WO2017128997A1 - 业务处理方法、数据处理方法及装置 - Google Patents
业务处理方法、数据处理方法及装置 Download PDFInfo
- Publication number
- WO2017128997A1 WO2017128997A1 PCT/CN2017/071409 CN2017071409W WO2017128997A1 WO 2017128997 A1 WO2017128997 A1 WO 2017128997A1 CN 2017071409 W CN2017071409 W CN 2017071409W WO 2017128997 A1 WO2017128997 A1 WO 2017128997A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- keyword
- news message
- news
- resource category
- message
- 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.)
- Ceased
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
- G06F16/285—Clustering or classification
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06Q30/0251—Targeted advertisements
- G06Q30/0252—Targeted advertisements based on events or environment, e.g. weather or festivals
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
Definitions
- the present application relates to the field of Internet technologies, and in particular, to a service processing method, a data processing method, and an apparatus.
- the existing business process mainly depends on the attribute information of the network resource itself.
- business processes may be affected by outside information.
- sales of some products are often affected by hot news and information. Therefore, the existing business processing methods are relatively simple and the processing effect is not good, so a new business processing method is needed.
- aspects of the present application provide a service processing method, a data processing method, and a device, which are used to provide a new service processing method, improve service processing quality, and enrich service processing methods.
- An aspect of the present application provides a service processing method, including:
- a data processing method including:
- the news message that meets the preset requirement is captured from the network platform
- a service processing method including:
- the news message that meets the preset requirement is captured from the network platform
- a service processing apparatus includes:
- a first determining module configured to determine a target resource category to which the network resource to be processed belongs
- An obtaining module configured to acquire a target news message that matches the target resource category
- a service module configured to perform service processing on the to-be-processed network resource according to the target news message.
- a data processing apparatus including:
- the capture module is configured to capture a news message that meets the preset requirement from the network platform according to a preset crawling period;
- a calculation module configured to calculate a similarity between the news message and each resource category in the resource category library
- a determining module configured to determine a resource category that meets a preset first similarity condition with a similarity between the news message
- a establishing module configured to establish a matching relationship between the news message and the determined resource category.
- a service processing apparatus includes:
- the capture module is configured to capture a news message that meets the preset requirement from the network platform according to a preset crawling period;
- a determining module configured to determine a target resource category that matches the news message
- a service module configured to perform service processing on network resources in the target resource category.
- determining a target resource category to which a network resource to be processed belongs acquiring a target news message that matches the target resource category, performing processing on the resource network to be processed according to the target news message, or capturing a news message, determining and The target resource category matched by the news message is processed according to the target resource category, and a service processing method based on the matching relationship between the news message and the resource category is provided, and the influence of the news message on the business process is fully utilized, and the effect is improved.
- Business processing accuracy while enriching business processing methods.
- FIG. 1 is a schematic flowchart of a service processing method according to an embodiment of the present application
- FIG. 2a is a schematic flowchart of a data processing method according to another embodiment of the present application.
- FIG. 2b and 2c are schematic diagrams of a system architecture for performing the method shown in FIG. 2a according to another embodiment of the present application;
- 2d is a schematic diagram of a matching relationship between a news message and a resource category listed in another embodiment of the present application;
- FIG. 3 is a schematic flowchart of a service processing method according to another embodiment of the present disclosure.
- FIG. 4 is a schematic structural diagram of a service processing apparatus according to another embodiment of the present disclosure.
- FIG. 5 is a schematic structural diagram of a service processing apparatus according to another embodiment of the present disclosure.
- FIG. 6 is a schematic structural diagram of a data processing apparatus according to another embodiment of the present disclosure.
- FIG. 7 is a schematic structural diagram of a data processing apparatus according to another embodiment of the present disclosure.
- FIG. 8 is a schematic structural diagram of a service processing apparatus according to another embodiment of the present disclosure.
- FIG. 1 is a schematic flowchart diagram of a service processing method according to an embodiment of the present application. As shown in Figure 1, The method includes:
- the network resource is processed to perform business processing.
- the embodiment provides a service processing method, which can be executed by a service processing device to implement a new service processing process, improve service processing quality, and enrich service processing manner.
- the inventor of the present application provides a new service processing method, the main principle of which is: performing service processing based on a matching relationship between a resource category and a news message.
- a news message, a service process, and a network resource are involved.
- the network resource involved in the service processing method of the present application is referred to as a network resource to be processed, and the resource category to which the network resource belongs is referred to as a target resource category, and the news message matching the target resource category is referred to as a target news message.
- the embodiment of the present application does not limit the content of the news message, and may include, for example, at least one of a news event, a hot topic, a character dynamic, a product information, and the like; in addition, the implementation format of the news message is not limited, for example, the text may be included. At least one of a picture, a video, and the like.
- the resource category in the embodiment of the present application refers to a category to which the network resource belongs.
- the embodiment of the present application does not limit the type of network resources. In different application scenarios, network resources will be different, and the categories to which network resources belong will vary. for example:
- network resources can be various goods and services provided by sellers.
- resource categories can be categories to which network resources belong, such as womenswear, menswear, shoes, life, learning, sports, outdoor, maternal and child. Wait. It should be noted that the embodiment of the present application does not limit the category level, that is, in the embodiment of the present application, the resource category may include various levels of categories.
- the service processing method in this embodiment specifically includes:
- the resource category to which the network resource to be processed belongs is determined as the target resource category; then, the news message matching the target resource category is acquired as the target news message, and then the network resource to be processed is processed according to the target news message.
- the specific process of processing the network resources for processing the service according to the target news message may be different.
- some of the business scenarios in the e-commerce field are taken as an example for example.
- the process of processing the network resources according to the target news message in other application scenarios may be implemented. .
- the B-type users here refer to users in the B-type trade scene. Such users purchase goods not for their own consumption, but for re-transactions, such as sales or processing.
- the C-type users here refer to users in the C-type trade scene. These users are ordinary consumers, and their purchases are mainly used for their own consumption. Specifically, the category to which the item to be recommended belongs belongs is determined as the target category, and the news message matching the target category is obtained as the target news message, and the recommended item is recommended according to the target news message.
- the recommendation process for the recommended product includes, but is not limited to, the following processing:
- Determining whether the item to be recommended has a recommended value according to the target news message that is, determining whether to recommend the item to be recommended to the user; for example, the category of the anti-fog mask and the related "Chai Jing Publishing" under the dome”
- This hot news match according to the hot news about "Chai Jing published” under the dome, you can determine the anti-fog mask has recommended value, so you can recommend anti-fog masks to users;
- the brand of the product to be recommended ie, which brand of the product is recommended
- the place of origin the product of which place is recommended
- the price range the product in which price range is recommended
- At least one of seller information which seller's products are recommended
- pictures and text information used for recommendation.
- e-commerce platforms provide purchases to users (here, sellers). Decision making is also a relatively major business scenario. Since hot news and information can affect the price and popularity of the goods, the method provided in this embodiment can be used to provide the seller with more accurate purchasing decisions. Specifically, the category to which the various commodities belong is determined as the target category, and the news message matching the target category is acquired as the target news message, and the purchase strategy of the user is generated for each commodity according to the target news message.
- the foregoing procurement policy for generating various users according to the target news message includes but is not limited to the following processing:
- the target news message For each product, according to the target news message, determine whether the product has a purchasing value for the user, that is, whether the user needs to purchase the product; for example, the category of the anti-fog mask and the related "Chai Jing published" dome Under the hot news of the 'speech', according to the hot news about the "Chai Jing published” under the dome, we can confirm that the recent sales of anti-fog masks will rise sharply, and determine the anti-fog masks with purchasing value. Therefore, it is possible to determine the purchase of anti-fog masks;
- At least one of the purchase quantity, the purchase price, the purchase cycle, the purchase merchant, and the like may be determined.
- the embodiment first determines a target resource category to which the network resource to be processed belongs, acquires a target news message that matches the target resource category, and performs processing on the service network according to the target news message to provide a news-based service.
- the business processing method of the matching relationship between the message and the resource category fully utilizes the influence of the news message on the business process, improves the processing precision of the business, and enriches the business processing mode.
- the matching relationship between the resource category and the news message may be established in advance. Based on this, the foregoing step 102, that is, acquiring the target news message that matches the target resource category, is specifically: querying, according to the target resource category, a matching relationship between the pre-established resource category and the news message, to obtain a match with the target resource category. News message as the target news message.
- FIG. 2a is a schematic flowchart of a data processing method according to another embodiment of the present application.
- the data processing method is used to pre-establish a matching relationship between a resource category and a news message.
- the matching relationship between the resource category and the news message may be pre-established by using the method shown in FIG. 2a, and then the matching relationship between the resource category and the news message is queried according to the target resource category, and the target is obtained.
- the target news message that matches the resource category.
- the matching relationship between the resource category and the news message established by the method described in FIG. 2a can be applied to various application scenarios that require the matching relationship, and is not only applicable to the foregoing implementation manner.
- the method includes:
- the news message that meets the preset requirement is captured from the network platform.
- the method flow shown in Figure 2a can be performed using, but not limited to, the system architecture of Figures 2b and 2c.
- the crawling engine shown in FIG. 2b can capture the news message that meets the preset requirement from the network platform according to the preset crawling period; the news message captured by the crawling engine can be stored in the message shown in FIG. 2b.
- a data storage system the data storage system can be implemented by using a mysql relational database, but is not limited thereto.
- the information extraction platform shown in FIG. 2c performs information extraction, and completes the establishment of a matching relationship between the news message and the determined resource category according to the extracted information.
- the capture period in step 201 can be adaptively set according to the application scenario, for example, one day, one week, three days, five days, and the like.
- the requirements are set in advance, and the news that meets the preset requirements is specifically captured. Messages, which can reduce the number of news messages and improve processing efficiency.
- the preset requirement may be that the heat is greater than a specified heat threshold (so that a hot news message may be obtained), or the time of occurrence is later than a specified time (so that a recent news message may be obtained).
- the crawling engine can use crawlers to capture hot news from major news sites (such as Sina.com, whose website is www.sina.com.cn, Sohu.com, whose website is www.sohu.com, etc.). .
- the so-called hot news is also a hot news message, for example, it may be the top news news on major news websites, such as headlines.
- the crawler herein may employ Jsoup directed gripping technology, but is not limited thereto.
- For the captured news message calculating the similarity between the news message and each resource category in the resource category library, and determining the resource category that satisfies the preset first similarity condition with the similarity of the news message as The news message matches the resource category, and then establishes a matching relationship between the news message and the determined resource category.
- the matching relationship between the foregoing news message and the resource category may be stored in the data storage system, but is not limited to the database.
- an implementation manner of the foregoing step 202 includes:
- the similarity between the news message and each resource category is calculated according to the keyword of the news message and the keyword of each resource category.
- the crawling engine can be subdivided into an engine management module, a news crawling module, a comment crawling module, and a data interface module.
- the engine management module is responsible for managing URLs on the network (referred to as URL management) and managing URLs that need to be crawled (referred to as crawl point management).
- News messages are used to describe a fact that the body and title of a news message can express the main meaning of a news message.
- the news capture module can capture the news message, and store the captured news message through the data interface module into the news information table in the data storage system.
- the commentary information of the news message can reflect the concern of the network user (which can be referred to as the netizen). For example, who was in the incident of a female driver in Chengdu? In this news, the full text does not mention the driving recorder. The news itself cannot be used to dig into the driving recorder. However, in the following comments, many people have mentioned the importance of the driving recorder. .
