WO2019141109A1 - 内容推荐方法及装置 - Google Patents

内容推荐方法及装置 Download PDF

Info

Publication number
WO2019141109A1
WO2019141109A1 PCT/CN2019/070830 CN2019070830W WO2019141109A1 WO 2019141109 A1 WO2019141109 A1 WO 2019141109A1 CN 2019070830 W CN2019070830 W CN 2019070830W WO 2019141109 A1 WO2019141109 A1 WO 2019141109A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
location
content
service
location data
Prior art date
Application number
PCT/CN2019/070830
Other languages
English (en)
French (fr)
Inventor
刘阳阳
Original Assignee
阿里巴巴集团控股有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 阿里巴巴集团控股有限公司 filed Critical 阿里巴巴集团控股有限公司
Publication of WO2019141109A1 publication Critical patent/WO2019141109A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/953Querying, e.g. by the use of web search engines
    • G06F16/9535Search customisation based on user profiles and personalisation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/953Querying, e.g. by the use of web search engines
    • G06F16/9537Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/50Network services
    • H04L67/52Network services specially adapted for the location of the user terminal
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00Network arrangements or protocols for supporting network services or applications
    • H04L67/50Network services
    • H04L67/55Push-based network services

Definitions

  • the embodiments disclosed in the present specification relate to the field of Internet technologies, and in particular, to a content recommendation method and apparatus.
  • the present specification describes a content recommendation method and apparatus for determining content recommendation information recommended to a user according to the location information of the user and the feature information of each content information in the content information library, thereby recommending more accurate content information to the user in time.
  • a content recommendation method includes:
  • the content recommendation information recommended to the user is determined based at least on the location information and the feature information of each content information in the content information repository.
  • the location mapping model is obtained through the following steps:
  • the determining location information corresponding to the location data includes:
  • the location information is determined according to the location label corresponding to the determined cluster.
  • the determining content recommendation information recommended to the user includes:
  • the feature information includes the service category information of the content information
  • the determining content information related to the service category information from the content information database includes:
  • the content information corresponding to the service category information is determined from the content information library.
  • the feature information of each content information includes predetermined location information corresponding to the content information
  • the determining content recommendation information recommended to the user includes:
  • the content information is used as the content recommendation information.
  • the feature information is determined based on keyword information and location information extracted from the content information.
  • the location data includes at least one of wireless fidelity WiFi fingerprint data and latitude and longitude data.
  • the location information includes at least one of a building name, a merchant name, and a business circle information.
  • a content recommendation device in a second aspect, includes:
  • An obtaining unit configured to acquire location data of the user
  • a determining unit configured to determine location information corresponding to the location data according to a location mapping model, where the location mapping model is obtained based on pre-acquired service information, where the service information includes service location data and a corresponding service location data;
  • the processing unit is configured to determine content recommendation information recommended to the user according to at least the location information and the feature information of each content information in the content information library.
  • the location mapping model is obtained by the determining unit by the following steps:
  • the determining unit specifically includes:
  • a first determining subunit configured to determine, from the plurality of clusters, a cluster corresponding to the location data of the user
  • a second determining subunit configured to determine the location information according to the location label corresponding to the determined cluster.
  • the processing unit specifically includes:
  • a first processing subunit configured to determine service category information corresponding to the location information
  • a second processing sub-unit configured to determine, according to the feature information of the respective content information, content information related to the service category information from the content information library, and use the content information as the content recommendation information.
  • the feature information in the second processing sub-unit includes the service category information of the content information, and the second processing sub-unit is specifically configured to:
  • the content information corresponding to the service category information is determined from the content information library.
  • the feature information of each content information included in the processing unit includes predetermined location information corresponding to the content information, and the processing unit is specifically configured to:
  • the content information is used as the content recommendation information.
  • the feature information included in the processing unit is determined based on keyword information and location information extracted from the content information.
  • the location data acquired by the acquiring unit includes at least one of wireless fidelity WiFi fingerprint data and latitude and longitude data.
  • the location information determined by the determining unit includes at least one of a building name, a merchant name, and a business circle information.
  • a computer readable storage medium having a computer program stored thereon is provided.
  • the computer program is executed in a computer, the computer is caused to perform the method provided by any of the above-described first aspects.
  • a computing device comprising a memory and a processor.
  • An executable code is stored in the memory, and when the processor executes the executable code, the method provided by any one of the foregoing first aspects is implemented.
  • the content recommendation method and device provided by the present specification determine the location information corresponding to the location data by acquiring the location data of the user and according to the location mapping model. Then, based on the location information and the feature information of each content information in the content information repository, content recommendation information recommended to the user is determined, so that more accurate content information is recommended to the user in time.
  • FIG. 1 is a schematic diagram of an application scenario of a content recommendation method according to an embodiment of the present disclosure
  • FIG. 2 is a flowchart of a content recommendation method according to an embodiment of the disclosure
  • FIG. 3 is a schematic diagram of clustering service location data into multiple clusters according to an embodiment of the disclosure
  • FIG. 4 is a flowchart of a method for determining feature information provided by an embodiment of the present disclosure
  • FIG. 5 is a schematic diagram of a content recommendation apparatus according to an embodiment of the disclosure.
  • FIG. 1 is a schematic diagram of an application scenario of a content recommendation method according to an embodiment of the present disclosure.
  • the execution body of the recommendation method may be a server.
  • a content recommendation platform for example, an Alipay application platform
  • a terminal for example, the terminal can be a mobile phone, a tablet, a wearable smart device, etc.
  • the server acquires location data of the user (eg, latitude and longitude data), and according to the location mapping model (eg, the location mapping model may include a mapping relationship between the location data and the location information), determining location information corresponding to the location data (eg, Car 4S shop).
  • the feature information may include service category information
  • content recommendation information eg, car maintenance knowledge, etc.
  • the recommendation method of the business function determines the location information corresponding to the location data by acquiring the location data of the user and according to the location mapping model. Then, based on the location information and the feature information of each content information in the content information repository, content recommendation information recommended to the user is determined, so that more accurate content information is recommended to the user in time.
  • FIG. 2 is a flowchart of a content recommendation method according to an embodiment of the disclosure.
  • the execution body of the method may be a device having processing capabilities: a server or a system or device, for example, a server as shown in FIG. As shown in FIG. 2, the method specifically includes:
  • Step S210 acquiring location data of the user.
  • the location data may be data collected from a user's terminal by a Location Based Service (LBS).
  • LBS includes multiple positioning methods, such as Global Positioning System (GPS) positioning, base station positioning, and Wireless Fidelity (WiFi) positioning.
  • GPS Global Positioning System
  • WiFi Wireless Fidelity
  • the location data may include latitude and longitude data obtained by GPS positioning or base station positioning.
  • the latitude and longitude data included in the position data is: north latitude 39° 54'25.70′′ and east longitude 116°23′28.49′′.
  • the location data may include WiFi fingerprint data obtained by WiFi location.
  • the WiFi fingerprint data may include an address of a WiFi Access Point (AP) (eg, a Media Access Control (MAC) address) and a corresponding signal strength.
  • AP WiFi Access Point
  • MAC Media Access Control
  • Step S220 determining location information corresponding to the location data of the user according to the location mapping model.
  • the location information is different from the location data in that the location information is location information that is perceptible by the user and has clear semantics.
  • the venue information may include a building name (eg, a new Zhongguan Building), a merchant name (eg, a Hamm toy store), and a business circle information (eg, Sanlitun Taikooli).
  • the venue information may also include service category information (eg, food, clothing, beauty, movies, etc.) associated with the venue (eg, merchant).
