WO2018040310A1 - Artificial intelligence-based recommended data acquisition method, apparatus and device, and non-volatile computer storage medium - Google Patents

Artificial intelligence-based recommended data acquisition method, apparatus and device, and non-volatile computer storage medium Download PDF

Info

Publication number
WO2018040310A1
WO2018040310A1 PCT/CN2016/107077 CN2016107077W WO2018040310A1 WO 2018040310 A1 WO2018040310 A1 WO 2018040310A1 CN 2016107077 W CN2016107077 W CN 2016107077W WO 2018040310 A1 WO2018040310 A1 WO 2018040310A1
Authority
WO
WIPO (PCT)
Prior art keywords
entity
recommended
data
feature data
recommendation
Prior art date
Application number
PCT/CN2016/107077
Other languages
French (fr)
Chinese (zh)
Inventor
刘凯
吕雅娟
Original Assignee
北京百度网讯科技有限公司
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 北京百度网讯科技有限公司 filed Critical 北京百度网讯科技有限公司
Publication of WO2018040310A1 publication Critical patent/WO2018040310A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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
    • G06F16/957—Browsing optimisation, e.g. caching or content distillation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2216/00—Indexing scheme relating to additional aspects of information retrieval not explicitly covered by G06F16/00 and subgroups
    • G06F2216/03—Data mining