- the comment crawling module can capture the comment information of the news message, and store the captured comment information into the news comment table in the data storage system through the data interface module.
- the information extraction platform may obtain at least one type of information of the text, the title, and the comment information of the news message from the data storage system; and then perform key on at least one of the text, the title, and the comment information of the news message.
- a word extraction process to obtain at least one of a body keyword, a title keyword, and a comment keyword; combining and de-duxing at least one of a body keyword, a title keyword, and a comment keyword to obtain a news The key word of the message.
- the information extraction platform shown in FIG. 2c may be subdivided into a topic word extraction module, a title word segmentation module, and a merge and deduplication module.
- the method for performing keyword extraction processing may be: subject word extraction processing by the keyword extraction module; title of the news message In fact, because of its relatively simple, it can be closed by the title word segmentation module.
- the key word extraction processing may be: word segmentation processing; for the comment information of the news message, because of the large amount of information, the method for extracting the keyword may be: the keyword extraction module may be It performs subject word extraction processing.
- the deduplication processing herein may be implemented by using a clustering algorithm. Specifically, at least one of a text keyword, a title keyword, and a comment keyword is clustered, and a keyword grouped into one type is replaced with one keyword.
- vector keywords can be used to describe these keywords, and these keywords are clustered by a clustering hierarchical clustering algorithm to classify similar keywords into the same category.
- the keywords that can be extracted include “explosion”, “fire”, “firefighter”, “Binhai New Area”, “Hazardous Chemicals”, “Environmental Pollution”, “Dead Injury”, etc.
- the keywords “fire”, “explosion” and “firefighter” are highly similar to the sub-categories of “fire extinguishers”, so it is necessary to group these keywords into one category and then choose one of them.
- the keyword represents all the keywords in the cluster.
- the text, the title and the comment information of the news message can be simultaneously used to obtain the keyword of the news message. Then, a text, a title, and a comment information of the news message are simultaneously used, and the specific implementation manner of obtaining the keyword of the news message includes: extraction processing, filtering processing, and merging and deduplication processing.
- the extraction process refers to: performing subject word extraction processing on the text and comment information of the news message, obtaining the body keyword and the comment keyword, and performing word segmentation processing on the title of the news message to obtain the title keyword.
- the three types of information may be stored in two tables, for example, the text and the title of the news message are stored in the news information table shown in FIG. 2b, and the news is The comment information of the message is stored in the news comment table shown in Figure 2b to facilitate separate processing.
- the TF-IDF model can be used to extract the attention point of the netizen as the comment keyword.
- the female driver in Chengdu was In the news news, a large number of driving recorders appeared in the comments of netizens, and the commentary keyword of the driving recorder can be quickly discovered through the TF-IDF model.
- the filtering process refers to filtering the text keyword, the title keyword, and the comment keyword separately to remove the words such as stop words, person names, place names, and time.
- the words such as stop words, person names, place names, and time.
- the title of the news was divided into the following keywords: “Engineering Fellows”, “Parade”, “Blue”, “After”, The words “Beijing-Tianjin-Hebei”, “contaminant”, “significant” and “rise”, the “post” here is a stop word, which is removed.
- the merging and de-duplication processing refers to combining and de-duplicating the filtered text keywords, title keywords, and comment keywords to obtain keywords of news messages.
- the similarity between the news message and each resource category may be calculated according to the keyword of the news message and the keyword of each resource category.
- An implementation method for calculating the similarity between a news message and each resource category according to a keyword of a news message and a keyword of each resource category includes:
- the similarity between the news message and each resource category is calculated based on the word vector of the keyword of the news message and the word vector of the keyword of each resource category.
- the Word2Vec model can be used to calculate the similarity between news messages and resource categories.
- the Word2Vec model needs to use a corpus.
- the corpus can be composed of a large number of news messages related to network resources, network resources provided by a network resource provider, details thereof, comment information of news messages, resource category information, and the like.
- an implementation manner of calculating a similarity between a news message and each resource category according to a word vector of a keyword of a news message and a word vector of a keyword of each resource category includes:
- n keywords For each of the n keywords, the similarity between the word vector of the keyword and the word vector of each of the m keywords is calculated to obtain n*m similarities;
- the n*m similarities are averaged, and the average of the n*m similarities is used as the similarity between the news message corresponding to the n keywords and the resource category corresponding to the m keywords.
- the similarity between a news message and each resource category can be calculated, and then the resource category that satisfies the preset first similarity condition is selected as the resource category that matches the news message.
- a matching relationship between a news message and a resource category is shown in Figure 2d.
- the left side is the hot news of "Chai Jing's "under the dome” speech
- the right side is the category of anti-fog masks that match the news.
- the step 102 that is, acquiring the target news message that matches the target resource category, is specifically: calculating a similarity between each news message and the target resource category in the news corpus; acquiring the target resource category A news message in which the similarity satisfies the preset second similarity condition as the target news message.
- the foregoing implementation manner of calculating the similarity between each news message in the news corpus and the target resource category includes:
- the title and the comment information of the news message For each news message, according to at least one type of information in the body text, the title and the comment information of the news message, the keyword of the news message is obtained, and the news message is calculated according to the keyword of the news message and the keyword of the target resource category.
- the similarity between target resource categories is obtained.
- the similarity between the news message and the target resource category is calculated according to the keyword of the news message and the keyword of the target resource category, including:
- the similarity between the news message and the target resource category is calculated based on the word vector of the keyword of the news message and the word vector of the keyword of the target resource category.
- the Word2Vec model can be used to calculate the similarity between news messages and target resource categories.
- the number of keywords of the news message is recorded as l, and the key of the target resource category is selected.
- the number of words is denoted by k, where l and k are natural numbers greater than one.
- the l*k similarities are averaged, and the average of the l*k similarities is used as the similarity between the news message corresponding to the l keywords and the target resource category corresponding to the k keywords.
- the similarity between each news message and the target resource category in the news corpus can be calculated, and then the news message satisfying the preset second similarity condition can be selected as the news message matching the target resource category.
- the matching relationship may be stored in the data storage system shown in FIG. 2c, and specifically, the matching may be stored in the data storage system shown in FIG. 2c. In the result list.
- the above method embodiment is to find a matching news message from the perspective of network resources, and then perform service processing on the network resource based on the matched news message.
- the following method embodiment will find a resource category that matches the news message from the perspective of the news message, and then perform business processing on the network resource under the resource category that matches the news message.
- FIG. 3 is a schematic flowchart diagram of a service processing method according to another embodiment of the present disclosure. As shown in FIG. 3, the method includes:
- the news message that meets the preset requirement is captured from the network platform.
- the embodiment provides a service processing method, which can be executed by a service processing device to implement a new service processing process, improve service processing quality, and enrich service processing manner.
- Hot news causing people to pay attention to the air quality around them, thus promoting anti-fog
- the sales volume of such masks and other products increased; the hot news of "Chengdu female drivers were beaten" mentioned the driving recorder, which caused everyone to pay attention to driving recorders, and the related products of driving recorders also ushered in sales peaks.
- the inventor of the present application provides a new service processing method, the main principle of which is: performing service processing based on a matching relationship between a resource category and a news message.
- the resource category matching the news message in the business processing method of the present application is referred to as a target resource category.
- the embodiment of the present application does not limit the content of the news message, and may include, for example, at least one of a news event, a hot topic, a character dynamic, a product information, and the like; in addition, the implementation format of the news message is not limited, for example, the text may be included. At least one of a picture, a video, and the like.
- the resource category in the embodiment of the present application refers to a category to which the network resource belongs.
- the embodiment of the present application does not limit the type of network resources. In different application scenarios, network resources will be different, and the categories to which network resources belong will vary. for example:
- network resources can be various goods and services provided by sellers.
- resource categories can be categories to which network resources belong, such as womenswear, menswear, shoes, life, learning, sports, outdoor, maternal and child. Wait. It should be noted that the embodiment of the present application does not limit the category level, that is, in the embodiment of the present application, the resource category may include various levels of categories.
- the service processing method in this embodiment specifically includes:
- the news message that meets the preset requirement is captured from the network platform.
- the crawling period in the above steps may be adaptively set according to the application scenario, and may be, for example, one day, one week, three days, five days, and the like.
- the requirements are set in advance, and the news that meets the preset requirements is specifically captured. Messages, which can reduce the number of news messages and improve processing efficiency.
- the preset requirement may be that the heat is greater than a specified heat threshold (so that a hot news message may be obtained), or the time of occurrence is later than a specified time (so that a recent news message may be obtained).
- hot news is also a hot news message, for example, it may be the top news news on major news websites, such as headlines.
- the crawler herein may employ Jsoup directed gripping technology, but is not limited thereto.
- the captured news message can be stored in a data storage system, the data storage system It can be implemented by mysql relational database, but it is not limited to this.
- the target resource category that matches the captured news message is determined.
- An alternative implementation of determining the target resource category described above includes:
- a resource category whose similarity with the news message satisfies the preset first similarity condition is determined as the target resource category.
- the foregoing implementation manner of calculating the similarity between the news message and each resource category in the resource category library includes:
- the similarity between the news message and each resource category is calculated according to the keyword of the news message and the keyword of each resource category.
- the foregoing implementation manner of acquiring keywords of the news message according to at least one of the text, the title, and the comment information of the news message includes:
- At least one of a body keyword, a title keyword, and a comment keyword is combined and de-duplicated to obtain a keyword of a news message.
- the foregoing implementation manner of calculating the similarity between the news message and each resource category according to the keyword of the news message and the keyword of each resource category includes:
- the similarity between the news message and each resource category is calculated based on the word vector of the keyword of the news message and the word vector of the keyword of each resource category.
- the target resource class Do After determining the target resource category that matches the captured news message, the target resource class Do not use the network resources for business processing.
- the specific process for performing network processing on network resources under the target resource category may be different.
- some of the business scenarios in the e-commerce field are taken as an example for example.
- the process of performing service processing on network resources under the target resource category in other application scenarios can be implemented on the basis of the following examples. .
- the method provided in this embodiment can be used to recommend the product to the user. Specifically, according to the preset crawling period, the news message that meets the preset requirement is captured from the network platform; the target resource category that matches the news message is determined; and the product under the target resource category is recommended for processing.
- the above recommendation processing on the products under the target resource category includes but is not limited to the following processing:
- the present embodiment determines the target resource category that matches the news message by grabbing the news message, and then performs service processing on the network resource under the target resource category, and provides a matching based on the news message and the resource category.
- the business processing method of the relationship makes full use of the impact of news messages on the business process, improves the processing precision of the business, and enriches the business processing method.
- the method provided by the foregoing embodiment of the present application can be applied to the field of e-commerce, where the network resource can be a commodity on the e-commerce platform, and the category to which the network resource belongs can be a commodity category, and the hot news and the commodity category can be established.
- the matching relationship between the eyes is beneficial for buyers and sellers to grasp the first-hand industry hotspot information, so that buyers and sellers can conduct business processing or decision based on the matching relationship.
- the technical solution of the present application can automatically capture hot news without real intervention, and achieve intelligent, automatic, and high efficiency.
- FIG. 4 is a schematic structural diagram of a service processing apparatus according to another embodiment of the present disclosure. As shown in FIG. 4, the apparatus includes: a first determining module 41, an obtaining module 42, and a service module 43.
- the first determining module 41 is configured to determine a target resource category to which the network resource to be processed belongs.
- the obtaining module 42 is configured to acquire a target news message that matches the target resource category determined by the first determining module 41.