  • the conversion from the location data to the location information can be performed according to a previously obtained location mapping model.
  • the location mapping model includes a mapping relationship between GPS coordinate data and a location tag, the mapping relationship being obtained in advance by manual acquisition.
  • the location mapping model may be trained to be obtained based on the service location data and the corresponding service location data included in the pre-acquired service information.
  • the service corresponding to the service information may include a payment service (eg, payment using an payment application) and a location service (eg, locating the current location when the social platform sends status information).
  • the service corresponding to the service information may include a payment service, and the corresponding service information may include payment information.
  • Obtaining the payment information may include: the server may collect the payment information when detecting that the user submits the order through the terminal or pays the order, specifically including collecting the location data of the terminal as the service location data in the payment information, and the submitted order Obtain the service location information corresponding to the service location data.
  • a payment application eg, Alipay
  • the server of the payment application can collect the location data of the terminal at this time through WiFi positioning (eg, WiFi fingerprint data).
  • service location data included in the payment information eg, McDonald's College Road Shop).
  • the service corresponding to the service information may include a location service, and the corresponding service information may include location service information.
  • the obtaining the location service information may include: when the server detects that the user uses the location service by using the terminal, the server may obtain the location service information, where the location data of the collection terminal is used as the service location data in the location service information, and the location service is obtained by the user.
  • Location information selected or created In an example, the user uses a terminal that supports GPS positioning to log in to a work application (eg, a nail) to perform a work punching operation. At this time, the location data (eg, latitude and longitude data) of the collection terminal and the punch information may be collected by GPS. Service location data (eg, California gourmet restaurant).
  • the user uses the terminal connected to the cellular network to publish the status including the location information on the social platform.
  • the location data eg, latitude and longitude data
  • the location data of the collection terminal can be collected by the base station and the user selects (or creates ) service location data (eg, one or two cafes).
  • the location mapping model may be obtained through the following steps: clustering the service location data included in the service information to obtain a plurality of clusters. Then, the location label corresponding to the plurality of clusters is determined according to the service location data corresponding to the service location data in the service information. In this way, a mapping relationship between multiple clusters and multiple place tags can be established.
  • the clustering algorithm may be used to cluster the service location data and obtain multiple clusters.
  • the clustering algorithm may be a GEOSHASH algorithm or a DBSCAN algorithm, which is not limited herein.
  • the service location data includes a plurality of latitude and longitude data
  • the GEOSHASH algorithm can be used to convert the data into a GEOHASH grid, and the grid number of each latitude and longitude data is determined.
  • the service location data includes 100 latitude and longitude coordinates, and after converting the 100 latitude and longitude coordinates into a GEOHASH mesh by the GEOHASH algorithm, as shown in FIG. 3, each 20 latitude and longitude coordinates have the same mesh number, and the obtained 5 meshes are obtained.
  • the numbers are: WX4G01, WX4H02, WX4I03, WX4J04, WX4K05.
  • the above 100 latitude and longitude coordinates are clustered into 5 clusters.
  • the service location data includes a plurality of WiFi fingerprint data.
  • the data may be clustered by using a DBSCAN algorithm to obtain a plurality of clusters having different numbers.
  • the DBSCAN algorithm is a density-based clustering algorithm. Unlike the partitioning and hierarchical clustering methods, it defines the cluster as the largest set of points with density connections, can divide the area with sufficient high density into clusters, and can find clusters of arbitrary shapes in the spatial database of noise. Specifically, in the DBSCAN algorithm, all position points are first marked as core points, boundary points, or noise points, and noise points are deleted. Then assign an edge between all the core points within the preset parameters, each group of connected core points form a cluster, and assign each boundary point to a cluster of core points associated with it, thereby completing Clustering of location points.
  • the precision of the cluster obtained by the clustering can be controlled by adjusting the parameters of the clustering algorithm.
  • the accuracy of the cluster can be controlled by controlling the number of coded bits in the GEOHASH algorithm that encodes the service location data. More specifically, the more coded bits of service location data, the more accurate the resulting cluster range.
  • the accuracy of the cluster can be controlled by controlling the size of the neighborhood radius ⁇ in the DBSCAN algorithm. More specifically, the smaller the value of the input domain radius ⁇ , the more accurate the resulting cluster size range.
  • the plurality of clusters obtained by the cluster may determine the location labels corresponding to the clusters in the plurality of clusters according to the service location data corresponding to the service location data in the service information.
  • the service location data corresponding to the service location data includes the name of the service site (for example, a happiness supermarket)
  • the name of the service site may be used as a location tag corresponding to the cluster in which the service location data is located.
  • a plurality of service location data may be corresponding to a single cluster, and a range (eg, a merchant or a business circle) identified by a plurality of service location data corresponding to the plurality of service location data may be different.
  • a single cluster can have multiple place tags, such as a merchant tag and a business circle tag.
  • a place label for a cluster may include a merchant label (eg, West MCA) and a business circle label (eg, Dazhong Temple).
  • determining the location information corresponding to the location data of the user may include: determining a cluster corresponding to the location data of the user from the plurality of clusters obtained by the cluster. Next, the location information is determined based on the location tag corresponding to the determined cluster.
  • the location mapping model includes location data of a center point of each of the plurality of clusters, and a mapping relationship between the plurality of clusters and the plurality of location labels.
  • determining the cluster corresponding to the location data of the user from the plurality of clusters obtained by the cluster may include: calculating a distance between the location data of the user and the location data of each center point, and calculating a minimum value of the distance The corresponding cluster is used as a cluster corresponding to the location data of the user.
  • determining the location information according to the location label corresponding to the determined cluster may include: mapping the relationship between the plurality of clusters and the plurality of location labels, and determining the cluster corresponding to the location data of the user. And determining a place label corresponding to the cluster, thereby determining location information.
  • the determined location tag includes a merchant tag (eg, West MCA), and the location information including the merchant name (eg, West MCA) may be determined accordingly.
  • a merchant tag eg, West MCA
  • the location information including the merchant name eg, West MCA
  • step S230 content recommendation information recommended to the user is determined based on at least the location information and the feature information of each content information in the content information repository.
  • determining content recommendation information recommended to the user may include: determining service category information corresponding to the location information, and determining content information related to the service category information from the content information base according to the feature information of each content information, and This content information is used as content recommendation information.
  • determining the content information related to the service category information from the content information base may include: calculating, according to the feature information of each content information, the relevance of each content information and the service category information, and calculating the correlation according to the Determine the content information related to the service category information.
  • content information whose relevance is within a predetermined range may be determined as content information related to the service category information.
  • the content information may be ranked according to the relevance, and the content information within the predetermined range (eg, the predetermined range may be the top five) may be determined as the content information related to the service category information.
  • the location information determined in step S220 is "automobile 4S shop", and accordingly, it may be determined that the service category information corresponding to the location information is "car". Accordingly, the content information of the service category information of "car” can be determined from the content information base, and this content information is used as the content recommendation information recommended to the user.
  • the feature information of each content information may include predetermined location information corresponding to the content information.
  • determining the content recommendation information recommended to the user may include: when the affiliation relationship between the location information corresponding to the respective content information and the location information corresponding to the location data conforms to the preset rule, the content information As content recommendation information.