Definitions

  • the present invention relates to query technology, and in particular, to a method, an apparatus, a device and a non-volatile computer storage medium for acquiring recommended data based on artificial intelligence.
  • Artificial Intelligence abbreviated as AI in English. It is a new technical science that studies and develops theories, methods, techniques, and applications for simulating, extending, and extending human intelligence. Artificial intelligence is a branch of computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that responds in a manner similar to human intelligence. Research in this area includes robotics, speech recognition, image recognition, Natural language processing and expert systems.
  • the terminal integrates more and more functions, so that the system function list of the terminal contains more and more corresponding applications (Application, APP).
  • Some applications involve some artificial intelligence-based recommendation data acquisition services, such as Baidu maps, Baidu glutinous rice, etc., users can select the entities corresponding to the recommended data according to the recommendation data, such as restaurants, movies, and so on.
  • the recommendation data such as restaurants, movies, and so on.
  • aspects of the present invention provide a method, an apparatus, and a device for acquiring recommendation data based on artificial intelligence, and a non-volatile computer storage medium for improving the reliability of entity recommendation.
  • An aspect of the present invention provides a method for acquiring recommendation data based on artificial intelligence, including:
  • the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and global feature data of the entity class to which the entity to be recommended belongs; Operation copy information characteristic data of the entity category to which the entity to be recommended belongs; and operation copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs;
  • the recommendation data of the entity to be recommended is obtained according to the specified number of description data with the highest recommendation score.
  • the acquiring physical feature data of the entity to be recommended includes at least one of the following steps:
  • the operating copy information of the entity category to which the entity to be recommended belongs belongs to obtain the physical feature data of the entity to be recommended
  • the operation copy information of the other entity categories other than the entity category to which the entity to be recommended belongs is mined to obtain the entity feature data of the entity to be recommended.
  • the at least one description is obtained according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended Recommended scores for each description data in the data, including:
  • the foregoing aspect, and any possible implementation manner further provide an implementation manner, where the entity feature data according to the to-be-recommended entity, at least one description data of the to-be-recommended entity, and each description data Emotion data, obtaining a recommendation score for each of the at least one description data, including:
  • the obtaining the recommendation data of the entity to be recommended according to the specified number of description data with the highest recommendation score including:
  • Another aspect of the present invention provides an apparatus for acquiring recommendation data based on artificial intelligence, including:
  • An acquiring unit configured to acquire entity feature data of the entity to be recommended;
  • the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; Global feature data; operational copy information characteristic data of the entity category to which the entity to be recommended belongs; and operational copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs;
  • a scoring unit configured to obtain, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data
  • a recommendation unit configured to obtain recommendation data of the entity to be recommended according to the specified number of description data with the highest recommendation score.
  • the obtaining unit is specifically configured to perform at least one of the following steps:
  • the operating copy information of the entity category to which the entity to be recommended belongs belongs to obtain the physical feature data of the entity to be recommended
  • the operation copy information of the other entity categories other than the entity category to which the entity to be recommended belongs is mined to obtain the entity feature data of the entity to be recommended.
  • an apparatus comprising:
  • One or more processors are One or more processors;
  • One or more programs the one or more programs being stored in the memory, when executed by the one or more processors:
  • the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and global feature data of the entity class to which the entity to be recommended belongs; Operation copy information characteristic data of the entity category to which the entity to be recommended belongs; and operation copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs;
  • the recommendation data of the entity to be recommended is obtained according to the specified number of description data with the highest recommendation score.
  • a nonvolatile computer storage medium stores one or more programs that, when executed by a device, cause the device to:
  • the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and global feature data of the entity class to which the entity to be recommended belongs; Operation copy information characteristic data of the entity category to which the entity to be recommended belongs; and operation copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs;
  • the recommendation data of the entity to be recommended is obtained according to the specified number of description data with the highest recommendation score.
  • the embodiment of the present invention obtains the entity feature data of the entity to be recommended; the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; Global feature data of the entity category to which the entity to be recommended belongs; the business copy information feature data of the entity category to which the entity to be recommended belongs; and the operational copy information characteristic data of the entity category other than the entity category to which the entity to be recommended belongs, and further Obtaining, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data, so that the specified number of description data according to the recommendation score is the highest
  • the recommendation data of the entity to be recommended is obtained, and the recommendation data of the entity to be recommended is generated by using the entity feature data of the entity to be recommended, so that all aspects related to the entity to be recommended are comprehensively considered, so that the entity to be recommended is very Easy to be adopted by users, thus Recommended entity Rel
  • the erroneous transmission of the negative information can be controlled from the source.
  • FIG. 1 is a schematic flowchart of a method for acquiring recommendation data based on artificial intelligence according to an embodiment of the present invention
  • FIG. 2 is a schematic structural diagram of an apparatus for acquiring recommendation data based on artificial intelligence according to another embodiment of the present invention.
  • the terminals involved in the embodiments of the present invention may include, but are not limited to, a mobile phone, a personal digital assistant (PDA), a wireless handheld device, a tablet computer, and a personal computer (Personal Computer, PC). ), MP3 player, MP4 player, wearable device (for example, smart glasses, smart watches, smart bracelets, etc.).
  • PDA personal digital assistant
  • PC Personal Computer
  • FIG. 1 is a schematic flowchart of a method for acquiring recommendation data based on artificial intelligence according to an embodiment of the present invention, as shown in FIG. 1 .
  • the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and global feature of the entity class to which the entity to be recommended belongs Data; operational copy information characteristic data of the entity category to which the entity to be recommended belongs; and operational copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs.
  • execution entities of 101 to 103 may be applications located in the local terminal, or may be plug-ins or software development kits (SDKs) installed in applications located in the local terminal.
  • the functional unit may also be a processing engine located in the network side server, or may be a distributed system located on the network side, which is not specifically limited in this embodiment.
  • the application may be a local application (nativeApp) installed on the terminal, or may be a web application (webApp) of the browser on the terminal, which is not limited in this embodiment.
  • the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and a global class of the entity class to which the entity to be recommended belongs Feature data; operation copy information feature data of the entity category to which the entity to be recommended belongs; and operation copy information feature data of other entity categories other than the entity category to which the entity to be recommended belongs, and further according to the entity of the entity to be recommended Feature data and at least one description data of the entity to be recommended, obtaining a recommendation score of each description data in the at least one description data, so that the entity to be recommended is obtained according to a specified number of description data with the highest recommendation score
  • the recommended data because the entity feature data of the entity to be recommended is used to generate the recommendation data of the entity to be recommended, so that all aspects related to the entity to be recommended can be comprehensively considered, so that the entity to be recommended is very It is easily adopted by users, which improves the reliability of entity recommendations.
  • the view information is used to ensure that the recommended data generated by the mining is actually related to the entity to be recommended, for example, the vocabulary information amount, the N-gram frequency (N-Gram) frequency, the low-dimensional feature information of the recommended entity, or the comment dimension , topic information (Topic) information, semantic vector information, context logic relationship information waiting for high-dimensional feature information of the recommended entity, or may also be a popular keyword of the entity to be recommended, and the like;
  • the entity feature data of the entity to be recommended that is, the global feature data of the entity class to which the entity to be recommended belongs
  • the comment dimension information and the comment differentiation information for example, the vocabulary information amount, the N-gram frequency (N-Gram) frequency, waiting for the low-dimensional feature information of the entity category to which the recommendation entity belongs, or may also be the comment dimension, the topic (Topic) information, semantic vector information, context logic relationship information waiting for high-dimensional feature information of the entity category to which the recommendation entity belongs, and the like;
  • the operating copy information of the entity category to which the entity to be recommended belongs is obtained to obtain the entity feature data of the entity to be recommended, that is, the operating copy information characteristic data of the entity category to which the entity to be recommended belongs, which is mainly used for mining the highlight copy
  • Similar comments of the style are used as recommendation data, for example, low-dimensional feature information of a vocabulary information amount, an N-gram frequency, or the like, or may be a topic (Topic) information, semantic vector information, and context.
  • High-dimensional feature information of operational copy files such as logical relationship information, or lexical information, Basic information about operational documents such as syntactic information, etc.;
  • the operational document information feature data of the entity category is mainly used to mine similar comments with a highlight copy style as recommendation data, for example, low-dimensional feature information of the operational copy such as vocabulary information quantity and N-gram frequency (N-Gram) frequency. Or, it may be high-dimensional feature information of a business copy such as Topic information, semantic vector information, and context logic relationship information, or may also be basic feature information of a business copy such as lexical information, syntax information, and the like.
  • the sentiment data of each of the description data may be specifically acquired, and then, according to the entity feature data of the entity to be recommended, At least one of the description data of the entity to be recommended and the sentiment data of each of the description data obtains a recommendation score for each of the at least one description data.
  • the so-called sentiment data describing the data is used to indicate the sentimental tendency of the description data, which may include, but is not limited to, a positive emotional tendency and a negative emotional tendency, or may also include different degrees of positive emotional tendency and negative emotional tendency, this embodiment No particular limitation is imposed.
  • the description data corresponding to the sentiment data indicating the positive sentiment tendency may be selected according to the sentiment data of each of the description data, and then, according to the entity feature data of the to-be-recommended entity and the selected description. Data, obtaining a recommendation score for each of the description data in the selected description data. In this way, by performing subsequent analysis processing by selecting the description data corresponding to the emotion data indicating the positive emotion tendency, it is possible to control the erroneous transmission of the negative information from the source.
  • a pre-built bright spot recommendation model may be specifically used to perform the specific step of 102, that is, according to the physical feature data and the location of the entity to be recommended. At least one description data of the recommended entity is referred to, and a recommendation score of each of the at least one description data is obtained.
  • the acquired entity feature data of the to-be-recommended entity may be used to perform corresponding feature mining and calculation for each description data to generate corresponding entity features. Then, the generated entity features and their corresponding weights are input into the highlight recommendation model, and analyzed by the highlight recommendation model to output a recommendation score for each description data.
  • the constructed bright spot recommendation model may include, but is not limited to, a log-linear model, a support vector machine (SVM) model, a regression tree/forest model, and this embodiment does not Special restrictions are made.
  • SVM support vector machine
  • model parameters of the highlight recommendation model may be adjusted according to the execution result of 102 to improve the effect of the highlight recommendation model.
  • a pre-built recommended text model may be specifically used to perform the specific step of 103, that is, according to the specified number of description data with the highest recommended score.
  • the recommendation data of the entity to be recommended may be specifically used to perform the specific step of 103, that is, according to the specified number of description data with the highest recommended score.
  • the recommended text model may include, but is not limited to, a log-linear model, a support vector machine (SVM) model, a maximum entropy (MaxEnt) model, and a regression tree.
  • SVM support vector machine
  • MaxEnt maximum entropy
  • Forest model this embodiment is not particularly limited.
  • model parameters of the recommended text model may be adjusted according to the execution result of 103 to improve the effect of the recommended text model.
  • the at least one recommended text is selected according to the specified number of description data with the highest recommendation score. Further, the recommendation data of the to-be-recommended entity may be generated according to the at least one recommended text.
  • the entity feature data of the entity to be recommended is obtained by the at least one of the following data: the local feature data of the entity to be recommended; the entity to be recommended by the entity to be recommended Global signature data of the category; operation copy information feature data of the entity category to which the entity to be recommended belongs; and operational copy information feature data of other entity categories other than the entity category to which the entity to be recommended belongs, and then according to the to-be-recommended Obtaining, by the entity feature data of the entity and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data, so that the to-be-obtained can be obtained according to the specified number of description data with the highest recommendation score.
  • the recommendation data of the recommendation entity is used to generate the recommendation data of the entity to be recommended by using the entity feature data of the entity to be recommended, so that all aspects related to the entity to be recommended can be comprehensively considered, so that the entity to be recommended is easily adopted by the user. , thereby improving the reliability of the entity recommendation
  • the erroneous transmission of the negative information can be controlled from the source.
  • FIG. 2 is a schematic structural diagram of an apparatus for acquiring recommendation data based on artificial intelligence according to another embodiment of the present invention, as shown in FIG. 2 .
  • the apparatus for acquiring artificial intelligence-based recommendation data of the present embodiment may include an acquisition unit 21, a scoring unit 22, and a recommendation unit 23.
  • the acquiring unit 21 is configured to acquire the entity feature data of the entity to be recommended, and the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; the entity to be recommended Global signature data of the entity category; operational copy information feature data of the entity category to which the entity to be recommended belongs; and operational copy information feature data of other entity categories other than the entity category to which the entity to be recommended belongs; scoring unit 22, And obtaining, by the entity feature data of the to-be-recommended entity and at least one description data of the to-be-recommended entity, a recommendation score of each description data in the at least one description data; and a recommendation unit 23, configured to use the highest recommendation score A specified number of description data, obtaining recommendation data of the entity to be recommended.
  • part or all of the apparatus for acquiring the artificial intelligence-based recommendation data provided by the embodiment may be an application located in the local terminal, or may be a plug-in or a software development tool disposed in an application located in the local terminal.
  • Software Development The functional unit such as Kit, SDK, or the like may be a processing engine located in the network side server, or may be a distributed system located on the network side, which is not specifically limited in this embodiment.
  • the application may be a local application (nativeApp) installed on the terminal, or may be a web application (webApp) of the browser on the terminal, which is not limited in this embodiment.
  • the acquiring unit 21 may be specifically configured to perform at least one of the following steps:
  • the operating copy information of the entity category to which the entity to be recommended belongs belongs to obtain the physical feature data of the entity to be recommended
  • the operation copy information of the other entity categories other than the entity category to which the entity to be recommended belongs is mined to obtain the entity feature data of the entity to be recommended.
  • the scoring unit 22 may be specifically configured to acquire sentiment data of each of the description data; and according to physical feature data of the entity to be recommended, At least one description data of the recommendation entity and the sentiment data of each of the description data are referred to, and a recommendation score of each of the at least one description data is obtained.
  • the scoring unit 22 may be specifically configured to: according to the sentiment data of each of the description data, select description data corresponding to the sentiment data indicating the positive sentiment tendency; and the entity feature data according to the entity to be recommended And selected description data to get selected The recommended score for each description data in the description data.
  • the recommending unit 23 may be specifically configured to select at least one recommended text according to the specified number of description data with the highest recommended score; and according to the at least one recommendation Text, generating recommendation data of the entity to be recommended.
  • the method in the embodiment corresponding to FIG. 1 can be implemented by the apparatus for acquiring artificial intelligence based recommendation data provided by this embodiment.
  • the apparatus for acquiring artificial intelligence based recommendation data provided by this embodiment.
  • the entity feature data of the entity to be recommended is obtained by the acquiring unit; the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; the entity to be recommended Global feature data of the entity class; the business copy information feature data of the entity category to which the entity to be recommended belongs; and the operational copy information feature data of the other entity categories other than the entity category to which the entity to be recommended belongs, and further by the scoring unit Obtaining, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data, so that the recommendation unit can be according to the specified number of the highest recommendation score Descriptive data, obtaining recommendation data of the entity to be recommended, and using the entity feature data of the entity to be recommended to generate recommendation data of the entity to be recommended, so that all aspects related to the entity to be recommended can be comprehensively considered, so that the to-be-recommended Entities are easily adopted by users.
  • the erroneous transmission of the negative information can be controlled from the source.
  • 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 invention may be integrated into one processing unit, or each unit may exist physically separately, or may have two or more unit sets. In 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 above software functional unit 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 of the various embodiments of the present invention. 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)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The present invention provides an artificial intelligence-based recommended data acquisition method, apparatus and device, and a non-volatile computer storage medium. The method provided by an embodiment of the present invention comprises: obtaining entity feature data of an entity to be recommended, the entity feature data of the entity to be recommended comprising at least one of the following data: local feature data of the entity to be recommended, global feature data of an entity category to which the entity to be recommended belongs, operation copy information feature data of the entity category to which the entity to be recommended belongs, and operation copy information feature data of other entity categories except the entity category to which the entity to be recommended belongs; and then obtaining, according to the entity feature data of the entity to be recommended and at least one piece of descriptive data of the entity to be recommended, a recommendation score of each piece of descriptive data in the at least one piece of descriptive data, so that recommended data of the entity to be recommended can be obtained according to a specified number of pieces of descriptive data having the highest recommendation scores. Because the recommended data of the entity to be recommended is generated by using the entity feature data of the entity to be recommended, related various information of the entity to be recommended can be comprehensively considered, so that the entity to be recommended can be easily adopted by a user, and the reliability of entity recommendation is improved.

