CN110189171B - Feature data generation method, device and equipment - Google Patents
Feature data generation method, device and equipment Download PDFInfo
- Publication number
- CN110189171B CN110189171B CN201910444856.0A CN201910444856A CN110189171B CN 110189171 B CN110189171 B CN 110189171B CN 201910444856 A CN201910444856 A CN 201910444856A CN 110189171 B CN110189171 B CN 110189171B
- Authority
- CN
- China
- Prior art keywords
- data
- type
- feature
- characteristic
- target
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0242—Determining effectiveness of advertisements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0277—Online advertisement
Abstract
According to the method, the device and the equipment for generating the feature data, provided by the embodiment of the invention, after a generation instruction for indicating generation of the feature data is received, the requirement information corresponding to the generation instruction is obtained; the requirement information is information capable of indicating characteristic requirements of the target characteristic data; the target characteristic data is used for training a click through rate prediction model; extracting a first data identifier of basic data from the acquired demand information, wherein the basic data comprises historical characteristic data; searching basic data corresponding to the first data identification from pre-stored basic data; and constructing target characteristic data based on the searched basic data. The effect of improving the convenience of generating the characteristic data can be achieved.
Description
Technical Field
The invention relates to the field of click through rate prediction, in particular to a method, a device and equipment for generating feature data.
Background
In internet advertisement delivery, the CTR (Click-Through-Rate), i.e. the ratio of the actual number of clicks of the internet advertisement to the advertisement display amount, is an important index for measuring the advertisement delivery effect. Internet advertisements with different characteristics can generate different click through rates, so that in order to reasonably utilize advertisement positions, a preset click through rate prediction model is required to predict the click through rate of the internet advertisements to be delivered. The preset click through rate prediction model is a model obtained by training by utilizing sample characteristic data and a prediction result label of the sample characteristic data in advance, and the characteristics reflected by the sample characteristic data are the same as those of the internet advertisement to be predicted. The characteristics of the internet advertisement can include advertisement types, such as picture advertisement, video advertisement and the like; the release type, such as release time and release location, etc.; and user characteristics such as the area where the user is located and the age of the user, etc.
In the related technology, sample characteristic data are compiled by maintenance personnel according to characteristic requirements, corresponding alternative sample characteristic data are tested, and the alternative sample characteristic data passing the test are determined as sample characteristic data for training a click through rate prediction model.
However, different internet advertisements to be tested have different characteristics, and accordingly, different characteristic requirements exist for sample characteristic data for different internet advertisements to be tested, and the internet advertisements to be tested are often massive, so that the generation mode of the sample characteristic data needs to be manually compiled and tested many times, and the generation of the sample characteristic data is complicated and not convenient.
Disclosure of Invention
The embodiment of the invention aims to provide a method, a device and equipment for generating feature data, so as to achieve the effect of improving the convenience of feature data generation. The specific technical scheme is as follows:
in a first aspect, an embodiment of the present invention provides a method for generating feature data, where the method includes:
after a generation instruction for indicating generation of characteristic data is received, acquiring demand information corresponding to the generation instruction; the requirement information is information capable of indicating characteristic requirements of the target characteristic data; the target characteristic data is used for training a click through rate prediction model;
extracting a first data identifier of basic data from the acquired demand information, wherein the basic data comprises historical characteristic data;
searching basic data corresponding to the first data identification from pre-stored basic data;
and constructing target characteristic data based on the searched basic data.
Optionally, the requirement information further includes: the construction type of the target characteristic data; the construction type is divided according to the splicing form of the basic data and the target characteristic data;
after the obtaining of the demand information corresponding to the generation instruction, the method further includes:
extracting the construction type from the requirement information;
the constructing of the target feature data based on the found basic data comprises:
and constructing the searched basic data into target characteristic data according to the extracted construction type.
Optionally, a plurality of ordered first data identifiers exist in the requirement information;
the constructing the searched basic data into the target feature data according to the splicing form indicated by the construction type includes:
if the construction type is the first type, splicing the searched basic data according to the sequence of the plurality of first data identifiers to obtain target characteristic data; wherein the first type comprises: the splicing form is to splice the multiple basic data according to the sequence of the multiple basic data in the demand information.
Optionally, the generating instruction includes: a second data identification of the target feature data;
after receiving a generation instruction for instructing generation of feature data, acquiring demand information corresponding to the generation instruction, including:
extracting a second data identifier of the target characteristic data from the generation instruction;
and searching the demand information corresponding to the extracted second data identifier from the stored multiple demand information according to the corresponding relation between the preset second data identifier and the demand information.
Optionally, the requirement information further includes a feature type of the basic data; the feature type is a type obtained by dividing according to the type of the feature reflected by the basic data;
after the obtaining of the demand information corresponding to the generation instruction, the method further includes:
extracting the feature type from the requirement information;
searching the feature weight corresponding to the extracted feature type from the corresponding relation between the preset feature type and the feature weight;
the constructing of the target feature data based on the found basic data comprises:
and constructing target characteristic data carrying the searched characteristic weight based on the searched basic data.
In a second aspect, an embodiment of the present invention provides an apparatus for generating feature data, where the apparatus includes:
the system comprises a demand information acquisition module, a characteristic data generation module and a characteristic data generation module, wherein the demand information acquisition module is used for acquiring demand information corresponding to a generation instruction after receiving the generation instruction for indicating generation of characteristic data; the requirement information is information capable of indicating characteristic requirements of the target characteristic data; the target characteristic data is used for training a click through rate prediction model;
the data identification extraction module is used for extracting a first data identification of basic data from the acquired demand information, wherein the basic data comprises historical characteristic data;
the basic data searching module is used for searching basic data corresponding to the first data identifier from pre-stored basic data;
and the target data generation module is used for constructing target characteristic data based on the searched basic data.