  • the preset rules may be determined by the business party based on the attributes of the business (eg, the accuracy requirements of the business for the location of the venue). For example, when the service is to push the merchant coupon to the user, the preset rule may include that the location information corresponding to the content recommendation information is identical to the location information corresponding to the location data. When the service is to recommend a similar place in the vicinity to the user, the preset rule may include that the place information corresponding to the content recommendation information is the same as the place information portion corresponding to the position data.
  • the attributes of the business eg, the accuracy requirements of the business for the location of the venue.
  • the preset rule may include that the location information corresponding to the content recommendation information is identical to the location information corresponding to the location data.
  • the preset rule may include that the place information corresponding to the content recommendation information is the same as the place information portion corresponding to the position data.
  • the preset rule includes that the location information corresponding to the content recommendation information is identical to the location information corresponding to the location data, and the location information corresponding to the location data includes “Old Delicious Hot Pot” and “Sanlitun”.
  • content information such as merchant discount activity, merchant public number information, and merchant evaluation information
  • the location information "old and delicious fragrant pot” (merchant) and "Sanlitun” can be used as content recommendation information. .
  • the preset rule may include that the location information corresponding to the content recommendation information is the same as the location information portion corresponding to the location data, and the location information corresponding to the location data includes “Uniqlo” (business), “Sanlitun” (business) ring).
  • content information corresponding to the location information "Sanlitun” for example, a clothing store, a restaurant, etc. in the Sanlitun business district
  • Sanlitun for example, a clothing store, a restaurant, etc. in the Sanlitun business district
  • the method further includes: sending the content recommendation information to the user. For example, when the user opens the corresponding application App, or opens the recommended channel, or when it detects that the location of the user changes, the latest location information of the user is acquired, and the content recommendation information recommended to the user is determined according to the content recommendation information, and the user is sent to the user.
  • Content recommendation information when the user opens the corresponding application App, or opens the recommended channel, or when it detects that the location of the user changes, the latest location information of the user is acquired, and the content recommendation information recommended to the user is determined according to the content recommendation information, and the user is sent to the user.
  • the method further includes: determining service category information corresponding to the location information.
  • the location information may include service category information corresponding to the location in addition to the name of the location, and accordingly, the service category information corresponding to the location information may be directly determined according to the location information.
  • the service category information corresponding to the location information may be determined according to the correspondence between the pre-stored plurality of location information and the plurality of service category information.
  • the identified location tag includes a merchant tag (eg, West MCA), which can be used to determine a business name (eg, West MCA) and service category information (eg, food) corresponding to the merchant name.
  • the feature information mentioned in step S230 may be determined based on the keyword information and/or the location information extracted from the content information, and may be specifically determined by the method as shown in FIG. 4:
  • Step S410 preprocessing the content information.
  • the pre-processing may include structured analysis, word segmentation processing, de-stop word processing, word-of-speech (postag), and the like.
  • the structured analysis may include analyzing the paragraph structure in the content information, for example, determining the title and the body in the content information;
  • the word segmentation processing may include a unigram, a bigram, and a trigram.
  • removing the stop word may include removing the stop word in the content information according to the preset stop word table (eg, a functional word having no practical meaning: this, that,);
  • the part of speech tagging refers to the content information.
  • the part of speech (such as nouns, adverbs, adjectives, etc.) is marked.
  • Step S420 extracting keyword information according to the pre-processed content information.
  • the position of the word obtained by the preprocessing in the content information for example, in the title or in the text
  • the part of the tag for example, the text
  • the weight of the text using the TextRank algorithm or the TF-IDF (Term Frequency–inverse Document Frequency) algorithm Keyword information for example, in the title or in the text
  • the TextRank algorithm for example, in the title or in the text
  • the TF-IDF Term Frequency–inverse Document Frequency
  • Step S430 extracting location information according to the pre-processed content information.
  • the location information may be extracted according to the location tag included in the content information, and the location information may include location information.
  • the location tag may be a location tag for the producer of the content information when posting the content information, such as Beijing, Hangzhou, and the like.
  • NER Named Entity Recognition
  • a method of Named Entity Recognition may be used to identify a place name, an institution name, and the like in the content information. For example, it can be identified that the name of the content information is Wutai Mountain, and the institution name is Haidian District Civil Affairs Bureau (belonging to the location information).
  • the server may pre-store a location information database, and based on the location information, the location information matching the information in the location information database may be extracted from the pre-processed content information.
  • Step S440 input the initially extracted keyword information and location information into the pre-trained feature extraction model to determine the feature information.
  • the feature extraction model may be an offline training of large-scale content corpus data, and a word embedding model or a NER model based on Recurrent Neural Networks (RNN).
  • RNN Recurrent Neural Networks
  • the determination of the above feature information may be performed within a predetermined time (e.g., 5 min or 10 min) after the content information is generated.
  • the location information corresponding to the location data is determined by acquiring the location data of the user and according to the location mapping model. Then, based on the location information and the feature information of each content information in the content information repository, content recommendation information recommended to the user is determined, so that more accurate content information is recommended to the user in time.
  • the multiple embodiments disclosed in the present specification further provide a content recommendation device.
  • the device 500 includes:
  • the obtaining unit 510 is configured to acquire location data of the user.
  • a determining unit 520 configured to determine, according to the location mapping model, location information corresponding to the location data, where the location mapping model is obtained based on the pre-acquired service information, where the service information includes the service location data and the corresponding service location data;
  • the processing unit 530 is configured to determine content recommendation information recommended to the user according to at least the location information and the feature information of each content information in the content information repository.
  • the location mapping model is obtained by the determining unit 520 through the following steps:
  • the location label corresponding to the plurality of clusters is determined according to the service location data corresponding to the service location data in the service information.
  • the determining unit 520 specifically includes:
  • a first determining subunit 521 configured to determine a cluster corresponding to the location data of the user from the plurality of clusters
  • the second determining subunit 522 is configured to determine the location information according to the location label corresponding to the determined cluster.
  • the processing unit 530 specifically includes:
  • a first processing sub-unit 531 configured to determine service category information corresponding to the location information
  • the second processing sub-unit 532 is configured to determine content information related to the service category information from the content information base according to the feature information of each content information, and use the content information as the content recommendation information.
  • the feature information in the second processing sub-unit 532 includes service category information of the content information, and the second processing sub-unit 532 is specifically configured to:
  • Content information corresponding to the service category information is determined from the content information base.
  • the feature information of each content information included in the processing unit 530 includes predetermined location information corresponding to the content information, and the processing unit 530 is specifically configured to:
  • the content information is used as the content recommendation information.
  • the feature information included in the processing unit 530 is determined based on the keyword information and the location information extracted from the content information.
  • the location data acquired by the obtaining unit 510 includes at least one of wireless fidelity WiFi fingerprint data and latitude and longitude data.
  • the location information determined by the determining unit 520 includes at least one of a building name, a merchant name, and a business circle information.
  • the acquisition unit 510 acquires the location data of the user, and the determination unit 520 determines the location information corresponding to the location data according to the location mapping model.
  • the processing unit 530 determines the content recommendation information recommended to the user according to the location information and the feature information of each content information in the content information library, so as to promptly recommend more accurate content information to the user.