Description

基于人工智能的推荐数据的获取方法、装置、设备及非易失性计算机存储介质Method, device, device and non-volatile computer storage medium for acquiring recommendation data based on artificial intelligence
本申请要求了申请日为2016年09月05日，申请号为201610801268.4发明名称为“基于人工智能的推荐数据的获取方法及装置”的中国专利申请的优先权。The present application claims the priority of the Chinese patent application whose filing date is September 5, 2016, and whose application number is 201610801268.4, which is entitled "Acquisition Method and Apparatus for Recommendation Data Based on Artificial Intelligence".
技术领域Technical field
本发明涉及查询技术，尤其涉及一种基于人工智能的推荐数据的获取方法、装置、设备及非易失性计算机存储介质。The present invention relates to query technology, and in particular, to a method, an apparatus, a device and a non-volatile computer storage medium for acquiring recommended data based on artificial intelligence.
背景技术Background technique
人工智能(Artificial Intelligence)，英文缩写为AI。它是研究、开发用于模拟、延伸和扩展人的智能的理论、方法、技术及应用系统的一门新的技术科学。人工智能是计算机科学的一个分支，它企图了解智能的实质，并生产出一种新的能以人类智能相似的方式做出反应的智能机器，该领域的研究包括机器人、语言识别、图像识别、自然语言处理和专家系统等。Artificial Intelligence, abbreviated as AI in English. It is a new technical science that studies and develops theories, methods, techniques, and applications for simulating, extending, and extending human intelligence. Artificial intelligence is a branch of computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that responds in a manner similar to human intelligence. Research in this area includes robotics, speech recognition, image recognition, Natural language processing and expert systems.
随着通信技术的发展，终端集成了越来越多的功能，从而使得终端的系统功能列表中包含了越来越多相应的应用(Application，APP)。有些应用中会涉及一些基于人工智能的推荐数据的获取服务，例如，百度地图、百度糯米等，用户可以根据推荐数据，选择推荐数据所对应的实体，例如，餐厅、电影等。目前，通常可以挖掘用户评论信息中的高频文本，将这些高频文本整理成完成的推荐数据。 With the development of communication technology, the terminal integrates more and more functions, so that the system function list of the terminal contains more and more corresponding applications (Application, APP). Some applications involve some artificial intelligence-based recommendation data acquisition services, such as Baidu maps, Baidu glutinous rice, etc., users can select the entities corresponding to the recommended data according to the recommendation data, such as restaurants, movies, and so on. At present, it is usually possible to mine the high frequency text in the user comment information and organize the high frequency text into the completed recommendation data.
然而，由于用户评论信息中的高频文本的表述一般具有普遍性和局限性，使得依据这些高频文本所生成的推荐数据被用户所采纳的可能性并不是很高，从而导致了实体推荐的可靠性的降低。However, since the representation of high frequency text in the user comment information is generally universal and limited, the possibility that the recommendation data generated based on the high frequency text is adopted by the user is not very high, resulting in the entity recommendation. Reduced reliability.
发明内容Summary of the invention
本发明的多个方面提供一种基于人工智能的推荐数据的获取方法、装置、设备及非易失性计算机存储介质，用以提高实体推荐的可靠性。Aspects of the present invention provide a method, an apparatus, and a device for acquiring recommendation data based on artificial intelligence, and a non-volatile computer storage medium for improving the reliability of entity recommendation.
本发明的一方面，提供一种基于人工智能的推荐数据的获取方法，包括：An aspect of the present invention provides a method for acquiring recommendation data based on artificial intelligence, including:
获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据；Acquiring the entity feature data of the entity to be recommended; the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and global feature data of the entity class to which the entity to be recommended belongs; Operation copy information characteristic data of the entity category to which the entity to be recommended belongs; and operation copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs;
根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分；Obtaining, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data;
根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据。The recommendation data of the entity to be recommended is obtained according to the specified number of description data with the highest recommendation score.
如上所述的方面和任一可能的实现方式，进一步提供一种实现方式，所述获取待推荐实体的实体特征数据，包括下列步骤中的至少一项：The aspect as described above, and any possible implementation manner, further provide an implementation manner, where the acquiring physical feature data of the entity to be recommended includes at least one of the following steps:
对所述待推荐实体的全部用户评论信息，进行挖掘，以获得所述待 推荐实体的实体特征数据；Performing excavation on all user comment information of the entity to be recommended to obtain the Entity entity data of the recommendation entity;
对所述待推荐实体所属实体类别下的全部实体的全部用户评论信息，进行挖掘，以获得所述待推荐实体的实体特征数据；Performing all the user comment information of all the entities in the entity category to which the entity to be recommended belongs to obtain the entity feature data of the entity to be recommended;
对所述待推荐实体所属实体类别的运营文案信息，以获得所述待推荐实体的实体特征数据；以及The operating copy information of the entity category to which the entity to be recommended belongs belongs to obtain the physical feature data of the entity to be recommended;
对除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息，进行挖掘，以获得所述待推荐实体的实体特征数据。The operation copy information of the other entity categories other than the entity category to which the entity to be recommended belongs is mined to obtain the entity feature data of the entity to be recommended.
如上所述的方面和任一可能的实现方式，进一步提供一种实现方式，所述根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分，包括：The foregoing aspect, and any possible implementation manner, further provide an implementation, where the at least one description is obtained according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended Recommended scores for each description data in the data, including:
获取所述每个描述数据的情感数据；Obtaining the sentiment data of each of the description data;
根据所述待推荐实体的实体特征数据、所述待推荐实体的至少一个描述数据和所述每个描述数据的情感数据，获得所述至少一个描述数据中每个描述数据的推荐得分。And obtaining, according to the entity feature data of the to-be-recommended entity, the at least one description data of the to-be-recommended entity, and the sentiment data of each of the description data, a recommendation score of each description data in the at least one description data.
如上所述的方面和任一可能的实现方式，进一步提供一种实现方式，所述根据所述待推荐实体的实体特征数据、所述待推荐实体的至少一个描述数据和所述每个描述数据的情感数据，获得所述至少一个描述数据中每个描述数据的推荐得分，包括：The foregoing aspect, and any possible implementation manner, further provide an implementation manner, where the entity feature data according to the to-be-recommended entity, at least one description data of the to-be-recommended entity, and each description data Emotion data, obtaining a recommendation score for each of the at least one description data, including:
根据所述每个描述数据的情感数据，选择指示正面情感倾向的情感数据所对应的描述数据； Determining, according to the sentiment data of each of the description data, description data corresponding to the sentiment data indicating the positive sentiment tendency;
根据所述待推荐实体的实体特征数据和所选择的描述数据，获得所选择的描述数据中每个描述数据的推荐得分。And obtaining, according to the entity feature data of the entity to be recommended and the selected description data, a recommendation score of each description data in the selected description data.
如上所述的方面和任一可能的实现方式，进一步提供一种实现方式，所述根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据，包括：The aspect as described above, and any possible implementation manner, further providing an implementation manner, the obtaining the recommendation data of the entity to be recommended according to the specified number of description data with the highest recommendation score, including:
根据推荐得分最高的指定数量的描述数据，选择至少一个推荐文本；Selecting at least one recommended text based on the specified number of description data with the highest recommended score;
根据所述至少一个推荐文本，生成所述待推荐实体的推荐数据。And generating recommendation data of the entity to be recommended according to the at least one recommended text.
本发明的另一方面，提供一种基于人工智能的推荐数据的获取装置，包括：Another aspect of the present invention provides an apparatus for acquiring recommendation data based on artificial intelligence, including:
获取单元，用于获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据；An acquiring unit, configured to acquire entity feature data of the entity to be recommended; the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; Global feature data; operational copy information characteristic data of the entity category to which the entity to be recommended belongs; and operational copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs;
评分单元，用于根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分；a scoring unit, configured to obtain, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data;
推荐单元，用于根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据。a recommendation unit, configured to obtain recommendation data of the entity to be recommended according to the specified number of description data with the highest recommendation score.
如上所述的方面和任一可能的实现方式，进一步提供一种实现方式， 所述获取单元，具体用于执行下列步骤中的至少一项：An aspect of the above, and any possible implementation, further providing an implementation manner, The obtaining unit is specifically configured to perform at least one of the following steps:
对所述待推荐实体的全部用户评论信息，进行挖掘，以获得所述待推荐实体的实体特征数据；Performing mining on all user comment information of the entity to be recommended to obtain physical feature data of the entity to be recommended;
对所述待推荐实体所属实体类别下的全部实体的全部用户评论信息，进行挖掘，以获得所述待推荐实体的实体特征数据；Performing all the user comment information of all the entities in the entity category to which the entity to be recommended belongs to obtain the entity feature data of the entity to be recommended;
对所述待推荐实体所属实体类别的运营文案信息，以获得所述待推荐实体的实体特征数据；以及The operating copy information of the entity category to which the entity to be recommended belongs belongs to obtain the physical feature data of the entity to be recommended;
对除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息，进行挖掘，以获得所述待推荐实体的实体特征数据。The operation copy information of the other entity categories other than the entity category to which the entity to be recommended belongs is mined to obtain the entity feature data of the entity to be recommended.
如上所述的方面和任一可能的实现方式，进一步提供一种实现方式，所述评分单元，具体用于The aspect as described above and any possible implementation manner further provide an implementation manner, where the scoring unit is specifically used for
获取所述每个描述数据的情感数据；以及Obtaining the sentiment data of each of the described data;
根据所述待推荐实体的实体特征数据、所述待推荐实体的至少一个描述数据和所述每个描述数据的情感数据，获得所述至少一个描述数据中每个描述数据的推荐得分。And obtaining, according to the entity feature data of the to-be-recommended entity, the at least one description data of the to-be-recommended entity, and the sentiment data of each of the description data, a recommendation score of each description data in the at least one description data.
如上所述的方面和任一可能的实现方式，进一步提供一种实现方式，所述评分单元，具体用于The aspect as described above and any possible implementation manner further provide an implementation manner, where the scoring unit is specifically used for
根据所述每个描述数据的情感数据，选择指示正面情感倾向的情感数据所对应的描述数据；以及Determining, according to the sentiment data of each of the description data, description data corresponding to the sentiment data indicating the positive sentiment tendency;
根据所述待推荐实体的实体特征数据和所选择的描述数据，获得所选择的描述数据中每个描述数据的推荐得分。 And obtaining, according to the entity feature data of the entity to be recommended and the selected description data, a recommendation score of each description data in the selected description data.
如上所述的方面和任一可能的实现方式，进一步提供一种实现方式，所述推荐单元，具体用于An aspect of the foregoing, and any possible implementation manner, further providing an implementation manner, where the recommendation unit is specifically used to
根据推荐得分最高的指定数量的描述数据，选择至少一个推荐文本；以及Select at least one recommended text based on a specified number of description data with the highest recommended score;