- the service module 43 is configured to perform processing on the network resource to be processed according to the target news message acquired by the obtaining module 42.
- the obtaining module 42 is specifically configured to:
- the matching relationship between the pre-established resource category and the news message is queried to obtain the target news message.
- the apparatus further includes: a capture module 51, a calculation module 52, a second determination module 53, and an establishment module 54.
- the capture module 51 is configured to capture a news message that meets the preset requirement from the network platform according to a preset crawling period.
- the calculating module 52 is configured to calculate a similarity between the news message and each resource category in the resource category library.
- the second determining module 53 is configured to determine a resource category that meets a preset first similarity condition with a similarity between the news message.
- the establishing module 54 is configured to establish a matching relationship between the news message and the resource category determined by the second determining module 53.
- an implementation structure of the calculation module 52 includes an acquisition unit 521, a word segmentation unit 522, and a calculation unit 523.
- the obtaining unit 521 is configured to obtain a keyword of the news message according to at least one type of information of the text, the title, and the comment information of the news message.
- the word segmentation unit 522 is configured to perform word segmentation processing on each resource category to obtain keywords of each resource category.
- the calculating unit 523 is configured to calculate the similarity between the news message and the resource categories according to the keyword of the news message acquired by the obtaining unit 521 and the keyword of each resource category obtained by the word segmentation unit 522.
- the obtaining unit 521 is specifically configured to:
- At least one of a body keyword, a title keyword, and a comment keyword is combined and de-duplicated to obtain a keyword of a news message.
- the calculating unit 523 is specifically configured to:
- the similarity between the news message and each resource category is calculated based on the word vector of the keyword of the news message and the word vector of the keyword of each resource category.
- the obtaining module 42 is specifically configured to:
- a news message that satisfies the preset second similarity condition with the similarity between the target resource categories is acquired as the target news message.
- the obtaining module 42 is specifically used to:
- the title and the comment information of the news message is obtained, and the news message and the target resource are calculated according to the keyword of the news message and the keyword of the target resource category.
- the similarity between categories is
- the obtaining module 42 is configured to: when calculating the similarity between the news message and the target resource category according to the keyword of the news message and the keyword of the target resource category, specifically:
- the similarity between the news message and the target resource category is calculated based on the word vector of the keyword of the news message and the word vector of the keyword of the target resource category.
- the to-be-processed network resource may be an item
- the target resource category may be an item to which the item belongs. Category.
- the matching relationship between the news message and the resource category is specifically a matching relationship between the news message and the product category.
- the service processing module determines a target resource category to which the network resource to be processed belongs, acquires a target news message that matches the target resource category, performs service processing on the processing resource network according to the target news message, and implements the news based on the news.
- Business processing of the matching relationship between messages and resource categories giving full play to the impact of news messages on the business process, improving business processing accuracy, and enriching business processing methods.
- FIG. 6 is a schematic structural diagram of a data processing apparatus according to another embodiment of the present disclosure. As shown in FIG. 6, the apparatus includes a capture module 61, a calculation module 62, a determination module 63, and an establishment module 64.
- the capture module 61 is configured to capture, according to a preset crawling period, a news message that meets a preset requirement from the network platform.
- the calculating module 62 is configured to calculate a similarity between the news message and each resource category in the resource category library.
- the determining module 63 is configured to determine a resource category that meets a preset first similarity condition with a similarity between the news message.
- the establishing module 64 is configured to establish a matching relationship between the news message and the resource category determined by the determining module 63.
- an implementation structure of the computing module 62 includes: an obtaining unit 621, a word segmentation unit 622, and a computing unit 623.
- the obtaining unit 621 is configured to obtain a keyword of the news message according to at least one type of information of the text, the title, and the comment information of the news message.
- the word segmentation unit 622 is configured to perform word segmentation processing on each resource category to obtain keywords of each resource category.
- the calculating unit 623 is configured to calculate a similarity between the news message and each resource category according to the keyword of the news message and the keyword of each resource category.
- the obtaining unit 621 is specifically configured to:
- At least one of a body keyword, a title keyword, and a comment keyword is combined and de-duplicated to obtain a keyword of a news message.
- the calculating unit 623 is specifically configured to:
- the similarity between the news message and each resource category is calculated based on the word vector of the keyword of the news message and the word vector of the keyword of each resource category.
- the resource category here may be the category to which the item belongs.
- the matching relationship between the news message and the resource category is specifically a matching relationship between the news message and the product category.
- the data processing apparatus calculates the similarity between the news message and each resource category by grabbing the news message, determines the resource category matching the news message according to the similarity, and establishes the news message and the determined resource category.
- the matching relationship between the two provides conditions for subsequent network resource-based business processing.
- FIG. 8 is a schematic structural diagram of a service processing apparatus according to another embodiment of the present disclosure. As shown in FIG. 8, the device includes a capture module 81, a determination module 82, and a service module 83.
- the capture module 81 is configured to capture a news message that meets the preset requirement from the network platform according to a preset crawling period.
- the determining module 82 is configured to determine a target resource category that matches the above news message.
- the service module 83 is configured to perform service processing on the network resources in the target resource category.
- an implementation structure of the determination module 82 includes a computing unit and a determining unit.
- a calculating unit configured to calculate a similarity between the foregoing news message and each resource category in the resource category library
- a determining unit configured to determine, as the target resource category, the resource category that meets the preset first similarity condition with the similarity between the above news messages.
- calculation unit is specifically used to:
- the similarity between the news message and each resource category is calculated according to the keyword of the news message and the keyword of each resource category.
- the calculating unit is configured to: when acquiring the keyword of the news message according to at least one type of information in the body text, the title, and the comment information of the news message, specifically:
- At least one of a body keyword, a title keyword, and a comment keyword is combined and de-duplicated to obtain a keyword of a news message.
- the calculation unit when calculating the similarity between the news message and each resource category according to the keyword of the news message and the keyword of each resource category, the calculation unit is specifically used to:
- the similarity between the news message and each resource category is calculated based on the word vector of the keyword of the news message and the word vector of the keyword of each resource category.
- the service processing apparatus determines the target resource category that matches the news message by fetching the news message, and then performs service processing on the network resource under the target resource category, and provides a message based on the news message and the resource category.
- the business processing method of the matching relationship fully utilizes the influence of news messages on the business process, improves the processing precision of the business, and enriches the business processing mode.
- the disclosed system, apparatus, and method may be implemented in other manners.
- the device embodiments described above are merely illustrative.
- the division of the unit is only a logical function division.
- there may be another division manner for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored or not executed.
- the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.
- the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.
- each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
- the above integrated unit can be implemented in the form of hardware or in the form of hardware plus software functional units.
- the above-described integrated unit implemented in the form of a software functional unit can be stored in a computer readable storage medium.
- the software functional unit described above is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor to perform the methods described in various embodiments of the present application. Part of the steps.
- the foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like, which can store program codes. .
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Finance (AREA)
- Strategic Management (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Entrepreneurship & Innovation (AREA)
- General Engineering & Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Economics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Environmental & Geological Engineering (AREA)
- Quality & Reliability (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Information Transfer Between Computers (AREA)
- Communication Control (AREA)
Abstract
本申请提供一种业务处理方法、数据处理方法及装置。业务处理方法包括:确定待处理网络资源所属的目标资源类别;获取与所述目标资源类别相匹配的目标新闻消息;根据所述目标新闻消息,对所述待处理网络资源进行业务处理。本申请提供一种新的业务处理方法,可以提高业务处理质量,丰富业务处理方式。
Description
本申请涉及互联网技术领域,尤其涉及一种业务处理方法、数据处理方法及装置。
随着互联网技术的发展,网络资源越来越多。依赖于网络资源的业务处理也越来越多,例如与网络资源相关的信息推送、网络资源的上传/下载、网络资源的获取、以及网络资源管理等。
在现有业务处理过程中,主要依赖网络资源自身的属性信息。在某些情况下,业务处理过程可能会收到外界信息的影响,例如,在电子商务领域,一些商品的销售量往往会受热点新闻和资讯的影响。因此,现有业务处理方式的比较单一,处理效果不佳,因此需要一种新的业务处理方法。
【发明内容】
本申请的多个方面提供一种业务处理方法、数据处理方法及装置,用以提供一种新的业务处理方法,提高业务处理质量,丰富业务处理方式。
本申请的一方面,提供一种业务处理方法,包括:
确定待处理网络资源所属的目标资源类别;
获取与所述目标资源类别相匹配的目标新闻消息;
根据所述目标新闻消息,对所述待处理网络资源进行业务处理。
本申请的另一方面,提供一种数据处理方法,包括:
按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息;
计算所述新闻消息与资源类别库中各资源类别之间的相似度;
确定与所述新闻消息之间的相似度满足预设第一相似度条件的资源类别;
建立所述新闻消息和所述确定的资源类别之间的匹配关系。
本申请的又一方面,提供一种业务处理方法,包括:
按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息;
确定与所述新闻消息相匹配的目标资源类别;
对所述目标资源类别下的网络资源进行业务处理。
本申请的又一方面,提供一种业务处理装置,包括:
第一确定模块,用于确定待处理网络资源所属的目标资源类别;
获取模块,用于获取与所述目标资源类别相匹配的目标新闻消息;
业务模块,用于根据所述目标新闻消息,对所述待处理网络资源进行业务处理。
本申请的又一方面,提供一种数据处理装置,包括:
抓取模块,用于按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息;
计算模块,用于计算所述新闻消息与资源类别库中各资源类别之间的相似度;
确定模块,用于确定与所述新闻消息之间的相似度满足预设第一相似度条件的资源类别;
建立模块,用于建立所述新闻消息和所述确定的资源类别之间的匹配关系。
本申请的又一方面,提供一种业务处理装置,包括:
抓取模块,用于按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息;
确定模块,用于确定与所述新闻消息相匹配的目标资源类别;
业务模块,用于对所述目标资源类别下的网络资源进行业务处理。
在本申请中,确定待处理网络资源所属的目标资源类别,获取与该目标资源类别相匹配的目标新闻消息,根据目标新闻消息,对待处理资源网络进行业务处理,或者抓取新闻消息,确定与新闻消息相匹配的目标资源类别,根据目标资源类别进行业务处理,提供一种基于新闻消息与资源类别之间的匹配关系的业务处理方法,充分发挥新闻消息对业务处理过程的影响,提高
业务处理精度,同时丰富业务处理方式。
为了更清楚地说明本申请实施例中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作一简单地介绍,显而易见地,下面描述中的附图是本申请的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1为本申请一实施例提供的业务处理方法的流程示意图;