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Information Transfer Between Computers (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本说明书披露的实施例提供一种内容推荐方法,该方法包括:获取用户的位置数据,并根据位置映射模型,确定与位置数据对应的场所信息。然后,至少根据确定的场所信息,以及内容信息库中各个内容信息的特征信息,确定向所述用户推荐的内容推荐信息。

Description

内容推荐方法及装置 技术领域
本说明书披露的多个实施例涉及互联网技术领域,尤其涉及一种内容推荐方法及装置。
背景技术
随着互联网技术的发展,人们越来越频繁地浏览网络平台提供的内容信息。例如,在网络购物平台中浏览商品信息,或者在新闻平台浏览热点信息,或者在理财平台浏览理财资讯等。
不同用户在使用同一网络平台时,对其提供的内容信息的需求有着或多或少的差异。另一方面,网络平台中信息的海量增长也常常让用户难以选择。目前,向用户推荐的内容信息由于存在不够精准、不够及时等不足,难以满足用户的要求。因此,需要提供一种合理的方法,以满足用户浏览网络平台中提供的内容信息的多种需求。
发明内容
本说明书描述了一种内容推荐方法及装置,根据用户的场所信息和内容信息库中各个内容信息的特征信息确定向用户推荐的内容推荐信息,从而及时地向用户推荐更加精准的内容信息。
第一方面,提供了一种内容推荐方法。该方法包括:
获取用户的位置数据;
根据位置映射模型,确定与所述位置数据对应的场所信息,所述位置映射模型基于预先获取的服务信息而训练获得,所述服务信息中包含服务位置数据和对应的服务场所数据;
至少根据所述场所信息,以及内容信息库中各个内容信息的特征信息,确定向所述用户推荐的内容推荐信息。
在一种可能的实施方式中,所述位置映射模型通过以下步骤训练获得:
对所述服务信息中包含的服务位置数据进行聚类,获得多个类簇;
根据所述服务信息中与所述服务位置数据对应的服务场所数据,确定所述多个类簇对应的场所标签。
在一种可能的实施方式中,所述确定与所述位置数据对应的场所信息包括:
从所述多个类簇中确定与所述用户的位置数据对应的类簇;
根据确定出的类簇所对应的场所标签,确定所述场所信息。
在一种可能的实施方式中,所述确定向所述用户推荐的内容推荐信息,包括:
确定与所述场所信息对应的服务类别信息;
根据所述各个内容信息的特征信息,从所述内容信息库中确定与所述服务类别信息相关的内容信息,并将所述内容信息作为所述内容推荐信息。
在一种可能的实施方式中,所述特征信息包括所述内容信息的所述服务类别信息,所述从所述内容信息库中确定与所述服务类别信息相关的内容信息,包括:
从所述内容信息库中确定与所述服务类别信息对应的所述内容信息。
在一种可能的实施方式中,所述各个内容信息的特征信息包括预先确定的与该内容信息对应的场所信息,所述确定向所述用户推荐的内容推荐信息,包括:
当与所述各个内容信息对应的场所信息和与所述位置数据对应的场所信息之间的从属关系符合预设规则时,将该内容信息作为所述内容推荐信息。
在一种可能的实施方式中,所述特征信息基于从所述内容信息中提取出的关键词信息和场所信息而确定。
在一种可能的实施方式中,所述位置数据包括无线保真WiFi指纹数据和经纬度数据中的至少一种。
在一种可能的实施方式中,所述场所信息包括建筑名称、商户名称和商圈信息中的至少一种。
第二方面,提供了一种内容推荐装置。该装置包括:
获取单元,用于获取用户的位置数据;
确定单元,用于根据位置映射模型,确定与所述位置数据对应的场所信息,所述位置映射模型基于预先获取的服务信息而训练获得,所述服务信息中包含服务位置数据和对应的服务场所数据;
处理单元,用于至少根据所述场所信息,以及内容信息库中各个内容信息的特征信息,确定向所述用户推荐的内容推荐信息。
在一种可能的实施方式中,所述位置映射模型由所述确定单元通过以下步骤训练获得:
对所述服务信息中包含的服务位置数据进行聚类,获得多个类簇;
根据所述服务信息中与所述服务位置数据对应的服务场所数据,确定所述多个类簇对应的场所标签。
在一种可能的实施方式中,所述确定单元具体包括:
第一确定子单元,用于从所述多个类簇中确定与所述用户的位置数据对应的类簇;
第二确定子单元,用于根据确定出的类簇所对应的场所标签,确定所述场所信息。
在一种可能的实施方式中,所述处理单元具体包括:
第一处理子单元,用于确定与所述场所信息对应的服务类别信息;
第二处理子单元,用于根据所述各个内容信息的特征信息,从所述内容信息库中确定与所述服务类别信息相关的内容信息,并将所述内容信息作为所述内容推荐信息。
在一种可能的实施方式中,所述第二处理子单元中的特征信息包括所述内容信息的所述服务类别信息,所述第二处理子单元具体用于:
从所述内容信息库中确定与所述服务类别信息对应的所述内容信息。
在一种可能的实施方式中,所述处理单元中包括的各个内容信息的特征信息包括预先确定的与该内容信息对应的场所信息,所述处理单元具体用于:
当与所述各个内容信息对应的场所信息和与所述位置数据对应的场所信息之间的从属关系符合预设规则时,将该内容信息作为所述内容推荐信息。
在一种可能的实施方式中,所述处理单元中包括的特征信息基于从所述内容信息中提取出的关键词信息和场所信息而确定。
在一种可能的实施方式中,所述获取单元获取的位置数据包括无线保真WiFi指纹数据和经纬度数据中的至少一种。
在一种可能的实施方式中,所述确定单元确定的场所信息包括建筑名称、商户名称和商圈信息中的至少一种。
第三方面,提供了一种计算机可读存储介质,其上存储有计算机程序。当所述计算机程序在计算机中执行时,令计算机执行上述第一方面中任一种实施方式提供的方法。
第四方面,提供了一种计算设备,包括存储器和处理器。所述存储器中存储有可执行代码,所述处理器执行所述可执行代码时,实现上述第一方面中任一种实施方式提供的方法。
本说明书提供的一种内容推荐方法及装置,通过获取用户的位置数据,并根据位置映射模型,确定与位置数据对应的场所信息。然后,至少根据该场所信息,以及内容信息库中各个内容信息的特征信息,确定向所述用户推荐的内容推荐信息,从而及时地向用户推荐更加精准的内容信息。
附图说明
为了更清楚地说明本说明书披露的多个实施例的技术方案,下面将对实施例描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本说明书披露的多个实施例,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它的附图。
图1为本说明书披露的一个实施例提供的一种内容推荐方法的应用场景示意图;
图2为本说明书披露的一个实施例提供的一种内容推荐方法的流程图;
图3为本说明书披露的一个实施例提供的一种将服务位置数据聚类成多个类簇的示意图;
图4为本说明书披露的一个实施例提供的一种特征信息确定方法的流程图;
图5为本说明书披露的一个实施例提供的一种内容推荐装置的示意图。
具体实施方式
下面结合附图,对本说明书披露的多个实施例进行描述。
图1为本说明书披露的一个实施例提供的一种内容推荐方法的应用场景示意图。所述推荐方法的执行主体可以为服务器。当用户通过终端(如,终端可以为手机、平板电脑、可穿戴智能设备等)登录内容推荐平台(如,支付宝应用平台)时,可以采用本说明书披露的多个实施例提供的业务功能的推荐方法,服务器获取用户的位置数据(如, 经纬度数据),并根据位置映射模型(如,位置映射模型可以包括位置数据与场所信息的映射关系),确定与该位置数据对应的场所信息(如,汽车4S店)。然后,根据该场所信息,以及内容信息库中各个内容信息的特征信息(如,特征信息可以包括服务类别信息),确定向用户推荐的内容推荐信息(如,汽车保养知识等)。
本说明书披露的多个实施例提供的业务功能的推荐方法,通过获取用户的位置数据,并根据位置映射模型,确定与位置数据对应的场所信息。然后,至少根据该场所信息,以及内容信息库中各个内容信息的特征信息,确定向所述用户推荐的内容推荐信息,从而及时地向用户推荐更加精准的内容信息。
图2为本说明书披露的一个实施例提供的一种内容推荐方法的流程图。所述方法的执行主体可以为具有处理能力的设备:服务器或者系统或者装置,例如,如图1所示的服务器。如图2所示,所述方法具体包括:
步骤S210,获取用户的位置数据。
具体地,位置数据可以为通过基于位置的服务(Location Based Service,简称LBS)从用户的终端采集的数据。LBS中包括多种定位方式,如全球定位系统(Global Positioning System,简称GPS)定位、基站定位、无线保真(Wireless Fidelity,简称WiFi)定位等。
在一个实施例中,位置数据可以包括通过GPS定位或基站定位得到的经纬度数据。例如,位置数据中包括的经纬度数据为:北纬39°54′25.70″和东经116°23′28.49″。
在另一个实施例中,位置数据可以包括通过WiFi定位得到的WiFi指纹数据。在一个例子中,WiFi指纹数据可以包括WiFi接入点(Access Point,简称AP)的地址(如,媒体访问控制(Media Access Control,简称MAC)地址)和相应的信号强度等。
步骤S220,根据位置映射模型,确定与用户的位置数据对应的场所信息。
具体地,场所信息与位置数据的不同之处在于,场所信息是用户可以感知到的、具有明确语义的位置信息。在一个实施例中,场所信息可以包括建筑名称(如,新中关大厦)、商户名称(如,哈姆玩具店)和商圈信息(如,三里屯太古里)等。在另一个实施例中,场所信息还可以包括与场所(如,商户)相关的服务类别信息(如,美食、服装、美容、电影等)。
从位置数据到场所信息的转换可以根据预先获得的位置映射模型来执行。在一个实施例中,上述位置映射模型包括GPS坐标数据与场所标签之间的映射关系,该映射关 系预先通过人工采集而获得。
在另一实施例中,上述位置映射模型可以基于预先获取的服务信息中包含的服务位置数据和对应的服务场所数据而训练获得。与服务信息对应的服务可以包括支付服务(如,使用支付应用对订单进行支付)和定位服务(如,在社交平台发送状态信息时对当前位置进行定位)等。
其中,服务信息可以通过多种方式获取。在一个实施例中,与服务信息对应的服务可以包括支付服务,相应的服务信息可以包括支付信息。获取支付信息,可以包括:服务器在检测到用户通过终端提交订单或对订单进行支付时,可以采集支付信息,具体包括通过采集终端的位置数据作为支付信息中的服务位置数据,以及从提交的订单中获取与服务位置数据相应的服务场所信息。在一个例子中,用户使用已连接WiFi的终端上的支付应用(如,支付宝)进行支付操作时,该支付应用的服务器可以通过WiFi定位采集此时终端的位置数据,(如,WiFi指纹数据)和支付信息中包括的服务场所数据(如,麦当劳学院路店)。
在另一个实施例中,与服务信息对应的服务可以包括定位服务,相应的服务信息可以包括定位服务信息。获取定位服务信息,可以包括:服务器在检测到用户通过终端使用定位服务时,可以获取定位服务信息,具体包括采集终端的位置数据作为定位服务信息中的服务位置数据,以及获取用户在使用定位服务时选定或创建的位置信息。在一个例子中,用户使用支持GPS定位的终端登录工作应用(如,钉钉)进行上班打卡操作,此时,可以通过GPS定位采集终端的位置数据(如,经纬度数据)和打卡信息中包括的服务场所数据(如,加州美食餐厅)。在另一个例子中,用户使用连接有蜂窝网的终端在社交平台上发布包括场所信息的状态,此时,可以通过基站定位采集终端的位置数据(如,经纬度数据)和用户选定(或创建)的服务场所数据(如,一二咖啡馆)。
进一步地,基于预先获取的服务信息,位置映射模型可以通过以下步骤训练获得:对服务信息中包含的服务位置数据进行聚类,获得多个类簇。然后,根据服务信息中与服务位置数据对应的服务场所数据,确定多个类簇对应的场所标签。如此,可以建立多个类簇与多个场所标签之间的映射关系。
在一个实施例中,可以采用聚类算法对服务位置数据进行聚类,并获得多个类簇。其中聚类算法可以为GEOHASH算法或DBSCAN算法,在此不作限定。
在一个例子中,服务位置数据包括多个经纬度数据,此时可以采用GEOHASH算法 将该数据转换成GEOHASH网格,并确定各个经纬度数据的网格编号。例如,服务位置数据包括100个经纬度坐标,采用GEOHASH算法将这100经纬度坐标转换成GEOHASH网格后,如图3所示,每20个经纬度坐标具有相同的网格编号,得到的5个网格编号分别为:WX4G01、WX4H02、WX4I03、WX4J04、WX4K05。如此,以上100个经纬度坐标被聚类为5个类簇。
在另一个例子中,服务位置数据包括多个WiFi指纹数据,此时可以采用DBSCAN算法对该数据进行聚类,获取多个具有不同编号的类簇。DBSCAN算法是一种基于密度的聚类算法。与划分和层次聚类方法不同,它将簇定义为密度相连的点的最大集合,能够把具有足够高密度的区域划分为簇,并可在噪声的空间数据库中发现任意形状的聚类。具体而言,在DBSCAN算法中,首先将所有位置点标记为核心点、边界点或噪声点,删除其中的噪声点。然后为距离在预设参数之内的所有核心点之间赋予一条边,每组连通的核心点形成一个簇,将每个边界点指派到一个与之关联的核心点的簇中,由此完成位置点的聚类。
可以理解的是,可以通过调整聚类算法的参数,控制聚类得到的类簇的精度(如,类簇的范围大小)。例如,可以通过控制GEOHASH算法中对服务位置数据进行编码的编码位数,控制类簇的精度。更具体地,服务位置数据的编码位数越多,得到的类簇范围越精确。又例如,可以通过控制DBSCAN算法中邻域半径ε的大小,控制类簇的精度。更具体地,输入的领域半径ε的值越小,得到的类簇范围越精确。
在一个实施例中,对聚类获得的多个类簇,可以根据服务信息中与服务位置数据对应的服务场所数据,确定多个类簇中各个类簇所对应的场所标签。
在一个例子中,与服务位置数据对应的服务场所数据中包括服务场所的名称(如,幸福超市),则可以将此服务场所的名称作为该服务位置数据所在类簇对应的场所标签。
在另一个例子中,单个类簇中可以对应多个服务位置数据,而与多个服务位置数据对应的多个服务场所数据所标识的范围(如,商户或商圈)可能不同。相应地,单个类簇可以具有多个场所标签,如,商户标签和商圈标签等。例如,某个类簇的场所标签可以包括商户标签(如,西部马华)和商圈标签(如,大钟寺)等。
此步骤中,确定与用户的位置数据对应的场所信息,可以包括:从聚类获得的多个类簇中确定与用户的位置数据对应的类簇。接着,根据确定出的类簇所对应的场所标签,确定场所信息。
在一个实施例中,位置映射模型中包括多个类簇中各个类簇的中心点的位置数据,以及多个类簇与多个场所标签之间的映射关系。如此,从聚类获得的多个类簇中确定与用户的位置数据对应的类簇,可以包括:计算用户的位置数据与各中心点的位置数据之间距离,将计算出的距离的最小值所对应的类簇作为与用户的位置数据对应的类簇。
进一步地,根据确定出的类簇所对应的场所标签,确定场所信息,可以包括:根据多个类簇与多个场所标签之间的映射关系,以及确定出的与用户的位置数据对应类簇,确定与该类簇对应的场所标签,进而确定场所信息。
在一个实施例中,确定出的场所标签包括商户标签(如,西部马华),可以据此确定出包括商户名称(如,西部马华)的场所信息。
在步骤S220中确定出与用户的位置数据对应的场所信息后,接着,在步骤S230,至少根据场所信息,以及内容信息库中各个内容信息的特征信息,确定向用户推荐的内容推荐信息。
具体地,确定向用户推荐的内容推荐信息,可以包括:确定与场所信息对应的服务类别信息,以及根据各个内容信息的特征信息,从内容信息库中确定与服务类别信息相关的内容信息,并将该内容信息作为内容推荐信息。
在一个实施例中,从内容信息库中确定与服务类别信息相关的内容信息,可以包括:根据各个内容信息的特征信息,计算各个内容信息与服务类别信息的相关度,并根据计算出的相关度确定与服务类别信息相关的内容信息。
在一个例子中,可以将相关度在预定范围内(如,预定范围可以为大于0.6)的内容信息确定为与服务类别信息相关的内容信息。在另一个例子中,可以根据相关度对内容信息进行排名,并将名次在预定范围(如,预定范围可以为前五名)内的内容信息确定为与服务类别信息相关的内容信息。
在另一个实施例中,特征信息可以包括内容信息的服务类别信息。从内容信息库中确定与服务类别信息相关的内容信息,可以包括:从内容信息库中确定与服务类别信息对应的内容信息。
在一个例子中,步骤S220中确定的场所信息为“汽车4S店”,相应地,可以确定出与该场所信息对应的服务类别信息为“汽车”。相应地,可以从内容信息库中确定服务类别信息为“汽车”的内容信息,并将此内容信息作为向用户推荐的内容推荐信息。
或者,各个内容信息的特征信息中可以包括预先确定的与该内容信息对应的场所信 息。在这样的情况下,确定向用户推荐的内容推荐信息,可以包括:当与各个内容信息对应的场所信息和与位置数据对应的场所信息之间的从属关系符合预设规则时,将该内容信息作为内容推荐信息。
在一个实施例中,预设规则可以由业务方根据业务的属性(如,该业务对场所范围的精度要求)确定。例如,当业务为向用户推送商家优惠券时,预设规则可以包括要求与内容推荐信息对应的场所信息与位置数据对应的场所信息完全相同。当业务为向用户推荐附近的类似的场所时,预设规则可以包括要求与内容推荐信息对应的场所信息与位置数据对应的场所信息部分相同。
在一个例子中,预设规则包括要求与内容推荐信息对应的场所信息与位置数据对应的场所信息完全相同,与位置数据对应的场所信息包括“老好吃香锅”、“三里屯”。相应地,可以将与场所信息“老好吃香锅”(商户)、“三里屯”(商圈)对应的内容信息(如,商户优惠活动、商户公众号信息、商户评价信息)作为内容推荐信息。
在另一个例子中,预设规则可以包括要求与内容推荐信息对应的场所信息与位置数据对应的场所信息部分相同,与位置数据对应的场所信息包括“优衣库”(商户)、“三里屯”(商圈)。相应地,可以将与场所信息“三里屯”对应的内容信息(如,三里屯商圈内的服装店、餐厅等)作为内容推荐信息。
在通过上述步骤确定向用户推荐的内容推荐信息后,还可以包括:向用户发送内容推荐信息。例如,当用户打开相应的应用App,或打开推荐频道时,或当检测到用户的位置发生变化时,获取用户最新的位置信息,据此确定向用户推荐的内容推荐信息,并向用户发送该内容推荐信息。
需要说明的是,在步骤S220中,还可以包括:确定与场所信息对应的服务类别信息。在一个实施例中,场所信息中除了包括场所的名称外,还可以包括与场所对应的服务类别信息,相应地,可以据此直接确定出与场所信息对应的服务类别信息。
在另一个实施例中,可以在确定场所信息后,根据预存的多个场所信息和多个服务类别信息的对应关系,确定该场所信息对应的服务类别信息。在一个例子中,确定出的场所标签包括商户标签(如,西部马华),可以据此确定出包括商户名称(如,西部马华)、以及与商户名称对应的服务类别信息(如,美食)。
在步骤S230中提及的特征信息可以基于从内容信息中提取的关键词信息和/或位置信息而确定,具体可以采用如图4所示的方法进行确定:
步骤S410,对内容信息进行预处理。
具体地,预处理可以包括结构化分析、分词处理、去停用词处理、词性标注(postag)等。其中,结构化分析可以包括对内容信息中段落结构的分析,例如,判断出内容信息中的标题和正文;分词处理可以包括一元分词(unigram)、二元分词(bigram)、三元分词(trigram)等;去停用词可以包括根据预设的停用词表去除内容信息中的停用词(如,无实际意义的功能词:这、那、的);词性标注是指对内容信息中的词语的词性(如,名词、副词、形容词等)进行标注。
步骤S420,根据预处理后的内容信息,提取关键词信息。
具体地,可以根据预处理得到的词语在内容信息中的位置(如,位于标题中或位于正文中)、标注的词性,采用TextRank算法或TF-IDF(Term Frequency–inverse Document Frequency)算法加权识别出关键词信息。
步骤S430,根据预处理后的内容信息,提取位置信息。
具体地,可以根据内容信息中包括的位置标签提取位置信息,位置信息中可以包括场所信息。在一个实施例中,该位置标签可以为内容信息的生产者在发布该内容信息时,为其所贴的位置标签,如,北京、杭州等。
或者,可以采用命名实体识别(Named Entity Recognition,简称NER)的方法,识别内容信息中的地名、机构名等。例如,可以识别出内容信息中的地名为五台山,机构名为海淀区民政局(属于场所信息)。
又或者,服务器中可以预先存储有位置信息库,可以据此从预处理后的内容信息中提取出与所述位置信息库中的信息匹配的位置信息。
步骤S440,将初步提取的关键词信息和位置信息输入预先训练的特征提取模型中,确定出特征信息。
具体地,特征提取模型可以为对大规模内容语料数据进行离线训练,而产出的词嵌入(Word Embedding)模型或基于双向循环神经网络(Recurrent Neural Networks,简称RNN)的NER模型。
此外,上述特征信息的确定可以在内容信息产生以后的预定时间(如,5min或10min)内进行。
由上可知,在本说明书披露的多个实施例提供的内容推荐方法中,通过获取用 户的位置数据,并根据位置映射模型,确定与位置数据对应的场所信息。然后,至少根据该场所信息,以及内容信息库中各个内容信息的特征信息,确定向所述用户推荐的内容推荐信息,从而及时地向用户推荐更加精准的内容信息。
与内容推荐方法对应地,本说明书披露的多个实施例还提供一种内容推荐装置,如图5所示,该装置500包括:
获取单元510,用于获取用户的位置数据;
确定单元520,用于根据位置映射模型,确定与位置数据对应的场所信息,该位置映射模型基于预先获取的服务信息而训练获得,该服务信息中包含服务位置数据和对应的服务场所数据;
处理单元530,用于至少根据场所信息,以及内容信息库中各个内容信息的特征信息,确定向用户推荐的内容推荐信息。
在一种可能的实施方式中,位置映射模型由确定单元520通过以下步骤训练获得:
对服务信息中包含的服务位置数据进行聚类,获得多个类簇;
根据服务信息中与服务位置数据对应的服务场所数据,确定多个类簇对应的场所标签。
在一种可能的实施方式中,确定单元520具体包括:
第一确定子单元521,用于从多个类簇中确定与用户的位置数据对应的类簇;
第二确定子单元522,用于根据确定出的类簇所对应的场所标签,确定场所信息。
在一种可能的实施方式中,处理单元530具体包括:
第一处理子单元531,用于确定与场所信息对应的服务类别信息;
第二处理子单元532,用于根据各个内容信息的特征信息,从内容信息库中确定与服务类别信息相关的内容信息,并将内容信息作为内容推荐信息。
在一种可能的实施方式中,第二处理子单元532中的特征信息包括内容信息的服务类别信息,第二处理子单元532具体用于:
从内容信息库中确定与服务类别信息对应的内容信息。
在一种可能的实施方式中,处理单元530中包括的各个内容信息的特征信息包 括预先确定的与该内容信息对应的场所信息,处理单元530具体用于:
当与各个内容信息对应的场所信息和与位置数据对应的场所信息之间的从属关系符合预设规则时,将该内容信息作为内容推荐信息。
在一种可能的实施方式中,处理单元530中包括的特征信息基于从内容信息中提取出的关键词信息和场所信息而确定。
在一种可能的实施方式中,其特征在于,获取单元510获取的位置数据包括无线保真WiFi指纹数据和经纬度数据中的至少一种。
在一种可能的实施方式中,确定单元520确定的场所信息包括建筑名称、商户名称和商圈信息中的至少一种。
由上可知,在本说明书披露的多个实施例提供的内容推荐装置中,获取单元510获取用户的位置数据,确定单元520根据位置映射模型,确定与位置数据对应的场所信息。处理单元530至少根据该场所信息,以及内容信息库中各个内容信息的特征信息,确定向所述用户推荐的内容推荐信息,从而及时地向用户推荐更加精准的内容信息。
本领域技术人员应该可以意识到,在上述一个或多个示例中,本说明书披露的多个实施例所描述的功能可以用硬件、软件、固件或它们的任意组合来实现。当使用软件实现时,可以将这些功能存储在计算机可读介质中或者作为计算机可读介质上的一个或多个指令或代码进行传输。
以上所述的具体实施方式,对本说明书披露的多个实施例的目的、技术方案和有益效果进行了进一步详细说明,所应理解的是,以上所述仅为本说明书披露的多个实施例的具体实施方式而已,并不用于限定本说明书披露的多个实施例的保护范围,凡在本说明书披露的多个实施例的技术方案的基础之上,所做的任何修改、等同替换、改进等,均应包括在本说明书披露的多个实施例的保护范围之内。

Claims (18)

  1. 一种内容推荐方法,其特征在于,包括:
    获取用户的位置数据;
    根据位置映射模型,确定与所述位置数据对应的场所信息,所述位置映射模型基于预先获取的服务信息而训练获得,所述服务信息中包含服务位置数据和对应的服务场所数据;
    至少根据所述场所信息,以及内容信息库中各个内容信息的特征信息,确定向所述用户推荐的内容推荐信息。
  2. 根据权利要求1所述的方法,其特征在于,所述位置映射模型通过以下步骤训练获得:
    对所述服务信息中包含的服务位置数据进行聚类,获得多个类簇;
    根据所述服务信息中与所述服务位置数据对应的服务场所数据,确定所述多个类簇对应的场所标签。
  3. 根据权利要求2所述的方法,其特征在于,所述确定与所述位置数据对应的场所信息包括:
    从所述多个类簇中确定与所述用户的位置数据对应的类簇;
    根据确定出的类簇所对应的场所标签,确定所述场所信息。
  4. 根据权利要求1所述的方法,其特征在于,所述确定向所述用户推荐的内容推荐信息,包括:
    确定与所述场所信息对应的服务类别信息;
    根据所述各个内容信息的特征信息,从所述内容信息库中确定与所述服务类别信息相关的内容信息,并将所述内容信息作为所述内容推荐信息。
  5. 根据权利要求4所述的方法,其特征在于,所述特征信息包括所述内容信息的所述服务类别信息,所述从所述内容信息库中确定与所述服务类别信息相关的内容信息,包括:
    从所述内容信息库中确定与所述服务类别信息对应的所述内容信息。
  6. 根据权利要求1所述的方法,其特征在于,所述各个内容信息的特征信息包括预先确定的与该内容信息对应的场所信息,所述确定向所述用户推荐的内容推荐信息,包括:
    当与所述各个内容信息对应的场所信息和与所述位置数据对应的场所信息之间的从属关系符合预设规则时,将该内容信息作为所述内容推荐信息。
  7. 根据权利要求1-6任一项所述的方法,其特征在于,所述特征信息基于从所述内容信息中提取出的关键词信息和场所信息而确定。
  8. 根据权利要求1-6中任一项所述的方法,其特征在于,所述位置数据包括无线保真WiFi指纹数据和经纬度数据中的至少一种。
  9. 根据权利要求1-6任一项所述的方法,其特征在于,所述场所信息包括建筑名称、商户名称和商圈信息中的至少一种。
  10. 一种内容推荐装置,其特征在于,包括:
    获取单元,用于获取用户的位置数据;
    确定单元,用于根据位置映射模型,确定与所述位置数据对应的场所信息,所述位置映射模型基于预先获取的服务信息而训练获得,所述服务信息中包含服务位置数据和对应的服务场所数据;
    处理单元,用于至少根据所述场所信息,以及内容信息库中各个内容信息的特征信息,确定向所述用户推荐的内容推荐信息。
  11. 根据权利要求10所述的装置,其特征在于,所述位置映射模型由所述确定单元通过以下步骤训练获得:
    对所述服务信息中包含的服务位置数据进行聚类,获得多个类簇;
    根据所述服务信息中与所述服务位置数据对应的服务场所数据,确定所述多个类簇对应的场所标签。
  12. 根据权利要求11所述的装置,其特征在于,所述确定单元具体包括:
    第一确定子单元,用于从所述多个类簇中确定与所述用户的位置数据对应的类簇;
    第二确定子单元,用于根据确定出的类簇所对应的场所标签,确定所述场所信息。
  13. 根据权利要求10所述的装置,其特征在于,所述处理单元具体包括:
    第一处理子单元,用于确定与所述场所信息对应的服务类别信息;
    第二处理子单元,用于根据所述各个内容信息的特征信息,从所述内容信息库中确定与所述服务类别信息相关的内容信息,并将所述内容信息作为所述内容推荐信息。
  14. 根据权利要求13所述的装置,其特征在于,所述第二处理子单元中的特征信息包括所述内容信息的所述服务类别信息,所述第二处理子单元具体用于:
    从所述内容信息库中确定与所述服务类别信息对应的所述内容信息。
  15. 根据权利要求10所述的装置,其特征在于,所述处理单元中包括的各个内容信息的特征信息包括预先确定的与该内容信息对应的场所信息,所述处理单元具体用于:
    当与所述各个内容信息对应的场所信息和与所述位置数据对应的场所信息之间的从属关系符合预设规则时,将该内容信息作为所述内容推荐信息。
  16. 根据权利要求10-15任一项所述的装置,其特征在于,所述处理单元中包括的特征信息基于从所述内容信息中提取出的关键词信息和场所信息而确定。
  17. 根据权利要求10-15中任一项所述的装置,其特征在于,所述获取单元获取的位置数据包括无线保真WiFi指纹数据和经纬度数据中的至少一种。
  18. 根据权利要求10-15任一项所述的装置,其特征在于,所述确定单元确定的场所信息包括建筑名称、商户名称和商圈信息中的至少一种。
PCT/CN2019/070830 2018-01-17 2019-01-08 内容推荐方法及装置 WO2019141109A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810043114.2A CN108363733A (zh) 2018-01-17 2018-01-17 内容推荐方法及装置
CN201810043114.2 2018-01-17

Publications (1)

Publication Number Publication Date
WO2019141109A1 true WO2019141109A1 (zh) 2019-07-25

Family

ID=63006224

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/070830 WO2019141109A1 (zh) 2018-01-17 2019-01-08 内容推荐方法及装置

Country Status (3)

Country Link
CN (1) CN108363733A (zh)
TW (1) TWI703862B (zh)
WO (1) WO2019141109A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113807916A (zh) * 2021-09-02 2021-12-17 支付宝(杭州)信息技术有限公司 服务推荐处理方法及装置

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108363733A (zh) * 2018-01-17 2018-08-03 阿里巴巴集团控股有限公司 内容推荐方法及装置
CN110909250B (zh) * 2018-09-14 2023-05-02 阿里巴巴集团控股有限公司 信息处理方法及装置、存储介质、处理器
CN109409959A (zh) * 2018-10-31 2019-03-01 广州品唯软件有限公司 一种用户信息分析方法、装置、设备及介质
CN111723959B (zh) * 2019-03-19 2023-12-12 腾讯科技(深圳)有限公司 区域的划分方法、装置、存储介质及电子装置
CN110377195B (zh) * 2019-07-15 2022-09-30 腾讯科技(深圳)有限公司 展示交互功能的方法和装置
CN112395486B (zh) * 2019-08-12 2023-11-03 中国移动通信集团重庆有限公司 一种宽带业务推荐方法、系统、服务器和存储介质
CN111815361A (zh) * 2020-07-10 2020-10-23 北京思特奇信息技术股份有限公司 区域边界计算方法、装置、电子设备及存储介质
KR102412057B1 (ko) * 2021-06-07 2022-06-23 쿠팡 주식회사 스토어 정보 제공을 위한 전자 장치의 동작 방법 및 이를 지원하는 전자 장치

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090094189A1 (en) * 2007-10-08 2009-04-09 At&T Bls Intellectual Property, Inc. Methods, systems, and computer program products for managing tags added by users engaged in social tagging of content
CN103023977A (zh) * 2012-11-19 2013-04-03 华南理工大学 基于位置信息的推荐系统及推荐方法
CN106294489A (zh) * 2015-06-08 2017-01-04 北京三星通信技术研究有限公司 内容推荐方法、装置及系统
CN108363733A (zh) * 2018-01-17 2018-08-03 阿里巴巴集团控股有限公司 内容推荐方法及装置

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8229458B2 (en) * 2007-04-08 2012-07-24 Enhanced Geographic Llc Systems and methods to determine the name of a location visited by a user of a wireless device
US9699491B1 (en) * 2014-10-10 2017-07-04 ThinkAnalytics Content recommendation engine
CN106776776A (zh) * 2016-11-11 2017-05-31 广东小天才科技有限公司 一种运动场所信息的推荐方法及装置
CN107391605A (zh) * 2017-06-30 2017-11-24 北京奇虎科技有限公司 基于地理位置的信息推送方法、装置及移动终端
CN107545052A (zh) * 2017-08-23 2018-01-05 广东欧珀移动通信有限公司 信息推荐方法、装置、移动终端及存储介质

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090094189A1 (en) * 2007-10-08 2009-04-09 At&T Bls Intellectual Property, Inc. Methods, systems, and computer program products for managing tags added by users engaged in social tagging of content
CN103023977A (zh) * 2012-11-19 2013-04-03 华南理工大学 基于位置信息的推荐系统及推荐方法
CN106294489A (zh) * 2015-06-08 2017-01-04 北京三星通信技术研究有限公司 内容推荐方法、装置及系统
CN108363733A (zh) * 2018-01-17 2018-08-03 阿里巴巴集团控股有限公司 内容推荐方法及装置

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113807916A (zh) * 2021-09-02 2021-12-17 支付宝(杭州)信息技术有限公司 服务推荐处理方法及装置

Also Published As

Publication number Publication date
CN108363733A (zh) 2018-08-03
TWI703862B (zh) 2020-09-01
TW201933879A (zh) 2019-08-16

Similar Documents

Publication Publication Date Title
WO2019141109A1 (zh) 内容推荐方法及装置
CN109145219B (zh) 基于互联网文本挖掘的兴趣点有效性判断方法和装置
JP6759844B2 (ja) 画像を施設に対して関連付けるシステム、方法、プログラム及び装置
CN107315759B (zh) 归类关键字的方法、装置和处理系统、分类模型生成方法
US10694324B2 (en) Method and apparatus for matching wireless hotspot with POI
CN110472066B (zh) 一种城市地理语义知识图谱的构建方法
CN108153824B (zh) 目标用户群体的确定方法及装置
US20150120723A1 (en) Methods and systems for processing speech queries
US10176244B2 (en) Text characterization of trajectories
CN109255564B (zh) 一种取件点地址推荐方法及装置
JP2017045435A (ja) ソーシャルメディアメッセージ及び施設の間のリンクを推定する方法、コンピュータシステム、及びプログラム
CN111488468B (zh) 地理信息知识点抽取方法、装置、存储介质及计算机设备
US20130310078A1 (en) Method and system for associating context information with defined location data
CN111291158B (zh) 一种信息查询方法、装置、电子设备及存储介质
US11675845B2 (en) Identifying merchant data associated with multiple data structures
JP2018537760A (ja) アドレス情報に基づいたアカウントマッピングの方法及び装置
US11082800B2 (en) Method and system for determining an occurrence of a visit to a venue by a user
WO2016101881A1 (zh) 通过计算机实现的信息推荐方法、装置和计算机设备
Lee et al. Spatio-temporal provenance: Identifying location information from unstructured text
CN112819544A (zh) 基于大数据的广告投放方法、装置、设备及存储介质
US9361633B1 (en) Determining present venue for a user device
CN110674208B (zh) 用于确定用户的职住地信息的方法和装置
CN114253990A (zh) 数据库查询方法、装置、计算机设备和存储介质
US10970774B1 (en) Systems and methods for locating items
US20140330838A1 (en) Content Distribution

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: 19740988

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19740988

Country of ref document: EP

Kind code of ref document: A1