根据所述至少一个推荐文本，生成所述待推荐实体的推荐数据。And generating recommendation data of the entity to be recommended according to the at least one recommended text.
本发明的另一方面，提供一种设备，包括：In another aspect of the invention, an apparatus is provided, comprising:
一个或者多个处理器；One or more processors;
存储器；Memory
一个或者多个程序，所述一个或者多个程序存储在所述存储器中，当被所述一个或者多个处理器执行时：One or more programs, the one or more programs being stored in the memory, when executed by the one or more processors:
获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据；Acquiring the entity feature data of the entity to be recommended; the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and global feature data of the entity class to which the entity to be recommended belongs; Operation copy information characteristic data of the entity category to which the entity to be recommended belongs; and operation copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs;
根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分；Obtaining, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data;
根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据。The recommendation data of the entity to be recommended is obtained according to the specified number of description data with the highest recommendation score.
本发明的另一方面，提供一种非易失性计算机存储介质，所述非易 失性计算机存储介质存储有一个或者多个程序，当所述一个或者多个程序被一个设备执行时，使得所述设备：In another aspect of the present invention, a nonvolatile computer storage medium is provided, A cryptographic computer storage medium stores one or more programs that, when executed by a device, cause the device to:
获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据；Acquiring the entity feature data of the entity to be recommended; the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and global feature data of the entity class to which the entity to be recommended belongs; Operation copy information characteristic data of the entity category to which the entity to be recommended belongs; and operation copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs;
根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分；Obtaining, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data;
根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据。The recommendation data of the entity to be recommended is obtained according to the specified number of description data with the highest recommendation score.
由上述技术方案可知，本发明实施例通过获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据，进而根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分，使得能够根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据，由于采用了待推荐实体的实体特征数据，来生成待推荐实体的推荐数据，使得能够综合考虑待推荐实体相关的各方面信息，使得该待推荐实体很容易被用户所采纳，从而提高了实体推荐 的可靠性。According to the foregoing technical solution, the embodiment of the present invention obtains the entity feature data of the entity to be recommended; the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; Global feature data of the entity category to which the entity to be recommended belongs; the business copy information feature data of the entity category to which the entity to be recommended belongs; and the operational copy information characteristic data of the entity category other than the entity category to which the entity to be recommended belongs, and further Obtaining, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data, so that the specified number of description data according to the recommendation score is the highest The recommendation data of the entity to be recommended is obtained, and the recommendation data of the entity to be recommended is generated by using the entity feature data of the entity to be recommended, so that all aspects related to the entity to be recommended are comprehensively considered, so that the entity to be recommended is very Easy to be adopted by users, thus Recommended entity Reliability.
另外，采用本发明所提供的技术方案，通过综合考虑待推荐实体的局部评论信息和待推荐实体所属实体类别的全局评论信息，能够挖掘出与全局存在较大差异的局部描述数据，作为待推荐实体的推荐数据。In addition, by adopting the technical solution provided by the present invention, by comprehensively considering the local comment information of the entity to be recommended and the global comment information of the entity class to which the entity to be recommended belongs, it is possible to mine the local description data that is largely different from the global, as the to-be-recommended The recommended data for the entity.
另外，采用本发明所提供的技术方案，通过引入和学习外部优质的运营文案信息，能够挖掘出低频但高质量的描述数据，作为待推荐实体的推荐数据。In addition, by adopting the technical solution provided by the present invention, by introducing and learning external high-quality operational copy file information, low-frequency but high-quality description data can be mined as recommendation data of the entity to be recommended.
另外，采用本发明所提供的技术方案，通过选择指示正面情感倾向的情感数据所对应的描述数据进行后续的分析处理，能够从源头上控制负面信息的错误传递。In addition, according to the technical solution provided by the present invention, by performing subsequent analysis processing by selecting the description data corresponding to the sentiment data indicating the positive sentiment tendency, the erroneous transmission of the negative information can be controlled from the source.
附图说明DRAWINGS
为了更清楚地说明本发明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发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the embodiments or the prior art description will be briefly described below. Obviously, the drawings in the following description are the present invention. For some embodiments, other drawings may be obtained from those of ordinary skill in the art in light of the inventive workability.
图1为本发明一实施例提供的基于人工智能的推荐数据的获取方法的流程示意图；FIG. 1 is a schematic flowchart of a method for acquiring recommendation data based on artificial intelligence according to an embodiment of the present invention;
图2为本发明另一实施例提供的基于人工智能的推荐数据的获取装置的结构示意图。 FIG. 2 is a schematic structural diagram of an apparatus for acquiring recommendation data based on artificial intelligence according to another embodiment of the present invention.
具体实施方式detailed description
为使本发明实施例的目的、技术方案和优点更加清楚，下面将结合本发明实施例中的附图，对本发明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发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的全部其他实施例，都属于本发明保护的范围。The technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. It is a partial embodiment of the invention, and not all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.
需要说明的是，本发明实施例中所涉及的终端可以包括但不限于手机、个人数字助理(Personal Digital Assistant，PDA)、无线手持设备、平板电脑(Tablet Computer)、个人电脑(Personal Computer，PC)、MP3播放器、MP4播放器、可穿戴设备(例如，智能眼镜、智能手表、智能手环等)等。It should be noted that the terminals involved in the embodiments of the present invention may include, but are not limited to, a mobile phone, a personal digital assistant (PDA), a wireless handheld device, a tablet computer, and a personal computer (Personal Computer, PC). ), MP3 player, MP4 player, wearable device (for example, smart glasses, smart watches, smart bracelets, etc.).
另外，本文中术语“和/或”，仅仅是一种描述关联对象的关联关系，表示可以存在三种关系，例如，A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B这三种情况。另外，本文中字符“/”，一般表示前后关联对象是一种“或”的关系。In addition, the term "and/or" herein is merely an association relationship describing an associated object, indicating that there may be three relationships, for example, A and/or B, which may indicate that A exists separately, and A and B exist at the same time. There are three cases of B alone. In addition, the character "/" in this article generally indicates that the contextual object is an "or" relationship.
图1为本发明一实施例提供的基于人工智能的推荐数据的获取方法的流程示意图，如图1所示。FIG. 1 is a schematic flowchart of a method for acquiring recommendation data based on artificial intelligence according to an embodiment of the present invention, as shown in FIG. 1 .
101、获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据。 The entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and global feature of the entity class to which the entity to be recommended belongs Data; operational copy information characteristic data of the entity category to which the entity to be recommended belongs; and operational copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs.
102、根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分。102. Obtain, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data.
103、根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据。103. Obtain recommendation data of the to-be-recommended entity according to the specified number of description data with the highest recommended score.
需要说明的是，101～103的执行主体的部分或全部可以为位于本地终端的应用，或者还可以为设置在位于本地终端的应用中的插件或软件开发工具包(Software Development Kit，SDK)等功能单元，或者还可以为位于网络侧服务器中的处理引擎，或者还可以为位于网络侧的分布式系统，本实施例对此不进行特别限定。It should be noted that some or all of the execution entities of 101 to 103 may be applications located in the local terminal, or may be plug-ins or software development kits (SDKs) installed in applications located in the local terminal. The functional unit may also be a processing engine located in the network side server, or may be a distributed system located on the network side, which is not specifically limited in this embodiment.
可以理解的是，所述应用可以是安装在终端上的本地程序(nativeApp)，或者还可以是终端上的浏览器的一个网页程序(webApp)，本实施例对此不进行限定。It is to be understood that the application may be a local application (nativeApp) installed on the terminal, or may be a web application (webApp) of the browser on the terminal, which is not limited in this embodiment.
这样，通过获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据，进而根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分，使得能够根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据，由于采用了待推荐实体的实体特征数据，来生成待推荐实体的推荐数据，使得能够综合考虑待推荐实体相关的各方面信息，使得该待推荐实体很 容易被用户所采纳，从而提高了实体推荐的可靠性。The entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and a global class of the entity class to which the entity to be recommended belongs Feature data; operation copy information feature data of the entity category to which the entity to be recommended belongs; and operation copy information feature data of other entity categories other than the entity category to which the entity to be recommended belongs, and further according to the entity of the entity to be recommended Feature data and at least one description data of the entity to be recommended, obtaining a recommendation score of each description data in the at least one description data, so that the entity to be recommended is obtained according to a specified number of description data with the highest recommendation score The recommended data, because the entity feature data of the entity to be recommended is used to generate the recommendation data of the entity to be recommended, so that all aspects related to the entity to be recommended can be comprehensively considered, so that the entity to be recommended is very It is easily adopted by users, which improves the reliability of entity recommendations.
可选地，在本实施例的一个可能的实现方式中，在101中，具体可以执行下列步骤中的至少一项：Optionally, in a possible implementation manner of this embodiment, in 101, at least one of the following steps may be specifically performed:
对所述待推荐实体的全部用户评论信息，进行挖掘，以获得所述待推荐实体的实体特征数据，即所述待推荐实体的局部特征数据，其主要用于挖掘与待推荐实体相关的不同观点信息，以保证挖掘生成的推荐数据与待推荐实体切实相关，例如，可以为词汇信息量、N元组(N-Gram)频次等待推荐实体的低维特征信息，或者，还可以为评论维度、主题(Topic)信息、语义向量信息、上下文逻辑关系信息等待推荐实体的高维特征信息，或者还可以为待推荐实体的热门关键词，等等；Performing all the user comment information of the to-be-recommended entity to obtain the entity feature data of the entity to be recommended, that is, the local feature data of the entity to be recommended, which is mainly used to mine different related entities to be recommended The view information is used to ensure that the recommended data generated by the mining is actually related to the entity to be recommended, for example, the vocabulary information amount, the N-gram frequency (N-Gram) frequency, the low-dimensional feature information of the recommended entity, or the comment dimension , topic information (Topic) information, semantic vector information, context logic relationship information waiting for high-dimensional feature information of the recommended entity, or may also be a popular keyword of the entity to be recommended, and the like;
对所述待推荐实体所属实体类别下的全部实体的全部用户评论信息，进行挖掘，以获得所述待推荐实体的实体特征数据，即所述待推荐实体所属实体类别的全局特征数据，其主要用于挖掘评论维度信息以及评论差异化信息，例如，可以为词汇信息量、N元组(N-Gram)频次等待推荐实体所属实体类别的低维特征信息，或者，还可以为评论维度、主题(Topic)信息、语义向量信息、上下文逻辑关系信息等待推荐实体所属实体类别的高维特征信息，等等；Performing all the user comment information of all the entities in the entity class to which the entity to be recommended belongs to obtain the entity feature data of the entity to be recommended, that is, the global feature data of the entity class to which the entity to be recommended belongs, For mining the comment dimension information and the comment differentiation information, for example, the vocabulary information amount, the N-gram frequency (N-Gram) frequency, waiting for the low-dimensional feature information of the entity category to which the recommendation entity belongs, or may also be the comment dimension, the topic (Topic) information, semantic vector information, context logic relationship information waiting for high-dimensional feature information of the entity category to which the recommendation entity belongs, and the like;
对所述待推荐实体所属实体类别的运营文案信息，以获得所述待推荐实体的实体特征数据，即所述待推荐实体所属实体类别的运营文案信息特征数据，其主要用于挖掘具有亮点文案风格的相似评论作为推荐数据，例如，可以为词汇信息量、N元组(N-Gram)频次等运营文案的低维特征信息，或者，还可以为主题(Topic)信息、语义向量信息、上下文逻辑关系信息等运营文案的高维特征信息，或者还可以为词法信息、 句法信息等运营文案的基本特征信息，等等；以及The operating copy information of the entity category to which the entity to be recommended belongs is obtained to obtain the entity feature data of the entity to be recommended, that is, the operating copy information characteristic data of the entity category to which the entity to be recommended belongs, which is mainly used for mining the highlight copy Similar comments of the style are used as recommendation data, for example, low-dimensional feature information of a vocabulary information amount, an N-gram frequency, or the like, or may be a topic (Topic) information, semantic vector information, and context. High-dimensional feature information of operational copy files such as logical relationship information, or lexical information, Basic information about operational documents such as syntactic information, etc.;
对除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息，进行挖掘，以获得所述待推荐实体的实体特征数据，即除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据，其主要用于挖掘具有亮点文案风格的相似评论作为推荐数据，例如，可以为词汇信息量、N元组(N-Gram)频次等运营文案的低维特征信息，或者，还可以为主题(Topic)信息、语义向量信息、上下文逻辑关系信息等运营文案的高维特征信息，或者还可以为词法信息、句法信息等运营文案的基本特征信息，等等。Performing excavation on the operation copy information of the entity class other than the entity class to which the entity to be recommended belongs, to obtain the entity feature data of the entity to be recommended, that is, other than the entity category to which the entity to be recommended belongs The operational document information feature data of the entity category is mainly used to mine similar comments with a highlight copy style as recommendation data, for example, low-dimensional feature information of the operational copy such as vocabulary information quantity and N-gram frequency (N-Gram) frequency. Or, it may be high-dimensional feature information of a business copy such as Topic information, semantic vector information, and context logic relationship information, or may also be basic feature information of a business copy such as lexical information, syntax information, and the like.
可选地，在本实施例的一个可能的实现方式中，在102中，具体可以获取所述每个描述数据的情感数据，进而，则可以根据所述待推荐实体的实体特征数据、所述待推荐实体的至少一个描述数据和所述每个描述数据的情感数据，获得所述至少一个描述数据中每个描述数据的推荐得分。Optionally, in a possible implementation manner of the embodiment, in 102, the sentiment data of each of the description data may be specifically acquired, and then, according to the entity feature data of the entity to be recommended, At least one of the description data of the entity to be recommended and the sentiment data of each of the description data obtains a recommendation score for each of the at least one description data.
所谓的描述数据的情感数据，用于指示描述数据的情感倾向，可以包括但不限于正面情感倾向和负面情感倾向，或者还可以包括不同程度的正面情感倾向和负面情感倾向，本实施例对此不进行特别限定。The so-called sentiment data describing the data is used to indicate the sentimental tendency of the description data, which may include, but is not limited to, a positive emotional tendency and a negative emotional tendency, or may also include different degrees of positive emotional tendency and negative emotional tendency, this embodiment No particular limitation is imposed.
具体来说，具体可以根据所述每个描述数据的情感数据，选择指示正面情感倾向的情感数据所对应的描述数据，然后，则可以根据所述待推荐实体的实体特征数据和所选择的描述数据，获得所选择的描述数据中每个描述数据的推荐得分。这样，通过选择指示正面情感倾向的情感数据所对应的描述数据进行后续的分析处理，能够从源头上控制负面信息的错误传递。 Specifically, the description data corresponding to the sentiment data indicating the positive sentiment tendency may be selected according to the sentiment data of each of the description data, and then, according to the entity feature data of the to-be-recommended entity and the selected description. Data, obtaining a recommendation score for each of the description data in the selected description data. In this way, by performing subsequent analysis processing by selecting the description data corresponding to the emotion data indicating the positive emotion tendency, it is possible to control the erroneous transmission of the negative information from the source.
可选地，在本实施例的一个可能的实现方式中，在102中，具体可以采用预先构建的亮点推荐模型，来执行102的具体步骤，即根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分。具体地，具体可以利用所获取的所述待推荐实体的实体特征数据，分别对每个描述数据进行相应的特征挖掘和计算，生成对应的实体特征。然后，将所生成的实体特征及其对应的权重，输入亮点推荐模型，由亮点推荐模型进行分析处理，从而输出每个描述数据的推荐得分。Optionally, in a possible implementation manner of the embodiment, in 102, a pre-built bright spot recommendation model may be specifically used to perform the specific step of 102, that is, according to the physical feature data and the location of the entity to be recommended. At least one description data of the recommended entity is referred to, and a recommendation score of each of the at least one description data is obtained. Specifically, the acquired entity feature data of the to-be-recommended entity may be used to perform corresponding feature mining and calculation for each description data to generate corresponding entity features. Then, the generated entity features and their corresponding weights are input into the highlight recommendation model, and analyzed by the highlight recommendation model to output a recommendation score for each description data.
具体来说，所构建的亮点推荐模型，可以包括但不限于对数线性(Log-linear)模型、支持向量机(Support Vector Machine，SVM)模型、回归树/森林模型，本实施例对此不进行特别限定。Specifically, the constructed bright spot recommendation model may include, but is not limited to, a log-linear model, a support vector machine (SVM) model, a regression tree/forest model, and this embodiment does not Special restrictions are made.
进一步可选地，可以根据102的执行结果，调整亮点推荐模型的模型参数，以提升该亮点推荐模型的效果。Further optionally, the model parameters of the highlight recommendation model may be adjusted according to the execution result of 102 to improve the effect of the highlight recommendation model.
可选地，在本实施例的一个可能的实现方式中，在103中，具体可以采用预先构建的推荐文本模型，来执行103的具体步骤，即根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据。Optionally, in a possible implementation manner of the embodiment, in 103, a pre-built recommended text model may be specifically used to perform the specific step of 103, that is, according to the specified number of description data with the highest recommended score. The recommendation data of the entity to be recommended.
具体来说，所构建的推荐文本模型，可以包括但不限于对数线性(Log-linear)模型、支持向量机(Support Vector Machine，SVM)模型、最大熵(Maximum Entropy，MaxEnt)模型、回归树/森林模型，本实施例对此不进行特别限定。Specifically, the recommended text model may include, but is not limited to, a log-linear model, a support vector machine (SVM) model, a maximum entropy (MaxEnt) model, and a regression tree. / Forest model, this embodiment is not particularly limited.
进一步可选地，可以根据103的执行结果，调整推荐文本模型的模型参数，以提升该推荐文本模型的效果。Further optionally, the model parameters of the recommended text model may be adjusted according to the execution result of 103 to improve the effect of the recommended text model.
可选地，在本实施例的一个可能的实现方式中，在103中，具体可 以根据推荐得分最高的指定数量的描述数据，选择至少一个推荐文本，进而，则可以根据所述至少一个推荐文本，生成所述待推荐实体的推荐数据。Optionally, in a possible implementation manner of this embodiment, in 103, specifically The at least one recommended text is selected according to the specified number of description data with the highest recommendation score. Further, the recommendation data of the to-be-recommended entity may be generated according to the at least one recommended text.
本实施例中，通过获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据，进而根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分，使得能够根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据，由于采用了待推荐实体的实体特征数据，来生成待推荐实体的推荐数据，使得能够综合考虑待推荐实体相关的各方面信息，使得该待推荐实体很容易被用户所采纳，从而提高了实体推荐的可靠性。In this embodiment, the entity feature data of the entity to be recommended is obtained by the at least one of the following data: the local feature data of the entity to be recommended; the entity to be recommended by the entity to be recommended Global signature data of the category; operation copy information feature data of the entity category to which the entity to be recommended belongs; and operational copy information feature data of other entity categories other than the entity category to which the entity to be recommended belongs, and then according to the to-be-recommended Obtaining, by the entity feature data of the entity and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data, so that the to-be-obtained can be obtained according to the specified number of description data with the highest recommendation score. The recommendation data of the recommendation entity is used to generate the recommendation data of the entity to be recommended by using the entity feature data of the entity to be recommended, so that all aspects related to the entity to be recommended can be comprehensively considered, so that the entity to be recommended is easily adopted by the user. , thereby improving the reliability of the entity recommendation
另外，采用本发明所提供的技术方案，通过综合考虑待推荐实体的局部评论信息和待推荐实体所属实体类别的全局评论信息，能够挖掘出与全局存在较大差异的局部描述数据，作为待推荐实体的推荐数据。In addition, by adopting the technical solution provided by the present invention, by comprehensively considering the local comment information of the entity to be recommended and the global comment information of the entity class to which the entity to be recommended belongs, it is possible to mine the local description data that is largely different from the global, as the to-be-recommended The recommended data for the entity.
另外，采用本发明所提供的技术方案，通过引入和学习外部优质的运营文案信息，能够挖掘出低频但高质量的描述数据，作为待推荐实体的推荐数据。In addition, by adopting the technical solution provided by the present invention, by introducing and learning external high-quality operational copy file information, low-frequency but high-quality description data can be mined as recommendation data of the entity to be recommended.
另外，采用本发明所提供的技术方案，通过选择指示正面情感倾向的情感数据所对应的描述数据进行后续的分析处理，能够从源头上控制负面信息的错误传递。 In addition, according to the technical solution provided by the present invention, by performing subsequent analysis processing by selecting the description data corresponding to the sentiment data indicating the positive sentiment tendency, the erroneous transmission of the negative information can be controlled from the source.
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，本发明并不受所描述的动作顺序的限制，因为依据本发明，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发明所必须的。It should be noted that, for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations, but those skilled in the art should understand that the present invention is not limited by the described action sequence. Because certain steps may be performed in other sequences or concurrently in accordance with the present invention. In addition, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.
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。In the above embodiments, the descriptions of the various embodiments are different, and the details that are not detailed in a certain embodiment can be referred to the related descriptions of other embodiments.
图2为本发明另一实施例提供的基于人工智能的推荐数据的获取装置的结构示意图，如图2所示。本实施例的基于人工智能的推荐数据的获取装置可以包括获取单元21、评分单元22和推荐单元23。其中，获取单元21，用于获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据；评分单元22，用于根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分；推荐单元23，用于根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据。2 is a schematic structural diagram of an apparatus for acquiring recommendation data based on artificial intelligence according to another embodiment of the present invention, as shown in FIG. 2 . The apparatus for acquiring artificial intelligence-based recommendation data of the present embodiment may include an acquisition unit 21, a scoring unit 22, and a recommendation unit 23. The acquiring unit 21 is configured to acquire the entity feature data of the entity to be recommended, and the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; the entity to be recommended Global signature data of the entity category; operational copy information feature data of the entity category to which the entity to be recommended belongs; and operational copy information feature data of other entity categories other than the entity category to which the entity to be recommended belongs; scoring unit 22, And obtaining, by the entity feature data of the to-be-recommended entity and at least one description data of the to-be-recommended entity, a recommendation score of each description data in the at least one description data; and a recommendation unit 23, configured to use the highest recommendation score A specified number of description data, obtaining recommendation data of the entity to be recommended.
需要说明的是，本实施例所提供的基于人工智能的推荐数据的获取装置的部分或全部可以为位于本地终端的应用，或者还可以为设置在位于本地终端的应用中的插件或软件开发工具包(Software Development  Kit，SDK)等功能单元，或者还可以为位于网络侧服务器中的处理引擎，或者还可以为位于网络侧的分布式系统，本实施例对此不进行特别限定。It should be noted that part or all of the apparatus for acquiring the artificial intelligence-based recommendation data provided by the embodiment may be an application located in the local terminal, or may be a plug-in or a software development tool disposed in an application located in the local terminal. Software Development The functional unit such as Kit, SDK, or the like may be a processing engine located in the network side server, or may be a distributed system located on the network side, which is not specifically limited in this embodiment.
可以理解的是，所述应用可以是安装在终端上的本地程序(nativeApp)，或者还可以是终端上的浏览器的一个网页程序(webApp)，本实施例对此不进行限定。It is to be understood that the application may be a local application (nativeApp) installed on the terminal, or may be a web application (webApp) of the browser on the terminal, which is not limited in this embodiment.
可选地，在本实施例的一个可能的实现方式中，所述获取单元21，具体可以用于执行下列步骤中的至少一项：Optionally, in a possible implementation manner of the embodiment, the acquiring unit 21 may be specifically configured to perform at least one of the following steps:
对所述待推荐实体的全部用户评论信息，进行挖掘，以获得所述待推荐实体的实体特征数据；Performing mining on all user comment information of the entity to be recommended to obtain physical feature data of the entity to be recommended;
对所述待推荐实体所属实体类别下的全部实体的全部用户评论信息，进行挖掘，以获得所述待推荐实体的实体特征数据；Performing all the user comment information of all the entities in the entity category to which the entity to be recommended belongs to obtain the entity feature data of the entity to be recommended;
对所述待推荐实体所属实体类别的运营文案信息，以获得所述待推荐实体的实体特征数据；以及The operating copy information of the entity category to which the entity to be recommended belongs belongs to obtain the physical feature data of the entity to be recommended;
对除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息，进行挖掘，以获得所述待推荐实体的实体特征数据。The operation copy information of the other entity categories other than the entity category to which the entity to be recommended belongs is mined to obtain the entity feature data of the entity to be recommended.
可选地，在本实施例的一个可能的实现方式中，所述评分单元22，具体可以用于获取所述每个描述数据的情感数据；以及根据所述待推荐实体的实体特征数据、所述待推荐实体的至少一个描述数据和所述每个描述数据的情感数据，获得所述至少一个描述数据中每个描述数据的推荐得分。Optionally, in a possible implementation manner of the embodiment, the scoring unit 22 may be specifically configured to acquire sentiment data of each of the description data; and according to physical feature data of the entity to be recommended, At least one description data of the recommendation entity and the sentiment data of each of the description data are referred to, and a recommendation score of each of the at least one description data is obtained.
具体来说，所述评分单元22，具体可以用于根据所述每个描述数据的情感数据，选择指示正面情感倾向的情感数据所对应的描述数据；以及根据所述待推荐实体的实体特征数据和所选择的描述数据，获得所选 择的描述数据中每个描述数据的推荐得分。Specifically, the scoring unit 22 may be specifically configured to: according to the sentiment data of each of the description data, select description data corresponding to the sentiment data indicating the positive sentiment tendency; and the entity feature data according to the entity to be recommended And selected description data to get selected The recommended score for each description data in the description data.
可选地，在本实施例的一个可能的实现方式中，所述推荐单元23，具体可以用于根据推荐得分最高的指定数量的描述数据，选择至少一个推荐文本；以及根据所述至少一个推荐文本，生成所述待推荐实体的推荐数据。Optionally, in a possible implementation manner of the embodiment, the recommending unit 23 may be specifically configured to select at least one recommended text according to the specified number of description data with the highest recommended score; and according to the at least one recommendation Text, generating recommendation data of the entity to be recommended.
需要说明的是，图1对应的实施例中方法，可以由本实施例提供的基于人工智能的推荐数据的获取装置实现。详细描述可以参见图1对应的实施例中的相关内容，此处不再赘述。It should be noted that the method in the embodiment corresponding to FIG. 1 can be implemented by the apparatus for acquiring artificial intelligence based recommendation data provided by this embodiment. For details, refer to related content in the embodiment corresponding to FIG. 1, and details are not described herein again.
本实施例中，通过获取单元获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据，进而由评分单元根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分，使得推荐单元能够根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据，由于采用了待推荐实体的实体特征数据，来生成待推荐实体的推荐数据，使得能够综合考虑待推荐实体相关的各方面信息，使得该待推荐实体很容易被用户所采纳，从而提高了实体推荐的可靠性。In this embodiment, the entity feature data of the entity to be recommended is obtained by the acquiring unit; the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; the entity to be recommended Global feature data of the entity class; the business copy information feature data of the entity category to which the entity to be recommended belongs; and the operational copy information feature data of the other entity categories other than the entity category to which the entity to be recommended belongs, and further by the scoring unit Obtaining, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data, so that the recommendation unit can be according to the specified number of the highest recommendation score Descriptive data, obtaining recommendation data of the entity to be recommended, and using the entity feature data of the entity to be recommended to generate recommendation data of the entity to be recommended, so that all aspects related to the entity to be recommended can be comprehensively considered, so that the to-be-recommended Entities are easily adopted by users. The recommendation to improve the reliability of the entity.
另外，采用本发明所提供的技术方案，通过综合考虑待推荐实体的局部评论信息和待推荐实体所属实体类别的全局评论信息，能够挖掘出与全局存在较大差异的局部描述数据，作为待推荐实体的推荐数据。 In addition, by adopting the technical solution provided by the present invention, by comprehensively considering the local comment information of the entity to be recommended and the global comment information of the entity class to which the entity to be recommended belongs, it is possible to mine the local description data that is largely different from the global, as the to-be-recommended The recommended data for the entity.
另外，采用本发明所提供的技术方案，通过引入和学习外部优质的运营文案信息，能够挖掘出低频但高质量的描述数据，作为待推荐实体的推荐数据。In addition, by adopting the technical solution provided by the present invention, by introducing and learning external high-quality operational copy file information, low-frequency but high-quality description data can be mined as recommendation data of the entity to be recommended.
另外，采用本发明所提供的技术方案，通过选择指示正面情感倾向的情感数据所对应的描述数据进行后续的分析处理，能够从源头上控制负面信息的错误传递。In addition, according to the technical solution provided by the present invention, by performing subsequent analysis processing by selecting the description data corresponding to the sentiment data indicating the positive sentiment tendency, the erroneous transmission of the negative information can be controlled from the source.
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。A person skilled in the art can clearly understand that for the convenience and brevity of the description, the specific working process of the system, the device and the unit described above can refer to the corresponding process in the foregoing method embodiment, and details are not described herein again.
在本发明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。In the several embodiments provided by the present invention, it should be understood that the disclosed system, apparatus, and method may be implemented in other manners. For example, the device embodiments described above are merely illustrative. For example, the division of the unit is only a logical function division. In actual implementation, 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. In addition, 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.
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。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.
另外，在本发明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。In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or may have two or more unit sets. In one unit. The above integrated unit can be implemented in the form of hardware or in the form of hardware plus software functional units.
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装置(可以是个人计算机，服务器，或者网络装置等)或处理器(processor)执行本发明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。The above-described integrated unit implemented in the form of a software functional unit can be stored in a computer readable storage medium. The above software functional unit 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 of the various embodiments of the present invention. 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. .
最后应说明的是：以上实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述实施例对本发明进行了详细的说明，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发明各实施例技术方案的精神和范围。 It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not limited thereto; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that The technical solutions described in the foregoing embodiments are modified, or the equivalents of the technical features are replaced. The modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (12)

  1. 一种基于人工智能的推荐数据的获取方法，其特征在于，包括：A method for acquiring recommendation data based on artificial intelligence, comprising:
    获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据；Acquiring the entity feature data of the entity to be recommended; the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and global feature data of the entity class to which the entity to be recommended belongs; Operation copy information characteristic data of the entity category to which the entity to be recommended belongs; and operation copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs;
    根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分；Obtaining, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data;
    根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据。The recommendation data of the entity to be recommended is obtained according to the specified number of description data with the highest recommendation score.
  2. 根据权利要求1所述的方法，其特征在于，所述获取待推荐实体的实体特征数据，包括下列步骤中的至少一项：The method according to claim 1, wherein the acquiring the entity feature data of the entity to be recommended comprises at least one of the following steps:
    对所述待推荐实体的全部用户评论信息，进行挖掘，以获得所述待推荐实体的实体特征数据；Performing mining on all user comment information of the entity to be recommended to obtain physical feature data of the entity to be recommended;
    对所述待推荐实体所属实体类别下的全部实体的全部用户评论信息，进行挖掘，以获得所述待推荐实体的实体特征数据；Performing all the user comment information of all the entities in the entity category to which the entity to be recommended belongs to obtain the entity feature data of the entity to be recommended;
    对所述待推荐实体所属实体类别的运营文案信息，以获得所述待推荐实体的实体特征数据；以及The operating copy information of the entity category to which the entity to be recommended belongs belongs to obtain the physical feature data of the entity to be recommended;
    对除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息，进行挖掘，以获得所述待推荐实体的实体特征数据。The operation copy information of the other entity categories other than the entity category to which the entity to be recommended belongs is mined to obtain the entity feature data of the entity to be recommended.
  3. 根据权利要求1所述的方法，其特征在于，所述根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述 至少一个描述数据中每个描述数据的推荐得分，包括：The method according to claim 1, wherein the obtaining the information according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended At least one recommended score describing each of the descriptive data in the data, including:
    获取所述每个描述数据的情感数据；Obtaining the sentiment data of each of the description data;
    根据所述待推荐实体的实体特征数据、所述待推荐实体的至少一个描述数据和所述每个描述数据的情感数据，获得所述至少一个描述数据中每个描述数据的推荐得分。And obtaining, according to the entity feature data of the to-be-recommended entity, the at least one description data of the to-be-recommended entity, and the sentiment data of each of the description data, a recommendation score of each description data in the at least one description data.
  4. 根据权利要求3所述的方法，其特征在于，所述根据所述待推荐实体的实体特征数据、所述待推荐实体的至少一个描述数据和所述每个描述数据的情感数据，获得所述至少一个描述数据中每个描述数据的推荐得分，包括：The method according to claim 3, wherein the obtaining is performed according to the entity feature data of the entity to be recommended, the at least one description data of the entity to be recommended, and the sentiment data of each of the description data. At least one recommended score describing each of the descriptive data in the data, including:
    根据所述每个描述数据的情感数据，选择指示正面情感倾向的情感数据所对应的描述数据；Determining, according to the sentiment data of each of the description data, description data corresponding to the sentiment data indicating the positive sentiment tendency;
    根据所述待推荐实体的实体特征数据和所选择的描述数据，获得所选择的描述数据中每个描述数据的推荐得分。And obtaining, according to the entity feature data of the entity to be recommended and the selected description data, a recommendation score of each description data in the selected description data.
  5. 根据权利要求1～4任一权利要求所述的方法，其特征在于，所述根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据，包括：The method according to any one of claims 1 to 4, wherein the obtaining the recommendation data of the entity to be recommended according to the specified number of description data with the highest recommendation score comprises:
    根据推荐得分最高的指定数量的描述数据，选择至少一个推荐文本；Selecting at least one recommended text based on the specified number of description data with the highest recommended score;
    根据所述至少一个推荐文本，生成所述待推荐实体的推荐数据。And generating recommendation data of the entity to be recommended according to the at least one recommended text.
  6. 一种基于人工智能的推荐数据的获取装置，其特征在于，包括：An apparatus for acquiring recommendation data based on artificial intelligence, comprising:
    获取单元，用于获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所 属实体类别之外的其他实体类别的运营文案信息特征数据；An acquiring unit, configured to acquire entity feature data of the entity to be recommended; the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; Global feature data; the operational copy information feature data of the entity category to which the entity to be recommended belongs; and the entity to be recommended Operational copy information characteristic data of other entity categories other than the entity category;
    评分单元，用于根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分；a scoring unit, configured to obtain, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data;
    推荐单元，用于根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据。a recommendation unit, configured to obtain recommendation data of the entity to be recommended according to the specified number of description data with the highest recommendation score.
  7. 根据权利要求6所述的装置，其特征在于，所述获取单元，具体用于执行下列步骤中的至少一项：The device according to claim 6, wherein the obtaining unit is specifically configured to perform at least one of the following steps:
    对所述待推荐实体的全部用户评论信息，进行挖掘，以获得所述待推荐实体的实体特征数据；Performing mining on all user comment information of the entity to be recommended to obtain physical feature data of the entity to be recommended;
    对所述待推荐实体所属实体类别下的全部实体的全部用户评论信息，进行挖掘，以获得所述待推荐实体的实体特征数据；Performing all the user comment information of all the entities in the entity category to which the entity to be recommended belongs to obtain the entity feature data of the entity to be recommended;
    对所述待推荐实体所属实体类别的运营文案信息，以获得所述待推荐实体的实体特征数据；以及The operating copy information of the entity category to which the entity to be recommended belongs belongs to obtain the physical feature data of the entity to be recommended;
    对除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息，进行挖掘，以获得所述待推荐实体的实体特征数据。The operation copy information of the other entity categories other than the entity category to which the entity to be recommended belongs is mined to obtain the entity feature data of the entity to be recommended.
  8. 根据权利要求6所述的装置，其特征在于，所述评分单元，具体用于The apparatus according to claim 6, wherein said scoring unit is specifically used for
    获取所述每个描述数据的情感数据；以及Obtaining the sentiment data of each of the described data;
    根据所述待推荐实体的实体特征数据、所述待推荐实体的至少一个描述数据和所述每个描述数据的情感数据，获得所述至少一个描述数据中每个描述数据的推荐得分。And obtaining, according to the entity feature data of the to-be-recommended entity, the at least one description data of the to-be-recommended entity, and the sentiment data of each of the description data, a recommendation score of each description data in the at least one description data.
  9. 根据权利要求8所述的装置，其特征在于，所述评分单元，具体 用于The apparatus according to claim 8, wherein said scoring unit is specific Used for
    根据所述每个描述数据的情感数据，选择指示正面情感倾向的情感数据所对应的描述数据；以及Determining, according to the sentiment data of each of the description data, description data corresponding to the sentiment data indicating the positive sentiment tendency;
    根据所述待推荐实体的实体特征数据和所选择的描述数据，获得所选择的描述数据中每个描述数据的推荐得分。And obtaining, according to the entity feature data of the entity to be recommended and the selected description data, a recommendation score of each description data in the selected description data.
  10. 根据权利要求6～9任一权利要求所述的装置，其特征在于，所述推荐单元，具体用于The device according to any one of claims 6 to 9, wherein the recommendation unit is specifically used for
    根据推荐得分最高的指定数量的描述数据，选择至少一个推荐文本；以及Select at least one recommended text based on a specified number of description data with the highest recommended score;
    根据所述至少一个推荐文本，生成所述待推荐实体的推荐数据。And generating recommendation data of the entity to be recommended according to the at least one recommended text.
  11. 一种设备，包括：A device that includes:
    一个或者多个处理器；One or more processors;
    存储器；Memory
    一个或者多个程序，所述一个或者多个程序存储在所述存储器中，当被所述一个或者多个处理器执行时：One or more programs, the one or more programs being stored in the memory, when executed by the one or more processors:
    获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据；Acquiring the entity feature data of the entity to be recommended; the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and global feature data of the entity class to which the entity to be recommended belongs; Operation copy information characteristic data of the entity category to which the entity to be recommended belongs; and operation copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs;
    根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分；Obtaining, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data;
    根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的 推荐数据。Obtaining the to-be-recommended entity according to the specified number of description data with the highest recommendation score Recommended data.
  12. 一种非易失性计算机存储介质，所述非易失性计算机存储介质存储有一个或者多个程序，当所述一个或者多个程序被一个设备执行时，使得所述设备：A non-volatile computer storage medium storing one or more programs, when the one or more programs are executed by a device, causing the device to:
    获取待推荐实体的实体特征数据；所述待推荐实体的实体特征数据包括下列数据中的至少一项：所述待推荐实体的局部特征数据；所述待推荐实体所属实体类别的全局特征数据；所述待推荐实体所属实体类别的运营文案信息特征数据；以及除了所述待推荐实体所属实体类别之外的其他实体类别的运营文案信息特征数据；Acquiring the entity feature data of the entity to be recommended; the entity feature data of the entity to be recommended includes at least one of the following data: local feature data of the entity to be recommended; and global feature data of the entity class to which the entity to be recommended belongs; Operation copy information characteristic data of the entity category to which the entity to be recommended belongs; and operation copy information characteristic data of other entity categories other than the entity category to which the entity to be recommended belongs;
    根据所述待推荐实体的实体特征数据和所述待推荐实体的至少一个描述数据，获得所述至少一个描述数据中每个描述数据的推荐得分；Obtaining, according to the entity feature data of the entity to be recommended and the at least one description data of the entity to be recommended, a recommendation score of each description data in the at least one description data;
    根据推荐得分最高的指定数量的描述数据，获得所述待推荐实体的推荐数据。 The recommendation data of the entity to be recommended is obtained according to the specified number of description data with the highest recommendation score.
PCT/CN2016/107077 2016-09-05 2016-11-24 Artificial intelligence-based recommended data acquisition method, apparatus and device, and non-volatile computer storage medium WO2018040310A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610801268.4A CN106383865B (en) 2016-09-05 2016-09-05 Artificial intelligence based recommended data acquisition method and device
CN201610801268.4 2016-09-05

Publications (1)

Publication Number Publication Date
WO2018040310A1 true WO2018040310A1 (en) 2018-03-08

Family

ID=57938999

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/107077 WO2018040310A1 (en) 2016-09-05 2016-11-24 Artificial intelligence-based recommended data acquisition method, apparatus and device, and non-volatile computer storage medium

Country Status (2)

Country Link
CN (1) CN106383865B (en)
WO (1) WO2018040310A1 (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110209933A (en) * 2019-05-20 2019-09-06 山东科技大学 A kind of biasing tensor resolution method based on regression tree contextual feature autocoding
CN112765486B (en) * 2021-01-22 2022-04-05 重庆邮电大学 Knowledge graph fused attention mechanism movie recommendation method

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102866992A (en) * 2011-07-04 2013-01-09 阿里巴巴集团控股有限公司 Method and device for displaying product information in webpage
CN103810162A (en) * 2012-11-05 2014-05-21 腾讯科技（深圳）有限公司 Method and system for recommending network information
CN103886090A (en) * 2014-03-31 2014-06-25 北京搜狗科技发展有限公司 Content recommendation method and device based on user favorites

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102473183A (en) * 2009-07-14 2012-05-23 索尼公司 Content recommendation system, content recommendation method, content recommendation device, and information recording medium
CN103258045B (en) * 2013-05-24 2016-12-28 百度在线网络技术（北京）有限公司 Content recommendation determines system and method
CN105786936A (en) * 2014-12-23 2016-07-20 阿里巴巴集团控股有限公司 Search data processing method and device

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102866992A (en) * 2011-07-04 2013-01-09 阿里巴巴集团控股有限公司 Method and device for displaying product information in webpage
CN103810162A (en) * 2012-11-05 2014-05-21 腾讯科技（深圳）有限公司 Method and system for recommending network information
CN103886090A (en) * 2014-03-31 2014-06-25 北京搜狗科技发展有限公司 Content recommendation method and device based on user favorites

Also Published As

Publication number Publication date
CN106383865B (en) 2020-03-27
CN106383865A (en) 2017-02-08

Similar Documents

Publication Publication Date Title
US20210049198A1 (en) Methods and Systems for Identifying a Level of Similarity Between a Filtering Criterion and a Data Item within a Set of Streamed Documents
US11714602B2 (en) Methods and systems for identifying a level of similarity between a plurality of data representations
CN111177569B (en) Recommendation processing method, device and equipment based on artificial intelligence
US10394851B2 (en) Methods and systems for mapping data items to sparse distributed representations
US10915577B2 (en) Constructing enterprise-specific knowledge graphs
US11138212B2 (en) Natural language response recommendation clustering for rapid retrieval
US9373075B2 (en) Applying a genetic algorithm to compositional semantics sentiment analysis to improve performance and accelerate domain adaptation
US10762297B2 (en) Semantic hierarchical grouping of text fragments
US11164026B2 (en) Graphical user interface generation based on image analysis
US10108698B2 (en) Common data repository for improving transactional efficiencies of user interactions with a computing device
US20190332619A1 (en) Methods and systems for mapping data items to sparse distributed representations
US20180285448A1 (en) Producing personalized selection of applications for presentation on web-based interface
US10817576B1 (en) Systems and methods for searching an unstructured dataset with a query
CN112182255A (en) Method and apparatus for storing media files and for retrieving media files
WO2018040310A1 (en) Artificial intelligence-based recommended data acquisition method, apparatus and device, and non-volatile computer storage medium
WO2022073341A1 (en) Disease entity matching method and apparatus based on voice semantics, and computer device
JP6996190B2 (en) Compound word generator, program and compound word generation method

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: 16914872

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: 16914872

Country of ref document: EP

Kind code of ref document: A1