Optionally, the requirement information further includes: the construction type of the target characteristic data; the construction type is divided according to the splicing form of the basic data and the target characteristic data;
the device further comprises: the construction type extracting module is used for extracting the construction type from the requirement information after the requirement information acquiring module acquires the requirement information corresponding to the generation instruction;
the target data generation module is specifically configured to:
and constructing the searched basic data into target characteristic data according to the extracted construction type.
Optionally, a plurality of ordered first data identifiers exist in the requirement information;
the target data generation module is specifically configured to:
if the construction type is the first type, splicing the searched basic data according to the sequence of the plurality of first data identifiers to obtain target characteristic data; wherein the first type comprises: the splicing form is to splice the multiple basic data according to the sequence of the multiple basic data in the demand information.
Optionally, the generating instruction includes: a second data identification of the target feature data;
the demand information acquisition module is specifically configured to:
extracting a second data identifier of the target characteristic data from the generation instruction;
and searching the demand information corresponding to the extracted second data identifier from the stored multiple demand information according to the corresponding relation between the preset second data identifier and the demand information.
Optionally, the requirement information further includes a feature type of the basic data; the feature type is a type obtained by dividing according to the type of the feature reflected by the basic data;
the device further comprises: the characteristic type extracting module is used for extracting the characteristic type from the requirement information after the requirement information acquiring module acquires the requirement information corresponding to the generating instruction;
the device further comprises: the characteristic weight searching module is used for searching the characteristic weight corresponding to the extracted characteristic type from the corresponding relation between the preset characteristic type and the characteristic weight;
the target data generation module is specifically configured to:
and constructing target characteristic data carrying the searched characteristic weight based on the searched basic data.
In a third aspect, an embodiment of the present invention provides an electronic device, including:
the system comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete mutual communication through the bus; a memory for storing a computer program; and a processor configured to execute the program stored in the memory to implement the steps of the feature data generation method according to the first aspect.
In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, where a computer program is stored in the storage medium, and when the computer program is executed by a processor, the computer program implements the steps of the feature data generation method provided in the first aspect.
In the scheme provided by the embodiment of the invention, after a generation instruction for indicating generation of characteristic data is received, the requirement information corresponding to the generation instruction is obtained; the requirement information is information capable of indicating the characteristic requirement of the target characteristic data; therefore, the first data identification of the basic data for constructing the target feature data can be extracted from the acquired requirement information; further searching basic data corresponding to the first data identification from the pre-stored basic data; and the basic data comprises historical characteristic data, so that target characteristic data used for training a click through rate prediction model can be directly constructed on the basis of the searched basic data, and automatic generation of the target characteristic data is realized. Therefore, compared with the manual generation of the feature data, the manual compiling and feature testing operation is saved, and the effect of improving the convenience of feature data generation is realized.
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.
Fig. 1 is a schematic flow chart of a method for generating feature data according to an embodiment of the present invention;
fig. 2 is a schematic flow chart of a method for generating feature data according to another embodiment of the present invention;
fig. 3 is a schematic structural diagram of a feature data generation apparatus according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a feature data generation apparatus according to another embodiment of the present invention;
fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make those skilled in the art better understand the technical solution of the present invention, the technical solution 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 only a part of the embodiments of the present invention, and not all of the embodiments. 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.
First, a method for generating feature data according to an embodiment of the present invention will be described below.
The method for generating feature data provided in the embodiment of the present invention may be applied to an electronic device capable of performing data processing, where the device may include a desktop computer, a portable computer, an internet television, an intelligent mobile terminal, a wearable intelligent terminal, a server, and the like, and is not limited herein, and any electronic device that can implement the embodiment of the present invention belongs to the protection scope of the embodiment of the present invention.
As shown in fig. 1, a flow of a method for generating feature data according to an embodiment of the present invention may include:
s101, after receiving a generation instruction for indicating generation of feature data, acquiring demand information corresponding to the generation instruction; the requirement information is information capable of indicating characteristic requirements of the target characteristic data; the target feature data is used for training a click through rate prediction model.
Among them, the source of the generation instruction indicating the generation of the feature data may be various. For example, the generation instruction may be an instruction input by a maintenance person, or may be a trigger signal generated when a training algorithm for training the click through rate prediction model is executed to the step of acquiring the sample feature data.
In order to generate target feature data for training a click through rate prediction model, it is necessary to acquire requirement information that can indicate a feature requirement of the target feature data. The demand information may be various. For example, the requirement information may be a configuration file having, as content, a feature description describing a feature requirement of the target feature data, or may be a feature description itself describing a feature requirement of the target feature data. Or, for example, the requirement information may be a requirement document obtained by storing the feature description input and selected by the maintenance personnel through the presentation interface. Any information that can indicate the feature requirement of the target feature data can be used as the requirement information of the present invention, which is not limited in this embodiment.
Because different internet advertisements to be tested have different characteristic requirements on sample characteristic data, the click through rate prediction models for predicting the different internet advertisements to be tested need to use sample data with different characteristic requirements, and the generation instruction is generated by aiming at the trained click through rate prediction model; therefore, the acquired requirement information needs to correspond to the generation instruction to ensure that the characteristic requirement reflected by the acquired requirement information is suitable for the trained click through rate prediction model. For ease of understanding, the following describes the manner in which the requirement information is obtained in an alternative embodiment.
In an alternative embodiment, generating the instructions may include: a storage location of the demand information;
correspondingly, the acquiring the demand information corresponding to the generation instruction may specifically include: and extracting the storage position of the demand information in the generating instruction, and acquiring the demand information corresponding to the generating instruction according to the acquired storage position.
In a specific application, the requirement information is determined in advance according to the characteristic requirement of the target characteristic data and then stored, and the storage position of the requirement information can be added in the generation instruction. Further, since the storage locations of the demand information corresponding to the different feature demands are different from each other, the demand information stored in the storage location of the generation instruction corresponds to the generation instruction, and the reflected feature demands are applied to the trained click through rate prediction model.
For example, the generation instruction O1 is generated for the click through rate prediction model M1, and the click through rate prediction model M1 is a model for predicting the picture advertisement, so that the generation instruction O1 carries the storage address a1 storing the demand information C1, and the demand information C1 reflects the characteristic demand of the picture advertisement. The generation instruction O2 is generated for the click through rate prediction model M2, and the click through rate prediction model M2 is a model for predicting video advertisements, so that the generation instruction O2 carries the storage address a2 storing the demand information C2, and the demand information C2 reflects the characteristic demand of the video advertisements.
Compared with the mode of acquiring the demand information by using the identification of the demand information, the mode of acquiring the demand information by using the storage position of the demand information can save the step of establishing the corresponding relation between the identification of the demand information and the demand information, and is favorable for improving the acquisition efficiency of the demand information.
In another alternative embodiment, generating the instructions may include: a second data identification of the target feature data; correspondingly, the step of acquiring the requirement information corresponding to the generation instruction after receiving the generation instruction for instructing generation of the feature data may specifically include the following steps:
extracting a second data identifier of the target characteristic data from the generation instruction;
and searching the demand information corresponding to the extracted second data identifier from the stored multiple demand information according to the corresponding relation between the preset second data identifier and the demand information.
Wherein the second data identity may be manifold. Illustratively, the second data identification may be a name of the target feature data, e.g., video ad feature Va1, picture ad feature Pa1, and so on. Alternatively, the second data identity may be a marker symbol of the target characteristic data, such as Va1 and Pa1, for example.
In addition, the extraction manner of the second data identifier may be various. For example, since the generating instruction has a fixed content format, the second data identifier may be extracted from the generating instruction by using a regular expression algorithm, where the regular expression algorithm is an algorithm for constructing a logic according to a preset character string and identifying a character string in the text that meets the construction logic. Or, for example, if the second data identifier has a preset flag character in the generation instruction, the second data identifier may be extracted by recognizing the preset flag character in the generation instruction. Alternatively, for example, if the generation instruction has a fixed content format and content length, the second data identifier is located at a fixed location of the generation instruction, and the second data identifier may be extracted from the fixed location. Any extraction method capable of extracting the second data identifier can be used in the present invention, and this embodiment does not limit this.
The preset correspondence relationship between the second data identifier and the requirement information may be various. For example, the preset correspondence relationship between the second data identifier and the requirement information may be a pointer indicating requirement information corresponding to the second data identifier. Or, for example, the preset correspondence between the second data identifier and the demand information may be a correspondence table between the second data identifier and the demand information, where the second data identifier and the demand information are stored in the correspondence table, or the storage addresses of the second data identifier and the demand information are stored in the correspondence table.
In the two optional embodiments, the storage address of the demand information may be complex and a certain format needs to be followed, for example, C/: folder 1/: folder 2, and the like, which are relatively more concise than the identification of the demand information, so that compared with the method for acquiring the demand information by using the storage location of the demand information, the method for acquiring the demand information by using the identification of the demand information can simplify the generation instruction, thereby improving the convenience of generating the target feature data, and can reduce the storage cost occupied by the generation instruction by using the identification with relatively less data amount.
Any method capable of acquiring the requirement information can be used in the present invention, and the present embodiment does not limit this.
S102, extracting a first data identifier of basic data from the acquired demand information, wherein the basic data comprises historical characteristic data.
The first data identity may be manifold, similar to the second data identity. Illustratively, the first data identification may be a name of the underlying data, e.g., underlying data bd1, or the like. Alternatively, the first data identification may be, for example, a tag of the underlying data, such as bd1 or the like.
The extraction of the first data identity may be manifold, similar to the extraction of the second data identity. Illustratively, if the requirement information has a fixed content format, the first data identification may be extracted from the requirement information using a regular expression algorithm. Or, for example, if the first data identifier has a preset flag character in the requirement information, the first data identifier may be extracted by identifying the preset flag character in the requirement information. Alternatively, for example, if the requirement information has a fixed content format and content length, the first data identifier is located at a fixed position of the requirement information, and the first data identifier may be extracted from the fixed position. Any extraction method capable of extracting the first data identifier can be used in the present invention, and this embodiment does not limit this.
In order to save manual operation steps and realize automation of feature data generation, the basic data for constructing the target feature data can comprise historical feature data which can be directly obtained and does not need to be compiled manually. The source of the historical characteristic data may be various. Illustratively, the source of the historical characteristic data may include at least one of the following: the characteristic data which passes the test in the historical artificially generated characteristic data, the characteristic data which is currently used for training the click through rate prediction model in the artificially generated characteristic data, and the characteristic data which is used by the trained prediction model.
The content of the historical characteristic data may be varied in a particular application. Illustratively, the content of the historical characteristic data may be the characteristic itself, such as "advertisement type: video, release time: late 20:00, user area: beijing, gender of the user: a woman ". Or, for example, the content of the historical feature data may be a feature value obtained by processing a feature, such as a conversion feature "advertisement type: video, release time: late 20:00, user area: beijing, gender of the user: woman ", the resulting vector V1, etc. When the content of the historical characteristic data is the characteristic value, the content of the target characteristic data obtained based on the historical characteristic data, namely the basic data, can be ensured to be the characteristic value, so that the processing operation of the click through rate prediction model needing to be trained on the historical characteristic data is saved, and the training efficiency of the click through rate prediction model is improved. And compared with the characteristic, the characteristic value occupies less storage space, so that the storage cost of the basic data can be saved.
In addition, the underlying data may be stored in a variety of forms. Illustratively, the basic data itself may be directly stored, or the basic data may be converted into a Key-Value form for storage by using a preset Hash algorithm, such as a Hash algorithm. For the basic data stored in the form of Key-Value, the Key can be used as the index of the basic data, so that it is ensured that the corresponding basic data can be directly searched by using the index in the subsequent step S103, the searching efficiency is improved, and the storage cost is reduced. In addition, the above-described method using an index can use the index as the first data identifier of the base data, and can reduce the cost of the storage space compared to the case of using the name of the base data.
S103, searching basic data corresponding to the first data identification from the pre-stored basic data.
Corresponding to the storage form of the basic data in step S102, the search mode of the basic data may be various. For example, if the basic data is directly stored and has a fixed storage area, such as a basic database, the basic data corresponding to the first data identifier may be searched in the basic database. Or, for example, if the basic data is stored in a Key-Value form, the corresponding basic data stored in Value may be obtained by using the first data identifier as an index Key.
Any search mode capable of searching the basic data corresponding to the first data identifier can be used in the present invention, and this embodiment does not limit this.
And S104, constructing target characteristic data based on the searched basic data.
The target feature data may be constructed in various ways. For example, the found basic data may be directly used as the target feature data. Or, for example, a plurality of basic data may be spliced according to a certain splicing relationship between the basic data and the target feature data to obtain the target feature data. The splicing form may include splicing the basic data according to a preset sequence, or splicing the basic data according to any sequence, and so on.
In the scheme provided by the embodiment of the invention, after a generation instruction for indicating generation of characteristic data is received, the requirement information corresponding to the generation instruction is obtained; the requirement information is information capable of indicating the characteristic requirement of the target characteristic data; therefore, the first data identification of the basic data for constructing the target feature data can be extracted from the acquired requirement information; further searching basic data corresponding to the first data identification from the pre-stored basic data; and the basic data comprises historical characteristic data, so that target characteristic data used for training a click through rate prediction model can be directly constructed on the basis of the searched basic data, and automatic generation of the target characteristic data is realized. Therefore, compared with the manual generation of the feature data, the manual compiling and feature testing operation is saved, and the effect of improving the convenience of feature data generation is realized.
Optionally, the requirement information may further include a feature type of the basic data; the feature type is a type obtained by dividing according to the type of the feature reflected by the target feature data;
correspondingly, after the obtaining of the requirement information corresponding to the generation instruction, the method for generating the feature data provided by the embodiment of the present invention may further include the following steps:
extracting the characteristic type of basic data from the demand information;
searching the feature weight corresponding to the extracted feature type from the corresponding relation between the preset feature type and the feature weight;
correspondingly, the constructing the target feature data based on the found basic data may include:
and constructing target characteristic data carrying the searched characteristic weight based on the searched basic data.
In the above alternative embodiments, the characteristic types of the underlying data may be various. Exemplary, feature types may include: advertisement type, impression time type, impression form type, geographic location type, gender type, and the like. By way of example, the advertisement types may include: video, pictures, and text, among others; the release time types may include: gold time periods, holidays, late-night time periods, and the like; the geographical location types may include: first line cities, second line cities, highlands, coastal areas, and the like; gender types may include: women and men.
Similar to the extraction of the second data identifier, the feature type of the basic data is also extracted from the requirement information, and therefore, similarly, the extraction of the feature type of the basic data may be various. For example, if the requirement information has a fixed content format, a regular expression algorithm may be used to extract the feature type of the basic data from the requirement information. Or, for example, if the feature type of the basic data has a preset flag bit character in the requirement information, the feature type of the basic data may be extracted by identifying the preset flag bit character in the requirement information. Alternatively, for example, if the requirement information has a fixed content format and a fixed content length, the feature type of the base data is located at a fixed position of the requirement information, and the feature type of the base data may be extracted from the fixed position. Any extraction method capable of extracting the feature type of the basic data can be used in the present invention, and this embodiment does not limit this.
In addition, the preset corresponding relationship between the feature type and the feature weight may be various. For example, the preset correspondence between the feature type and the feature weight may be a pointer indicating the feature weight corresponding to the feature type. Or, for example, the preset correspondence between the feature type and the feature weight may be a correspondence table between the second feature type and the feature weight, where the feature type and the feature weight are stored in the correspondence table.
In addition, the preset corresponding relationship between the feature type and the feature weight may be set in various ways, and specifically, the degree of contribution of the historical click through rate may be set according to the historical feature type corresponding to the historical click through rate. For example, according to the historical feature type corresponding to the historical click through rate, the degree of contribution to the impression time type is relatively highest, and the degree of contribution to the advertisement type is relatively low, then the preset corresponding relationship between the feature type and the feature weight may be set as: the type of time of delivery corresponds to the characteristic weight "1", and the type of advertisement corresponds to the characteristic weight "0.8".
After the feature weight is obtained, target feature data carrying the found feature weight can be constructed based on the found basic data. When the constructed target characteristic data carrying the searched characteristic weight is used for training the click through rate prediction model, the trained click through rate prediction model can directly use the characteristic weight carried by the target characteristic data, the characteristic weight corresponding to each characteristic does not need to be searched or set, and the training efficiency of the click through rate prediction model can be improved.
In addition, the specific implementation manner of constructing the target feature data carrying the searched feature weight based on the searched basic data may be various. Illustratively, the found basic data can be labeled with corresponding feature weights; and constructing target characteristic data by using the basic data marked with the characteristic weight. Or, for example, the target feature data may be constructed by using the found basic data, and corresponding feature weights may be labeled for the basic data in the constructed target feature data.
As shown in fig. 2, a flow of a method for generating feature data according to another embodiment of the present invention may include:
s201, after receiving a generation instruction for indicating generation of feature data, acquiring demand information corresponding to the generation instruction; the demand information comprises a first data identifier of the basic data and a construction type of the target characteristic data; the construction type is divided according to the splicing form of the basic data and the target characteristic data.
S201 is a similar step to S101 in the embodiment of fig. 1 of the present invention, except that the requirement information in S201 may further include a construction type of the target feature data in addition to the first data identifier, and for the same parts, details are not repeated here, and refer to the description of the embodiment of fig. 1 of the present invention.
The splicing form of the basic data and the target characteristic data can be various. Illustratively, the splicing form may include at least one of the following forms: in a first form: splicing the plurality of basic data according to the sequence of the plurality of basic data in the demand information to obtain target characteristic data; a second form: splicing the plurality of basic data according to the size sequence of the characteristic weights corresponding to the plurality of basic data; a third form: taking the basic data as target characteristic data, and not splicing; and a fourth form: and randomly sequencing the plurality of basic data by using a preset random sequencing algorithm, and splicing the plurality of basic data according to the sequence of the randomly sequenced basic data.
Correspondingly, the construction types obtained by dividing according to the splicing form of the basic data and the target characteristic data can be various. For example, the first type may correspond to a splicing form described in the first form, the second type may correspond to a splicing form described in the second form, the third type may correspond to a splicing form described in the third form, and the fourth type may correspond to a splicing form described in the fourth form.
Also, the build type content may be various. For example, the type identifier may be used as the content of the build type, for example, the type identifier "diff" may be used as the content of the first type, and the type identifier "comm" may be used as the content of the third type. Or, for example, the splicing form itself corresponding to the building type may be used as the content of the building type, for example, the splicing form "splices a plurality of pieces of basic data according to the sequence of the plurality of pieces of basic data in the requirement information, and obtains the target feature data" as the content of the building type. It can be understood that, compared with the case where the splicing form itself corresponding to the construction type is used as the construction type, the case where the type identifier is used as the construction type can simplify the construction type, reduce the cost of the storage space occupied by the required information, and increase the generation efficiency of the feature data by using the data processing pressure of the electronic device for generating the feature data.
In addition, in the splicing form, the multiple basic data are spliced in the order of the demand information, so that the obtained target characteristic data can be ensured to accord with the characteristic expression habit of maintenance personnel, and the effect of meeting the humanized demand is achieved. The feature weights corresponding to the basic data can be determined according to the arrangement sequence of the basic data in the target feature data, the feature weights do not need to be marked in the target feature data, the generation efficiency of the target feature data can be improved, and the cost of a storage space occupied by the target feature data is reduced.
S202, extracting a first data identifier of the basic data and a construction type of the target characteristic data from the acquired demand information.
S202 is a similar step to S102 in the embodiment of fig. 1, except that S202 extracts a construction type in addition to the first data identifier, and for the same parts, details are not repeated here, which is described in the embodiment of fig. 1.
Similar to the first data identification, the extraction manner of the construction type of the target feature data may be various. For example, if the requirement information has a fixed content format, the construction type of the target feature data may be extracted from the requirement information by using a regular expression algorithm. Or, for example, if the construction type of the target feature data has a preset flag bit character in the requirement information, the construction type of the target feature data may be extracted by identifying the preset flag bit character in the requirement information. Alternatively, for example, if the requirement information has a fixed content format and a fixed content length, the construction type of the target feature data is located at a fixed position of the requirement information, and the construction type of the target feature data may be extracted from the fixed position. Any construction type extraction method capable of extracting the target feature data can be used in the present invention, and this embodiment does not limit this.
S203, searching basic data corresponding to the first data identification from the pre-stored basic data.
S203 is the same as S103 in the embodiment of fig. 1, and is not repeated herein, for details, see the description of the embodiment of fig. 1.
And S204, constructing the searched basic data into target characteristic data according to the extracted construction type.
With the above step S204, the construction of the target feature data may be various corresponding to different construction types. For ease of understanding, the following description is in the form of alternative embodiments.
In an optional embodiment, a plurality of ordered first data identifiers exist in the requirement information;
correspondingly, the constructing the searched basic data into the target feature data according to the splicing form indicated by the construction type may specifically include:
if the construction type is the first type, splicing the searched basic data according to the sequence of the plurality of first data identifiers to obtain target characteristic data; wherein the first type includes: the splicing form is to splice the multiple basic data according to the sequence of the multiple basic data in the demand information.
Wherein, the order of the first data identification in the requirement information can be various. For example, the first data id may be arranged in the requirement information in the order "di 2, di 1" in the requirement information, and the first data id may be arranged in the order "di 2" at the first position and "di 1" at the second position. Or, for example, the sequence number of the first data identifier in the requirement information may be, for example, the sequence number of the first data identifier "di 2" is 2, the sequence number of the first data identifier "di 1" is 1, then the first data identifier "di 2" is at the second bit, and the first data identifier "di 1" is at the first bit.
For example, the arrangement order of the first data identifier in the demand information is "di 2, di 1", the basic data corresponding to the first data "di 2" is "golden time period", the basic data corresponding to the first data identifier "di 1" is "video advertisement", the searched basic data are spliced according to the order of the first data identifier, and the obtained target feature data is "golden time period video advertisement".
For the optional implementation mode, the multiple basic data are spliced in the order of the demand information, so that the obtained target characteristic data can be ensured to accord with the characteristic expression habit of maintainers, and the effect of meeting humanized requirements is achieved. And compared with the sequence taking the sequence number of the first data identifier as the first data identifier, the sequence of the first data identifier in the demand information can be a process of marking the sequence number for the first data identifier, and the sequence number does not need to be carried and read, so that the cost of the storage space occupied by the demand information can be saved, and the generation efficiency of the target characteristic data is improved.
Similarly, in another alternative embodiment, if the build type is any one of the second type, the third type, and the fourth type described above, the target feature data may be built in a similar step to that when the build type is the first type described above. The difference lies in that when the construction types are different, the following splicing forms are different, the same parts are not described herein again, and the description of the construction target feature data is described when the construction type is the first type.
In addition, in the embodiment of fig. 2, the type identifier is used as the content of the construction type, and compared with the splicing form itself corresponding to the construction type as the content of the construction type, if the source of the demand information is input by the maintenance personnel through the display interface, the checkable type identifier can be provided in the display interface, and the maintenance personnel can directly check the type identifier of the required construction type without entering relatively more text descriptions, so that the operation can be simplified, and the efficiency can be improved.
Corresponding to the above method embodiment, an embodiment of the present invention further provides a device for generating feature data.
As shown in fig. 3, the structure of the apparatus for generating feature data according to an embodiment of the present invention may include:
a demand information obtaining module 301, configured to obtain demand information corresponding to a generation instruction after receiving the generation instruction instructing to generate feature data; the requirement information is information capable of indicating characteristic requirements of the target characteristic data; the target characteristic data is used for training a click through rate prediction model;
a data identifier extracting module 302, configured to extract a first data identifier of basic data from the acquired demand information, where the basic data includes historical feature data;
a basic data searching module 303, configured to search basic data corresponding to the first data identifier from pre-stored basic data;
and the target data generation module 304 is configured to construct target feature data based on the found basic data.
In the scheme provided by the embodiment of the invention, after a generation instruction for indicating generation of characteristic data is received, the requirement information corresponding to the generation instruction is obtained; the requirement information is information capable of indicating the characteristic requirement of the target characteristic data; therefore, the first data identification of the basic data for constructing the target feature data can be extracted from the acquired requirement information; further searching basic data corresponding to the first data identification from the pre-stored basic data; and the basic data comprises historical characteristic data, so that target characteristic data used for training a click through rate prediction model can be directly constructed on the basis of the searched basic data, and automatic generation of the target characteristic data is realized. Therefore, compared with the manual generation of the feature data, the manual compiling and feature testing operation is saved, and the effect of improving the convenience of feature data generation is realized.
Optionally, the requirement information further includes a feature type of the basic data; the feature type is a type obtained by dividing according to the type of the feature reflected by the basic data;
the device further comprises: a feature type extracting module, configured to extract the feature type from the demand information after the demand information acquiring module 301 acquires the demand information corresponding to the generation instruction;
the device further comprises: the characteristic weight searching module is used for searching the characteristic weight corresponding to the extracted characteristic type from the corresponding relation between the preset characteristic type and the characteristic weight;
the target data generation module 304 is specifically configured to:
and constructing target characteristic data carrying the searched characteristic weight based on the searched basic data.
As shown in fig. 4, in the structure of the apparatus for generating feature data according to another embodiment of the present invention, the requirement information further includes: the construction type of the target characteristic data; the construction type is divided according to the splicing form of the basic data and the target characteristic data; the apparatus may include:
a requirement information obtaining module 401, configured to obtain requirement information corresponding to a generation instruction after receiving the generation instruction instructing to generate feature data; the requirement information is information capable of indicating characteristic requirements of the target characteristic data; the target characteristic data is used for training a click through rate prediction model;
a data identifier extracting module 402, configured to extract a first data identifier of basic data from the acquired demand information, where the basic data includes historical feature data;
a basic data searching module 403, configured to search basic data corresponding to the first data identifier from pre-stored basic data;
a construction type extracting module 404, configured to extract the construction type from the requirement information after the requirement information obtaining module obtains the requirement information corresponding to the generation instruction;
and a target data generating module 405, configured to construct the found basic data into target feature data according to the extracted construction type.
Optionally, a plurality of ordered first data identifiers exist in the requirement information;
the target data generation module 405 is specifically configured to:
if the construction type is the first type, splicing the searched basic data according to the sequence of the plurality of first data identifiers to obtain target characteristic data; wherein the first type comprises: the splicing form is to splice the multiple basic data according to the sequence of the multiple basic data in the demand information.
Optionally, the generating instruction includes: a second data identification of the target feature data;
the requirement information obtaining module 401 is specifically configured to:
extracting a second data identifier of the target characteristic data from the generation instruction;
and searching the demand information corresponding to the extracted second data identifier from the stored multiple demand information according to the corresponding relation between the preset second data identifier and the demand information.
Corresponding to the above embodiment, an embodiment of the present invention further provides an electronic device, as shown in fig. 5, where the electronic device may include:
the system comprises a processor 501, a communication interface 502, a memory 503 and a communication bus 504, wherein the processor 501, the communication interface 502 and the memory complete mutual communication through the communication bus 504 through the 503;
a memory 503 for storing a computer program;
the processor 501 is configured to implement the steps of the method for generating feature data according to any one of the embodiments when executing the computer program stored in the memory 503.
In the scheme provided by the embodiment of the invention, after a generation instruction for indicating generation of characteristic data is received, the requirement information corresponding to the generation instruction is obtained; the requirement information is information capable of indicating the characteristic requirement of the target characteristic data; therefore, the first data identification of the basic data for constructing the target feature data can be extracted from the acquired requirement information; further searching basic data corresponding to the first data identification from the pre-stored basic data; and the basic data comprises historical characteristic data, so that target characteristic data used for training a click through rate prediction model can be constructed directly on the basis of the searched basic data, and automatic generation of the target characteristic data is realized. Therefore, compared with the manual generation of the feature data, the manual compiling and feature testing operation is saved, and the effect of improving the convenience of feature data generation is realized.
The Memory may include a RAM (Random Access Memory) or an NVM (Non-Volatile Memory), such as at least one disk Memory. Optionally, the memory may also be at least one memory device located remotely from the processor.
The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array) or other Programmable logic device, discrete Gate or transistor logic device, discrete hardware component.
The computer-readable storage medium provided by an embodiment of the present invention is included in an electronic device, and a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the method for generating feature data in any of the above embodiments are implemented.
In the scheme provided by the embodiment of the invention, after a generation instruction for indicating generation of characteristic data is received, the requirement information corresponding to the generation instruction is obtained; the requirement information is information capable of indicating the characteristic requirement of the target characteristic data; therefore, the first data identification of the basic data for constructing the target feature data can be extracted from the acquired requirement information; further searching basic data corresponding to the first data identification from the pre-stored basic data; and the basic data comprises historical characteristic data, so that target characteristic data used for training a click through rate prediction model can be constructed directly on the basis of the searched basic data, and automatic generation of the target characteristic data is realized. Therefore, compared with the manual generation of the feature data, the manual compiling and feature testing operation is saved, and the effect of improving the convenience of feature data generation is realized.
In a further embodiment of the present invention, there is also provided a computer program product containing instructions which, when run on a computer, cause the computer to perform the method of generating feature data as described in any of the above embodiments.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in or transmitted from a computer-readable storage medium to another computer-readable storage medium, for example, from a website, computer, server, or data center, over a wired (e.g., coaxial cable, fiber optic, DSL (Digital Subscriber Line), or wireless (e.g., infrared, radio, microwave, etc.) network, to another website, computer, server, or data center, to any available medium that is accessible by a computer or that is a data storage device including one or more integrated servers, data centers, etc. the available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD (Digital Versatile Disc, digital versatile disc)), or a semiconductor medium (e.g.: SSD (Solid State Disk)), etc.
In this document, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, 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 an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the device and electronic apparatus embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and reference may be made to some descriptions of the method embodiments for relevant points.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.
Claims (12)
1. A method for generating feature data, the method comprising:
after a generation instruction for indicating generation of characteristic data is received, acquiring demand information corresponding to the generation instruction; the generating instructions may include: the storage position of the demand information or a second data identifier of the target characteristic data; the requirement information is information capable of indicating characteristic requirements of the target characteristic data; the target characteristic data is used for training a click through rate prediction model;
extracting a first data identifier of basic data from the acquired demand information, wherein the basic data comprises historical characteristic data;
searching basic data corresponding to the first data identification from pre-stored basic data;
and constructing target characteristic data based on the searched basic data.
2. The method of claim 1, wherein the demand information further comprises: the construction type of the target characteristic data; the construction type is divided according to the splicing form of the basic data and the target characteristic data;
after the obtaining of the demand information corresponding to the generation instruction, the method further includes:
extracting the construction type from the requirement information;
the constructing of the target feature data based on the found basic data comprises:
and constructing the searched basic data into target characteristic data according to the extracted construction type.
3. The method of claim 2, wherein there are a plurality of ordered first data identifications in the demand information;
the constructing the searched basic data into the target feature data according to the splicing form indicated by the construction type includes:
if the construction type is the first type, splicing the searched basic data according to the sequence of the plurality of first data identifiers to obtain target characteristic data; wherein the first type comprises: the splicing form is to splice the multiple basic data according to the sequence of the multiple basic data in the demand information.
4. The method of any of claims 1-3, wherein generating the instruction comprises: a second data identification of the target feature data;
after receiving a generation instruction for instructing generation of feature data, acquiring demand information corresponding to the generation instruction, including:
extracting a second data identifier of the target characteristic data from the generation instruction;
and searching the demand information corresponding to the extracted second data identifier from the stored multiple demand information according to the corresponding relation between the preset second data identifier and the demand information.
5. The method of claim 1, wherein the requirements information further includes a characteristic type of the underlying data; the feature type is a type obtained by dividing according to the type of the feature reflected by the basic data;
after the obtaining of the demand information corresponding to the generation instruction, the method further includes:
extracting the feature type from the requirement information;
searching the feature weight corresponding to the extracted feature type from the corresponding relation between the preset feature type and the feature weight;
the constructing of the target feature data based on the found basic data comprises:
and constructing target characteristic data carrying the searched characteristic weight based on the searched basic data.
6. An apparatus for generating feature data, the apparatus comprising:
the system comprises a demand information acquisition module, a characteristic data generation module and a characteristic data generation module, wherein the demand information acquisition module is used for acquiring demand information corresponding to a generation instruction after receiving the generation instruction for indicating generation of characteristic data; the generating instructions may include: the storage position of the demand information or a second data identifier of the target characteristic data; the requirement information is information capable of indicating characteristic requirements of the target characteristic data; the target characteristic data is used for training a click through rate prediction model;
the data identification extraction module is used for extracting a first data identification of basic data from the acquired demand information, wherein the basic data comprises historical characteristic data;
the basic data searching module is used for searching basic data corresponding to the first data identifier from pre-stored basic data;
and the target data generation module is used for constructing target characteristic data based on the searched basic data.
7. The apparatus of claim 6, wherein the demand information further comprises: the construction type of the target characteristic data; the construction type is divided according to the splicing form of the basic data and the target characteristic data;
the device further comprises: the construction type extracting module is used for extracting the construction type from the requirement information after the requirement information acquiring module acquires the requirement information corresponding to the generation instruction;
the target data generation module is specifically configured to:
and constructing the searched basic data into target characteristic data according to the extracted construction type.
8. The apparatus of claim 7, wherein there are a plurality of ordered first data identifications in the requirement information;
the target data generation module is specifically configured to:
if the construction type is the first type, splicing the searched basic data according to the sequence of the plurality of first data identifiers to obtain target characteristic data; wherein the first type comprises: the splicing form is to splice the multiple basic data according to the sequence of the multiple basic data in the demand information.
9. The apparatus of any of claims 6-8, wherein the generating instructions comprise: a second data identification of the target feature data;
the demand information acquisition module is specifically configured to:
extracting a second data identifier of the target characteristic data from the generation instruction;
and searching the demand information corresponding to the extracted second data identifier from the stored multiple demand information according to the corresponding relation between the preset second data identifier and the demand information.
10. The apparatus of claim 6, wherein the requirements information further comprises a feature type of the base data; the feature type is a type obtained by dividing according to the type of the feature reflected by the basic data;
the device further comprises: the characteristic type extracting module is used for extracting the characteristic type from the requirement information after the requirement information acquiring module acquires the requirement information corresponding to the generating instruction;
the device further comprises: the characteristic weight searching module is used for searching the characteristic weight corresponding to the extracted characteristic type from the corresponding relation between the preset characteristic type and the characteristic weight;
the target data generation module is specifically configured to:
and constructing target characteristic data carrying the searched characteristic weight based on the searched basic data.
11. An electronic device is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor and the communication interface are used for realizing the communication between the processor and the memory through the bus; a memory for storing a computer program; a processor for executing a program stored in the memory to perform the method steps of any of claims 1 to 5.
12. A computer-readable storage medium, characterized in that a computer program is stored in the storage medium, which computer program, when being executed by a processor, carries out the method steps of any one of the claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910444856.0A CN110189171B (en) | 2019-05-27 | 2019-05-27 | Feature data generation method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910444856.0A CN110189171B (en) | 2019-05-27 | 2019-05-27 | Feature data generation method, device and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110189171A CN110189171A (en) | 2019-08-30 |
| CN110189171B true CN110189171B (en) | 2021-08-06 |
Family
ID=67717877
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910444856.0A Active CN110189171B (en) | 2019-05-27 | 2019-05-27 | Feature data generation method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110189171B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111144735B (en) * | 2019-12-23 | 2023-08-04 | 北京欧珀通信有限公司 | Data processing method and device, electronic equipment and storage medium |
| CN114098729B (en) * | 2020-08-27 | 2023-11-10 | 中国科学院心理研究所 | Heart interval-based emotion state objective measurement method |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9396277B2 (en) * | 2011-12-09 | 2016-07-19 | Microsoft Technology Licensing, Llc | Access to supplemental data based on identifier derived from corresponding primary application data |
| US9699485B2 (en) * | 2012-08-31 | 2017-07-04 | Facebook, Inc. | Sharing television and video programming through social networking |
| CN105824876A (en) * | 2016-03-01 | 2016-08-03 | 乐视网信息技术（北京）股份有限公司 | Data querying method and device |
| CN108122122A (en) * | 2016-11-29 | 2018-06-05 | 腾讯科技（深圳）有限公司 | Advertisement placement method and system |
| CN108536553A (en) * | 2018-04-17 | 2018-09-14 | 北京星震维度信息技术有限公司 | Backup data store method for determining position, device, equipment and system |
| CN108805332B (en) * | 2018-05-07 | 2022-12-02 | 北京奇艺世纪科技有限公司 | Feature evaluation method and device |
-
2019
- 2019-05-27 CN CN201910444856.0A patent/CN110189171B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN110189171A (en) | 2019-08-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107315824B (en) | Method and device for generating thermodynamic diagram | |
| CN110019616B (en) | POI (Point of interest) situation acquisition method and equipment, storage medium and server thereof | |
| CN103377287A (en) | Method and device for putting in item information | |
| US11244153B2 (en) | Method and apparatus for processing information | |
| CN110674423A (en) | Address positioning method and device, readable storage medium and electronic equipment | |
| CN113220734A (en) | Course recommendation method and device, computer equipment and storage medium | |
| EP3961426A2 (en) | Method and apparatus for recommending document, electronic device and medium | |
| CN113592535B (en) | Advertisement recommendation method and device, electronic equipment and storage medium | |
| CN109767257B (en) | Advertisement putting method and system based on big data analysis and electronic equipment | |
| CN110189171B (en) | Feature data generation method, device and equipment | |
| CN109377284B (en) | Method and electronic equipment for pushing information | |
| CN110059172B (en) | Method and device for recommending answers based on natural language understanding | |
| CN113297287B (en) | Automatic user policy deployment method and device and electronic equipment | |
| CN113590756A (en) | Information sequence generation method and device, terminal equipment and computer readable medium | |
| KR20230150239A (en) | A method and a device for providing recommendation information for affiliated stores | |
| CN111415168B (en) | Transaction alarm method and device | |
| CN114265777B (en) | Application program testing method and device, electronic equipment and storage medium | |
| US20220164377A1 (en) | Method and apparatus for distributing content across platforms, device and storage medium | |
| CN106651408B (en) | Data analysis method and device | |
| CN115080824A (en) | Target word mining method and device, electronic equipment and storage medium | |
| CN109901892A (en) | A kind of method and apparatus of dynamic attribute verifying | |
| CN109308299B (en) | Method and apparatus for searching information | |
| CN111125272B (en) | Regional characteristic acquisition method, regional characteristic acquisition device, computer equipment and medium | |
| CN108268545A (en) | A kind of method and apparatus of the establishment in the user tag library of classification | |
| CN107368597B (en) | Information output method and device |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |