CN110941714A - Classification rule base construction method, application classification method and device - Google Patents
Classification rule base construction method, application classification method and device Download PDFInfo
- Publication number
- CN110941714A CN110941714A CN201811108770.2A CN201811108770A CN110941714A CN 110941714 A CN110941714 A CN 110941714A CN 201811108770 A CN201811108770 A CN 201811108770A CN 110941714 A CN110941714 A CN 110941714A
- Authority
- CN
- China
- Prior art keywords
- application
- classification
- classification rule
- attribute information
- classified
- 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.)
- Withdrawn
Links
Images
Abstract
The embodiment of the invention provides a classification rule base construction method, an application classification method and a device, wherein the application classification method comprises the following steps: and matching the attribute information of the application to be classified with the classification rules in the classification rule base to determine the category of the application to be classified. The method for constructing the classification rule base comprises the following steps: determining a category and attribute information of each of a plurality of first known applications, the attribute information including application description information and/or code description information; and constructing a classification rule base, wherein each classification rule comprises attribute information and a category corresponding to the attribute information. The invention adopts the application description information and/or the code description information with rich content types as the classification rule base constructed by the attribute information, the classification rules are rich and diverse, and when in classification, the classification of the application to be classified can be realized by matching any one of the attribute information with rich content types to be classified and applied to the classification rules, thereby improving the possibility of classifying the application to be classified and improving the classification coverage.
Description
Technical Field
The embodiment of the invention relates to the technical field of classification, in particular to a classification rule base construction method, an application classification method and an application classification device.
Background
The category of the application can be used for analyzing and acquiring related data to serve the public, for example, the user analysis is performed by using the category of the application, the preference of the user of the application can be acquired, and the user attribute of the application can be deduced, such as: gender, occupation, age, the region of the user and the like, so that a user portrait is constructed for recommendation in online shopping or obtaining of suspect information in criminal investigation.
The application category on the application market is generally manually selected for the application in the application classification catalog when the application is on the shelf, and the application category determination needs to be carried out through a certain method for the category without being on the shelf. A common method is to construct a classification model for class determination based on single attribute information common to applications, such as application names. The elements for classification in the method are single, and the degree of freedom of application name naming is extremely high, so that many applications cannot determine the classification.
Disclosure of Invention
The embodiment of the invention provides a classification rule base construction method, an application classification method and a device, which are used for solving the problems that the existing application classification method has limited classification coverage and is difficult to classify all applications in all application markets.
The embodiment of the invention provides a method for constructing a classification rule base, which comprises the following steps: determining a category and attribute information of each of a plurality of first known applications, the attribute information including application description information and/or code description information; and constructing a classification rule base, wherein each classification rule comprises attribute information and a category corresponding to the attribute information.
The embodiment of the invention provides a classification rule base construction device, which comprises: a determining module, configured to determine a category and attribute information of each of a plurality of first known applications, where the attribute information includes application description information and/or code description information; and the construction module is used for constructing a classification rule base, wherein each classification rule comprises attribute information and a category corresponding to the attribute information.
The embodiment of the invention provides an application classification method, which comprises the following steps: and matching the attribute information of the application to be classified with the classification rules in a classification rule base to determine the category of the application to be classified, wherein the classification rule base is constructed according to the construction method of the classification rule base.
An embodiment of the present invention provides an application classification apparatus, including: and the classification module is used for matching the attribute information of the application to be classified with the classification rules in the classification rule base to determine the category of the application to be classified, and the classification rule base is constructed according to the construction method of the classification rule base.
An embodiment of the present invention provides a computer device, including: a processor; and a memory for storing a computer program, the processor being configured to execute the computer program stored on the memory to implement the classification rule base construction method and the application classification method as described above.
An embodiment of the present invention provides a computer storage medium, in which a computer program is stored, and when the computer program is executed by a processor, the method for constructing a classification rule base and the method for applying a classification rule base are implemented.
According to the method and device for constructing the classification rule base, the application description information and/or the code description information with rich content types are used as the attribute information for constructing the classification rule base, so that the classification rules of the classification rule base are rich and diverse, when the application to be classified is classified, any information of the attribute information with rich content types to be classified and applied is matched with the classification rules in the classification rule base, the classification of the application to be classified can be realized, the possibility of classifying the application to be classified is improved, and the classification coverage is further improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a flowchart of a method for constructing a classification rule base according to a first embodiment of the present invention;
FIG. 2 is a flowchart of a method for constructing a classification rule base according to a second embodiment of the present invention;
fig. 3 is a flowchart of a method for constructing a classification rule base according to a third embodiment of the present invention;
fig. 4 is a flowchart of a method for constructing a classification rule base according to a fourth embodiment of the present invention;
fig. 5 is a flowchart of a method for constructing a classification rule base according to a fifth embodiment of the present invention;
FIG. 6 is a flowchart of an application classification method according to a seventh embodiment of the present invention;
fig. 7 is a flowchart of an application classification method according to an eighth embodiment of the present invention;
FIG. 8 is a flowchart of an application classification method according to a ninth embodiment of the present invention;
FIG. 9 is a diagram illustrating a method for constructing a classification rule base according to an embodiment of the present invention;
FIG. 10 is a diagram illustrating a method for constructing a classification rule base according to a second embodiment of the present invention;
fig. 11 is a schematic diagram of a method for constructing a classification rule base according to a third embodiment of the present invention;
fig. 12 is a schematic diagram of a method for constructing a classification rule base according to a fourth embodiment of the present invention;
fig. 13 is a schematic diagram of a method for constructing a classification rule base according to a fifth embodiment of the present invention;
FIG. 14 is a diagram illustrating an application classification method according to a seventh embodiment of the present invention;
FIG. 15 is a diagram illustrating an application classification method according to an eighth embodiment of the present invention;
fig. 16 is a schematic diagram of an application classification method according to a ninth embodiment of the apparatus of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a flowchart of a classification rule base construction method according to an embodiment of the present invention. As shown in fig. 1, in the present embodiment, the method includes:
step S101, determining the category and attribute information of each known application in a plurality of first known applications, wherein the attribute information comprises application description information and/or code description information;
in this embodiment, the first known application is an application of which the category is known, and can be crawled from various application markets. The application description information refers to application basic attribute information or description information of functions, contents or developers of the application, and is characterized from the visual cognition of the user on the application, such as application program names, package names, certificates, icons, market application comments and the like. The code description information is information extracted from the application code level that can characterize an application, such as class name, method name, built-in sdk, sensitive strings, library call relations, etc. used in developing the first known application. The attribute information may be selected from a number of application description information and/or a number of code description information. For example, the attribute information may include: program name, package name, certificate, class name, and library call relationship.
Step S103, a classification rule base is established, wherein each classification rule comprises attribute information and a category corresponding to the attribute information.
In this embodiment, for each first known application, a mapping relationship is formed between each type of application description information or code description information included in the attribute information and the category of the first known application to obtain a plurality of classification rules. For example, in the first known application, attribute information includes wechat (application name), com.tencent.mm (package name), CN ═ Tencent, OU ═ Tencent Guangzhou Research and Development Center, O ═ Tencent Technology (Shenzhen) Company Limited, L ═ Shenzhen, ST ═ Guangdong, C ═ 86 (certificate), and corresponding categories are social communications, then three classification rules are formed: WeChat-social communication, com.tencent.mm-social communication, CN ═ Tencent, OU ═ Tencent Guangzhou Research and Development Center, O ═ tenncent technology (Shenzhen) Company Limited, L ═ Shenzhen, ST ═ Guangdong, C ═ 86-social communication.
When the attribute information is the market application comment, the market application comment content of each first known application is segmented, and meaningful words are taken out to serve as keywords, so that a group of keywords is obtained. Wherein duplicate words are not present in the set of keywords. And forming a mapping relation between the group of keywords and the category of the first known application to obtain a classification rule. The market application comment of the first known application is the sum of all feedback information of the user on the first known application, namely, one first known application corresponds to one market application comment. The word segmentation method and strategy can be selected from the existing word segmentation methods and are not detailed here. The classification rule is formed, specifically, for example, if a set of keywords corresponding to the first known application is high definition, fast, and resource rich, and the corresponding category is video playing, the classification rule is (high definition, fast, resource rich) -video playing.
The classification rules corresponding to each first known application are used together to form a classification rule base.
In the first embodiment of the method, the application description information and/or the code description information with rich content types is used as the attribute information to construct the classification rule base, so that the classification rules of the classification rule base are rich and diverse, and when the application to be classified is classified, any information in the attribute information with rich content types of the application to be classified is matched with the classification rules in the classification rule base, so that the application to be classified can be classified, the possibility of classifying the application to be classified is improved, and the classification coverage is further improved.
Fig. 2 is a flowchart of a classification rule base construction method according to a second embodiment of the present invention. As shown in fig. 2, in the present embodiment, on the basis of the first method embodiment, the method further includes:
step S205, if the application description information is a market application comment, the classification rule base further includes a weight of each valid word in each category, and the weight of a certain valid word in a certain category is determined according to the number of times that a certain valid word appears in all market application comments of a certain category and the number of categories corresponding to a certain valid word, where a valid word is a keyword included in all market application comments of a first known application of any category, and the number of categories corresponding to a certain valid word is the number of categories of a first known application in which a certain valid word appears in a market application comment.
In this embodiment, each first known application corresponds to a category and a market application review. All market application reviews of a certain category are the sum of the market application reviews corresponding to each first known application of the category. And segmenting all market application comments of each category to obtain a series of vocabularies, and taking out the meaningful vocabularies as keywords, namely effective words, so as to obtain all the effective words. Counting the occurrence frequency of each effective word in all market application comments of each category, determining the category with the occurrence frequency not being zero, and counting the number of the categories with the occurrence frequency not being zero as the number of the categories corresponding to the effective words.
The weight of each valid word on each category can be calculated according to TF-IDF (Term Frequency-inverse document Frequency). The word frequency TF is the frequency of a certain effective word appearing in the market comment corresponding to a certain category, and the larger the TF value is, the higher the importance of the certain effective word in the market comment corresponding to the certain category is; the inverse document frequency IDF is a measure of the popularity of the certain effective word, and if the value is smaller, the certain effective word is considered to be very popular among market reviews corresponding to various categories, and if the value is large, the certain effective word is considered to be rarely appeared among market reviews corresponding to various categories except the certain category.
Specifically, the weight W of the ith valid word in all valid words in a certain categoryiThe calculation formula is as follows:
Wi＝TFi*IDFi(1)
in the above formula (1), TFiAnd IDFiRespectively the word frequency and the reverse file frequency of the ith effective word; in the above formula (2), niFor the number of times the ith valid word appears in all market application reviews for that category, njThe number of times that the jth vocabulary in all market application comments of the category appears in all market application comments of the category is k, and the k is the number of vocabularies obtained by segmenting all market application comments of the category; in the above equation (3), N is the number of classes of all the first known applications, and N isi∈mThe number of the corresponding categories of the ith valid word.
The greater the weight of the ith valid word on the category, the more important the ith valid word is for the category.
In the embodiment of the method, if the application description information is the market application comment, the classification rule base further comprises the weight of each effective word in each category, and when the application to be classified is classified, the category of the application to be classified can be determined according to the weight of each effective word in the market application comment to be classified in the application to be classified and the weight of each effective word in the classification rule base. Due to the fact that the importance of each effective word in the market application comment of the application to be classified to each category in the application to be classified and the classification rule base is considered, the application to be classified is classified more accurately.
Fig. 3 is a flowchart of a classification rule base construction method provided by the third embodiment of the method of the present invention. As shown in fig. 3, in the present embodiment, on the basis of the first method embodiment, the method further includes:
step S302, if a certain attribute information corresponds to multiple categories, when a variance between the numbers of the first known applications corresponding to the categories is greater than a preset threshold, the category corresponding to the certain attribute information is specifically the category with the largest number of the corresponding first known applications.
In this embodiment, the attribute information of a certain first known application is selected from several kinds of application description information and/or several kinds of code description information. Each attribute information of a certain first known application corresponds to a category of the certain first known application. Certain attribute information of different classes of first known applications may be the same. Therefore, the same attribute information may correspond to a plurality of categories.
In this case, for a plurality of categories corresponding to the same attribute information, if the variance between the numbers of the first known applications corresponding to the respective categories is large, it indicates that the distribution of the attribute information is not uniform among the categories, and the larger the variance is, the larger the distribution difference of the attribute information among the categories is, the larger the possibility that a certain distribution is larger than other distributions is, that is, it indicates that most of the first known applications having the attribute information may correspond to one or a few categories. Therefore, when a certain attribute information corresponds to a plurality of categories, the number of the first known applications corresponding to each category is counted, and the variance of the number of the first known applications corresponding to each category is calculated. And if the variance is larger than a preset threshold value, taking the category with the maximum number corresponding to the first known application as the category of the attribute information. The first threshold may be determined by: determining a plurality of categories corresponding to the plurality of attribute information respectively; and calculating the variance among the number of the first known applications corresponding to each category for each category corresponding to each attribute information, arranging the variances corresponding to each attribute information in a descending order, and selecting a proper variance value or value range through manual check so that the number of the first known applications under a certain category in each category corresponding to each attribute information is far larger than that of other categories when the variance corresponding to each attribute information is larger than the variance value or value range. If the variance is less than or equal to the preset threshold, it indicates that the distribution of the categories is not greatly different, and no processing is required.
The embodiment of the method of the invention improves the classification accuracy of the classification rule corresponding to the attribute information by determining a category with the highest proportion from a plurality of categories of the attribute information as the unique category of the attribute information.
In addition, step S302 is further included on the basis of each of the above-mentioned method embodiments other than the first and third method embodiments, and the resulting method embodiment is also within the scope of the present invention and will not be described in detail herein.
Fig. 4 is a flowchart of a classification rule base construction method according to a fourth embodiment of the present invention. As shown in fig. 4, in the present embodiment, on the basis of the first method embodiment, the method further includes:
step S405, the classification rule base further comprises the priority of each classification rule, and the priority is determined according to the classification accuracy of the attribute information corresponding to the classification rule to a plurality of second known applications.
In this embodiment, for each classification rule in the classification rule base, the attribute information of the classification rule is matched with the attribute information of each second known application in the plurality of second known applications, the category of the classification rule is used as the classification category of the second known application which is successfully matched, and the number of the second known applications which are successfully matched with the classification rule is obtained. Wherein, successful matching means that the attribute information is consistent.
And comparing the classification category of each second known application successfully matched with the real category to obtain the number of second known applications with correct classification in the second known applications successfully matched with the classification rule. And taking the ratio of the number of second known applications which are classified correctly in the second known applications which are successfully matched with the classification rule to the number of second known applications which are successfully matched with the classification rule as the classification accuracy of the classification rule.
And sequencing the classification rules from large to small according to the classification accuracy. And determining the priority of each classification rule according to the sequence, wherein the priority is high when the sequence is first, and the priority is low when the sequence is not first.
The embodiment of the method determines the priority of each classification rule according to the classification accuracy, and when the number of the matched rules to be classified is at least two, the classification of the classification rule with the highest priority is taken as the classification to be classified, so that the classification accuracy can be improved.
In addition, step S405 is further included on the basis of each of the above method embodiments other than the first and fourth method embodiments, and the resulting method embodiment is also within the scope of the present invention and will not be described in detail herein.
Fig. 5 is a flowchart of a classification rule base construction method according to a fifth embodiment of the present invention. As shown in fig. 5, in the present embodiment, on the basis of the first method embodiment, the method includes:
step S505, the classification rule base is divided into a plurality of classification rule sub-bases, the number of the classification rule sub-bases is the same as the types of the attribute information in the classification rule base, and each classification rule sub-base comprises a classification rule corresponding to a single attribute information;
in this embodiment, the classification rules corresponding to the same type of attribute information are classified into one classification rule sub-library, for example, the classification rule whose attribute information is the name of the application program is classified into one classification rule sub-library. For any attribute information of the applications to be classified, only the classification rule sub-library corresponding to the attribute information is used for classifying the applications to be classified, so that the attribute information of the applications to be classified is prevented from being matched with different types of attribute information in the classification rule sub-library, and the calculation amount is reduced.
Step S507, classifying a plurality of second known applications by utilizing each classification rule sub-library to obtain the classification accuracy of each classification rule sub-library;
in this embodiment, for each classification rule sub-library in the classification rule library, matching the attribute information of the classification rule sub-library with the same-type attribute information of each second known application in a plurality of second known applications, taking the class corresponding to the successfully matched attribute information in the classification rule sub-library as the classification class of the second known application, and obtaining the number of the successfully matched second known applications in the classification rule sub-library. Wherein, successful matching means that the attribute information is consistent.
And comparing the classification category of the second known application with the real category to obtain the number of the second known applications with correct classification in the second known applications successfully matched with the classification rule sub-library. And taking the ratio of the number of second known applications which are classified correctly in the second known applications successfully matched with the classification rule sub-library to the number of second known applications successfully matched with the classification rule sub-library as the classification accuracy of the classification rule sub-library.
Step S509, determining the priority of each classification rule sub-library according to the classification accuracy of each classification rule sub-library.
And sequencing each classification rule sub-library from large to small according to the classification accuracy. And determining the priority of the corresponding sub-library according to the sequence, wherein the priority is high when the sequence is first, and the priority is low when the sequence is not first. When the attribute information to be classified and applied is at least two kinds of attribute information in the preset number of kinds, classifying the application to be classified and applied according to the sequence of the priorities of the classification rule sub-libraries from high to low by adopting the classification rule sub-libraries corresponding to the at least two kinds of attribute information respectively. For example, when the attribute information of the application to be classified is an application program name and a group of market comment words, and the priority of the classification rule sub-library corresponding to the application program name is higher than that of the classification rule sub-library corresponding to the market comment words, the classification application to be classified is firstly classified by using the classification rule sub-library corresponding to the application program name, and if the classification is successful, the classification is stopped; and if the classification cannot be carried out, classifying the application to be classified by adopting a classification rule sub-library corresponding to the market comment words.
The embodiment of the method can reduce the calculated amount in the classification process by dividing the classification rule base into a plurality of classification rule sub-bases, and can improve the classification accuracy by determining the priority of each classification rule sub-base according to the classification accuracy and adopting the classification rule sub-base with higher accuracy when classifying the application to be classified.
In addition, on the basis of each method embodiment except the method embodiments one and five, step S505, step S507 and step S509 are further included, and the formed method embodiment is also within the protection scope of the present invention, and will not be described in detail herein.
On the basis of each of the above method embodiments, the method for constructing the classification rule base may further include:
when a plurality of new known applications appear in each application market, determining the category and preset category attribute information of each new known application; and constructing a new classification rule according to the new known application categories and the attribute information of the preset categories, and adding the new classification rule into a classification rule base.
Among these, the new known applications are applications newly developed by software vendors. The method for determining the attribute information of the category and the preset category of each new known application is the same as the method for determining the attribute information of the category and the preset category of each first known application in step S102 of the method embodiment, and will not be repeated here.
Constructing a new classification rule according to the new class of the known applications and the attribute information of the preset class; comparing each new classification rule with each classification rule in the classification rule base, if the new classification rule is the same as any classification rule in the classification rule base, not processing, if the attribute information of the new classification rule is different from the attribute information of each classification rule in the classification rule base, adding the new classification rule into the classification rule base, and if the attribute information of the new classification rule is the same as the attribute information of any classification rule in the classification rule base and the category of the new classification rule is different from the category of any classification rule, pushing to manually judge whether to add the new classification rule base.
The classification rules are built according to new applications appearing in the application market, and the classification rule base is added, so that the coverage of the application classification rule base is further expanded, and the application market change can be adapted.
Fig. 6 is a flowchart of an application classification method according to a sixth embodiment of the present invention. As shown in fig. 6, in the present embodiment, the method includes:
step S601, matching the attribute information of the application to be classified with the classification rules in the classification rule base, and determining the category of the application to be classified, where the classification rule base is constructed according to the method for constructing the classification rule base described in any one of the first to fourth method embodiments.
In this embodiment, the determination of the category of the application to be classified may be specifically classified into three cases:
in the first case: if the classification rule base is constructed according to the method for constructing the classification rule base described in the first or third embodiment of the method, when the applications to be classified are classified, the matching sequence of the attribute information is determined first, and then the categories of the applications to be classified are determined.
The method for determining the matching sequence of the attribute information comprises the following steps: if only one kind of attribute information is applied to be classified, directly matching the attribute information with each attribute information in a classification rule base; if the attribute information to be classified and applied is of a plurality of types, matching each attribute information with all attribute information in the classification rule base in sequence according to a certain sequence, for example, according to the acquisition sequence of various attribute information in the plurality of attribute information.
The method for determining the category of the application to be classified comprises the following steps: the matching can be stopped once the matching is successful, and the category corresponding to the successfully matched attribute information is used as the category of the application to be classified; the attribute information to be classified and applied can be matched with all attribute information in the classification rule base to obtain classification rules which are successfully matched, if the classification rule is one, the category of the classification rule is used as the category to be classified and applied, and if the classification rule is multiple, the category of any classification rule in the multiple classification rules is used as the category to be classified and applied.
In the second case: if the classification rule base is constructed according to the method for constructing the classification rule base in the second embodiment of the method, if the attribute information of the application to be classified is the market application comment, determining the weight of each effective word in the market application comment to be classified on the application to be classified, and determining the class of the application to be classified according to the weight of each keyword in the market application comment to be classified on the application to be classified and the weight of each class in the classification rule base; otherwise, the classification rule base is adopted as the corresponding classification method constructed according to the classification rule base construction method in the first or third method embodiment, and the category of the application to be classified is determined.
If the attribute information of the applications to be classified is the market application comment, the specific process for determining the classes of the applications to be classified is as follows: and segmenting the market application comments to be classified to obtain a series of words, and taking out the meaningful words to obtain effective words to be classified. Calculating the weight of each valid word of the application to be classified on the application to be classified may be calculated using TF-IDF. The word frequency TF is the frequency of a certain effective word to be classified and applied in the market comment to be classified and applied, and the larger the TF value is, the higher the importance of the certain effective word to be classified and applied in the market comment to be classified and applied is; the inverse document frequency IDF is a measure of the popularity of the certain valid word applied to be classified, if the value is smaller, the certain valid word applied to be classified is considered to be very popular among the market reviews of the application to be classified and the market reviews of the first known applications, and if the value is large, the certain valid word applied to be classified is considered to be rarely appeared in the market reviews of the first known applications other than the application to be classified.
The weight W of the ith valid word applied to be classified on the application to be classifiediThe calculation formula is as follows:
Wi＝TFi*IDFi(1)
in the above formula (1), TFiAnd IDFiThe word frequency and the reverse file frequency of the ith effective word are respectively. In the above formula (2), niFor the number of times the ith valid word appears in the market application review of the application to be classified, njAnd k is the number of the words obtained by segmenting the market application comments to be classified. In the above formula (3), N is the total number of applications to be classified and all first known applications; n isi∈mThe total number of applications containing the ith valid word in the market application review is the applications to be classified and the first known application.
The determination mode of the weight of each effective word to be classified and applied on each category in the classification rule base is as follows: if the market application comment of the first known application contains the effective word to be classified and applied, acquiring the weight of the effective word on each category in the classification rule base from the classification rule base; if not, the weight of the valid word in each category in the classification rule base can be set by itself, for example, to 0.
For each category in the classification rule base, determining the category of the application to be classified according to the weight of each keyword in the market application review of the application to be classified on the application to be classified and the weight of each category in the classification rule base, for example: each valid word in the market application review of the application to be classified, the weight of the valid word on the application to be classified, and the weights of the valid word in the categories of "online shopping", "social communication", and "financial management" in the classification rule base are shown in table 1.
| Valid words to be applied by classification | Applications to be classified | Online shopping | Social interaction of communications | Finance management |
| Shopping | 0.2 | 1 | 0.1 | 0.9 |
| Coupon | 0.8 | 0.4 | 0 | 0.6 |
| Red envelope | 0.5 | 0.6 | 0.5 | 0.5 |
| Payment | 0.5 | 0.9 | 0.5 | 0.4 |
| Chat | 0.1 | 0 | 1 | 0.3 |
TABLE 1
The weight applied to the category "shopping on the internet" to be classified is:
p1＝0.2*1+0.8*0.4+0.5*0.6+0.5*0.9+0.1*0＝1.27；
the weight applied to the category "social communication" to be classified is:
p2＝0.2*0.1+0.8*0+0.5*0.5+0.5*0.5+0.1*1＝0.62；
the weight applied to the category "financial financing" to be classified is:
p3＝0.2*0.9+0.8*0.6+0.5*0.5+0.5*0.4+0.1*0.3＝0.96。
and taking the category 'financial management' with the highest probability as the category to be classified and applied. Preferably, the significant words with higher weight on each category are used for the calculation.
In the third case: if the classification rule base is constructed according to the method for constructing the classification rule base described in the fourth embodiment of the method, when the applications to be classified are classified, the matching sequence of the attribute information is determined first, and then the categories of the applications to be classified are determined.
The method for determining the matching order of the attribute information is consistent with the method for determining the matching order of the attribute information when the classification rule base is constructed according to the method for constructing the classification rule base described in the first or third embodiment of the method.
The method for determining the category of the application to be classified comprises the following steps: and matching the attribute information to be classified and applied with all attribute information in the classification rule base according to the matching sequence of the attribute information to obtain the classification rule successfully matched. And if the classification rule is one, taking the class of the classification rule as the class to be classified and applied, and if the classification rule is multiple, taking the class of the classification rule with the highest priority as the class to be classified and applied according to the priority of the classification rule.
In the sixth embodiment of the method, the application description information and/or the code description information with rich content types is used as the attribute information to construct the classification rule base, so that the classification rules of the classification rule base are rich and diverse, and when the application to be classified is classified, any information in the attribute information with rich content types of the application to be classified is matched with the classification rules in the classification rule base, so that the application to be classified can be classified, the possibility of classifying the application to be classified is improved, and the classification coverage is further improved.
When the classification rule base is constructed by the method for constructing the classification rule base according to the second method embodiment, the step S302 is further included, and the classification rule base is used to determine the class of the application to be classified according to the classification rule base constructed by the method for constructing the classification rule base according to the second method embodiment.
When the classification rule base is constructed by the method for constructing the classification rule base according to the fourth embodiment, the step S302 is further included, and the classification rule base is used to determine the class of the application to be classified according to the classification method constructed according to the fourth embodiment.
If the attribute information is a market application comment, the classification rule base is constructed according to the classification rule base construction method described in the second method embodiment and the corresponding classification method determines the category of the application to be classified when the attribute information is a market application comment; and if the attribute information is not the market application comment, determining the category of the application to be classified according to the classification rule base constructed by the method for constructing the classification rule base in the fourth embodiment of the method by adopting the classification rule base.
Fig. 7 is a flowchart of an application classification method according to a seventh embodiment of the present invention. As shown in fig. 7, in the present embodiment, the method includes:
step S701, matching the attribute information of the application to be classified with the attribute information of the application to be classified of a known class, if the matching is successful, taking the class corresponding to the matched attribute information as the class of the application to be classified, and if the matching is unsuccessful, determining the class of the application to be classified according to the application classification method described in the seventh embodiment of the method; wherein the class of the known class to be classified is determined according to the application classification method described in method embodiment seven.
In this embodiment, the attribute information of the application to be classified is compared with the attribute information of the application to be classified of the known category, and if the attribute information of the application to be classified of the known category is consistent with the attribute information of the application to be classified of the known category, the category of the application to be classified is used as the category of the application to be classified, and the attribute information of the application to be classified does not need to be matched with the classification rule base, so that the classification speed is increased. And if the application types are inconsistent, determining the classes of the applications to be classified by adopting the seventh method embodiment.
Fig. 8 is a flowchart of an application classification method according to an eighth embodiment of the present invention. As shown in fig. 8, in the present embodiment, the method includes:
step S801, sequentially matching the attribute information to be classified and applied with the attribute information in each classification rule sub-library according to the order of the priority of each classification rule sub-library from high to low, and determining the category of the application to be classified, wherein the classification rule library is constructed according to the method for constructing the classification rule library described in the fifth method embodiment.
In this embodiment, the matching sequence of the attribute information and the corresponding classification rule sub-library are determined, and then each attribute information is matched with each attribute information in the classification rule sub-library corresponding to the same kind of attribute information to determine the category of the application to be classified.
The method for determining the matching sequence of the attribute information and the corresponding classification rule sub-library comprises the following steps: and when the applications to be classified are classified, if the attribute information of the applications to be classified is one, comparing the type of the attribute information with the type of the attribute information corresponding to each classification rule sub-library, and determining the classification rule sub-library corresponding to the attribute information. If the attribute information to be classified and applied is a plurality of types, comparing the type of each attribute information with the attribute information type corresponding to each classification rule sub-library, determining the classification rule sub-library corresponding to each attribute information, and taking the sequence of the priority levels of the classification rule sub-libraries corresponding to the plurality of attribute information respectively from high to low as the matching sequence of the plurality of attribute information.
The method for determining the category of the application to be classified by matching each attribute information with each attribute information in the classification rule sub-base corresponding to the same kind of attribute information is similar to the method for determining the category of the application to be classified when the classification rule base in the sixth embodiment is constructed according to the method for constructing the classification rule base in the first or third embodiment, and details are not described herein.
According to the method, the classification rule base comprising the classification rule sub-bases is adopted, when the application to be classified is classified, the attribute information of the application to be classified is matched with the classification rule sub-bases corresponding to the attribute information of the same type according to the type of the attribute information of the application to be classified, so that the matching amount is greatly reduced, and the classification speed is improved; by determining the priority for the classification rule sub-base and classifying the application to be classified by adopting the classification rule sub-base with higher accuracy, the classification accuracy can be improved.
In addition, the classification rule base is constructed according to a classification rule base construction method formed by further including step S505, step S507 and step S509 or further including step S302, step S505, step S507 and step S509 on the basis of the second method embodiment, and the classification of the applications to be classified is determined jointly by the method for determining the matching order of the attribute information and the corresponding classification rule sub-base in the eighth method embodiment and the method for determining the classes of the applications to be classified when the classification rule base in the sixth method embodiment is constructed according to the classification rule base construction method described in the second method embodiment.
In addition, the classification rule base is constructed according to a classification rule base construction method formed by further including step S505, step S507 and step S509 on the basis of the third method embodiment, and the application classification method described in the eighth method embodiment is adopted to determine the category of the application to be classified.
In addition, the classification rule base is constructed according to a classification rule base construction method formed by further including step S505, step S507 and step S509 or further including step S302, step S505, step S507 and step S509 on the basis of the fourth method embodiment, and the classification of the applications to be classified is determined jointly by the method for determining the matching order of the attribute information and the corresponding classification rule sub-base in the eighth method embodiment and the method for determining the classes of the applications to be classified when the classification rule base in the sixth method embodiment is constructed according to the classification rule base construction method described in the fourth method embodiment.
In addition, the classification rule base is constructed according to a classification rule base construction method formed by further including step S205 and step S405 or further including step S205, step S302 and step S405 on the basis of the fifth method embodiment, and the category of the application to be classified is determined by using the method for determining the matching order of the attribute information and the corresponding classification rule sub-base in the eighth method embodiment and the method for determining the category of the application to be classified when the classification rule base is constructed according to the second method embodiment. When the attribute information is not a market application comment, determining the category of the application to be classified by adopting the method for determining the category of the application to be classified when the classification rule base in the sixth method embodiment is constructed according to the method for constructing the classification rule base in the fourth method embodiment.
On the basis of the above embodiment of each application classification method, the application classification method may further include:
and if the rule is not matched, determining the category of the application to be classified according to the similarity between the application to be classified and the application with the known category.
In the embodiment of the invention, if the rule is not matched, similarity calculation can be carried out on the application to be classified and the first known application, and the class of the application with the highest similarity is taken as the class of the application to be tested. When calculating the similarity, any information capable of describing the application may be selected, for example, the code similarity, the icon similarity, and the like of the application may be calculated.
In addition, when the number of the applications to be classified is large, the batch processing system can be adopted to classify the applications to be classified in batches, so that the classification speed is improved. For example, the classification rules in the classification rules library are stored in the hbase database. When a plurality of applications to be classified need to be classified, the function provided by the hbase is utilized to store the attribute information of the applications to be classified in an array, the attribute information of each application to be classified is sequentially matched with the classification rule to obtain a classification result, and the classification result is output once.
Fig. 9 is a schematic diagram of a classification rule base building apparatus according to an embodiment of the present invention. As shown in fig. 9, in the present embodiment, the apparatus includes:
a determining module 101, configured to determine a category and attribute information of each of a plurality of first known applications, where the attribute information includes application description information and/or code description information;
the constructing module 103 is configured to construct a classification rule base, where each classification rule includes attribute information and a category corresponding to the attribute information.
In the first embodiment of the device, the application description information and/or the code description information with rich content types is used as the attribute information to construct the classification rule base, so that the classification rules of the classification rule base are rich and diverse, and when the application to be classified is classified, any information in the attribute information with rich content types of the application to be classified is matched with the classification rules in the classification rule base, so that the application to be classified can be classified, the possibility of classifying the application to be classified is improved, and the classification coverage is further improved.
Fig. 10 is a schematic diagram of a classification rule base building apparatus according to a second embodiment of the present invention. As shown in fig. 10, in the present embodiment, on the basis of the first embodiment of the apparatus, the apparatus includes:
the weight determining module 205 is configured to, if the application description information is a market application comment, further include a weight of each valid word in each category, where the weight of a certain valid word in a certain category is determined according to the number of times that a certain valid word appears in all market application comments of a certain category and the number of categories corresponding to the certain valid word, where the valid word is a keyword included in all market application comments of each category of known applications, and the number of categories corresponding to the certain valid word is the number of categories of known applications in which the certain valid word appears in the market application comment.
In the embodiment of the device, if the application description information is the market application comment, the classification rule base further comprises the weight of each effective word in each category, and when the application to be classified is classified, the category of the application to be classified can be determined according to the weight of each effective word in the market application comment to be classified in the application to be classified and the weight of each effective word in the classification rule base. Due to the fact that the importance of each effective word in the market application comment of the application to be classified to each category in the application to be classified and the classification rule base is considered, the application to be classified is classified more accurately.
Fig. 11 is a schematic diagram of a classification rule base building apparatus according to a third embodiment of the present invention. As shown in fig. 11, in the present embodiment, on the basis of the first embodiment of the apparatus, the apparatus includes:
the category determining module 302 is configured to, if a certain attribute information corresponds to multiple categories, specifically, when a variance between the numbers of the first known applications corresponding to the respective categories is greater than a preset threshold, the category corresponding to the certain attribute information is the category with the largest number of the corresponding first known applications.
The embodiment of the device of the invention improves the classification accuracy of the classification rule corresponding to the attribute information by determining a category with the highest proportion from a plurality of categories of the attribute information as the unique category of the attribute information.
In addition, on the basis of each of the above-mentioned device embodiments other than the first and third device embodiments, a category determination module 302 is further included, and the resulting device embodiment is also within the scope of the present invention and will not be described in detail herein.
Fig. 12 is a schematic diagram of a classification rule base building apparatus according to a fourth embodiment of the present invention. As shown in fig. 12, in this embodiment, on the basis of the first embodiment of the apparatus, the apparatus further includes:
the priority determining module 405 is configured to determine the priority of each classification rule in the classification rule base according to the classification accuracy of the attribute information corresponding to the classification rule to the second known applications.
The embodiment of the device determines the priority of each classification rule according to the classification accuracy, and when the number of the matched rules to be classified is at least two, the classification of the classification rule with the highest priority is taken as the classification to be classified, so that the classification accuracy can be improved.
In addition, on the basis of each of the above-mentioned device embodiments other than the first and fourth device embodiments, a priority determination module 405 is further included, and the resulting device embodiment is also within the scope of the present invention and will not be described in detail herein.
Fig. 13 is a schematic diagram of a classification rule base building apparatus according to a fifth embodiment of the present invention. As shown in fig. 13, in the present embodiment, on the basis of the first embodiment of the apparatus, the apparatus includes:
a dividing module 505, configured to divide the classification rule base into a plurality of classification rule sub-bases, where the number of the classification rule sub-bases is the same as the type of attribute information in the classification rule base, and each classification rule sub-base includes a classification rule corresponding to a single attribute information;
an accuracy obtaining module 507, configured to classify the plurality of second known applications by using each classification rule sub-library to obtain a classification accuracy of each classification rule sub-library;
a priority determining module 509, configured to determine the priority of each classification rule sub-library according to the classification accuracy of each classification rule sub-library.
The embodiment of the device can reduce the calculated amount in the classification process by dividing the classification rule base into a plurality of classification rule sub-bases, and can improve the classification accuracy by determining the priority of each classification rule sub-base according to the classification accuracy and adopting the classification rule sub-base with higher accuracy when classifying the application to be classified.
In addition, on the basis of each of the above device embodiments except for the first and fifth device embodiments, the device further includes a dividing module 505, an accuracy rate obtaining module 507 and a priority determining module 509, and the formed device embodiments are also within the protection scope of the present invention and will not be described in detail herein.
On the basis of each of the above device embodiments, the classification rule base building device may further include:
the adding module is used for determining the category and preset category attribute information of each new known application when a plurality of new known applications appear in each application market; and constructing a new classification rule according to the new known application categories and the attribute information of the preset categories, and adding the new classification rule into a classification rule base.
The classification rules are built according to new applications appearing in the application market, and the classification rule base is added, so that the coverage of the application classification rule base is further expanded, and the application market change can be adapted.
Fig. 14 is a schematic diagram of an application classification apparatus according to a sixth embodiment of the present invention. As shown in fig. 14, in the present embodiment, the apparatus includes:
the classification module 601 is configured to match attribute information of applications to be classified with classification rules in a classification rule base, and determine categories of the applications to be classified, where the classification rule base is constructed according to the method for constructing the classification rule base described in any one of the first to fourth method embodiments.
In the sixth embodiment of the device, the application description information and/or the code description information with rich content types is used as the attribute information to construct the classification rule base, so that the classification rules of the classification rule base are rich and diverse, and when the application to be classified is classified, any information in the attribute information with rich content types of the application to be classified is matched with the classification rules in the classification rule base, so that the application to be classified can be classified, the possibility of classifying the application to be classified is improved, and the classification coverage is further improved.
Fig. 15 is a schematic diagram of an application classification apparatus according to an eighth embodiment of the apparatus of the present invention. As shown in fig. 15, in the present embodiment, the apparatus includes:
the classification module 701 is configured to match the attribute information of the application to be classified with the attribute information of the application to be classified of a known category, if the matching is successful, use the category corresponding to the matched attribute information as the category of the application to be classified, and if the matching is unsuccessful, determine the category of the application to be classified according to the application classification method described in the seventh method embodiment; wherein the class of the known class to be classified is determined according to the application classification method described in method embodiment seven.
The eighth embodiment of the device of the invention compares the attribute information of the application to be classified with the attribute information of the application to be classified of the known class, if the attribute information of the application to be classified is consistent with the attribute information of the application to be classified of the known class, the class of the application to be classified of the known class is used as the class of the application to be classified, and the attribute information of the application to be classified is not required to be matched with the classification rule base, so that the classification speed is improved.
Fig. 16 is a schematic diagram of an application classification apparatus according to a ninth embodiment of the apparatus of the present invention. As shown in fig. 16, in the present embodiment, the apparatus includes:
the classification module 801 is configured to match, in order from high to low, attribute information to be classified and applied with attribute information in each classification rule sub-library in sequence according to the priority of each classification rule sub-library, and determine a category of the to-be-classified application, where the classification rule library is constructed according to the classification rule library construction method described in the fifth method embodiment.
In the ninth embodiment of the device, the classification rule base comprising the plurality of classification rule sub-bases is adopted, when the application to be classified is classified, the attribute information of the application to be classified is matched with the classification rule sub-bases corresponding to the attribute information of the same type according to the type of the attribute information of the application to be classified, so that the matching amount is greatly reduced, and the classification speed is improved; by determining the priority for the classification rule sub-base and classifying the application to be classified by adopting the classification rule sub-base with higher accuracy, the classification accuracy can be improved.
On the basis of the above embodiment of each application classification device, the application classification device may further include:
and the unmatched determining module is used for determining the category of the application to be classified according to the similarity between the application to be classified and the application with the known category if the rule is not matched.
An embodiment of the present invention provides a computer device, including: a processor; and a memory for storing a computer program, wherein the processor is configured to execute the computer program stored in the memory to implement the methods of the first to eighth embodiments of the methods of the present invention.
The embodiment of the invention provides a computer storage medium, wherein a computer program is stored in the computer storage medium, and when being executed by a processor, the computer program realizes the methods of the first to the eighth embodiments of the method of the invention.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for constructing a classification rule base is characterized by comprising the following steps:
determining a category and attribute information of each of a plurality of first known applications, the attribute information including application description information and/or code description information;
and constructing a classification rule base, wherein each classification rule comprises attribute information and a category corresponding to the attribute information.
2. The method of building a classification rule base according to claim 1, further comprising:
if the application description information is market application comments, the classification rule base further comprises the weight of each valid word in each category, and the weight of a certain valid word in a certain category is determined according to the frequency of the certain valid word appearing in all the market application comments of the certain category and the number of the categories corresponding to the certain valid word, wherein the valid word is a keyword contained in all the market application comments of a first known application of any category, and the number of the categories corresponding to the certain valid word is the number of the categories of the first known application in which the certain valid word appears in the market application comments.
3. The method according to claim 1, wherein if a certain attribute information corresponds to a plurality of categories, when a variance between the numbers of the first known applications corresponding to the respective categories is greater than a preset threshold, the category corresponding to the certain attribute information is specifically the category with the largest number of the corresponding first known applications.
4. The method of constructing a classification rule base according to claim 1, 2 or 3,
the classification rule base further comprises the priority of each classification rule, and the priority is determined according to the classification accuracy of the attribute information corresponding to the classification rule to a plurality of second known applications.
5. The method according to any one of claim 4, wherein the application description information includes at least one of application program name, package name, certificate, icon, and market application comment;
the code description information includes at least one of a class name, a method name, built-in sdk, a sensitive string, and a library call relationship.
6. A classification rule base building apparatus, comprising:
a determining module, configured to determine a category and attribute information of each of a plurality of first known applications, where the attribute information includes application description information and/or code description information;
and the construction module is used for constructing a classification rule base, wherein each classification rule comprises attribute information and a category corresponding to the attribute information.
7. An application classification method, comprising:
matching the attribute information of the application to be classified with the classification rules in a classification rule base to determine the class of the application to be classified, wherein the classification rule base is constructed according to the construction method of the classification rule base in any one of claims 1 to 5.
8. An application classification apparatus, comprising:
the classification module is used for matching the attribute information of the application to be classified with the classification rules in the classification rule base to determine the category of the application to be classified, wherein the classification rule base is constructed according to the construction method of the classification rule base in any one of claims 1 to 5.
9. A computer device, comprising:
a processor;
and a memory for storing a computer program, the processor being configured to execute the computer program stored in the memory to implement the classification rule base construction method according to claims 1 to 5 and the application classification method according to claim 7.
10. A computer storage medium, characterized in that a computer program is stored in the computer storage medium, which computer program, when being executed by a processor, implements the classification rule base construction method according to claims 1-5 and the application classification method according to claim 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811108770.2A CN110941714A (en) | 2018-09-21 | 2018-09-21 | Classification rule base construction method, application classification method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811108770.2A CN110941714A (en) | 2018-09-21 | 2018-09-21 | Classification rule base construction method, application classification method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110941714A true CN110941714A (en) | 2020-03-31 |
Family
ID=69905548
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811108770.2A Withdrawn CN110941714A (en) | 2018-09-21 | 2018-09-21 | Classification rule base construction method, application classification method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110941714A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111443900A (en) * | 2020-04-20 | 2020-07-24 | 广州城市信息研究所有限公司 | City design management and control rule base construction method, rule base and rule base calling method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH03235124A (en) * | 1990-02-13 | 1991-10-21 | Mitsubishi Electric Corp | Rule base constructing system |
| CN103186612A (en) * | 2011-12-30 | 2013-07-03 | 中国移动通信集团公司 | Lexical classification method and system and realization method |
| CN104965829A (en) * | 2014-05-14 | 2015-10-07 | 腾讯科技（深圳）有限公司 | Method, server and system for classifying terminal applications |
| CN105630975A (en) * | 2015-12-24 | 2016-06-01 | 联想(北京)有限公司 | Information processing method and electronic device |
| CN105956031A (en) * | 2016-04-25 | 2016-09-21 | 深圳市永兴元科技有限公司 | Text classification method and apparatus |
| CN107463935A (en) * | 2016-06-06 | 2017-12-12 | 工业和信息化部电信研究院 | Application class methods and applications sorter |
-
2018
- 2018-09-21 CN CN201811108770.2A patent/CN110941714A/en not_active Withdrawn
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH03235124A (en) * | 1990-02-13 | 1991-10-21 | Mitsubishi Electric Corp | Rule base constructing system |
| CN103186612A (en) * | 2011-12-30 | 2013-07-03 | 中国移动通信集团公司 | Lexical classification method and system and realization method |
| CN104965829A (en) * | 2014-05-14 | 2015-10-07 | 腾讯科技（深圳）有限公司 | Method, server and system for classifying terminal applications |
| CN105630975A (en) * | 2015-12-24 | 2016-06-01 | 联想(北京)有限公司 | Information processing method and electronic device |
| CN105956031A (en) * | 2016-04-25 | 2016-09-21 | 深圳市永兴元科技有限公司 | Text classification method and apparatus |
| CN107463935A (en) * | 2016-06-06 | 2017-12-12 | 工业和信息化部电信研究院 | Application class methods and applications sorter |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111443900A (en) * | 2020-04-20 | 2020-07-24 | 广州城市信息研究所有限公司 | City design management and control rule base construction method, rule base and rule base calling method |
| CN111443900B (en) * | 2020-04-20 | 2021-05-25 | 广州城市信息研究所有限公司 | City design management and control rule base construction method, rule base and rule base calling method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2022141861A1 (en) | Emotion classification method and apparatus, electronic device, and storage medium | |
| US10019442B2 (en) | Method and system for peer detection | |
| US8290927B2 (en) | Method and apparatus for rating user generated content in search results | |
| CN103679462B (en) | A kind of comment data treating method and apparatus, a kind of searching method and system | |
| EP2866421B1 (en) | Method and apparatus for identifying a same user in multiple social networks | |
| WO2021068610A1 (en) | Resource recommendation method and apparatus, electronic device and storage medium | |
| CN108717407B (en) | Entity vector determination method and device, and information retrieval method and device | |
| US8660901B2 (en) | Matching of advertising sources and keyword sets in online commerce platforms | |
| WO2021159738A1 (en) | Data recommendation method and device based on medical field, and server and storage medium | |
| CN104077723B (en) | A kind of social networks commending system and method | |
| CN111159563A (en) | Method, device and equipment for determining user interest point information and storage medium | |
| CN113407854A (en) | Application recommendation method, device and equipment and computer readable storage medium | |
| CN112560105B (en) | Joint modeling method and device for protecting multi-party data privacy | |
| CN109344232A (en) | A kind of public feelings information search method and terminal device | |
| CN106469163A (en) | A kind of public number recommends method and system | |
| CN110941714A (en) | Classification rule base construction method, application classification method and device | |
| CN115329207B (en) | Intelligent sales information recommendation method and system | |
| CN110941638B (en) | Application classification rule base construction method, application classification method and device | |
| CN109144999B (en) | Data positioning method, device, storage medium and program product | |
| US20220058658A1 (en) | Method of scoring and valuing data for exchange | |
| CN111353793A (en) | CRM (customer relationship management) service recommendation method and device | |
| CN110852078A (en) | Method and device for generating title | |
| US20140324523A1 (en) | Missing String Compensation In Capped Customer Linkage Model | |
| CN115238165A (en) | Information pushing method and device based on machine learning, storage medium and terminal | |
| CN110347905B (en) | Method, device and storage medium for determining information association degree and information recommendation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20200331 |