention has the beneficial effects that:
1. analyzing the browsing behaviors of the user to obtain interest recommendation indexes of the user, and deeply mining the overlapping degree between the resource data to judge the association relation between the resource data in each relevant field of the user to obtain preference parameters; the requirements and interests of the user can be more accurately understood, and support is provided for personalized and accurate resource recommendation;
2. screening out resource data related to the search word in an autonomous search mode through a neural network model and a TF-IDF algorithm, and calculating a recommended value, so that the resource data with higher relativity can be rapidly presented when a search result is displayed, and the search time of a user is saved;
3. Through overlapping degree sequencing and long-chain recommendation, the sequencing and recommendation sequence of the resource data is optimized, so that a user can find interested resources more easily, and the efficiency of information acquisition is improved.
Drawings
The invention is further described below with reference to the accompanying drawings.
FIG. 1 is a schematic diagram of a system module connection of the present invention.
Detailed Description
It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be combined with each other. The application will be described in detail below with reference to the drawings in connection with embodiments.
In order that those skilled in the art will better understand the present application, a technical solution in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, shall fall within the scope of the present application.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate in order to describe the embodiments of the application herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be noted that, relevant information (including, but not limited to, user browsing behavior information, resource data of a virtual library, etc.) and data (including, but not limited to, data for presentation, analyzed data, etc.) related to the present disclosure are information and data authorized by a user or sufficiently authorized by each party.
Referring to fig. 1, the present invention is an application scenario analysis system for a virtual library, including: the system comprises a database, a scene analysis module and a scene optimization module; the database is used for storing resource data of the virtual library and browsing behavior information of the user; it should be noted that, generally, one resource data includes text data, image data, and audio data, for example: a portion of the prose novel data comprising an electronic version of the prose novel, an image of the prose novel data, and audio of the prose novel;
The scene analysis module analyzes the frequency of using the virtual library by the user to judge the learning state of the user; meanwhile, overlapping degree analysis is carried out through the historical browsing record of the user so as to extract and obtain preference parameters of the user; sending the generated interest recommendation strategy and preference parameters to a scene optimization module; the method comprises the following steps:
Step one: retrieving user browsing behavior information, wherein the browsing behavior information comprises fields and keywords corresponding to resource data of each browsing behavior, and browsing starting time and browsing ending time; a time cross axis is established, and each resource data is marked on the time cross axis according to the corresponding browsing start time and browsing end time to obtain a browsing behavior track graph of the user; calculating the interval duration between two adjacent browsing behaviors by carrying out difference calculation on the browsing ending time of the previous browsing behavior and the browsing starting time of the next browsing behavior in the two adjacent browsing behaviors; comparing and analyzing the interval duration with a set interval, and accumulating a high interval when the interval duration is larger than the maximum value in the set interval; when the interval duration is within the set interval, accumulating the interval once; when the interval duration is smaller than the minimum value in the set interval, accumulating the low interval once; counting the number of high intervals, medium intervals and low intervals respectively, and marking the numbers as A1, A2 and A3 respectively; if A1 is more than or equal to A2+A3 or A2 is more than or equal to A1+A3 or A3 is more than or equal to A1+A2, the browsing behavior of the user is in a relatively regular state, the user is in a planning learning state, and A3 is more than or equal to A1+A2 in a high-frequency learning planning; when A1 is more than or equal to A2+A3, generating a low-frequency stable learning state; when A2 is more than or equal to A1+A3, generating an intermediate frequency stable learning state; when A3 is more than or equal to A1+A2, generating a high-frequency stable learning state; otherwise, the fact that the frequency of using the virtual library by the user is irregular is explained, and a general learning state is generated;
Step two: numbering the browsing behaviors according to the sequence of the corresponding browsing time (the browsing start time and the browsing end time) to obtain the corresponding browsing times of each browsing behavior, and calculating the time difference between the browsing start time and the browsing end time corresponding to the browsing behaviors to obtain the browsing duration, so that the browsing duration corresponding to each browsing time can be obtained; constructing a two-dimensional rectangular coordinate system by taking the marks as abscissa and the browsing time length as ordinate, inputting the browsing time length into the coordinate system according to the corresponding marks, and marking the position longer than the position in the coordinate system during browsing as a browsing point; sequentially connecting browsing points by adopting a smooth curve to obtain a line graph of the browsing duration changing along with time; drawing a curve tangent at the browsing points, calculating the slope of the tangent to obtain slope marks of all the browsing points as Xj, wherein j=1, 2,3 … … J, the value of J is a positive integer, J represents the total number of browsing behaviors, and J represents the label of any one of the browsing behaviors; summing the slopes smaller than zero, taking the absolute value, marking the browsing monotonic decreasing degree as H1, and summing the slopes larger than zero to obtain the browsing monotonic increasing degree as H2; using a set formula Calculating to obtain a browsing activity value Hb, wherein b1 and b2 are set proportionality coefficients respectively, and e is a natural constant;
Step three: setting each learning state to correspond to one learning frequency value as Bo, wherein o=1, 2,3 and 4; when o=1, B1 represents a learning frequency value corresponding to the high-frequency steady learning state; when o=2, B2 represents a learning frequency value corresponding to the intermediate frequency steady learning state; when o=3, B3 represents a learning frequency value corresponding to the low-frequency steady learning state; when o=4, B4 represents a learning frequency value corresponding to the general learning state; and B1 > B2 > B3 > B4 > 0; the learning frequency value Bo of the user can be obtained, and the learning frequency value Bo and the browsing activity value Hb are calculated through a set formula BH=ln (b3×e Bo+b4×eHb +1) to obtain a user interest recommendation index BH, wherein b3 and b4 are respectively set proportionality coefficients, and e is a natural constant; according to the formula, when the learning frequency of the user is higher, the learning rule of the user is described, the higher the learning frequency is, the user has a stable learning interest label in the virtual library, and the user can learn more continuously and deeply for resources in a specific field; when the browsing activity value of the user is larger, the exploration and browsing interest degree of the user on the resource data in the virtual library is described; therefore, the greater the interest recommendation index of the user, the more likely the recommendation of relevant resource data to the user according to the user history interest tags is the user demand resource data; comparing and analyzing the interest recommendation index with a set recommendation threshold, and generating an interest recommendation strategy when the interest recommendation index of the user is larger than the set recommendation threshold;
Step four: setting each resource data to correspond to one field, wherein the field corresponds to a plurality of keywords; therefore, the resource data can be classified and labeled, so that a user can conveniently search and screen resources; classifying the fields according to the fields of the resource data browsed by the user to obtain the fields which are recorded as the related fields of the user, wherein the related fields are the fields corresponding to the resource data browsed by the user;
Step five: extracting keywords of each resource data in each related field, counting the number of intersection keywords and the number of union keywords of any two resource data, and marking the intersection keywords and the number of union keywords as T1 and T2 respectively; it should be noted that, if the same keywords exist in any two resource data, the same keywords are marked as a pair of intersection keywords, and the number of union keywords can be obtained by traversing all the keywords of any two resource data;
calculating the weight of the keyword corresponding to each resource data in the resource data by using a TF-IDF (Term Frequency-I nverse Document Frequency) algorithm, thereby obtaining the weight corresponding to each keyword of each resource data; summing weights corresponding to a pair of intersection keywords to obtain intersection weights, so that the intersection weights corresponding to the intersection keywords of any two resource data can be obtained, and compared and analyzed with a set weight interval, and when the intersection weights are larger than the maximum value in the set weight interval, the intersection weights are marked as first-level intersection weights; when the intersection weight is within the set weight interval, the intersection weight is marked as a secondary intersection weight; when the intersection weight is smaller than the minimum value in the set weight interval, the intersection weight is marked as three-level intersection weight; counting the number of the first-level intersection weight, the second-level intersection weight and the third-level intersection weight respectively, and marking the number as Q1, Q2 and Q3 respectively; summing the first-level intersection weight, the second-level intersection weight and the third-level intersection weight to obtain a first-level intersection value, a second-level intersection value and a third-level intersection value, and marking the first-level intersection value, the second-level intersection value and the third-level intersection value as Q4, Q5 and Q6 respectively; using a set formula Calculating to obtain intersection degree values Qf of any two resource data, wherein f1, f2 and f3 are respectively set proportionality coefficients, e is a natural constant, and f1 is more than f2 is more than f3 is more than 0;
Passing the intersection keyword number T1, the union keyword number T2 and the intersection degree value Qf through a set formula Calculating to obtain the overlapping degree TQ of any two resource data, wherein f4 and f5 are respectively set proportionality coefficients; from the formula, when the number of intersection keywords is equal to the number of union keywords/>When the keywords of the two data resources are completely overlapped, the intersection degree value is larger; comparing and analyzing the overlapping degree with a set overlapping threshold value, and when the overlapping degree is larger than the set overlapping threshold value, indicating that the two resource data have an association relationship, and marking the two resource data as associated resources; extracting common keywords of all the associated resources as associated keywords of associated resource data; the related keywords of each related field can be obtained and recorded as preference parameters;
Analyzing the browsing behaviors of the user to judge the interest recommendation index of the user and deeply mining the overlapping degree between the resource data to judge the association relation between the resource data in each relevant field of the user so as to obtain preference parameters; the method and the device can be used for more accurately understanding the demands and interests of the users and providing support for personalized and accurate resource recommendation.
The scene optimization module provides personalized resource recommendation for the user according to the received interest recommendation strategy, so that the user search time is reduced, and the user experience is optimized; the method comprises the following steps:
Setting two search modes of the virtual library, wherein the two search modes are respectively autonomous search and recommendation assistance; the autonomous search refers to a behavior of directly selecting a required field to perform self-search by making a user have a relatively clear target; recommendation assistance refers to the behavior that a user does not need to search and the virtual library carries out self-recommendation according to the preference of the user; the user can freely switch between the two searching modes; the flexibility and the personalized selection are beneficial to improving the user experience and meeting the requirements and preferences of different users;
If the user selects the autonomous retrieval mode, the neural network model is utilized to screen out the resource data related to the retrieval words and record the resource data as related resource data, and the neural network model can carry out semantic analysis and context understanding on the retrieval words and the resource data of the user in a deep learning mode to screen out the resource data related to the retrieval words; the method comprises the steps of calling keywords of related resource data, comparing the keywords with preference parameters to obtain the number of the related keywords in the related resource data, marking the number as U1, calculating weights of the related keywords in the related resource data by using a TF-IDF algorithm, and summing the weights to obtain a total weight value which is marked as U2; calculating by using a set formula ua=a1×u1+a2×u2 to obtain a recommended value Ua, wherein a1 and a2 are set scaling factors respectively; sequencing relevant resource data according to corresponding recommended values from large to small and displaying the relevant resource data as a retrieval result; screening out resource data related to the search word in an autonomous search mode through a neural network model and a TF-IDF algorithm, and calculating a recommended value, so that the resource data with higher relativity can be rapidly presented when a search result is displayed, and the search time of a user is saved;
If the user selects the recommended auxiliary mode, K1: the resource data of the last browsing action of the user is called as initial resources, and K2: the method comprises the steps of calling resource data in the same field as an initial resource and overlapping degree between each resource data and the initial resource, and sequencing the resource data according to the sequence from high overlapping degree to low overlapping degree to be used as recommendation sequencing; k3: repeating the steps K1-K2 until the user closes the recommendation auxiliary mode, so as to realize the recommendation process of a long chain and continuously extend the recommendation range; through overlapping degree sequencing and long-chain recommendation, the sequencing and recommendation sequence of the resource data are optimized, so that a user can find interested resources more easily, and the efficiency of information acquisition is improved;
the personalized and efficient resource recommendation service is provided for the user by combining the autonomous retrieval mode and the recommendation auxiliary mode, so that multiple beneficial effects of improving user experience, reducing search time, increasing user satisfaction and the like are achieved.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises an element.
The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and variations of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. which come within the spirit and principles of the application are to be included in the scope of the claims of the present application.
Claims (6)
1. An application scenario analysis system for a virtual library, comprising: the scene analysis module and the scene optimization module;
The scene analysis module analyzes the learning state of the user through the browsing behavior of the user and judges whether to generate an interest recommendation strategy; meanwhile, carrying out overlapping analysis according to the browsing behaviors of the user so as to refine preference parameters of the user;
the scene optimization module optimizes the searching mode according to the preference parameters, and specifically comprises the following steps:
Setting two search modes in the virtual library, and screening out resource data related to the search word by using a neural network model to be recorded as related resource data if the user selects an autonomous search mode; the key words of the related resource data are called and compared with the preference parameters to obtain the quantity of the related key words in the related resource data, the weight of the related key words in the related resource data is calculated by utilizing a TF-IDF algorithm, and the weight total value is obtained by summing the weight total value; normalizing the number and the weight total value of the related keywords, taking the numerical value of the normalized number, analyzing the numerical value to obtain a recommended value, sequencing related resource data according to the corresponding recommended value from large to small, and displaying the recommended value as a retrieval result;
if the user selects the recommended auxiliary mode:
K1: the resource data of the last browsing action of the user is called as initial resources;
k2: the method comprises the steps of calling resource data in the same field as an initial resource and overlapping degree between each resource data and the initial resource, and sequencing the resource data according to the sequence from high overlapping degree to low overlapping degree to be used as recommendation sequencing;
K3: repeating steps K1-K2 until the user turns off the recommended auxiliary mode.
2. The application scenario analysis system of a virtual library according to claim 1, wherein the specific process of analyzing the learning state of the user is:
Retrieving user browsing behavior information, wherein the browsing behavior information comprises fields and keywords corresponding to resource data of each browsing behavior, and browsing starting time and browsing ending time; a time cross axis is established, and each resource data is marked on the time cross axis according to the corresponding browsing start time and browsing end time to obtain a browsing behavior track graph of the user;
Calculating the interval duration of two adjacent browsing behaviors; comparing and analyzing the interval duration with a set interval so as to divide the interval duration into a high interval, a medium interval and a low interval; counting the number of high intervals, medium intervals and low intervals respectively, and marking the numbers as A1, A2 and A3 respectively; when A1 is more than or equal to A2+A3, generating a low-frequency stable learning state; when A2 is more than or equal to A1+A3, generating an intermediate frequency stable learning state; when A3 is more than or equal to A1+A2, generating a high-frequency stable learning state; otherwise, a general learning state is generated.
3. The application scenario analysis system of claim 2, wherein the specific process of determining whether to generate the interest recommendation policy is:
Tracking the browsing behavior and the corresponding browsing time length to establish a graph of the browsing behavior changing along with time, and carrying out graph analysis on the graph to obtain a browsing activity value;
setting each learning state to be respectively corresponding to one learning frequency value as Bo, carrying out normalization processing on the learning frequency value and the browsing activity value, and taking the numerical value, and analyzing the numerical value to obtain a user interest recommendation index; and comparing and analyzing the interest recommendation index with a set recommendation threshold, and generating an interest recommendation strategy when the interest recommendation index of the user is larger than the set recommendation threshold.
4. The application scenario analysis system of claim 3, wherein the specific process of analyzing the graphics is:
Numbering the browsing behaviors according to the sequence of the corresponding browsing time to obtain the corresponding browsing times of each browsing behavior, and calculating the time difference between the browsing start time and the browsing end time corresponding to the browsing behaviors to obtain the browsing duration, so that the browsing duration corresponding to each browsing time can be obtained;
Constructing a two-dimensional rectangular coordinate system by taking the marks as abscissa and the browsing time length as ordinate, inputting the browsing time length into the coordinate system according to the corresponding marks, and marking the position longer than the position in the coordinate system during browsing as a browsing point; sequentially connecting browsing points by adopting a smooth curve to obtain a line graph of the browsing duration changing along with time; a curve tangent is made at the browsing points, and the slope of the tangent is calculated to obtain the slope of each browsing point; summing the slopes smaller than zero, taking absolute values, and summing the slopes larger than zero to obtain browsing monotonic increasing degrees;
and carrying out normalization processing on the slope, the browsing monotonic decreasing degree and the browsing monotonic increasing degree, taking the numerical value, and analyzing the numerical value to obtain the browsing activity value.
5. The system for analyzing application scenarios in a virtual library according to claim 1, wherein the specific process of performing overlay analysis to refine the preference parameters of the user according to the browsing behavior of the user is as follows:
Extracting keywords of each resource data in each related field, and counting the number of intersection keywords and the number of union keywords of any two resource data;
Performing deep analysis on the intersection weight of the keywords between any two resource data to obtain an intersection degree value;
normalizing the number of intersection keywords, the number of union keywords and the intersection degree value, taking the numerical value, and analyzing the numerical value to obtain the overlapping degree of any two resource data; comparing and analyzing the overlapping degree with a set overlapping threshold value, and recording the two resource data as related resources when the overlapping degree is larger than the set overlapping threshold value; extracting common keywords of all the associated resources as associated keywords of associated resource data; the associated keywords of each relevant field can be obtained and recorded as preference parameters of the user.
6. The application scene analysis system of a virtual library according to claim 5, wherein the specific process of performing the deep analysis on the intersection weight of the keywords between any two resource data is as follows:
Calculating the weight of the key words corresponding to the resource data in the resource data by using a TF-IDF algorithm, thereby obtaining the weight corresponding to the key words of the resource data; summing weights corresponding to a pair of intersection keywords to obtain intersection weights, so that intersection weights corresponding to any two intersection keywords of the resource data can be obtained, and comparing and analyzing the intersection weights with a set weight interval to divide the intersection weights into primary intersection weights, secondary intersection weights and tertiary intersection weights; respectively counting the number of the first-level intersection weight, the second-level intersection weight and the third-level intersection weight, and respectively summing the first-level intersection weight, the second-level intersection weight and the third-level intersection weight to obtain a first-level intersection value, a second-level intersection value and a third-level intersection value;
And carrying out normalization processing on the first-level intersection weight, the second-level intersection weight, the number of the third-level intersection weights, the first-level intersection value, the second-level intersection value and the third-level intersection value, taking the numerical values, and analyzing the numerical values to obtain the intersection degree value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410288311.6A CN118097078A (en) | 2024-03-14 | 2024-03-14 | Application scene analysis system of virtual library |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410288311.6A CN118097078A (en) | 2024-03-14 | 2024-03-14 | Application scene analysis system of virtual library |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118097078A true CN118097078A (en) | 2024-05-28 |
Family
ID=91151424
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410288311.6A Pending CN118097078A (en) | 2024-03-14 | 2024-03-14 | Application scene analysis system of virtual library |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118097078A (en) |
-
2024
- 2024-03-14 CN CN202410288311.6A patent/CN118097078A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Balog et al. | Transparent, scrutable and explainable user models for personalized recommendation | |
| US11645317B2 (en) | Recommending topic clusters for unstructured text documents | |
| US7953735B2 (en) | Information processing apparatus, method and program | |
| US20180225710A1 (en) | User segment identification based on similarity in content consumption | |
| US20140229476A1 (en) | System for Information Discovery & Organization | |
| CN101566997A (en) | Determining words related to given set of words | |
| CN105426550B (en) | Collaborative filtering label recommendation method and system based on user quality model | |
| CN111008321A (en) | Recommendation method and device based on logistic regression, computing equipment and readable storage medium | |
| CN102640146A (en) | Database searching method, system and controller | |
| CN113535963A (en) | Long text event extraction method and device, computer equipment and storage medium | |
| KR20120049551A (en) | Method for recommending content based on user preference and profile of content | |
| Chiny et al. | Netflix recommendation system based on TF-IDF and cosine similarity algorithms | |
| CN116257694B (en) | Intelligent search recommendation method and system based on online learning courses | |
| Wei et al. | Online education recommendation model based on user behavior data analysis | |
| D'Addio et al. | A collaborative filtering approach based on user's reviews | |
| Bonnel et al. | Effective organization and visualization of web search results | |
| KR101826594B1 (en) | Method and system for recommending e-book based on knowledge structure | |
| CN118097078A (en) | Application scene analysis system of virtual library | |
| KR101563625B1 (en) | Ontology based image contents recommendation system using relations of metadata and method for the same | |
| Hoxha et al. | Towards a modular recommender system for research papers written in albanian | |
| CN112445904A (en) | Knowledge retrieval method, knowledge retrieval device, knowledge retrieval equipment and computer readable storage medium | |
| Evert et al. | A distributional approach to open questions in market research | |
| Luo et al. | Integrating multi-modal content analysis and hyperbolic visualization for large-scale news video retrieval and exploration | |
| CN110795598A (en) | Intelligent column generation method based on VCA video tag analysis | |
| Sun et al. | CROA: A Content-Based Recommendation Optimization Algorithm for Personalized Knowledge Services |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination |