Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The terms first, second and the like in the description and in the claims, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged, as appropriate, such that embodiments of the present invention may be implemented in sequences other than those illustrated or described herein, and that the objects identified by "first," "second," etc. are generally of a type, and are not limited to the number of objects, such as the first object may be one or more.
The method for determining internet service information provided by the invention can be applied to various kinds of internet with consultation service function, and for convenience of explanation, in the following embodiment, the method for determining internet service information provided by the invention is exemplified by application of the method for determining internet service information provided by the invention to industrial internet.
In practice, the industrial internet involves industry segments, each with its own needs and expertise thresholds, so that users need to make a lot of expertise consultations when buying industrial products or obtaining solutions. How to better answer the consultation of the client and grasp the business machine is an important problem to be solved by the industrial Internet.
It should be noted that, in the related art, only the consultation problem of the user is simply recorded, and then the consultation problem is classified and solved, but how to timely and accurately notify the appropriate problem to the solver who can solve the problem is not a good solution.
In this regard, in the method for determining internet service information provided by the embodiment of the invention, the user behavior data is collected and organized into the available index, and then the behavior characteristics such as possible preference of the user are predicted by using the preset network model, and when the user submits the consultation request data in the foreground, the first service personnel are recommended to the user according to the behavior characteristics, so that the first service personnel can more efficiently solve the consultation request of the user.
In order to facilitate understanding of the internet service information determining method, the internet service information determining device and the electronic device provided by the invention, the invention is described below with reference to the accompanying drawings:
Referring to fig. 1, the method for determining internet service information provided by the embodiment of the application may include the following steps:
Step 101, a first input is received.
In implementations, the first input may be any input by a foreground user in a target internet page. For example, the first input may be one or more of opening a target Internet page, browsing in the target Internet page, clicking on a commodity in the target Internet page, entering advisory request data at an advisory service portal in the target Internet page, collecting a product, and the like, which are not intended to be exhaustive.
Step 102, responding to the first input, and acquiring behavior characteristic data of a foreground user in a target internet page, wherein the behavior characteristic data comprises consultation request data.
In an implementation, the behavior feature data may be any data and/or log and/or process generated based on the first input. In implementation, the behavior data of the foreground user can be collected by presetting a buried point at the front end.
The foregoing consultation request data may be consultation request data input by the foreground user at the consultation service portal in the target internet page, and in response thereto, the first service personnel corresponding to the requirement information may be matched in the background of the target internet through the following steps 103 to 105, so as to recommend the first service personnel to provide the consultation service for the user. For example, when the foreground user inputs the requirement information of the product to be purchased in the consultation service portal of the foreground of the target internet, the foreground user can find the professional B according to the behavior characteristic data of the user and inform the professional B of the requirement information, so that the professional B recommends a proper product according to the requirement information of the product to be purchased by the foreground user.
The above-mentioned consultation request data may be consultation data input by a foreground user in the consultation service portal, in an implementation, in order to prevent a maliciously submitted consultation request, the consultation content may be verified, and the corresponding consultation request data may be generated after the verification.
For example, as shown in fig. 3, in the case that the background administrator sets the target portal column to be consultable (that is, displays the consultation service portal in the webpage corresponding to the target portal column), the foreground user may click on the consultation service portal and input the consultation content in the consultation service portal, at this time, in response to the click operation of the foreground user on the consultation service portal, the service consultation platform may be requested to generate a verification code, return the verification code to the target internet page, and store the verification code, where the target internet page includes the above-mentioned consultation service portal, so that the foreground user also needs to input the verification code in the verification box of the consultation service portal, and when the foreground user submits the consultation content, the service consultation platform may match the verification code input by the foreground user with the stored verification code, and generate the consultation request data corresponding to the consultation content only when the matching is successful.
And 103, determining the value of each index in N-dimensional indexes according to the behavior characteristic data, wherein N is an integer greater than 1, and the N-dimensional indexes are used for quantifying the behavior characteristics from N dimensions.
In practice, the N-dimensional index may be an index that can quantify the behavior feature data from N dimensions, such as a flow time for a product, a page, a number of clicks on a product in a day, an hour, and a week, respectively, and so forth.
As an optional implementation manner, the determining the value of each index in the N-dimensional indexes according to the behavior feature data includes:
according to the behavior characteristic data, a preset M-dimensional index is determined, wherein M is an integer greater than N;
Screening the M-dimensional index based on a random forest algorithm to obtain an N-dimensional index;
And determining the N as the value of each index in the indexes.
In implementation, the M-dimensional index may be a preset index with a large number of dimensions, but the excessive index increases complexity of a subsequent preset network model, thereby increasing calculation overhead, and the partial index may have multiple collinearity, thereby generating interference on prediction accuracy of the preset network model. In the embodiment of the application, the M-dimensional index is screened through the random forest algorithm with higher robustness, so that the robustness of the screened N-dimensional index is higher than that of the M-dimensional index, and the prediction precision of the preset network model can be improved while the calculation cost is reduced.
For example, the behavior characteristic data can be stored in a spark big data platform, so that the excellent performance of spark on big data processing can be utilized, the data are firstly subjected to preliminary cleaning, then the data are calculated and arranged into multi-dimensional indexes according to operation statistics indexes, and finally 1500 multi-dimensional indexes are obtained through arrangement. For example, basic indexes (such as browsing times, clicking times and the like of a foreground user on a certain product), time sequence indexes (such as indexes of clicking amount in one day, clicking amount in two days and the like) obtained by disassembling according to time dimensions, and cross indexes (such as cross indexes of collected occupancy rate of a commodity after being clicked in one day, purchased occupancy rate of the commodity after being browsed in 3 days and the like) obtained by combining the dimension indexes.
Optionally, the filtering the M-dimensional index based on the random forest algorithm to obtain an N-dimensional index includes:
determining a characteristic importance coefficient of each index in the M-dimensional indexes based on a random forest algorithm, wherein the characteristic importance coefficient is positively correlated with a coefficient of a kunit of the corresponding index;
and selecting an N-dimensional index with the characteristic importance coefficient larger than or equal to a preset value from the M-dimensional index.
In practice, the feature importance coefficients described above may be calculated by the following formula:
Wherein Hj represents the characteristic importance coefficient of index j, VIMj represents the Kidney coefficient of index j, saidRepresenting the sum of the gains of all the indicators.
The selecting the N-dimensional index with the feature importance coefficient greater than or equal to the preset value from the M-dimensional index may be arranging the M-dimensional index in order from the high feature importance coefficient to the low feature importance coefficient, and selecting the N-dimensional index with the top N bits. The larger the feature importance coefficient is, the higher the importance degree of the corresponding index is. Therefore, the N-dimensional index with higher importance degree is selected from the M-dimensional indexes, so that the calculation cost is reduced, and the prediction precision of a preset network model can be improved.
And 104, converting the values of all indexes in the N-dimensional indexes into two-dimensional single-channel pictures.
In practice, the two-dimensional single-channel picture may be referred to as a two-dimensional gray-scale picture, and the storage form of the two-dimensional single-channel picture in a computer is actually a two-dimensional array, and the two-dimensional array may be converted into the two-dimensional single-channel picture through a python image library (Python Image Library, PIL), for example, assuming that N is equal to 784, the two-dimensional single-channel picture may be equivalent to a 28×28 two-dimensional array.
And 105, calculating the two-dimensional single-channel picture based on a preset network model so as to predict and obtain a first service personnel matched with the consultation request data.
The preset network model may be a pre-trained network model of the dream theft space (Inception), so that the possible preference of the user can be predicted by using the Inception network model with excellent effect in the field of image recognition. The network model Inception is first proposed in CHRISTIAN SZEGEDY et al, in its paper "good DEEPER WITH Convolutions", and is mainly used for two-dimensional image processing.
Of course, in implementation, the preset network model may be other network models, such as a neural network model, which is not limited herein.
In implementation, the N-dimensional index is converted into a two-dimensional single-channel picture, so that the data structure of the N-dimensional index can be simplified, and further the model complexity of the preset network model is simplified.
In an implementation, the two-dimensional single-channel picture is calculated through a preset network model, so that the first service personnel obtained through prediction can be service personnel matched with behavior habits, favorites and the like of the foreground user, and therefore the service personnel can provide more efficient consultation services for the foreground user.
As an optional implementation manner, the preset network model includes a preset Inception network model, and the calculating the two-dimensional single-channel picture based on the preset network model is performed to predict a first service person matched with the consultation request data, including:
Inputting the two-dimensional single-channel picture into a preset dream stealing space Inception network model;
And predicting to obtain the first service personnel matched with the consultation request data according to the output result of the preset Inception network model.
In an implementation, the preset Inception network model includes a convolution layer, a max-pooling layer, a first Inception model layer, a second Inception model layer, a global average pooling layer, and an output layer;
The convolution layer is used for extracting first features of the N-dimensional index, the maximum pooling layer is used for carrying out maximum pooling treatment on the first features, and the first features after the maximum pooling treatment are input into the first Inception model layer;
The first Inception model layer is used for extracting second features from the first features after the maximum pooling treatment;
the second Inception model layer is used for extracting a third feature from the second feature;
the global average pooling layer is used for carrying out global average pooling treatment on the second feature to obtain a first prediction result, and is also used for carrying out global average pooling treatment on the third feature to obtain a second prediction result;
the output layer is used for determining a first service person matched with the consultation request data according to the first prediction result and the second prediction result, and outputting identification information of the first service person.
With the Inception network model shown in fig. 2, the convolution layers may include 2 layers of 32 convolution kernels of 3*3 and one layer of 64 convolution kernels of 3*3 for simultaneous feature extraction using multi-dimensional convolution kernels. The first features extracted by the max-pooling convolutional layer are then max-pooled.
In addition, the first Inception model layer may include 4 columns of convolution groups, so that the first feature after the maximum pooling process is input to the 4 columns of convolution groups in the first Inception model layer to perform feature extraction together, where the first column of convolution groups performs feature extraction by using 64 convolution kernels of 1*1, the second column of convolution groups performs feature extraction by using 48 convolutions of 1*1 followed by 64 convolutions of 5*5, the third column of convolution groups performs feature extraction by using 64 convolutions of 1*1 followed by 96 convolutions of 3*3, and the fourth column of convolution groups performs feature extraction by using a convolution kernel of 3*3 followed by 32 convolutions of 1*1. And finally, merging the four columns of extracted second features, inputting the merged second features to a global average pooling layer, and inputting the second features to a second Inception model layer. The structure of the second Inception model layer is the same as that of the first Inception model layer, and will not be described in detail herein. After extracting a third feature from the second feature by the second Inception model layer, inputting the third feature to a global averaging pooling layer. The global average pooling layer is followed by an output layer (i.e., softmax layer), and the second feature (i.e., the first prediction result) after global average pooling and the third feature (i.e., the second prediction result) after global average pooling are combined at the softmax layer to obtain a prediction result (i.e., the first service personnel).
It can be seen from the above that, the preset Inception network model provided by the embodiment of the invention uses the multidimensional convolution kernel to simultaneously perform feature extraction, so as to promote the diversity of the extracted features, thereby enabling the preset Inception network model trained according to the diversity of the features and the prediction result calculated based on the preset Inception network model to be more reliable, and the implicit layer (namely the first Inception model layer) in the preset Inception network model can also output the result (namely the first prediction result) through the global average pooling layer, and can play the effect of model fusion by combining the first prediction result with the second prediction result, and in addition, the global average pooling layer is used to replace the full-connection layer, so that the parameter calculation amount can be reduced.
It should be noted that, the preset Inception network model may also predict a product that the foreground user may be interested in according to the value of each of the N-dimensional indexes, and the process of predicting the product that the foreground user may be interested in according to the value of each of the N-dimensional indexes by using the preset Inception network model is similar to the process of predicting the first service personnel according to the value of each of the N-dimensional indexes by using the preset Inception network model, which is not described herein. In this way, after a product which may be interested by a certain foreground user is stored, the recommendation information of the product can be displayed in the subsequent internet pages of the tactics of the foreground user, so as to improve the recommendation effect of the product.
As an optional implementation manner, after the calculating the two-dimensional single channel picture based on the preset network model to predict the first service personnel matching the consultation request data, the method further includes:
Acquiring a prestored first communication address corresponding to the first service personnel;
And sending first prompt information to the first communication address, wherein the first prompt information is used for prompting the first service personnel to respond to the consultation request data.
In the implementation, the communication address of each service personnel, such as a mobile phone number, a mailbox, an account name of a social application program and the like, can be stored in the industrial Internet database applying the Internet service information determining method provided by the invention, so that after the first service personnel is determined, a short message can be sent to the first service personnel, a call is made, a mail is sent, a prompt message is sent through the social application program and the like to remind the first service personnel to provide consultation service for a foreground user, and the consultation request of the foreground user can be responded in time.
As an optional implementation manner, the internet service information determining method further includes:
Acquiring a preconfigured first association relationship, wherein the first association relationship comprises an association relationship between each preconfigured role and a corresponding consultation column;
determining a target role corresponding to the target consultation column according to the first association relationship under the condition that the consultation request data is data input in the target consultation column;
determining a second attendant preconfigured with the attribute of the target role;
acquiring a pre-stored second communication address corresponding to the second service personnel;
and sending second prompt information to the second communication address, wherein the second prompt information is used for prompting the second service personnel to respond to the consultation request data.
In an implementation, the first association relationship may be an association relationship between each portal column configured by the background user and a corresponding role.
The portal column can be a product theme, a consultation theme and the like, the roles can be parameters for distinguishing authorities configured by a background user through a background management system, and each service personnel can have respective role attributes, for example, if the background user configures a role A, a role B and a role C, the configuration role A is associated with the consultation column A, the configuration role B is associated with the consultation column B and the configuration role C is associated with the consultation column C, when the foreground user submits consultation request data through the consultation column A, a second service personnel with the role A attribute can be recommended accordingly to provide consultation services for the foreground user. It should be noted that, in practical application, one role may be associated with one or at least two consultation columns, and/or one consultation column may be associated with one or at least two roles, and/or one service person may have one or at least two role attributes, which are not described herein.
For example, as shown in fig. 4, the configuration process of the character includes:
Step 1, an administrator logs in a background of the consultation service, adds a role A, and associates the role with related service personnel A, for example, binding is carried out through id, wherein the user center stores the association relation between each role and the service personnel associated with each role and the communication information of each service personnel;
step 2, an administrator configures consultation request data (namely a first association relation) of which consultation column is to be received by the role A, and stores the configuration;
Step 3, providing a piece of consultation request data submitted by a user, and storing the consultation request data by a background;
step 4, according to the consulting column consulted by the consulting request data, acquiring the corresponding configured role (assuming that the role is a role A);
Step 5, inquiring the user center according to the role A that the service personnel A is configured;
And 6, sending prompt information to the service personnel A to remind the service personnel A to process the consultation request online.
In the embodiment of the invention, the service personnel can only see the consultation content of the related columns (the related columns are configured to represent the columns with the same roles corresponding to the service personnel), but the consultation content of other columns is invisible, so that the data security can be improved.
In this embodiment, the consultation service is combined with the industrial internet portal, and by configuring a specific role for a portal column, and assigning respective service personnel to the roles, when a foreground user submits a consultation request based on a certain portal column, the service personnel of the role associated with the portal column will receive prompt information at the first time, and respond to the consultation request accordingly. Therefore, the first service personnel and the second service personnel can be selected to respond to the consultation request of the foreground user at the same time, so that the probability that the service personnel recommended for the foreground user can effectively answer the consultation request is improved, and the reliability of the consultation service is improved.
As an optional implementation manner, before the obtaining the behavior feature data of the foreground user in the target internet page, the behavior feature data includes consultation request data, the method further includes:
receiving a hypertext markup language (HyperText Markup Language, HTML) template of the target internet page;
Updating the target internet page according to the HTML template of the target internet page, wherein the updated target internet page comprises a consultation service entrance, and the consultation request data is data acquired in the consultation service entrance.
In practice, the HTML template can be uploaded to a server of the target internet by a background manager, and the consulting service portal in the target internet page can be updated (added/deleted) by uploading the HTML template. For example, a background administrator can set a target internet page for showing the consulting service portal at any time, upload an HTML template, and then, through refreshing a data structure (redis) cache server, the target internet page of the foreground can show the consulting service portal.
In the embodiment, a background manager can add or delete the consultation service inlet in any page of the industrial Internet portal in real time in a background configuration mode, so that the free combination of the consultation service inlet and any page is realized, and the user consultation related subject content of each subdivision industry in the industrial Internet is facilitated.
In order to facilitate understanding of the method for determining internet service information provided by the embodiment of the present invention, taking a flowchart as shown in fig. 5 as an example, the method for determining internet service information provided by the embodiment of the present invention, as shown in fig. 5, may include the following steps:
step 1, a precondition user enters a consultation service entrance in a target internet page, and a consultation module transmits a verification code to a foreground;
Step 2, providing a column capable of being consulted to the foreground through the portal module;
step 3, a foreground user submits consultation request data;
step 4, storing behavior characteristic data of a foreground user to a Spark big data platform through a front-end buried point;
Step 5, calculating behavior characteristic data of a foreground user in real time by a Spark big data platform to obtain available indexes (before the step, a manager finishes the association relationship among service personnel, roles and columns);
Step 6, inception, calculating by a network model algorithm engine according to data of a Spark big data platform, and predicting an interested product and a matched first service personnel;
Step 7, synchronizing the calculation result of Inception network model algorithm engine to the characteristic database;
Step 8, the background stores the consultation request data under the condition that the verification code input by the foreground user is verified to be correct;
step 9, inquiring the corresponding roles according to the columns corresponding to the consultation request data;
Step 10, inquiring the communication address of the service personnel with the role in the user center according to the role;
step 10, notifying a second service personnel to process the consultation request online based on the communication address;
and 11, inquiring the matched first service personnel from the characteristic database, and informing the first service personnel to process the consultation request online.
According to the invention, the binding among the service personnel, the roles and the consultation columns can be realized, the consultation of a specific sub-industry can be only notified to the relevant service personnel, the relevant service personnel can see the consultation content, the consultation processing efficiency and the consultation processing effect are improved, in addition, the invention predicts the interested products of the user and the relevant matched service personnel by utilizing the Spark big data platform and the algorithm engine based on Inception network model, and can better play the role of the service consultation system on industrial Internet popularization.
Referring to fig. 6, which is a block diagram of an internet service information determining apparatus according to an embodiment of the present invention, as shown in fig. 6, the internet service information determining apparatus 600 includes:
a receiving module 601 for receiving a first input;
a first obtaining module 602, configured to obtain, in response to the first input, behavior feature data of a foreground user in a target internet page, where the behavior feature data includes consultation request data;
A first determining module 603, configured to determine, according to the behavior feature data, a value of each index in N-dimensional indexes, where N is an integer greater than 1, and the N-dimensional indexes are used to quantify the behavior feature from N dimensions;
The conversion module 604 is configured to convert values of all the N-dimensional indexes into a two-dimensional single-channel picture;
the calculating module 605 is configured to calculate the two-dimensional single-channel picture based on a preset network model, so as to predict a first service person matching with the consultation request data.
Optionally, the internet service information determining apparatus 600 further includes:
the second acquisition module is used for acquiring a prestored first communication address corresponding to the first service personnel;
the first sending module is used for sending first prompt information to the first communication address, and the first prompt information is used for prompting the first service personnel to respond to the consultation request data.
Optionally, the internet service information determining apparatus 600 further includes:
The third acquisition module is used for acquiring a preconfigured first association relationship, wherein the first association relationship comprises an association relationship between each preconfigured role and a corresponding consultation column;
the second determining module is used for determining a target role corresponding to the target consultation column according to the first association relation under the condition that the consultation request data is data input in the target consultation column;
a third determining module, configured to determine a second attendant preconfigured with an attribute of the target role;
A fourth obtaining module, configured to obtain a second communication address corresponding to the second service person, where the second communication address is stored in advance;
The second sending module is used for sending second prompt information to the second communication address, and the second prompt information is used for prompting the second service personnel to respond to the consultation request data.
Optionally, the first determining module 603 includes:
The first determining unit is used for determining a preset M-dimensional index according to the behavior characteristic data, wherein M is an integer larger than N;
The screening unit is used for screening the M-dimensional index based on a random forest algorithm to obtain an N-dimensional index;
And the second determining unit is used for determining that the N is the value of each index in the indexes.
Optionally, the screening unit includes:
a determining subunit, configured to determine a feature importance coefficient of each index in the M-dimensional indexes based on a random forest algorithm, where the feature importance coefficient is positively related to a kunit coefficient of the corresponding index;
And the screening subunit is used for selecting the N-dimensional index with the characteristic importance coefficient larger than or equal to a preset value from the M-dimensional index.
Optionally, the preset network model includes a preset dream theft space Inception network model, and the computing module 605 includes:
The input unit is used for inputting the two-dimensional single-channel picture into a preset dream stealing space Inception network model;
and the prediction unit is used for predicting and obtaining the first service personnel matched with the consultation request data according to the output result of the preset Inception network model.
Optionally, the preset Inception network model includes a convolution layer, a max pooling layer, a first Inception model layer, a second Inception model layer, a global average pooling layer, and an output layer;
The convolution layer is used for extracting first features of the N-dimensional index, the maximum pooling layer is used for carrying out maximum pooling treatment on the first features, and the first features after the maximum pooling treatment are input into the first Inception model layer;
The first Inception model layer is used for extracting second features from the first features after the maximum pooling treatment;
the second Inception model layer is used for extracting a third feature from the second feature;
the global average pooling layer is used for carrying out global average pooling treatment on the second feature to obtain a first prediction result, and is also used for carrying out global average pooling treatment on the third feature to obtain a second prediction result;
the output layer is used for determining a first service person matched with the consultation request data according to the first prediction result and the second prediction result, and outputting identification information of the first service person.
Optionally, the internet service information determining apparatus 600 further includes:
the receiving module is used for receiving the hypertext markup language (HTML) template of the target internet page;
And the updating module is used for updating the target internet page according to the HTML template of the target internet page, wherein the updated target internet page comprises a consultation service entrance, and the consultation request data is data acquired in the consultation service entrance.
The internet service information determining apparatus 600 provided in the embodiment of the present invention can implement each process implemented by the method embodiment shown in fig. 1 or fig. 5, and can obtain the same beneficial effects, and for avoiding repetition, a detailed description is omitted here.
Optionally, as shown in fig. 7, an electronic device 700 is further provided according to an embodiment of the present invention, including a processor 701, a memory 702, and a program or an instruction stored in the memory 702 and capable of being executed on the processor 701, where the program or the instruction implements each process of the method embodiment shown in fig. 1 when executed by the processor 701, and the process achieves the same technical effects, and for avoiding repetition, a description is omitted herein.
The embodiment of the present invention further provides a computer readable storage medium, where a program or an instruction is stored, where the program or the instruction implements each process of the method embodiment shown in fig. 1 when executed by a processor, and the process can achieve the same technical effects, so that repetition is avoided, and no further description is given here.
Wherein the processor is a processor in the electronic device described in the above embodiment. The readable storage medium includes a computer readable storage medium such as a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk or an optical disk, and the like.
It should be noted that, in this document, 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 the element. Furthermore, it should be noted that the scope of the methods and apparatus in the embodiments of the present invention is not limited to performing the functions in the order shown or discussed, but may also include performing the functions in a substantially simultaneous manner or in an opposite order depending on the functions involved, e.g., the described methods may be performed in an order different from that described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) comprising instructions for causing a terminal (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the method according to the embodiments of the present invention.
The embodiments of the present invention have been described above with reference to the accompanying drawings, but the present invention is not limited to the above-described embodiments, which are merely illustrative and not restrictive, and many forms may be made by those having ordinary skill in the art without departing from the spirit of the present invention and the scope of the claims, which are to be protected by the present invention.