图2a为本申请另一实施例提供的数据处理方法的流程示意图;
图2b和图2c为本申请又一实施例提供的用于执行图2a所示方法的系统架构示意图;
图2d为本申请又一实施例列举的新闻消息与资源类目之间匹配关系的示意图;
图3为本申请又一实施例提供的业务处理方法的流程示意图;
图4为本申请又一实施例提供的业务处理装置的结构示意图;
图5为本申请又一实施例提供的业务处理装置的结构示意图;
图6为本申请又一实施例提供的数据处理装置的结构示意图;
图7为本申请又一实施例提供的数据处理装置的结构示意图;
图8为本申请又一实施例提供的业务处理装置的结构示意图。
为使本申请实施例的目的、技术方案和优点更加清楚,下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
图1为本申请一实施例提供的业务处理方法的流程示意图。如图1所示,
该方法包括:
101、确定待处理网络资源所属的目标资源类别。
102、获取与目标资源类别相匹配的目标新闻消息。
103、根据目标新闻消息,对待处理网络资源进行业务处理。
本实施例提供一种业务处理方法,可由业务处理装置来执行,用以实现新的业务处理流程,提高业务处理质量,丰富业务处理方式。
经过本申请发明人的分析研究,发现新闻消息与网络资源、以及依赖于网络资源的业务有着紧密的联系。本申请发明人最为直接的发现为:在电子商务领域,一些商品的销售量往往会受热点新闻和资讯的影响。举例说明,近期天津爆炸案相关的热点新闻,引发人们去注意消防安全及环境污染,从而促使灭火器、口罩、消毒水等这样的商品销售量上升;有关“柴静发表‘穹顶之下’演说”的热点新闻,引发人们去注意身边的空气质量,从而促使防雾霾口罩等商品销售量上升;“成都女司机被打”的热点新闻中提到了行车记录仪,引起了大家对行车记录仪的重视,行车记录仪相关产品也迎来销售高峰。
基于上述考虑,本申请发明人提供一种新的业务处理方法,其主要原理是:基于资源类别与新闻消息之间的匹配关系进行业务处理。在本申请提供的业务处理方法中,涉及新闻消息、业务处理、以及网络资源。为便于描述,将本申请业务处理方法涉及的网络资源称为待处理网络资源,将网络资源所属的资源类别称为目标资源类别,将与目标资源类别相匹配的新闻消息称为目标新闻消息。
首先说明,本申请实施例不限制新闻消息的内容,例如可以包括新闻事件、热点话题、人物动态、产品资讯等中的至少一种;另外,也不限制新闻消息的实现格式,例如可以包括文本、图片、视频等中的至少一种。
另外,本申请实施例中的资源类别是指网络资源所属的类别。本申请实施例不限定网络资源的类型。在不同应用场景中,网络资源会有所不同,网络资源所属的类别也会有所不同。举例说明:
在电子商务领域,网络资源可以是卖家提供的各种商品、服务等,相应的,资源类别可以是网络资源所属的类目,例如女装、男装、鞋子、生活、学习、运动、户外、母婴等。值得说明的是,本申请实施例并不限制类目等级,也就是说,在本申请实施例中,资源类别可以包括各种等级的类目。
基于上述介绍,本实施例的业务处理方法具体包括:
确定待处理网络资源所属的资源类别作为目标资源类别;然后,获取与目标资源类别相匹配的新闻消息作为目标新闻消息,之后,根据该目标新闻消息,对待处理网络资源进行业务处理。
值得说明的是,根据应用场景的不同,根据目标新闻消息,对待处理网络资源进行业务处理的具体流程也会有所不同。下面以电子商务领域中的部分业务场景为例进行举例说明,对本领域技术人员来说,在下述举例说明的基础上,可以实现其他应用场景中根据目标新闻消息对待处理网络资源进行业务处理的流程。
在电子商务领域中,电商平台向用户(这里的用户可以是B类用户,也可以是C类用户)进行商品推荐也是一种较为常见的业务场景。由于热门新闻和资讯会影响商品价格和热度,所以可以采用本实施例提供的方法可以向用户推荐一些与热门新闻和资讯有关的商品。这里的B类用户是指B类贸易场景中的用户,这类用户购买商品不是用于自己消费,而是用于再次交易,例如售卖或者加工生产。这里的C类用户是指C类贸易场景中的用户,这类用户是一般的消费者,其购买商品主要是用于自身消费。具体的,确定待推荐商品所属的类目作为目标类目,获取与目标类目相匹配的新闻消息作为目标新闻消息,根据该目标新闻消息,对待推荐商品进行推荐处理。
其中,上述根据该目标新闻消息,对待推荐商品进行推荐处理包括但不限于以下处理:
根据该目标新闻消息,确定该待推荐商品是否具有推荐价值,也就是确定是否向用户推荐该待推荐商品;例如,防雾霾口罩所属类目与有关“柴静发表‘穹顶之下’演说”这一热点新闻相匹配,根据有关“柴静发表‘穹顶之下’演说”这一热点新闻,可以确定防雾霾口罩具有推荐价值,因此可以向用户推荐防雾霾口罩;
进一步,若确定向用户推荐该待推荐商品,还可以确定待推荐商品的品牌(即推荐哪些品牌的商品)、产地(推荐哪些产地的商品)、价格区间(推荐位于哪些价格区间的商品)、卖家信息(推荐哪些卖家的商品)、图片以及推荐所使用的文字信息等中的至少一项。
另外,在电子商务领域中,电商平台向用户(这里是指卖家)提供采购
决策也是一种较为主要的业务场景。由于热门新闻和资讯会影响商品价格和热度,所以可以采用本实施例提供的方法可以向卖家提供更加精确的采购决策。具体的,确定各种商品所属的类目作为目标类目,获取与目标类目相匹配的新闻消息作为目标新闻消息,根据该目标新闻消息,对各种商品生成用户的采购策略。
其中,上述根据该目标新闻消息,对各种商品生成用户的采购策略包括但不限于以下处理:
对于每种商品,根据目标新闻消息,确定该商品对用户来说是否具有采购价值,也就是确定用户是否需要采购该商品;例如,防雾霾口罩所属类目与有关“柴静发表‘穹顶之下’演说”这一热点新闻相匹配,根据有关“柴静发表‘穹顶之下’演说”这一热点新闻,可以确定防雾霾口罩近期销量将大幅上升,确定防雾霾口罩具有采购价值,因此可以确定采购防雾霾口罩;
进一步,若确定用户需要采购该商品,还可以确定采购数量、采购价格、采购周期、采购商家等中的至少一项。
由上述可见,本实施例首先确定待处理网络资源所属的目标资源类别,获取与该目标资源类别相匹配的目标新闻消息,根据目标新闻消息,对待处理资源网络进行业务处理,提供一种基于新闻消息与资源类别之间的匹配关系的业务处理方法,充分发挥新闻消息对业务处理过程的影响,提高业务处理精度,同时丰富业务处理方式。
在一可选实施方式中,可以预先建立资源类别与新闻消息之间的匹配关系。基于此,上述步骤102,即获取与目标资源类别相匹配的目标新闻消息具体为:根据目标资源类别,查询预先建立的资源类别与新闻消息之间的匹配关系,以获取与目标资源类别相匹配的新闻消息作为目标新闻消息。
其中,图2a为本申请另一实施例提供的数据处理方法的流程示意图。该数据处理方法用于预先建立资源类别与新闻消息之间的匹配关系。例如,在上述实施方式中,可以采用图2a所示方法预先建立资源类别与新闻消息之间的匹配关系,然后根据目标资源类别,查询资源类别与新闻消息之间的匹配关系,获得与该目标资源类别相匹配的目标新闻消息。值得说明的是,采用图2a所述方法建立的资源类别与新闻消息之间的匹配关系可以应用于各种需要该匹配关系的应用场景中,并不仅仅适用于上述实施方式。如图2a所示,
该方法包括:
201、按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息。
202、计算上述新闻消息与资源类别库中各资源类别之间的相似度。
203、确定与新闻消息之间的相似度满足预设第一相似度条件的资源类别。
204、建立新闻消息和所确定的资源类别之间的匹配关系。
图2a所示方法流程可以采用但不限于图2b和2c系统架构来执行。具体的,图2b所示的抓取引擎可以按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息;抓取引擎抓取到的新闻消息可以存储在图2b所示的数据存储系统,所述数据存储系统可采用mysql关系型数据库实现,但不限于此。图2c所示信息提取平台进行信息提取,并根据提取出的信息完成新闻消息和所确定的资源类别之间的匹配关系的建立。
其中,步骤201中的抓取周期可以根据应用场景适应性设置,例如可以是一天、一周、三天、五天等。另外,考虑到网络平台上新闻消息的数量较多,并且这些新闻消息的价值会随着时间的增长而递减,所以在本实施例中,预先设定要求,具体抓取满足预设要求的新闻消息,这样可以降低新闻消息的数量,提高处理效率。所述预设要求可是热度大于指定热度阈值(这样可以获取热点新闻消息),或者是出现时间晚于指定时间(这样可以获取近期出现的新闻消息)。
例如,抓取引擎可以利用爬虫,将各大新闻网站(例如新浪网,其网址为www.sina.com.cn,搜狐网,其网址为www.sohu.com等)上的热点新闻抓取下来。所谓热点新闻也就是热度比较高的新闻消息,例如具体可能是各大新闻网站上比较靠前的新闻消息,如头条消息等。优选地,这里的爬虫可采用Jsoup定向抓取技术,但不限于此。
对于抓取到的新闻消息,计算该新闻消息与资源类别库中各资源类别之间的相似度,进而确定与该新闻消息之间的相似度满足预设第一相似度条件的资源类别作为与该新闻消息相匹配的资源类别,然后建立该新闻消息与所确定的资源类别之间的匹配关系。
值得说明的是,在每个抓取周期内一般会抓取到多个新闻消息,对每个新闻消息均采用上述方法进行处理。另外,随着抓取周期的递增,可以建立大量新闻消息与资源类别之间的匹配关系。
可选的,可以将上述新闻消息与资源类别之间的匹配关系存储到数据存储系统中,但并不限于数据库。
进一步,上述步骤202的一种实现方式包括:
根据新闻消息的正文、标题和评论信息中的至少一类信息,获取新闻消息的关键词;
对各资源类别分别进行分词处理,以获得各资源类别的关键词;
根据新闻消息的关键词与各资源类别的关键词,计算新闻消息与各资源类别之间的相似度。
进一步,抓取引擎可以细分为引擎管理模块、新闻抓取模块、评论抓取模块和数据接口模块。其中,引擎管理模块负责管理网络上的URL(记为URL管理)以及管理需要抓取的URL(简称为抓取点管理)。
新闻消息用于描述一个事实,新闻消息的正文、标题能够表达出新闻消息的主要含义。具体的,新闻抓取模块可以抓取新闻消息,并通过数据接口模块将抓取到的新闻消息存储到数据存储系统中的新闻信息表中。而新闻消息的评论信息则能体现网络用户(可简称为网友)的关注点。例如,在《成都女司机被打事件谁之过?》这个新闻消息中,全文没有提到行车记录仪,单从新闻消息本身是无法挖掘到行车记录仪这个信息,但是在下面的网友评论中,却有许多人都提到了行车记录仪的重要性。具体的,评论抓取模块可以抓取新闻消息的评论信息,并通过数据接口模块将抓取到评论信息存储到数据存储系统中的新闻评论表中。
基于上述,信息提取平台具体可以从数据存储系统中获取新闻消息的正文、标题和评论信息中的至少一类信息;然后,对新闻消息的正文、标题和评论信息中的至少一类信息进行关键词提取处理,以获得正文关键词、标题关键词和评论关键词中的至少一种;将正文关键词、标题关键词和评论关键词中的至少一种进行合并和去重处理,以获得新闻消息的关键词。
进一步,图2c所示信息提取平台可以细分为主题词提取模块、标题分词模块以及合并和去重模块。
可选的,对新闻消息的正文来说,由于其信息量较多,故对其进行关键词提取处理的方式可以为:可由主题词提取模块对其进行主题词提取处理;对新闻消息的标题来说,由于其相对简单,故可由标题分词模块对其进行关
键词提取处理的方式可以为:对其进行分词处理;对新闻消息的评论信息来说,由于其信息量较多,故对其进行关键词提取处理的方式可以为:可由主题词提取模块对其进行主题词提取处理。
这里之所以要进行去重去处理,是考虑到爬虫从各大新闻网站上抓取热点新闻可能存在重复。例如天津爆炸案,一时间是各大新闻网站的头条,所以很可能会从不同新闻网站上抓取到相同的新闻消息,所以可以出现重复或相似度极高的关键词,因此需要将重复或相似度极高(例如大于一定阈值)的关键词合并为一个。
可选的,这里的去重处理具体可以采用聚类算法实现。具体的,对正文关键词、标题关键词和评论关键词中的至少一种进行聚类处理,将聚为一类的关键词用其中一个关键词代替。例如,可以采用向量空间模型描述这些关键词,采用凝聚层次聚类算法对这些关键词进行聚类,从而将相似关键词归到同一类中。例如,对于天津爆炸案相关的新闻消息,可以提取出来的关键词包括“爆炸”、“火灾”、“消防员”、“滨海新区”、“危化物”、“环境污染”、“死伤”等,其中“火灾”、“爆炸”和“消防员”这几个关键词均与“灭火器”相关子类目的相似度极高,因此需要将这几个关键词聚为一类,然后从中选一个关键词代表该聚类中的所有关键词。
优选的,可以同时采用新闻消息的正文、标题和评论信息,获得新闻消息的关键词。则一种同时采用新闻消息的正文、标题和评论信息,获得新闻消息的关键词的具体实施方式包括:提取处理,过滤处理,以及合并和去重处理。
提取处理是指:对新闻消息的正文和评论信息分别进行主题词提取处理,获得正文关键词和评论关键词,对新闻消息的标题进行分词处理,以获得标题关键词。
可选的,为了便于对这三类信息分别进行处理,可以用两张表对这三类信息进行存储,例如,将新闻消息的正文和标题存储在图2b所示新闻信息表中,将新闻消息的评论信息存储在图2b所示新闻评论表中,以便于分别进行处理。
在对新闻消息的评论信息进行主题词提取处理的过程中,可以采用TF-IDF模型,提取出网友的关注点作为评论关键词。例如,在成都女司机被
打新闻消息中,网友的评论中大量出现行车记录仪,通过TF-IDF模型可以很快挖掘出行车记录仪这个评论关键词。
过滤处理是指:对正文关键词、标题关键词和评论关键词分别进行过滤,以去除其中的停用词、人名、地名、时间等词。例如对《工程院士:“阅兵蓝”后京津冀污染物显著上升》这篇新闻的标题进行分词处理得到的标题关键词包括“工程院士”、“阅兵”、“蓝”、“后”、“京津冀”、“污染物”、“显著”和“上升”等词,这里的“后”属于停用词,将其去除。
合并和去重处理是指:将过滤后的正文关键词、标题关键词和评论关键词进行合并和去重处理,以获得新闻消息的关键词。
在获得新闻消息的关键词之后,可以根据新闻消息的关键词与各资源类别的关键词,计算新闻消息与各资源类别之间的相似度。一种根据新闻消息的关键词与各资源类别的关键词,计算新闻消息与各资源类别之间的相似度的实施方式包括:
获取新闻消息的关键词的词向量和各资源类别的关键词的词向量;
根据新闻消息的关键词的词向量与各资源类别的关键词的词向量,计算新闻消息与各资源类别之间的相似度。
例如,在实际应用中,可以采用Word2Vec模型,计算新闻消息与各资源类别之间的相似度。Word2Vec模型需要使用语料库,在本实施例中,该语料库可由大量与网络资源相关的新闻消息、网络资源提供商提供的网络资源及其详情、新闻消息的评论信息、资源类别信息等组成。
可选的,考虑到新闻消息的关键词可能有多个,每个资源类目的关键词也可能有多个,则将新闻消息的关键词的个数记为n,将资源类目的关键词的个数记为m,其中,n和m是大于1的自然数。基于此,一种根据新闻消息的关键词的词向量与各资源类别的关键词的词向量,计算新闻消息与各资源类别之间的相似度的实施方式包括:
对于n个关键词中的每个关键词,分别计算该关键词的词向量与m个关键词中的每个关键词的词向量之间的相似度,获得n*m个相似度;
对n*m个相似度求平均,将n*m个相似度的平均值作为n个关键词所对应的新闻消息和m个关键词所对应的资源类目之间的相似度。
按照上述方式,可以计算出一新闻消息与各资源类目之间的相似度,进而从中选择满足预设第一相似度条件的资源类目,作为与该新闻消息相匹配的资源类别。以电子商务领域为例,一种新闻消息与资源类目之间的匹配关系如图2d所示。在图2d中,左侧为“柴静发表‘穹顶之下’演说”这一热点新闻,右侧是与该新闻消息相匹配的防雾霾口罩所属类目。
在另一可选实施方式中,上述步骤102,即获取与目标资源类别相匹配的目标新闻消息具体为:计算新闻语料库中各新闻消息与目标资源类别之间的相似度;获取与目标资源类别之间的相似度满足预设第二相似度条件的新闻消息作为目标新闻消息。
其中,上述计算新闻语料库中各新闻消息与所述目标资源类别之间的相似度的一种实施方式包括:
对目标资源类别进行分词处理,以获得目标资源类别的关键词;
对于每个新闻消息,根据该新闻消息的正文、标题和评论信息中的至少一类信息,获取该新闻消息的关键词,根据新闻消息的关键词与目标资源类别的关键词,计算新闻消息与目标资源类别之间的相似度。
值得说明的是,该实施方式与上述步骤202的实施方式相类似,其各步骤的具体实现可参见上述步骤202的具体实施方式中的相应描述,在此不再赘述。
相应的,上述根据新闻消息的关键词与目标资源类别的关键词,计算新闻消息与目标资源类别之间的相似度包括:
获取新闻消息的关键词的词向量和目标资源类别的关键词的词向量;
根据新闻消息的关键词的词向量与目标资源类别的关键词的词向量,计算新闻消息和目标资源类别之间的相似度。
例如,在实际应用中,可以采用Word2Vec模型,计算新闻消息与目标资源类别之间的相似度。
可选的,考虑到新闻消息的关键词可能有多个,目标资源类目的关键词也可能有多个,则将新闻消息的关键词的个数记为l,将目标资源类目的关键词的个数记为k,其中,l和k是大于1的自然数。基于此,一种根据新闻
消息的关键词的词向量与目标资源类别的关键词的词向量,计算新闻消息与各资源类别之间的相似度的实施方式包括:
对于l个关键词中的每个关键词,分别计算该关键词的词向量与k个关键词中的每个关键词的词向量之间的相似度,获得l*k个相似度;
对l*k个相似度求平均,将l*k个相似度的平均值作为l个关键词所对应的新闻消息和k个关键词所对应的目标资源类目之间的相似度。
按照上述方式,可以计算出新闻语料库中各新闻消息与目标资源类目之间的相似度,进而从中选择满足预设第二相似度条件的新闻消息,作为与目标资源类目匹配的新闻消息。
进一步可选的,在获得新闻消息与资源类别之间的匹配关系之后,可以将该匹配关系存储到图2c所示的数据存储系统中,具体可以存储到图2c所示数据存储系统中的匹配结果列表中。
以上方法实施例是从网络资源的角度出发,找到相匹配的新闻消息,进而基于相匹配的新闻消息对网络资源进行业务处理。以下方法实施例将从新闻消息的角度出发,找到与新闻消息相匹配的资源类别,然后对与所述新闻消息相匹配的资源类别下的网络资源进行业务处理。
图3为本申请又一实施例提供的业务处理方法的流程示意图。如图3所示,该方法包括:
301、按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息。
302、确定与新闻消息相匹配的目标资源类别。
303、对目标资源类别下的网络资源进行业务处理。
本实施例提供一种业务处理方法,可由业务处理装置来执行,用以实现新的业务处理流程,提高业务处理质量,丰富业务处理方式。
经过本申请发明人的分析研究,发现新闻消息与网络资源、以及依赖于网络资源的业务有着紧密的联系。本申请发明人最为直接的发现为:在电子商务领域,一些商品的销售量往往会受热点新闻和资讯的影响。举例说明,近期天津爆炸案相关的热点新闻,引发人们去注意消防安全及环境污染,从而促使灭火器、口罩、消毒水等这样的商品销售量上升;有关“柴静发表‘穹顶之下’演说”的热点新闻,引发人们去注意身边的空气质量,从而促使防雾
霾口罩等商品销售量上升;“成都女司机被打”的热点新闻中提到了行车记录仪,引起了大家对行车记录仪的重视,行车记录仪相关产品也迎来销售高峰。
基于上述考虑,本申请发明人提供一种新的业务处理方法,其主要原理是:基于资源类别与新闻消息之间的匹配关系进行业务处理。为便于描述,将本申请业务处理方法中与新闻消息相匹配的资源类别称为目标资源类别。
首先说明,本申请实施例不限制新闻消息的内容,例如可以包括新闻事件、热点话题、人物动态、产品资讯等中的至少一种;另外,也不限制新闻消息的实现格式,例如可以包括文本、图片、视频等中的至少一种。
另外,本申请实施例中的资源类别是指网络资源所属的类别。本申请实施例不限定网络资源的类型。在不同应用场景中,网络资源会有所不同,网络资源所属的类别也会有所不同。举例说明:
在电子商务领域,网络资源可以是卖家提供的各种商品、服务等,相应的,资源类别可以是网络资源所属的类目,例如女装、男装、鞋子、生活、学习、运动、户外、母婴等。值得说明的是,本申请实施例并不限制类目等级,也就是说,在本申请实施例中,资源类别可以包括各种等级的类目。
基于上述介绍,本实施例的业务处理方法具体包括:
首先,按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息。
上述步骤中的抓取周期可以根据应用场景适应性设置,例如可以是一天、一周、三天、五天等。另外,考虑到网络平台上新闻消息的数量较多,并且这些新闻消息的价值会随着时间的增长而递减,所以在本实施例中,预先设定要求,具体抓取满足预设要求的新闻消息,这样可以降低新闻消息的数量,提高处理效率。所述预设要求可是热度大于指定热度阈值(这样可以获取热点新闻消息),或者是出现时间晚于指定时间(这样可以获取近期出现的新闻消息)。
例如,可以利用爬虫,将各大新闻网站(例如新浪网,其网址为www.sina.com.cn,搜狐网,其网址为www.sohu.com等)上的热点新闻抓取下来。所谓热点新闻也就是热度比较高的新闻消息,例如具体可能是各大新闻网站上比较靠前的新闻消息,如头条消息等。优选地,这里的爬虫可采用Jsoup定向抓取技术,但不限于此。
另外,抓取到的新闻消息可以存储在数据存储系统,所述数据存储系统
可采用mysql关系型数据库实现,但不限于此。
在抓取到新闻消息之后,确定与抓取到的新闻消息相匹配的目标资源类别。
上述确定目标资源类别的一种可选实施方式包括:
计算新闻消息与资源类别库中各资源类别之间的相似度;
确定与新闻消息之间的相似度满足预设第一相似度条件的资源类别作为目标资源类别。
进一步,上述计算新闻消息与资源类别库中各资源类别之间的相似度的实施方式包括:
根据新闻消息的正文、标题和评论信息中的至少一类信息,获取新闻消息的关键词;
对各资源类别分别进行分词处理,以获得各资源类别的关键词;
根据新闻消息的关键词与各资源类别的关键词,计算新闻消息与各资源类别之间的相似度。
更进一步,上述根据新闻消息的正文、标题和评论信息中的至少一类信息,获取新闻消息的关键词的实施方式包括:
对新闻消息的正文、标题和评论消息中的至少一类信息进行关键词提取处理,以获得正文关键词、标题关键词和评论关键词中的至少一种;
将正文关键词、标题关键词和评论关键词中的至少一种进行合并和去重处理,以获得新闻消息的关键词。
更进一步,上述根据新闻消息的关键词与各资源类别的关键词,计算新闻消息与各资源类别之间的相似度的实施方式包括:
获取新闻消息的关键词的词向量和各资源类别的关键词的词向量;
根据新闻消息的关键词的词向量与各资源类别的关键词的词向量,计算新闻消息与各资源类别之间的相似度。
在此说明,上述各步骤的具体实施方式可参见图2a所示实施例中相应步骤的描述,在此不再赘述。
在确定与抓取到的新闻消息相匹配的目标资源类别之后,对目标资源类
别下的网络资源进行业务处理。
值得说明的是,根据应用场景的不同,对目标资源类别下的网络资源进行业务处理的具体流程也会有所不同。下面以电子商务领域中的部分业务场景为例进行举例说明,对本领域技术人员来说,在下述举例说明的基础上,可以实现其他应用场景中对目标资源类别下的网络资源进行业务处理的流程。
在电子商务领域中,电商平台向用户推荐商品是一种较为常见的业务场景。由于热门新闻和资讯会影响商品价格和热度,所以可以采用本实施例提供的方法向用户推荐商品。具体的,按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息;确定与新闻消息相匹配的目标资源类目;对目标资源类目下的商品进行推荐处理。
其中,上述对目标资源类目下的商品进行推荐处理包括但不限于以下处理:
确定目标资源类目下的商品是否具有推荐价值,以及在确定该商品具有推荐价值需要进行推荐时的推荐力度以及推荐方式等。
由上述可见,本实施例通过抓取新闻消息,确定与新闻消息相匹配的目标资源类别,进而对目标资源类别下的网络资源进行业务处理,提供一种基于新闻消息与资源类别之间的匹配关系的业务处理方法,充分发挥新闻消息对业务处理过程的影响,提高业务处理精度,同时丰富业务处理方式。
在此说明,本申请上述实施例提供的方法可以应用于电子商务领域,则网络资源可以是电商平台上的商品、网络资源所属的类别可以是商品类目,可以建立起热门新闻与商品类目之间的匹配关系,有利于让买家和卖家掌握一手的行业热点资讯,便于买家和卖家基于该匹配关系进行业务处理或决策。另外,本申请技术方案在实现上无需人工干预,能够自动捕捉热点新闻,达到智能化、自动化,效率较高。
需要说明的是,对于前述的各方法实施例,为了简单描述,故将其都表述为一系列的动作组合,但是本领域技术人员应该知悉,本申请并不受所描述的动作顺序的限制,因为依据本申请,某些步骤可以采用其他顺序或者同时进行。其次,本领域技术人员也应该知悉,说明书中所描述的实施例均属于优选实施例,所涉及的动作和模块并不一定是本申请所必须的。
在上述实施例中,对各个实施例的描述都各有侧重,某个实施例中没有详述的部分,可以参见其他实施例的相关描述。
图4为本申请又一实施例提供的业务处理装置的结构示意图。如图4所示,该装置包括:第一确定模块41、获取模块42和业务模块43。
第一确定模块41,用于确定待处理网络资源所属的目标资源类别。
获取模块42,用于获取与第一确定模块41所确定的目标资源类别相匹配的目标新闻消息。
业务模块43,用于根据获取模块42获取的目标新闻消息,对待处理网络资源进行业务处理。
在一可选实施方式中,获取模块42具体可用于:
根据目标资源类别,查询预先建立的资源类别与新闻消息之间的匹配关系,以获取目标新闻消息。
在一可选实施方式中,如图5所示,该装置还包括:抓取模块51、计算模块52、第二确定模块53和建立模块54。
抓取模块51,用于按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息。
计算模块52,用于计算新闻消息与资源类别库中各资源类别之间的相似度。
第二确定模块53,用于确定与新闻消息之间的相似度满足预设第一相似度条件的资源类别。
建立模块54,用于建立新闻消息和第二确定模块53所确定的资源类别之间的匹配关系。
进一步,如图5所示,计算模块52的一种实现结构包括:获取单元521、分词单元522和计算单元523。
获取单元521,用于根据新闻消息的正文、标题和评论信息中的至少一类信息,获取新闻消息的关键词。
分词单元522,用于对各资源类别分别进行分词处理,以获得各资源类别的关键词。
计算单元523,用于根据获取单元521获取的新闻消息的关键词与分词单元522获得的各资源类别的关键词,计算新闻消息与所述各资源类别之间的相似度。
进一步,获取单元521具体用于:
对新闻消息的正文、标题和评论消息中的至少一类信息进行关键词提取处理,以获得正文关键词、标题关键词和评论关键词中的至少一种;
将正文关键词、标题关键词和评论关键词中的至少一种进行合并和去重处理,以获得新闻消息的关键词。
进一步,计算单元523具体用于:
获取新闻消息的关键词的词向量和各资源类别的关键词的词向量;
根据新闻消息的关键词的词向量与各资源类别的关键词的词向量,计算新闻消息与各资源类别之间的相似度。
在一可选实施方式中,获取模块42具体用于:
计算新闻语料库中各新闻消息与目标资源类别之间的相似度;
获取与目标资源类别之间的相似度满足预设第二相似度条件的新闻消息作为目标新闻消息。
进一步地,获取模块42在计算新闻语料库中各新闻消息与目标资源类别之间的相似度时,具体用于:
对目标资源类别进行分词处理,以获得目标资源类别的关键词;
对于每个新闻消息,根据新闻消息的正文、标题和评论信息中的至少一类信息,获取新闻消息的关键词,根据新闻消息的关键词与目标资源类别的关键词,计算新闻消息与目标资源类别之间的相似度。
更进一步地,获取模块42在根据新闻消息的关键词与目标资源类别的关键词,计算新闻消息与目标资源类别之间的相似度时,具体用于:
获取新闻消息的关键词的词向量和目标资源类别的关键词的词向量;
根据新闻消息的关键词的词向量与目标资源类别的关键词的词向量,计算新闻消息和目标资源类别之间的相似度。
可选的,待处理网络资源可以是商品,目标资源类别可以是商品所属的
类目。相应的,新闻消息和资源类别之间的匹配关系具体为新闻消息与商品类目之间的匹配关系。
本实施例提供的业务处理模块,确定待处理网络资源所属的目标资源类别,获取与该目标资源类别相匹配的目标新闻消息,根据目标新闻消息,对待处理资源网络进行业务处理,实现了基于新闻消息与资源类别之间的匹配关系的业务处理,充分发挥新闻消息对业务处理过程的影响,提高业务处理精度,同时丰富业务处理方式。
图6为本申请又一实施例提供的数据处理装置的结构示意图。如图6所示,该装置包括:抓取模块61、计算模块62、确定模块63和建立模块64。
抓取模块61,用于按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息。
计算模块62,用于计算新闻消息与资源类别库中各资源类别之间的相似度。
确定模块63,用于确定与新闻消息之间的相似度满足预设第一相似度条件的资源类别。
建立模块64,用于建立新闻消息和确定模块63所确定的资源类别之间的匹配关系。
在一可选实施方式中,如图7所示,计算模块62的一种实现结构包括:获取单元621、分词单元622和计算单元623。
获取单元621,用于根据新闻消息的正文、标题和评论信息中的至少一类信息,获取新闻消息的关键词。
分词单元622,用于对各资源类别分别进行分词处理,以获得各资源类别的关键词。
计算单元623,用于根据新闻消息的关键词与各资源类别的关键词,计算新闻消息与各资源类别之间的相似度。
进一步地,获取单元621具体用于:
对新闻消息的正文、标题和评论消息中的至少一类信息进行关键词提取处理,以获得正文关键词、标题关键词和评论关键词中的至少一种;
将正文关键词、标题关键词和评论关键词中的至少一种进行合并和去重处理,以获得新闻消息的关键词。
更进一步地,计算单元623具体用于:
获取新闻消息的关键词的词向量和各资源类别的关键词的词向量;
根据新闻消息的关键词的词向量与各资源类别的关键词的词向量,计算新闻消息与各资源类别之间的相似度。
可选的,这里的资源类别可以是商品所属的类目。相应的,新闻消息和资源类别之间的匹配关系具体为新闻消息与商品类目之间的匹配关系。
本实施例提供的数据处理装置,通过抓取新闻消息,计算新闻消息与各资源类别之间的相似度,根据相似度确定与新闻消息相匹配的资源类别,建立新闻消息与所确定的资源类别之间的匹配关系,为后续基于网络资源的业务处理提供条件。
图8为本申请又一实施例提供的业务处理装置的结构示意图。如图8所示,该装置包括:抓取模块81、确定模块82和业务模块83。
抓取模块81,用于按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息。
确定模块82,用于确定与上述新闻消息相匹配的目标资源类别。
业务模块83,用于对上述目标资源类别下的网络资源进行业务处理。
在一可选实施方式中,确定模块82的一种实现结构包括:计算单元和确定单元。
计算单元,用于计算上述新闻消息与资源类别库中各资源类别之间的相似度;
确定单元,用于确定与上述新闻消息之间的相似度满足预设第一相似度条件的资源类别作为目标资源类别。
进一步,计算单元具体用于:
根据新闻消息的正文、标题和评论信息中的至少一类信息,获取新闻消息的关键词;
对各资源类别分别进行分词处理,以获得各资源类别的关键词;
根据新闻消息的关键词与各资源类别的关键词,计算新闻消息与各资源类别之间的相似度。
更进一步,计算单元在根据新闻消息的正文、标题和评论信息中的至少一类信息,获取新闻消息的关键词时,具体用于:
对新闻消息的正文、标题和评论消息中的至少一类信息进行关键词提取处理,以获得正文关键词、标题关键词和评论关键词中的至少一种;
将正文关键词、标题关键词和评论关键词中的至少一种进行合并和去重处理,以获得新闻消息的关键词。
更进一步,计算单元在根据新闻消息的关键词与各资源类别的关键词,计算新闻消息与各资源类别之间的相似度时,具体用于:
获取新闻消息的关键词的词向量和各资源类别的关键词的词向量;
根据新闻消息的关键词的词向量与各资源类别的关键词的词向量,计算新闻消息与各资源类别之间的相似度。
本实施例提供的业务处理装置,通过抓取新闻消息,确定与新闻消息相匹配的目标资源类别,进而对目标资源类别下的网络资源进行业务处理,提供一种基于新闻消息与资源类别之间的匹配关系的业务处理方法,充分发挥新闻消息对业务处理过程的影响,提高业务处理精度,同时丰富业务处理方式。
所属领域的技术人员可以清楚地了解到,为描述的方便和简洁,上述描述的系统,装置和单元的具体工作过程,可以参考前述方法实施例中的对应过程,在此不再赘述。
在本申请所提供的几个实施例中,应该理解到,所揭露的系统,装置和方法,可以通过其它的方式实现。例如,以上所描述的装置实施例仅仅是示意性的,例如,所述单元的划分,仅仅为一种逻辑功能划分,实际实现时可以有另外的划分方式,例如多个单元或组件可以结合或者可以集成到另一个系统,或一些特征可以忽略,或不执行。另一点,所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口,装置或单元的间接耦合或通信连接,可以是电性,机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的,作为单元显示的部件可以是或者也可以不是物理单元,即可以位于一个地方,或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外,在本申请各个实施例中的各功能单元可以集成在一个处理单元中,也可以是各个单元单独物理存在,也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现,也可以采用硬件加软件功能单元的形式实现。
上述以软件功能单元的形式实现的集成的单元,可以存储在一个计算机可读取存储介质中。上述软件功能单元存储在一个存储介质中,包括若干指令用以使得一台计算机设备(可以是个人计算机,服务器,或者网络设备等)或处理器(processor)执行本申请各个实施例所述方法的部分步骤。而前述的存储介质包括:U盘、移动硬盘、只读存储器(Read-Only Memory,ROM)、随机存取存储器(Random Access Memory,RAM)、磁碟或者光盘等各种可以存储程序代码的介质。
最后应说明的是:以上实施例仅用以说明本申请的技术方案,而非对其限制;尽管参照前述实施例对本申请进行了详细的说明,本领域的普通技术人员应当理解:其依然可以对前述各实施例所记载的技术方案进行修改,或者对其中部分技术特征进行等同替换;而这些修改或者替换,并不使相应技术方案的本质脱离本申请各实施例技术方案的精神和范围。
Claims (36)
- 一种业务处理方法,其特征在于,包括:确定待处理网络资源所属的目标资源类别;获取与所述目标资源类别相匹配的目标新闻消息;根据所述目标新闻消息,对所述待处理网络资源进行业务处理。
- 根据权利要求1所述的方法,其特征在于,所述获取与所述目标资源类别相匹配的目标新闻消息,包括:根据所述目标资源类别,查询预先建立的资源类别与新闻消息之间的匹配关系,以获取所述目标新闻消息。
- 根据权利要求2所述的方法,其特征在于,所述建立资源类别与新闻消息之间的匹配关系,包括:按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息;计算所述新闻消息与资源类别库中各资源类别之间的相似度;确定与所述新闻消息之间的相似度满足预设第一相似度条件的资源类别;建立所述新闻消息和所述确定的资源类别之间的匹配关系。
- 根据权利要求3所述的方法,其特征在于,所述计算新闻消息与资源类别库中各资源类别之间的相似度,包括:根据所述新闻消息的正文、标题和评论信息中的至少一类信息,获取所述新闻消息的关键词;对所述各资源类别分别进行分词处理,以获得所述各资源类别的关键词;根据所述新闻消息的关键词与所述各资源类别的关键词,计算所述新闻消息与所述各资源类别之间的相似度。
- 根据权利要求4所述的方法,其特征在于,所述根据所述新闻消息的正文、标题和评论信息中的至少一类信息,获取所述新闻消息的关键词,包括:对所述新闻消息的正文、标题和评论消息中的至少一类信息进行关键词提取处理,以获得正文关键词、标题关键词和评论关键词中的至少一种;将所述正文关键词、标题关键词和评论关键词中的至少一种进行合并和去 重处理,以获得所述新闻消息的关键词。
- 根据权利要求4或5所述的方法,其特征在于,所述根据所述新闻消息的关键词与所述各资源类别的关键词,计算所述新闻消息与所述各资源类别之间的相似度,包括:获取所述新闻消息的关键词的词向量和所述各资源类别的关键词的词向量;根据所述新闻消息的关键词的词向量与所述各资源类别的关键词的词向量,计算所述新闻消息与所述各资源类别之间的相似度。
- 根据权利要求1所述的方法,其特征在于,所述获取与所述目标资源类别相匹配的目标新闻消息,包括:计算新闻语料库中各新闻消息与所述目标资源类别之间的相似度;获取与所述目标资源类别之间的相似度满足预设第二相似度条件的新闻消息作为所述目标新闻消息。
- 根据权利要求7所述的方法,其特征在于,所述计算新闻语料库中各新闻消息与所述目标资源类别之间的相似度,包括:对所述目标资源类别进行分词处理,以获得所述目标资源类别的关键词;对于每个新闻消息,根据所述新闻消息的正文、标题和评论信息中的至少一类信息,获取所述新闻消息的关键词,根据所述新闻消息的关键词与所述目标资源类别的关键词,计算所述新闻消息与所述目标资源类别之间的相似度。
- 根据权利要求8所述的方法,其特征在于,所述根据所述新闻消息的关键词与所述目标资源类别的关键词,计算所述新闻消息与所述目标资源类别之间的相似度,包括:获取所述新闻消息的关键词的词向量和所述目标资源类别的关键词的词向量;根据所述新闻消息的关键词的词向量与所述目标资源类别的关键词的词向量,计算所述新闻消息和所述目标资源类别之间的相似度。
- 一种数据处理方法,其特征在于,包括:按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息;计算所述新闻消息与资源类别库中各资源类别之间的相似度;确定与所述新闻消息之间的相似度满足预设第一相似度条件的资源类别;建立所述新闻消息和所述确定的资源类别之间的匹配关系。
- 根据权利要求10所述的方法,其特征在于,所述计算新闻消息与资源类别库中各资源类别之间的相似度,包括:根据所述新闻消息的正文、标题和评论信息中的至少一类信息,获取所述新闻消息的关键词;对所述各资源类别分别进行分词处理,以获得所述各资源类别的关键词;根据所述新闻消息的关键词与所述各资源类别的关键词,计算所述新闻消息与所述各资源类别之间的相似度。
- 根据权利要求11所述的方法,其特征在于,所述根据所述新闻消息的正文、标题和评论信息中的至少一类信息,获取所述新闻消息的关键词,包括:对所述新闻消息的正文、标题和评论消息中的至少一类信息进行关键词提取处理,以获得正文关键词、标题关键词和评论关键词中的至少一种;将所述正文关键词、标题关键词和评论关键词中的至少一种进行合并和去重处理,以获得所述新闻消息的关键词。
- 根据权利要求11或12所述的方法,其特征在于,所述根据所述新闻消息的关键词与所述各资源类别的关键词,计算所述新闻消息与所述各资源类别之间的相似度,包括:获取所述新闻消息的关键词的词向量和所述各资源类别的关键词的词向量;根据所述新闻消息的关键词的词向量与所述各资源类别的关键词的词向量,计算所述新闻消息与所述各资源类别之间的相似度。
- 一种业务处理方法,其特征在于,包括:按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息;确定与所述新闻消息相匹配的目标资源类别;对所述目标资源类别下的网络资源进行业务处理。
- 根据权利要求14所述的方法,其特征在于,所述确定与所述新闻消息 相匹配的目标资源类别,包括:计算所述新闻消息与资源类别库中各资源类别之间的相似度;确定与所述新闻消息之间的相似度满足预设第一相似度条件的资源类别作为所述目标资源类别。
- 根据权利要求15所述的方法,其特征在于,所述计算新闻消息与资源类别库中各资源类别之间的相似度,包括:根据所述新闻消息的正文、标题和评论信息中的至少一类信息,获取所述新闻消息的关键词;对所述各资源类别分别进行分词处理,以获得所述各资源类别的关键词;根据所述新闻消息的关键词与所述各资源类别的关键词,计算所述新闻消息与所述各资源类别之间的相似度。
- 根据权利要求16所述的方法,其特征在于,所述根据所述新闻消息的正文、标题和评论信息中的至少一类信息,获取所述新闻消息的关键词,包括:对所述新闻消息的正文、标题和评论消息中的至少一类信息进行关键词提取处理,以获得正文关键词、标题关键词和评论关键词中的至少一种;将所述正文关键词、标题关键词和评论关键词中的至少一种进行合并和去重处理,以获得所述新闻消息的关键词。
- 根据权利要求16或17所述的方法,其特征在于,所述根据所述新闻消息的关键词与所述各资源类别的关键词,计算所述新闻消息与所述各资源类别之间的相似度,包括:获取所述新闻消息的关键词的词向量和所述各资源类别的关键词的词向量;根据所述新闻消息的关键词的词向量与所述各资源类别的关键词的词向量,计算所述新闻消息与所述各资源类别之间的相似度。
- 一种业务处理装置,其特征在于,包括:第一确定模块,用于确定待处理网络资源所属的目标资源类别;获取模块,用于获取与所述目标资源类别相匹配的目标新闻消息;业务模块,用于根据所述目标新闻消息,对所述待处理网络资源进行业务 处理。
- 根据权利要求19所述的装置,其特征在于,所述获取模块具体用于:根据所述目标资源类别,查询预先建立的资源类别与新闻消息之间的匹配关系,以获取所述目标新闻消息。
- 根据权利要求20所述的装置,其特征在于,还包括:抓取模块,用于按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息;计算模块,用于计算所述新闻消息与资源类别库中各资源类别之间的相似度;第二确定模块,用于确定与所述新闻消息之间的相似度满足预设第一相似度条件的资源类别;建立模块,用于建立所述新闻消息和所述确定的资源类别之间的匹配关系。
- 根据权利要求21所述的装置,其特征在于,所述计算模块包括:获取单元,用于根据所述新闻消息的正文、标题和评论信息中的至少一类信息,获取所述新闻消息的关键词;分词单元,用于对所述各资源类别分别进行分词处理,以获得所述各资源类别的关键词;计算单元,用于根据所述新闻消息的关键词与所述各资源类别的关键词,计算所述新闻消息与所述各资源类别之间的相似度。
- 根据权利要求22所述的装置,其特征在于,所述获取单元具体用于:对所述新闻消息的正文、标题和评论消息中的至少一类信息进行关键词提取处理,以获得正文关键词、标题关键词和评论关键词中的至少一种;将所述正文关键词、标题关键词和评论关键词中的至少一种进行合并和去重处理,以获得所述新闻消息的关键词。
- 根据权利要求22或23所述的装置,其特征在于,所述计算单元具体用于:获取所述新闻消息的关键词的词向量和所述各资源类别的关键词的词向量;根据所述新闻消息的关键词的词向量与所述各资源类别的关键词的词向量,计算所述新闻消息与所述各资源类别之间的相似度。
- 根据权利要求19所述的装置,其特征在于,所述获取模块具体用于:计算新闻语料库中各新闻消息与所述目标资源类别之间的相似度;获取与所述目标资源类别之间的相似度满足预设第二相似度条件的新闻消息作为所述目标新闻消息。
- 根据权利要求25所述的装置,其特征在于,所述获取模块具体用于:对所述目标资源类别进行分词处理,以获得所述目标资源类别的关键词;对于每个新闻消息,根据所述新闻消息的正文、标题和评论信息中的至少一类信息,获取所述新闻消息的关键词,根据所述新闻消息的关键词与所述目标资源类别的关键词,计算所述新闻消息与所述目标资源类别之间的相似度。
- 根据权利要求26所述的装置,其特征在于,所述获取模块具体用于:获取所述新闻消息的关键词的词向量和所述目标资源类别的关键词的词向量;根据所述新闻消息的关键词的词向量与所述目标资源类别的关键词的词向量,计算所述新闻消息和所述目标资源类别之间的相似度。
- 一种数据处理装置,其特征在于,包括:抓取模块,用于按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息;计算模块,用于计算所述新闻消息与资源类别库中各资源类别之间的相似度;确定模块,用于确定与所述新闻消息之间的相似度满足预设第一相似度条件的资源类别;建立模块,用于建立所述新闻消息和所述确定的资源类别之间的匹配关系。
- 根据权利要求28所述的装置,其特征在于,所述计算模块包括:获取单元,用于根据所述新闻消息的正文、标题和评论信息中的至少一类信息,获取所述新闻消息的关键词;分词单元,用于对所述各资源类别分别进行分词处理,以获得所述各资源 类别的关键词;计算单元,用于根据所述新闻消息的关键词与所述各资源类别的关键词,计算所述新闻消息与所述各资源类别之间的相似度。
- 根据权利要求29所述的装置,其特征在于,所述获取单元具体用于:对所述新闻消息的正文、标题和评论消息中的至少一类信息进行关键词提取处理,以获得正文关键词、标题关键词和评论关键词中的至少一种;将所述正文关键词、标题关键词和评论关键词中的至少一种进行合并和去重处理,以获得所述新闻消息的关键词。
- 根据权利要求28或29所述的装置,其特征在于,所述计算单元具体用于:获取所述新闻消息的关键词的词向量和所述各资源类别的关键词的词向量;根据所述新闻消息的关键词的词向量与所述各资源类别的关键词的词向量,计算所述新闻消息与所述各资源类别之间的相似度。
- 一种业务处理装置,其特征在于,包括:抓取模块,用于按照预设抓取周期,从网络平台上抓取满足预设要求的新闻消息;确定模块,用于确定与所述新闻消息相匹配的目标资源类别;业务模块,用于对所述目标资源类别下的网络资源进行业务处理。
- 根据权利要求32所述的装置,其特征在于,所述确定模块包括:计算单元,用于计算所述新闻消息与资源类别库中各资源类别之间的相似度;确定单元,用于确定与所述新闻消息之间的相似度满足预设第一相似度条件的资源类别作为所述目标资源类别。
- 根据权利要求33所述的装置,其特征在于,所述计算单元具体用于:根据所述新闻消息的正文、标题和评论信息中的至少一类信息,获取所述新闻消息的关键词;对所述各资源类别分别进行分词处理,以获得所述各资源类别的关键词;根据所述新闻消息的关键词与所述各资源类别的关键词,计算所述新闻消息与所述各资源类别之间的相似度。
- 根据权利要求34所述的装置,其特征在于,所述计算单元具体用于:对所述新闻消息的正文、标题和评论消息中的至少一类信息进行关键词提取处理,以获得正文关键词、标题关键词和评论关键词中的至少一种;将所述正文关键词、标题关键词和评论关键词中的至少一种进行合并和去重处理,以获得所述新闻消息的关键词。
- 根据权利要求34或35所述的装置,其特征在于,所述计算单元具体用于:获取所述新闻消息的关键词的词向量和所述各资源类别的关键词的词向量;根据所述新闻消息的关键词的词向量与所述各资源类别的关键词的词向量,计算所述新闻消息与所述各资源类别之间的相似度。
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2018539082A JP2019507425A (ja) | 2016-01-27 | 2017-01-17 | サービス処理方法と、データ処理方法及び装置 |
| US16/046,894 US20180330002A1 (en) | 2016-01-27 | 2018-07-26 | Service Processing Method, and Data Processing Method and Apparatus |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610055298.5 | 2016-01-27 | ||
| CN201610055298.5A CN107015976B (zh) | 2016-01-27 | 2016-01-27 | 业务处理方法、数据处理方法及装置 |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| US16/046,894 Continuation US20180330002A1 (en) | 2016-01-27 | 2018-07-26 | Service Processing Method, and Data Processing Method and Apparatus |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2017128997A1 true WO2017128997A1 (zh) | 2017-08-03 |
Family
ID=59397329
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2017/071409 Ceased WO2017128997A1 (zh) | 2016-01-27 | 2017-01-17 | 业务处理方法、数据处理方法及装置 |
Country Status (5)
| Country | Link |
|---|---|
| US (1) | US20180330002A1 (zh) |
| JP (1) | JP2019507425A (zh) |
| CN (1) | CN107015976B (zh) |
| TW (1) | TWI790990B (zh) |
| WO (1) | WO2017128997A1 (zh) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112650919A (zh) * | 2020-11-30 | 2021-04-13 | 北京百度网讯科技有限公司 | 实体资讯分析方法、装置、设备及存储介质 |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110633406B (zh) * | 2018-06-06 | 2023-08-01 | 北京百度网讯科技有限公司 | 事件专题的生成方法、装置、存储介质和终端设备 |
| CN111796925A (zh) * | 2019-04-09 | 2020-10-20 | Oppo广东移动通信有限公司 | 算法模型的筛选方法、装置、存储介质和电子设备 |
| CN113495954B (zh) * | 2020-03-20 | 2024-11-26 | 北京沃东天骏信息技术有限公司 | 一种文本数据确定方法和装置 |
| CN112328937B (zh) * | 2020-11-04 | 2024-01-30 | 支付宝(杭州)信息技术有限公司 | 信息投放方法及装置 |
| JP7287992B2 (ja) * | 2021-01-28 | 2023-06-06 | ヤフー株式会社 | 情報処理装置、情報処理システム、情報処理方法、及びプログラム |
| JP7284196B2 (ja) * | 2021-01-28 | 2023-05-30 | ヤフー株式会社 | 情報処理装置、情報処理方法、及びプログラム |
| CN116028720B (zh) * | 2023-03-30 | 2023-06-09 | 无锡五车人工智能科技有限公司 | 基于人工智能的目标资源处理方法、系统及存储介质 |
| CN117194659A (zh) * | 2023-08-30 | 2023-12-08 | 支付宝(杭州)信息技术有限公司 | 行业分类方法和系统 |
| CN116992111B (zh) * | 2023-09-28 | 2023-12-26 | 中国科学技术信息研究所 | 数据处理方法、装置、电子设备及计算机存储介质 |
| CN120744245B (zh) * | 2025-09-04 | 2025-11-11 | 天津大学 | 基于价值感知的双层注意力网络的新闻内容推荐方法 |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20020015198A (ko) * | 2000-08-21 | 2002-02-27 | 정회선 | 인터넷을 이용한 증권 정보 및/또는 뉴스의 실시간 문자및/또는 음성 서비스 방법 및 시스템 |
| CN103226554A (zh) * | 2012-12-14 | 2013-07-31 | 西藏同信证券有限责任公司 | 基于新闻数据的股票自动匹配分类方法和系统 |
| CN104115178A (zh) * | 2011-11-30 | 2014-10-22 | 汤姆森路透社全球资源公司 | 基于新闻和情绪分析来预测市场行为的方法和系统 |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007041721A (ja) * | 2005-08-01 | 2007-02-15 | Ntt Resonant Inc | 情報分類方法およびプログラム、装置および記録媒体 |
| JP4859892B2 (ja) * | 2008-08-12 | 2012-01-25 | ヤフー株式会社 | 商品広告配信装置、商品広告配信方法、及び商品広告配信制御プログラム |
| JP4915021B2 (ja) * | 2008-09-10 | 2012-04-11 | ヤフー株式会社 | 検索装置、および検索装置の制御方法 |
| CN101446959A (zh) * | 2008-12-30 | 2009-06-03 | 深圳市迅雷网络技术有限公司 | 一种基于互联网的新闻推荐方法和系统 |
| CN102929937B (zh) * | 2012-09-28 | 2015-09-16 | 福州博远无线网络科技有限公司 | 基于文本主题模型的商品分类的数据处理方法 |
-
2016
- 2016-01-27 CN CN201610055298.5A patent/CN107015976B/zh active Active
-
2017
- 2017-01-17 JP JP2018539082A patent/JP2019507425A/ja active Pending
- 2017-01-17 WO PCT/CN2017/071409 patent/WO2017128997A1/zh not_active Ceased
- 2017-01-23 TW TW106102459A patent/TWI790990B/zh not_active IP Right Cessation
-
2018
- 2018-07-26 US US16/046,894 patent/US20180330002A1/en not_active Abandoned
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20020015198A (ko) * | 2000-08-21 | 2002-02-27 | 정회선 | 인터넷을 이용한 증권 정보 및/또는 뉴스의 실시간 문자및/또는 음성 서비스 방법 및 시스템 |
| CN104115178A (zh) * | 2011-11-30 | 2014-10-22 | 汤姆森路透社全球资源公司 | 基于新闻和情绪分析来预测市场行为的方法和系统 |
| CN103226554A (zh) * | 2012-12-14 | 2013-07-31 | 西藏同信证券有限责任公司 | 基于新闻数据的股票自动匹配分类方法和系统 |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112650919A (zh) * | 2020-11-30 | 2021-04-13 | 北京百度网讯科技有限公司 | 实体资讯分析方法、装置、设备及存储介质 |
| CN112650919B (zh) * | 2020-11-30 | 2023-09-01 | 北京百度网讯科技有限公司 | 实体资讯分析方法、装置、设备及存储介质 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107015976A (zh) | 2017-08-04 |
| TW201732650A (zh) | 2017-09-16 |
| JP2019507425A (ja) | 2019-03-14 |
| CN107015976B (zh) | 2020-09-29 |
| TWI790990B (zh) | 2023-02-01 |
| US20180330002A1 (en) | 2018-11-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2017128997A1 (zh) | 业务处理方法、数据处理方法及装置 | |
| CN107861972B (zh) | 一种用户录入商品信息后显示商品全结果的方法及设备 | |
| TWI694401B (zh) | 融合用戶關係資料的搜尋方法和裝置 | |
| Zhao et al. | We know what you want to buy: a demographic-based system for product recommendation on microblogs | |
| CN110647631B (zh) | 案例推荐的方法及装置、存储介质、处理器 | |
| CN106294425B (zh) | 商品相关网络文章之自动图文摘要方法及系统 | |
| TW201428513A (zh) | 查詢詞融合方法,商品資訊發佈方法和搜尋方法及系統 | |
| WO2013159608A1 (zh) | 网络交易平台及其处理方法 | |
| CN116561446A (zh) | 多模态项目推荐方法、系统及设备和存储介质 | |
| CN105741134A (zh) | 跨数据源的营销人群在市场营销中应用的方法及装置 | |
| CN107292642B (zh) | 一种基于图像的商品推荐方法及系统 | |
| CN107730346A (zh) | 物品聚类的方法和装置 | |
| US10055741B2 (en) | Method and apparatus of matching an object to be displayed | |
| CN106354845A (zh) | 基于传播结构的微博谣言识别方法和系统 | |
| TWI705411B (zh) | 社交業務特徵用戶的識別方法和裝置 | |
| US10074032B2 (en) | Using images and image metadata to locate resources | |
| CN106604066A (zh) | 改进的视频app中个性化推荐方法及系统 | |
| TWI645348B (zh) | 商品相關網路文章之自動圖文摘要方法及系統 | |
| CN108959323B (zh) | 视频分类方法和装置 | |
| US20160314145A1 (en) | Using image recognition to locate resources | |
| CN116739626A (zh) | 商品数据挖掘处理方法、装置、电子设备及可读介质 | |
| CN118537091A (zh) | 一种商品信息的处理方法及商品搜索方法 | |
| CN116595241B (zh) | 新媒体信息展示方法、装置、电子设备及计算机可读介质 | |
| CN112132368A (zh) | 信息处理方法以及装置、计算设备、存储介质 | |
| CN104102662B (zh) | 一种用户兴趣偏好相似度确定方法及装置 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 17743612 Country of ref document: EP Kind code of ref document: A1 |
|
| ENP | Entry into the national phase |
Ref document number: 2018539082 Country of ref document: JP Kind code of ref document: A |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 17743612 Country of ref document: EP Kind code of ref document: A1 |