WO2020042776A1 - 推荐方法、装置、存储介质和终端设备 - Google Patents

推荐方法、装置、存储介质和终端设备 Download PDF

Info

Publication number
WO2020042776A1
WO2020042776A1 PCT/CN2019/095560 CN2019095560W WO2020042776A1 WO 2020042776 A1 WO2020042776 A1 WO 2020042776A1 CN 2019095560 W CN2019095560 W CN 2019095560W WO 2020042776 A1 WO2020042776 A1 WO 2020042776A1
Authority
WO
WIPO (PCT)
Prior art keywords
influence
factor
impact factor
degree
value
Prior art date
Application number
PCT/CN2019/095560
Other languages
English (en)
French (fr)
Inventor
吴东峰
孙凯
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
Priority to US16/957,260 priority Critical patent/US20210365776A1/en
Priority to EP19855724.1A priority patent/EP3678067A4/en
Priority to KR1020207019637A priority patent/KR102402918B1/ko
Priority to JP2020534513A priority patent/JP7150228B2/ja
Publication of WO2020042776A1 publication Critical patent/WO2020042776A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/08Learning methods
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/10Office automation; Time management
    • G06Q10/109Time management, e.g. calendars, reminders, meetings or time accounting
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/06Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063Operations research, analysis or management
    • G06Q10/0631Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/02Reservations, e.g. for tickets, services or events
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/06Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063Operations research, analysis or management
    • G06Q10/0631Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06313Resource planning in a project environment

Definitions

  • the present application relates to the field of computer technology, and in particular, to a recommended method, device, storage medium, and terminal device.
  • AI Artificial Intelligence
  • a conference room is reserved for a user in a conference room reservation system.
  • the meeting rooms desired by the users do not exist or have been reserved.
  • providing similar conference rooms can also meet users' expectations to a certain extent.
  • the meeting room is filtered from the meeting room result options. If the user wishes to reserve a conference room at 3 pm today, and enters a filter condition of "3 pm today" in the system, it is found that there is no screening result. Then, the user enters a filter condition of "this afternoon” into the system again. At this time, the system queries the meeting rooms this afternoon, including the meeting rooms at 3:30, 4 and 30:30. The system recommends this result to the user for selection.
  • Option 1 uses conditions to make logical judgments and has a large workload.
  • the user may be required to make further judgments based on the screening results provided by the system, and the user experience is poor.
  • Option 2 needs to collect a large amount of sample data for training, and the initial work cost is too high.
  • neural network algorithms have large uncertainties and poor adaptability. Even after training with a large amount of sample data, the recommended results obtained by the model during use may not be accurate. Therefore, using neural network algorithms to recommend, the accuracy is not high.
  • the embodiments of the present application provide a recommendation method, an apparatus, a storage medium, and a terminal device to solve or alleviate the above one or more technical problems in the prior art.
  • an embodiment of the present application provides a recommendation method, including: determining a weight and a change amount of an influence factor of a candidate object; obtaining an expected value and an actual value of the influence factor; and according to the expected value, the actual value, A change amount and a weight, calculating an influence degree of the influence factor on the candidate object; and calculating a recommendation degree of the candidate object according to the influence degree and weight of the influence factor.
  • the type of the influence factor is a numerical type
  • the change amount includes an upper limit change amount and a lower limit change amount
  • the influence factor pair is calculated.
  • the degree of influence of the candidate includes:
  • Equation 1 and Equation 2 Use Equation 1 and Equation 2 to calculate the degree of influence of the impact factor on the candidate object. If t> s, then If t ⁇ s, then Where t is the actual value of the impact factor, s is the expected value of the impact factor, sl is the degree of influence of the impact factor on the candidate object, and div is the difference between the actual value of the impact factor and the expected value The absolute value of; up is the upper limit change of the impact factor; down is the lower limit change of the impact factor, and q is the weight of the impact factor.
  • the method further includes: determining that an actual value of the impact factor falls between the lower limit value and the upper limit value And determine that the degree of influence of the influence factor on the candidate object is zero.
  • calculating the recommendation degree of the candidate object according to the influence degree and weight of the influence factor including: calculating a sum of the influence degree of the influence factor and The ratio of the sum of the weights to obtain the recommendation degree of the candidate object.
  • the type of the impact factor is a non-numerical type, and the calculation is based on the expected value, actual value, change amount, and weight of the impact factor.
  • the degree of influence of the influence factor on the candidate includes determining an expected value and an actual value of the influence factor as the expected conditions and actual conditions of the influence factor, and determining whether the expected conditions and actual conditions of the influence factor are The same; when it is determined that the expected condition of the impact factor is the same as the actual condition, it is determined that the impact degree of the impact factor on the candidate is 100%; and when the expected condition of the impact factor is determined to be different from the actual condition In the same case, it is determined that the influence degree of the influence factor on the candidate object is zero.
  • the method further includes: obtaining a plurality of candidate objects, according to the recommendation degree of the plurality of candidate objects, and Sorting the plurality of candidate objects to obtain a sorting result; and sending the sorting result to a user terminal.
  • an embodiment of the present application provides a recommendation device, including: a determining module for determining a weight and a change amount of an influence factor of a candidate object; an obtaining module for obtaining an expected value and an actual value of the influence factor; A calculation module for calculating an influence degree of the influence factor on the candidate object according to an expected value, an actual value, a change amount, and a weight of the influence factor; and a second calculation module for calculating the influence factor according to the influence factor Influence degree and weight, and calculate the recommendation degree of the candidate object.
  • the type of the influence factor is a numerical type
  • the change amount includes an upper limit change amount and a lower limit change amount
  • the first calculation module includes: a lower limit A calculation unit is configured to calculate a difference between an expected value of the influence factor and the lower limit change amount to obtain a lower limit value of the influence factor; an upper limit calculation unit is used to calculate the expected value of the influence factor and the upper limit change The difference value of the quantity to obtain the upper limit value of the impact factor; a first determining unit configured to determine that an actual value of the influence factor falls between the lower limit value and the upper limit value; and a first calculation unit , Configured to calculate an influence degree of the influence factor on the candidate object according to an expected value, an actual value, an upper limit change amount, a lower limit change amount, and a weight of the influence factor.
  • the first calculation unit is further configured to: use Expression 1 and Expression 2 to calculate the impact factor on the candidate object. Degree of influence, if t> s, then If t ⁇ s, then Where t is the actual value of the impact factor, s is the expected value of the impact factor, sl is the degree of influence of the impact factor on the candidate object, and div is the difference between the actual value of the impact factor and the expected value
  • the first calculation module further includes: a second calculation unit, configured to determine that the actual value of the impact factor falls in the Outside the lower limit value and the upper limit value, and determine that the degree of influence of the influence factor on the candidate object is zero.
  • the second calculation and calculation module is further configured to calculate a ratio of a sum of influence degrees of the influence factors and a sum of weights to obtain the candidate object Degree of recommendation.
  • the type of the impact factor is a non-numeric type
  • the first calculation module includes: an obtaining unit configured to set an expected value of the impact factor And the actual value are determined as the expected condition and the actual condition of the impact factor, and determine whether the expected condition and the actual condition of the impact factor are the same; a second determination unit, configured to determine the expected condition and the actual condition of the impact factor In the same case, determine that the influence degree of the influence factor on the candidate object is 100%; and a third determination unit, configured to determine the expected condition of the influence factor that is different from the actual condition The influence degree of the influence factor on the candidate object is zero.
  • the apparatus further includes: a sorting module, configured to obtain a plurality of candidate objects, and according to the recommendation of the plurality of candidate objects Degree, and sorting the plurality of candidate objects to obtain a sorting result; and a result pushing module, configured to send the sorting result to a user terminal.
  • a sorting module configured to obtain a plurality of candidate objects, and according to the recommendation of the plurality of candidate objects Degree, and sorting the plurality of candidate objects to obtain a sorting result
  • a result pushing module configured to send the sorting result to a user terminal.
  • the functions of the device may be implemented by hardware, or may be implemented by hardware executing corresponding software.
  • the hardware or software includes one or more modules corresponding to the functions described above.
  • the recommendation structure includes a processor and a memory, where the memory is used for the recommendation device to execute the above-mentioned recommendation program, and the processor is configured to execute the program stored in the memory.
  • the recommendation device may further include a communication interface for the recommendation device to communicate with other equipment or a communication network.
  • an embodiment of the present application further provides a computer-readable storage medium for computer software instructions used by a recommendation device, which includes a program for executing the foregoing recommendation method.
  • the embodiment of the present application can determine the impact factor and the weight and change amount of the impact factor during the recommendation process of the candidate object, realize the dynamic deployment of the impact factor and the personalized recommendation of the candidate object, and has high versatility.
  • the candidate condition is not selected one by one by using the logical judgment of the expected conditions, the workload is small, and the user experience is good.
  • FIG. 1 is a schematic flowchart of an embodiment of a recommendation method provided in this application
  • FIG. 2 is a schematic diagram of an embodiment of an impact factor provided by this application.
  • FIG. 3 is a schematic flowchart of an embodiment of a calculation process of an influence degree of an influence factor provided on a candidate object provided by the present application
  • FIG. 4 is a schematic flowchart of an embodiment of a similarity model provided in this application.
  • FIG. 5 is a schematic diagram of an embodiment of a calculation parameter relationship of an influence degree in the case of a numerical type influence factor provided by the present application.
  • FIG. 6 is a schematic flowchart of an embodiment of a calculation process of an influence degree in a case of a non-numeric type influence factor provided by the present application;
  • FIG. 7 is a schematic structural diagram of an embodiment of a recommendation device provided by this application.
  • FIG. 8 is a schematic structural diagram of an embodiment of a terminal device provided in this application.
  • an embodiment of the present application provides a recommendation method, which can be applied to a terminal device.
  • the terminal device may include a computer, a microcomputer, a mobile phone, a tablet, or a mobile phone.
  • the terminal device may run a recommendation system to implement the method of this embodiment.
  • This embodiment includes steps S100 to S400, as follows:
  • the influence factor is an element that determines the recommendation degree of the candidate object, and influences the degree of similarity between the candidate object and the user's expectations.
  • the impact factor may include the start time or scheduled time of the conference, the capacity of the conference room, the number of projectors (or whether there are projectors), and the like. If the meeting start time is important, you can determine that the meeting start time has a weight of 60. If the user expects the start time of the conference to allow fluctuation within 15 minutes, the change amount of the impact factor includes the upper limit change amount and the lower limit change amount, both of which are 15 minutes.
  • the impact factor can include the following attributes:
  • Type the type of influence factor.
  • difference (numeric) impact factor the type between the interval between the two times, the difference between the two lengths, etc.
  • matching impact factor the degree of matching of the two strings, the The same degree, etc.
  • the weight indicates the influence of the influence factor on the recommendation degree of the candidate object. The greater the weight, the greater the influence of the impact factor on the degree of recommendation of the candidate object;
  • the capacity impact factor may be determined as the impact factor of the recommended conference room through step S100. If the upper limit change and the lower limit change are 2 and 1, respectively, the conference room with a capacity between 5 people (baseline-lower limit change) and 7 people (baseline + upper limit change) can be used as a recommended meeting Room candidates.
  • the expected values of the same impact factor of different candidate objects are generally the same.
  • the user can input the expected value of the impact factor through the user terminal.
  • the actual value of the same impact factor for different candidates may be the same or different.
  • the recommendation system includes: the meeting start time (meeting room idle time) of the conference room A is 9 am; the meeting of the conference room B The start time is 11 am; the meeting in conference room C is 10:30 am.
  • the expected value is 10 am
  • the actual value of conference room A is 9 am
  • the actual value of conference room B is 11 am
  • the actual value of conference room C It's 10:30 in the morning.
  • the product of the similarity between the expected value and the actual value of the influence factor and the weight value may be used as the influence degree of the influence factor on the candidate object. For example, if the expected value of the start time of the meeting is 10 am, the actual value of the start time of the meeting room A is 9 am, and the similarity between the two values is a. If the weight value is 10, the start time of the meeting is The degree of influence of chamber A is the product of 10 and a.
  • the embodiment of the present application can adjust the influence factor and the weight and change amount of the influence factor during the execution of the recommendation method of the candidate object, realize the dynamic configuration of the influence factor, achieve personalized recommendation of the candidate object, and have high versatility.
  • the candidate condition is not selected by logical judgment using expected conditions, and the workload is small and the efficiency is high.
  • the calculation process of the influence degree of the influence factor on the candidate object in step S300 may include steps S310 to S340, as follows:
  • step S350 is further included.
  • the impact degree of the impact factor on the candidate object is zero.
  • the impact factor of the conference start time is 15:00, the lower limit change amount is 15 minutes, and the upper limit change amount is 30 minutes, the lower limit value is 14:45, and the upper limit value is The limit is 15:30. If the actual value of the conference start time of the conference room A is 15:15, the degree of influence of the conference start time on the conference room A is calculated in step S340. If the actual conference start time of conference room A is 16:00, which exceeds the upper and lower limits of the conference start time, the impact of conference start time on conference room A is zero, abruptly reduced, or negatively affected.
  • the degree of similarity between the expected value and the actual value of the impact factor may be calculated first, and then multiplied by the weight to obtain the impact degree of the impact factor on the candidate object.
  • the similarity interface may include numerical object similarity, character object similarity, set object similarity, and general object similarity.
  • the impact factors may include numerical impact factors and matching impact factors.
  • the matching impact factor is also called a non-numeric type impact factor. Therefore, the algorithm for calculating the influence degree of the influence factor on the candidate object may include a numerical similarity algorithm and a matching similarity algorithm.
  • the numerical similarity algorithm is used to calculate the similarity between the expected value of the numerical impact factor and the actual value; the matching similarity algorithm is used to calculate the degree of consistency between the expected impact factor and the actual situation. For example, the aforementioned character object similarity, set object similarity, and general object similarity.
  • t is the actual value of the impact factor
  • s is the expected value of the impact factor
  • div is the absolute value of the difference between the actual value of the impact factor and the expected value
  • up is the upper limit change of the impact factor
  • down is the lower limit of the impact factor The amount of change.
  • the degree of similarity between the expected value of the impact factor and the actual value simila is zero.
  • the above step S340 may include:
  • Equation 1 and Equation 2 Use Equation 1 and Equation 2 to calculate the degree of influence of the impact factor on the candidate object.
  • sl is the degree of influence of the influence factor on the candidate object
  • q is the weight of the influence factor
  • the above step S400 may include: calculating a ratio of the sum of the influence degrees of the influence factors and the sum of the weights to obtain a recommendation degree of the candidate object.
  • the process of calculating the impact degree of the impact factor on the candidate object may be as shown in FIG. 6 and includes steps S510 to S530, as follows:
  • S510 Determine an expected value and an actual value of the impact factor as the expected condition and the actual condition of the impact factor; and determine whether the expected condition and the actual condition of the impact factor are the same.
  • the impact factor is whether there is a projector in the conference room and the user expects that there is a projector in the conference room, but the conference room A does not have a projector.
  • the expected condition of the impact factor is not the same as the actual situation of the conference room A, so the presence of a projector in the conference room has a degree of influence on the conference A to zero. If there is a projector in conference room A, at this time, the expected condition of the impact factor is the same as the actual situation, and the impact degree of the impact factor on the candidate object is 100%.
  • the obtained recommendation result may be provided to the user terminal.
  • the multiple candidate objects are sorted according to the recommendation degree of the multiple candidate objects; then, the sorting result is sent to the user terminal.
  • a candidate whose recommendation degree satisfies the recommendation threshold is selected as the recommendation result.
  • a personalized recommendation model can be defined.
  • the recommendation models can be accumulated to obtain the similarity between complex candidate objects and expected objects.
  • the method of this embodiment can be applied to practical scenarios such as booking a conference room, a hotel room, or an airplane flight.
  • a user inputs a desired conference room into a user terminal in a voice or text manner, and the conference room recommendation system (software) in the user terminal will recommend a conference room recommendation result that meets the expectations according to the method of this embodiment.
  • the conference room recommendation system software
  • Table 1 to Table 5 as an example:
  • the projector impact factor belongs to the matching impact factor. If there is a projector in the conference room, the similarity is 100%; if there is no projector in the conference room, the similarity is zero. Therefore, there is no lower limit change and upper limit change of the impact factor of the projector.
  • the expected value of the corresponding impact factor can be converted as follows:
  • the influence degrees of the three influence factors of the conference room to be selected are calculated respectively, and then the similarity percentage (recommended degree) of the three conference rooms and the user's desired conference room is calculated.
  • the lower limit value in the table is the difference between the expected value and the lower limit change amount
  • the upper limit value is the difference between the expected value and the upper limit change amount
  • the similarity percentages of conference room B, conference room C, and the desired conference room are calculated according to the data in Table 4 and Table 5, respectively: 75% and 40%.
  • the recommended conference room ranking is B> A> C. This sorting result can be fed back to the user terminal, and the user can select a conference room from it.
  • This embodiment can better meet the needs of users, and can adjust the size of the impact factors and their impact values to provide comprehensive and comprehensive recommendation results.
  • an embodiment of the present application provides a recommendation device, including:
  • a determination module 100 is configured to determine a weight and a change amount of an impact factor of a candidate object; an acquisition module 200 is configured to acquire an expected value and an actual value of the impact factor; a first calculation module 300 is configured to calculate an expected value of the impact factor , The actual value, the amount of change, and the weight to calculate the degree of influence of the impact factor on the candidate object; and a second calculation module 400 for calculating a recommendation of the candidate object according to the degree of influence and weight of the impact factor degree.
  • the type of the influence factor is a numerical type
  • the change amount includes an upper limit change amount and a lower limit change amount
  • the first calculation module 300 includes a lower limit calculation unit for calculating all The difference between the expected value of the impact factor and the lower limit change amount is used to obtain the lower limit value of the influence factor
  • the upper limit calculation unit is configured to calculate the difference between the expected value of the influence factor and the upper limit change amount to obtain the The upper limit of the impact factor
  • a first calculating unit for The expected value, actual value, upper limit change amount, lower limit change amount, and weight of the factor are used to calculate the degree of influence of the influence factor on the candidate object.
  • the first calculation unit is further configured to calculate the degree of influence of the impact factor on the candidate object by using Equation 1 and Equation 2. If t> s, then If t ⁇ s, then Where t is the actual value of the impact factor, s is the expected value of the impact factor, sl is the degree of influence of the impact factor on the candidate object, and div is the difference between the actual value of the impact factor and the expected value The absolute value of; up is the upper limit change of the impact factor; down is the lower limit change of the impact factor, and q is the weight of the impact factor.
  • the first calculation module 300 further includes: a second calculation unit, configured to determine that an actual value of the impact factor falls outside the lower limit value and the upper limit value, And it is determined that the influence degree of the influence factor on the candidate object is zero.
  • the second calculation module is further configured to calculate a ratio of a sum of influence degrees of the influence factors and a sum of weights to obtain a recommendation degree of the candidate object.
  • the type of the impact factor is a non-numeric type
  • the first calculation module includes: an obtaining unit configured to determine an expected value and an actual value of the impact factor as the impact. The expected condition and actual condition of the factor, and determine whether the expected condition of the impact factor is the same as the actual condition; a second determining unit, configured to determine the expected condition of the impact factor is the same as the actual condition; The degree of influence of the impact factor on the candidate object is 100%; and a third determining unit, configured to determine that the impact factor on the candidate object is determined when the expected condition of the impact factor is different from the actual condition The degree of influence is zero.
  • the apparatus further includes: a ranking module, configured to obtain multiple candidate objects, rank the multiple candidate objects according to the recommendation degree of the multiple candidate objects, and obtain A ranking result; and a result pushing module, configured to send the ranking result to a user terminal.
  • a ranking module configured to obtain multiple candidate objects, rank the multiple candidate objects according to the recommendation degree of the multiple candidate objects, and obtain A ranking result
  • a result pushing module configured to send the ranking result to a user terminal.
  • the functions of the device may be implemented by hardware, or may be implemented by hardware executing corresponding software.
  • the hardware or software includes one or more modules corresponding to the functions described above.
  • the recommendation structure includes a processor and a memory, where the memory is used for the recommendation device to execute the recommended program in the first aspect, and the processor is configured to execute the program stored in the memory.
  • the recommendation device may further include a communication interface for the recommendation device to communicate with other equipment or a communication network.
  • the device includes a memory 21 and a processor 22.
  • the memory 21 stores a computer program that can be stored on the processor 22.
  • the processor 22 executes the computer program, the recommendation method in the above embodiment is implemented.
  • the number of the memory 21 and the processor 22 may be one or more.
  • the device also includes:
  • the communication interface 23 is used for communication between the processor 22 and an external device.
  • the memory 21 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), for example, at least one magnetic disk memory.
  • the bus may be an Industry Standard Architecture (ISA, Industry Standard Architecture) bus, an External Device Interconnect (PCI, Peripheral Component) bus, or an Extended Industry Standard Architecture (EISA, Extended Industry Standard Component) bus.
  • ISA Industry Standard Architecture
  • PCI External Device Interconnect
  • EISA Extended Industry Standard Component
  • the bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one thick line is used in FIG. 8, but it does not mean that there is only one bus or one type of bus.
  • the memory 21, the processor 22, and the communication interface 23 may complete communication with each other through an internal interface.
  • first and second are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as “first” and “second” may include at least one of these features explicitly or implicitly. In the description of the present application, the meaning of "a plurality” is two or more, unless specifically defined otherwise.
  • Any process or method description in a flowchart or otherwise described herein can be understood as a module, fragment, or portion of code that includes one or more executable instructions for implementing a particular logical function or step of a process
  • the scope of the preferred embodiments of this application includes additional implementations in which the functions may be performed out of the order shown or discussed, including performing the functions in a substantially simultaneous manner or in the reverse order according to the functions involved, which should It is understood by those skilled in the art to which the embodiments of the present application pertain.
  • Logic and / or steps represented in a flowchart or otherwise described herein, for example, a sequenced list of executable instructions that may be considered to implement a logical function, may be embodied in any computer-readable medium, For use by, or in combination with, an instruction execution system, device, or device (such as a computer-based system, a system that includes a processor, or another system that can fetch and execute instructions from an instruction execution system, device, or device) Or equipment.
  • a "computer-readable medium” may be any device that can contain, store, communicate, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device.
  • the computer-readable medium in the embodiment of the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the foregoing. More specific examples of computer-readable storage media include at least (non-exhaustive list) the following: electrical connections (electronic devices) with one or more wirings, portable computer disk cartridges (magnetic devices), random access memory (RAM ), Read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable read-only memory (CDROM).
  • the computer-readable storage medium may even be paper or other suitable media on which the program can be printed, because, for example, by optically scanning the paper or other media and then editing, interpreting, or other suitable means as necessary Process to obtain the program electronically and then store it in computer memory.
  • the computer-readable signal medium may include a data signal propagated in baseband or transmitted as a part of a carrier wave, which carries a computer-readable program code. Such a propagated data signal may take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing.
  • the computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, and the computer-readable medium may send, propagate, or transmit a program for use by or in connection with an instruction execution system, an input method, or a device. .
  • the program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, radio frequency (RF), or any suitable combination of the foregoing.
  • each part of the application may be implemented by hardware, software, firmware, or a combination thereof.
  • multiple steps or methods may be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system.
  • a suitable instruction execution system For example, if implemented in hardware, as in another embodiment, it may be implemented using any one or a combination of the following techniques known in the art: Discrete logic circuits, application-specific integrated circuits with suitable combinational logic gate circuits, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.
  • each functional unit in each embodiment of the present application may be integrated into one processing module, or each unit may exist separately physically, or two or more units may be integrated into one module.
  • the above integrated modules may be implemented in the form of hardware or software functional modules. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.
  • the storage medium may be a read-only memory, a magnetic disk, or an optical disk.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Tourism & Hospitality (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Data Mining & Analysis (AREA)
  • Development Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Educational Administration (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Health & Medical Sciences (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种推荐方法、装置、存储介质和终端设备,其中,所述方法包括:确定候选对象的影响因子的权重和变化量(S100);获取所述影响因子的期望值和实际值(S200);根据所述影响因子的期望值、实际值、变化量和权重,计算所述影响因子对所述候选对象的影响程度(S300);以及根据所述影响因子的影响程度和权重,计算所述候选对象的推荐程度(S400)。该方法可以提高推荐的准确度和效率。

Description

推荐方法、装置、存储介质和终端设备
本申请要求于2018年08月28日提交中国专利局、申请号为201810988166.7、发明名称为“推荐方法、装置、存储介质和终端设备”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及计算机技术领域,尤其涉及一种推荐方法、装置、存储介质和终端设备。
背景技术
随着AI(Artificial Intelligence,人工智能)技术的深度发展,人工智能的技术频繁应用于实际的生活场景。例如,在会议室预定系统中为用户预定会议室。但是,由于企业用户对于会议室的需求量大,可能存在用户想要的会议室不存在或者已被预定的情况。此时,提供相似的会议室可以也在一定程度上符合用户的预期。
对于会议室的筛选,目前有以下实现方案:
1、通过逻辑判断,筛选候选对象。例如,根据用户的需求,从会议室的结果选项中筛选出会议室。如果用户希望预定今天下午三点的会议室,并在系统中输入“今天下午三点”的筛选条件,发现没有筛选结果。然后,用户再次在系统中输入“今天下午”的筛选条件。此时,系统查询到今天下午的会议室,包括下午三点半、四点以及四点半的会议室。则系统将此结果推荐给用户进行选择。
2、通过神经网络算法,筛选候选对象。例如,将用户期望的会议室与系统筛选出来的会议室组合成样本数据,并标注两者的相符合的程度。利用标注好的样本数据构建神经网络模型。利用此模型筛选会议室时,可以自动判断筛选出来的会议室与用户期望的会议室的相符合的程度。
但是,上述方案仍然存在以下缺陷:
1、方案1利用条件进行逻辑判断,工作量大。并且,有可能需要用户根据系统提供的筛选结果进行进一步的判断,用户体验不佳。
2、方案2需要法采集大量的样本数据进行训练,前期工作成本过高。另外,神经网络算法不确定性大,适应性差,即使经过大量的样本数据的训练,模型在使用时得到的推荐结果也可能是不准确的。因而,利用神经网络算法来推荐,准确性不高。
发明内容
本申请实施例提供一种推荐方法、装置、存储介质和终端设备,以解决或缓解现有技术中的以上一个或多个技术问题。
第一方面,本申请实施例提供一种推荐方法,包括:确定候选对象的影响因子的权重和变化量;获取所述影响因子的期望值和实际值;根据所述影响因子的期望值、实际值、变化量和权重,计算所述影响因子对所述候选对象的影响程度;以及根据所述影响因子的影响程度和权重,计算所述候选对象的推荐程度。
结合第一方面,在第一方面的第一种实施方式,所述影响因子的类型为数值类型,所述变化量包括上限变化量和下限变化量,以及所述根据所述影响因子的期望值、实际值、变化量和权重,计算所述影响因子对所述候选对象的影响程度,包括:计算所述影响因子的期望值与所述下限变化量的差值,得到所述影响因子的下限值;计算所述影响因子的期望值与所述上限变化量的差值,得到所述影响因子的上限值;确定所述影响因子的实际值落在所述下限值与所述上限值之间;根据所述影响因子的期望值、实际值、上限变化量、下限变化量和权重,计算所述影响因子对所述候选对象的影响程度。
结合第一方面的第一种实施方式,在第一方面的第二种实施方式中,根据所述影响因子的期望值、实际值、上限变化量、下限变化量和权重,计算所述影响因子对所述候选对象的影响程度,包括:
采用式1和式2,计算所述影响因子对所述候选对象的影响程度,如 果t>s,则
Figure PCTCN2019095560-appb-000001
如果t<s,则
Figure PCTCN2019095560-appb-000002
其中,t为所述影响因子的实际值,s为所述影响因子的期望值,sl为所述影响因子对所述候选对象的影响程度,div为所述影响因子的实际值与期望值的差值的绝对值;up为所述影响因子的上限变化量;down为所述影响因子的下限变化量,q为所述影响因子的权重。
结合第一方面的第一种实施方式,在第一方面的第三种实施方式中,所述方法还包括:确定所述影响因子的实际值落在所述下限值与所述上限值之外,并且确定所述影响因子对所述候选对象的影响程度为零。
结合第一方面,在第一方面的第四种实施方式中,根据所述影响因子的影响程度和权重,计算所述候选对象的推荐程度,包括:计算所述影响因子的影响程度的总和与权重的总和的比值,得到所述候选对象的推荐程度。
结合第一方面,在第一方面的第五种实施方式中,所述影响因子的类型为非数值类型,以及,所述根据所述影响因子的期望值、实际值、变化量和权重,计算所述影响因子对所述候选对象的影响程度,包括:将所述影响因子的期望值和实际值确定为所述影响因子的期望条件和实际条件,并确定所述影响因子的期望条件与实际条件是否相同;在确定所述影响因子的期望条件与实际条件相同的情况下,确定所述影响因子对所述候选对象的影响程度为100%;以及在确定所述影响因子的期望条件与实际条件不相同的情况下,确定所述影响因子对所述候选对象的影响程度为零。
结合第一方面或其任意一种实施方式,在第一方面的第六种实施方式中,所述方法还包括:获取多个候选对象,按照所述多个候选对象的推荐程度大小,并且对所述多个候选对象进行排序,得到排序结果;以及将所述排序结果发送给用户终端。
第二方面,本申请实施例提供一种推荐装置,包括:确定模块,用于确定候选对象的影响因子的权重和变化量;获取模块,用于获取所述影响 因子的期望值和实际值;第一计算模块,用于根据所述影响因子的期望值、实际值、变化量和权重,计算所述影响因子对所述候选对象的影响程度;以及第二计算模块,用于根据所述影响因子的影响程度和权重,计算所述候选对象的推荐程度。
结合第二方面,在第二方面的第一种实施方式中,所述影响因子的类型为数值类型,所述变化量包括上限变化量和下限变化量,以及所述第一计算模块包括:下限计算单元,用于计算所述影响因子的期望值与所述下限变化量的差值,得到所述影响因子的下限值;上限计算单元,用于计算所述影响因子的期望值与所述上限变化量的差值,得到所述影响因子上限值;第一确定单元,用于确定所述影响因子的实际值落在所述下限值与所述上限值之间;以及第一计算单元,用于根据所述影响因子的期望值、实际值、上限变化量、下限变化量和权重,计算所述影响因子对所述候选对象的影响程度。
结合第二方面的第一种实施方式,在第二方面的第二种实施方式中,所述第一计算单元进一步用于:采用式1和式2,计算所述影响因子对所述候选对象的影响程度,如果t>s,则
Figure PCTCN2019095560-appb-000003
如果t<s,则
Figure PCTCN2019095560-appb-000004
其中,t为所述影响因子的实际值,s为所述影响因子的期望值,sl为所述影响因子对所述候选对象的影响程度,div为所述影响因子的实际值与期望值的差值的绝对值;up为所述影响因子的上限变化量;down为所述影响因子的下限变化量,q为所述影响因子的权重。
结合第二方面的第一种实施方式,在第二方面的第三种实施方式中,所述第一计算模块还包括:第二计算单元,用于确定所述影响因子的实际值落在所述下限值与所述上限值之外,并且确定所述影响因子对所述候选对象的影响程度为零。
结合第二方面,在第二方面的第四种实施方式,所述第二计算计算模块进一步用于:计算所述影响因子的影响程度的总和与权重的总和的比值,得到所述候选对象的推荐程度。
结合第二方面,在第二方面的第五种实施方式中,所述影响因子的类型为非数值类型,以及,所述第一计算模块包括:获取单元,用于将所述影响因子的期望值和实际值确定为所述影响因子的期望条件和实际条件,并确定所述影响因子的期望条件与实际条件是否相同;第二确定单元,用于在确定所述影响因子的期望条件与实际条件相同的情况下,确定所述影响因子对所述候选对象的影响程度为100%;以及第三确定单元,用于在确定所述影响因子的期望条件与实际条件不相同的情况下,确定所述影响因子对所述候选对象的影响程度为零。
结合第二方面或其任意一种实施方式,在第二方面的第六种实施方式中,所述装置还包括:排序模块,用于获取多个候选对象,按照所述多个候选对象的推荐程度大小,并且对所述多个候选对象进行排序,得到排序结果;以及结果推送模块,用于将所述排序结果发送给用户终端。
所述装置的功能可以通过硬件实现,也可以通过硬件执行相应的软件实现。所述硬件或软件包括一个或多个与上述功能相对应的模块。
在一个可能的设计中,推荐结构中包括处理器和存储器,所述存储器用于推荐装置执行上述推荐程序,所述处理器被配置为用于执行所述存储器中存储的程序。所述推荐装置还可以包括通信接口,用于推荐装置与其他设备或通信网络通信。
第三方面,本申请实施例还提供一种计算机可读存储介质,用于推荐装置所用的计算机软件指令,其中包括用于执行上述推荐方法所涉及的程序。
上述技术方案中的任意一个技术方案具有如下优点或有益效果:
本申请实施例可以在候选对象的推荐过程确定影响因子以及影响因子的权重和变化量的大小,实现影响因子的动态调配和候选对象的个性化推荐,通用性强。在候选对象的推荐程度的计算过程中,不利用期望条件进行逻辑判断的方式来逐一选择候选对象,工作量小,用户体验佳。
上述概述仅仅是为了说明书的目的,并不意图以任何方式进行限制。 除上述描述的示意性的方面、实施方式和特征之外,通过参考附图和以下的详细描述,本申请进一步的方面、实施方式和特征将会是容易明白的。
附图说明
在附图中,除非另外规定,否则贯穿多个附图相同的附图标记表示相同或相似的部件或元素。这些附图不一定是按照比例绘制的。应该理解,这些附图仅描绘了根据本发明公开的一些实施方式,而不应将其视为是对本发明范围的限制。
在附图中,除非另外规定,否则贯穿多个附图相同的附图标记表示相同或相似的部件或元素。这些附图不一定是按照比例绘制的。应该理解,这些附图仅描绘了根据本申请公开的一些实施方式,而不应将其视为是对本申请范围的限制。
图1是本申请提供的推荐方法的一个实施例的流程示意图;
图2是本申请提供的影响因子的一个实施例的示意图;
图3是本申请提供的影响因子对候选对象的影响程度的计算过程的一个实施例的流程示意图;
图4是本申请提供的相似度模型的一个实施例的流程示意图;
图5是本申请提供的在数值类型的影响因子情况下影响程度的计算参数关系的一个实施例的示意图;
图6是本申请提供的在非数值类型的影响因子情况下影响程度的计算过程的一个实施例的流程示意图;
图7是本申请提供的推荐装置的一个实施例的结构示意图;
图8是本申请提供的终端设备的一个实施例的结构示意图。
具体实施方式
在下文中,仅简单地描述了某些示例性实施例。正如本领域技术人员可认识到的那样,在不脱离本申请的精神或范围的情况下,可通过各种不同方式修改所描述的实施例。因此,附图和描述被认为本质上是示例性的而非限制性的。
请参阅图1,本申请实施例提供了一种推荐方法,可以应用于终端设备。终端设备可以包括计算机、微机、手机、平板、手机等。终端设备可以运行推荐系统,以实现本实施例的方法。本实施例包括步骤S100至步骤S400,具体如下:
S100,确定候选对象的影响因子的权重和变化量。
在本实施例中,影响因子是决定候选对象的推荐程度的要素,影响候选对象与用户期望之间相似程度。以预定会议室为例,影响因子可以包括会议开始时间或预定时间、会议室容量、投影仪数量(或者,是否有投影仪)等。如果会议开始时间比较重要,则可以确定会议开始时间的权重为60。如果用户期望的会议开始时间可以允许在其上下15分钟内的波动,则该影响因子的变化量包括上限变化量和下限变化量,均为15分钟。
一般来说,参见图2,影响因子可以包括以下几种属性:
(1)名称,影响因子的标识。例如:时间影响因子、容量影响因子。
(2)类型,影响因子的类型。例如:差值(数值)影响因子(两个时间之间的间隔、两个长度之间的差值等所属的类型)、匹配影响因子(两个字符串的匹配程度、两个集合中元素的相同程度等的类型)。
(3)权重,表示影响因子对候选对象的推荐程度的影响大小。权重越大,影响因子对于对候选对象的推荐程度的影响越大;
(4)下限变化量,影响因子在基准(标准值或期望值)之下的最小变化量;
(5)上限变化量:该影响因子在基准之上的最大变化量。
示例性地,在预定会议室的场景中,如果用户希望预定一个6人会议室,则可以通过步骤S100确定容量影响因子为推荐会议室的影响因子。如 果上限变化量和下限变化量分别为2和1,则会议室系统中容量在5人(基准-下限变化量)和7人(基准+上限变化量)之间的会议室,可以作为推荐会议室的候选对象。
S200,获取所述影响因子的期望值和实际值。
在本实施例中,在推荐过程中,不同候选对象的同一影响因子的期望值一般是相同的。用户可以通过用户终端输入影响因子的期望值。但不同候选对象的同一影响因子的实际值可以是相同的也可以是不相同的。例如,以影响因子会议开始时间为例,假设用户期望的会议开始时间为上午10点,推荐系统包括:会议室A的会议开始时间(会议室空闲时间)为上午9点;会议室B的会议开始时间为上午11点;会议室C的会议开始时间为上午10点半。在推荐过程中,对于会议开始时间这一个影响因子来说,期望值为上午10点,会议室A的实际值为上午9点、会议室B的实际值为上午11点,会议室C的实际值为上午10点半。
S300,根据影响因子的期望值、实际值、变化量和权重,计算影响因子对候选对象的影响程度。在本实施例中,可以将影响因子的期望值与实际值之间的相似程度与权重值的乘积,作为影响因子对候选对象的影响程度。例如,如果会议开始时间的期望值为上午10点,会议室A的会议开始时间的实际值为上午9点,得到两个数值的相似程度为a,如果权重值为10,则会议开始时间对会议室A的影响程度为10与a的乘积。
S400,根据影响因子的影响程度和权重,计算候选对象的推荐程度。
本申请实施例可以在候选对象的推荐方法的执行过程中调整影响因子以及影响因子的权重和变化量的大小,实现影响因子的动态配置,达到候选对象的个性化推荐,通用性强。在候选对象的推荐过程中,不利用期望条件进行逻辑判断来选择候选对象,工作量小、效率高。
在一种可能的实现方式中,如图3所示,如果影响因子为数值类型,可以上述步骤S300的影响因子对候选对象的影响程度的计算过程,可以包括步骤S310至步骤S340,如下:
S310,计算影响因子的期望值与下限变化量的差值,得到影响因子的下限值。
S320,计算影响因子的期望值与上限变化量的差值,得到影响因子的上限值。
S330,确定影响因子的实际值落在下限值与上限值之间。
S340,根据影响因子的期望值、实际值、上限变化量、下限变化量和权重,计算影响因子对候选对象的影响程度。
在一种可能的实现方式中,如图3所示,还包括步骤S350,确定影响因子的实际值落在下限值与上限值之外,则确定影响因子对候选对象的影响程度为零。
示例性地,以会议开始时间这一个影响因子为例,如果会议开始时间的期望为15:00,下限变化量为15分钟,上限变化量为30分钟,则下限值为14:45,上限值为15:30。如果会议室A的会议开始时间的实际值为15:15,则通过步骤S340计算会议开始时间对会议室A的影响程度。如果会议室A的会议开始时间的实际值为16:00,超出了会议开始时间的上下限值范围,则会议开始时间对会议室A的影响程度为零、骤减或负影响等。
在本实施例中,如果影响因子为数值,可以先计算影响因子的期望值与实际值的相似程度,再与权重相乘,得到影响因子对候选对象的影响程度。
如图4所示的相似度模型,如果存在期望对象A(代表影响因子的期望)和候选对象B(代表候选对象B的影响因子的实际值),可以将两者的数据输入相似度接口,进行计算两对象的相似度。最后可以计算得到候选对象B的推荐程度。相似度接口的计算可以包括数值对象相似度、字符对象相似度、集合对象相似度以及一般对象相似度等。
此外,影响因子可以包括数值影响因子和匹配影响因子。其中,匹配影响因子亦称非数值类型的影响因子。因而,计算影响因子对候选对象的影响程度的算法可以包括数值相似度算法和匹配相似度算法。其中,数值 相似度算法,用于计算数值影响因子的期望值与实际值的相似程度;匹配相似度算法,用于计算影响因子的期望与实际情况的相符程度。例如前述的字符对象相似度、集合对象相似度以及一般对象相似度。
参见图5,t为影响因子的实际值,s为影响因子的期望值,div为影响因子的实际值与期望值的差值的绝对值,up为影响因子的上限变化量;down为影响因子的下限变化量。如果影响因子的实际值在上限值与下限值之间的范围内,则采用式a和式b,计算影响因子的期望值与实际值的相似程度simila:
如果t>s,则
Figure PCTCN2019095560-appb-000005
如果t<s,则
Figure PCTCN2019095560-appb-000006
以及,如果影响因子的实际值不在上限值与下限值之间的范围内,则影响因子的期望值与实际值的相似程度simila为零。
示例性地,以影响因子会议开始时间为例,如果用户期望的会议开始时间为下午3点,查询到会议室A的会议开始时间(会议室空闲时间)为下午3点5分,时间可以看作是数值。根据式a进行计算,3点和3点5分的相似程度为:1-5/15=66.67%。如果会议开始时间的权重为60,则会议开始时间对候选对象会议室A的影响程度为:60*66.67%=40。
在本实施例中,将上述的式a和式b分别与权重相乘可以得到以下的式1和式2,即影响因子对候选对象的影响程度。那么,上述步骤S340可以包括:
采用式1和式2,计算影响因子对候选对象的影响程度,
如果t>s,则
Figure PCTCN2019095560-appb-000007
如果t<s,则
Figure PCTCN2019095560-appb-000008
其中,sl为影响因子对候选对象的影响程度,q为影响因子的权重。
在一种可能的实现方式中,上述步骤S400可以包括:计算影响因子的影响程度的总和与权重的总和的比值,得到候选对象的推荐程度。
示例性地,如果会议室A的影响因子包括会议开始时间、会议室容量 和是否存在投影仪,各影响因子对会议室A的影响程度分别为40、15和20,各影响因子的权重分别为60、20和20,则会议室A的推荐程度为(40+15+20)/(60+20+20)=75%。
在一种可能的实现方式中,如果影响因子的类型为非数值类型,则计算影响因子对候选对象的影响程度的过程,可以如图6所示,包括步骤S510至步骤S530,如下:
S510,将所述影响因子的期望值和实际值确定为所述影响因子的期望条件和实际条件;并确定所述影响因子的期望条件与实际条件是否相同。
S520,在确定所述影响因子的期望条件与实际条件相同的情况下,确定所述影响因子对所述候选对象的影响程度为100%。以及
S530,在确定所述影响因子的期望条件与实际条件不相同的情况下,确定所述影响因子对所述候选对象的影响程度为零。
示例性地,如果影响因子为会议室是否存在投影仪,而用户期望会议室有投影仪,但会议室A没有投影仪。此时,影响因子的期望条件与会议室A的实际情况不相同,则会议室是否存在投影仪对会议A的影响程度为零。如果会议室A有投影仪,此时,影响因子的期望条件与实际情况相同,则影响因子对候选对象的影响程度为100%。
在一种可能的实现方式中,如果有多个候选对象,在通过上述实施例计算得到多个候选对象的推荐程度之后,可以获取的推荐结果给用户终端。例如:按照多个候选对象的推荐程度大小,对多个候选对象进行排序;然后,将排序结果发送给用户终端。再如,选择推荐程度满足推荐阈值的候选对象作为推荐结果。
本实施例的优点包括以下:
1、通用性强。在确定好影响因子之后,可以定义个性化的推荐模型。
2、可扩展性好。在确定影响因子的过程中,增加或者删除影响因子。以及可以动态调配影响因子的数值。
3、实现多级推荐。可以将推荐模型进行累加,获取复杂的候选对象与 期望对象之间的相似度。
本实施例的方法可以应用于预定会议室、酒店客房或飞机航班等实际的场景。例如,用户通过语音或文字的方式将期望的会议室输入到用户终端中,用户终端中的会议室推荐系统(软件)将依据本实施例的方法推荐满足该期望的会议室推荐结果。可以有效地提高会议室推荐的准确度和效率。以下将结合表1至表5,以预定会议室这一应用场景为例,对会议室的推荐过程进行详细的描述:
1、确定会议室推荐系统推荐会议室的影响因子及相关属性。
定义的影响因子如表1所示:
表1:
名称 类型 权重 下限变化量 上限变化量
会议开始时间 数值影响因子 60 15 15
会议室容量 数值影响因子 20 2 4
是否有投影仪 匹配影响因子 20
其中,投影仪影响因子属于匹配影响因子。如果会议室有投影仪,则相似度为100%;如果会议室无投影仪,则相似度为零。因此,投影仪影响因子无下限变化量和上限变化量。
2、获取用户的期望条件,转换标准值(期望值)。
如果用户期望的会议室为:下午3点的可以容纳6个人且具有投影仪的会议室,则转换成相应的影响因子的期望值,可以如下:
会议开始时间的期望值:15:00;
会议室容量的期望值:6人;
是否有投影仪的期望:有。
3、获取待推荐的会议室的各影响因子的实际因子,分别计算各影响因子的相似度值,并最终计算待推荐的会议室与用户期望的会议室的相似度。
假设待推荐的会议室有三个,基本情况如表2所示:
表2:
会议室名称 会议开始时间 会议室容量 是否有投影仪
A 15:00 5
B 14:55 7
C 15:10 6
按照上述推荐方法,分别计算待选的会议室的三个影响因子的影响程度,进而算出三个会议室与用户期望会议室的相似度百分比(推荐程度)。
其中,会议室A的情况如表3所示:
表3:
属性 期望值 实际值 下限值 上限值 权重 影响程度
会议开始时间 15:00 15:00 14:45 15:15 60 60
会议室容量 6 5 4 10 20 10
是否有投影仪 - - 20 0
其中,表中下限值为期望值与下限变化量的差值,上限值为期望值与上限变化量的差值。
根据表3的数据可以计算得到,会议室A与期望的会议室的相似度百分比为:(60+10+0)/(60+20+20)=70%。
会议室B和C的情况分别如表4和表5所示:
表4:
属性 期望值 实际值 最小值 最大值 权重 影响程度
会议开始时间 15:00 14:55 14:45 15:15 60 40
会议室容量 6 7 4 10 20 15
是否有投影仪 - - 20 20
表5:
属性 期望值 实际值 最小值 最大值 权重 影响程度
会议开始时间 15:00 15:10 14:45 15:15 60 20
会议室容量 6 6 4 10 20 20
是否有投影仪 - - 20 0
按照上述计算会议室A的相似度百分比的计算原理,根据表4和表5的数据分别计算会议室B、会议室C与期望的会议室的相似度百分比,分别为:75%、40%。
因此,按照相似度百分比的大小,推荐的会议室的排序为B>A>C。可以将此排序结果反馈在用户终端,用户可以从中选择会议室。
本实施例能够较好地满足用户的需求,可以调整影响因子及其影响值的大小,提供综合全面的推荐结果。
参阅图7,本申请实施例提供一种推荐装置,包括:
确定模块100,用于确定候选对象的影响因子的权重和变化量;获取模块200,用于获取所述影响因子的期望值和实际值;第一计算模块300,用于根据所述影响因子的期望值、实际值、变化量和权重,计算所述影响因子对所述候选对象的影响程度;以及第二计算模块400,用于根据所述影响因子的影响程度和权重,计算所述候选对象的推荐程度。
在一种可能的实现方式中,所述影响因子的类型为数值类型,所述变化量包括上限变化量和下限变化量,以及所述第一计算模块300包括:下限计算单元,用于计算所述影响因子的期望值与所述下限变化量的差值,得到所述影响因子的下限值;上限计算单元,用于计算所述影响因子的期望值与所述上限变化量的差值,得到所述影响因子上限值;第一确定单元,用于确定所述影响因子的实际值落在所述下限值与所述上限值之间;以及第一计算单元,用于根据所述影响因子的期望值、实际值、上限变化量、下限变化量和权重,计算所述影响因子对所述候选对象的影响程度。
在一种可能的实现方式中,所述第一计算单元进一步用于:采用式1 和式2,计算所述影响因子对所述候选对象的影响程度,如果t>s,则
Figure PCTCN2019095560-appb-000009
如果t<s,则
Figure PCTCN2019095560-appb-000010
其中,t为所述影响因子的实际值,s为所述影响因子的期望值,sl为所述影响因子对所述候选对象的影响程度,div为所述影响因子的实际值与期望值的差值的绝对值;up为所述影响因子的上限变化量;down为所述影响因子的下限变化量,q为所述影响因子的权重。
在一种可能的实现方式中,所述第一计算模块300还包括:第二计算单元,用于确定所述影响因子的实际值落在所述下限值与所述上限值之外,并且确定所述影响因子对所述候选对象的影响程度为零。
在一种可能的实现方式中,第二计算模块进一步用于:计算所述影响因子的影响程度的总和与权重的总和的比值,得到所述候选对象的推荐程度。
在一种可能的实现方式中,所述影响因子的类型为非数值类型,以及,所述第一计算模块包括:获取单元,用于将所述影响因子的期望值和实际值确定为所述影响因子的期望条件和实际条件,并确定所述影响因子的期望条件与实际条件是否相同;第二确定单元,用于在确定所述影响因子的期望条件与实际条件相同的情况下,确定所述影响因子对所述候选对象的影响程度为100%;以及第三确定单元,用于在确定所述影响因子的期望条件与实际条件不相同的情况下,确定所述影响因子对所述候选对象的影响程度为零。
在一种可能的实现方式中,所述装置还包括:排序模块,用于获取多个候选对象,按照所述多个候选对象的推荐程度大小,并且对所述多个候选对象进行排序,得到排序结果;以及结果推送模块,用于将所述排序结果发送给用户终端。
所述装置的功能可以通过硬件实现,也可以通过硬件执行相应的软件实现。所述硬件或软件包括一个或多个与上述功能相对应的模块。
在一个可能的设计中,推荐结构中包括处理器和存储器,所述存储器 用于推荐装置执行上述第一方面中推荐程序,所述处理器被配置为用于执行所述存储器中存储的程序。所述推荐装置还可以包括通信接口,用于推荐装置与其他设备或通信网络通信。
本申请实施例还提供一种推荐终端设备,如图8所示,该设备包括:存储器21和处理器22,存储器21内存储有可在处理器22上的计算机程序。处理器22执行计算机程序时实现上述实施例中的推荐方法。存储器21和处理器22的数量可以为一个或多个。
该设备还包括:
通信接口23,用于处理器22与外部设备之间的通信。
存储器21可能包含高速RAM存储器,也可能还包括非易失性存储器(non-volatile memory),例如至少一个磁盘存储器。
如果存储器21、处理器22和通信接口23独立实现,则存储器21、处理器22和通信接口23可以通过总线相互连接并完成相互间的通信。总线可以是工业标准体系结构(ISA,Industry Standard Architecture)总线、外部设备互连(PCI,Peripheral Component)总线或扩展工业标准体系结构(EISA,Extended Industry Standard Component)总线等。总线可以分为地址总线、数据总线、控制总线等。为便于表示,图8中仅用一条粗线表示,但并不表示仅有一根总线或一种类型的总线。
可选的,在具体实现上,如果存储器21、处理器22及通信接口23集成在一块芯片上,则存储器21、处理器22及通信接口23可以通过内部接口完成相互间的通信。
在本说明书的描述中,参考术语“一个实施例”、“一些实施例”、“示例”、“具体示例”、或“一些示例”等的描述意指结合该实施例或示例描述的具体特征、结构、材料或者特点包含于本申请的至少一个实施例或示例中。而且,描述的具体特征、结构、材料或者特点可以在任一个或多个实施例或示例中以合适的方式结合。此外,在不相互矛盾的情况下,本领域的技术人员可以将本说明书中描述的不同实施例或示例以及不同实施例或 示例的特征进行结合和组合。
此外,术语“第一”、“第二”仅用于描述目的,而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此,限定有“第一”、“第二”的特征可以明示或隐含地包括至少一个该特征。在本申请的描述中,“多个”的含义是两个或两个以上,除非另有明确具体的限定。
流程图中或在此以其他方式描述的任何过程或方法描述可以被理解为,表示包括一个或更多个用于实现特定逻辑功能或过程的步骤的可执行指令的代码的模块、片段或部分,并且本申请的优选实施方式的范围包括另外的实现,其中可以不按所示出或讨论的顺序,包括根据所涉及的功能按基本同时的方式或按相反的顺序,来执行功能,这应被本申请的实施例所属技术领域的技术人员所理解。
在流程图中表示或在此以其他方式描述的逻辑和/或步骤,例如,可以被认为是用于实现逻辑功能的可执行指令的定序列表,可以具体实现在任何计算机可读介质中,以供指令执行系统、装置或设备(如基于计算机的系统、包括处理器的系统或其他可以从指令执行系统、装置或设备取指令并执行指令的系统)使用,或结合这些指令执行系统、装置或设备而使用。就本说明书而言,“计算机可读介质”可以是任何可以包含、存储、通信、传播或传输程序以供指令执行系统、装置或设备或结合这些指令执行系统、装置或设备而使用的装置。
本申请实施例的计算机可读介质可以是计算机可读信号介质或者计算机可读存储介质或者是上述两者的任意组合。计算机可读存储介质的更具体的示例至少(非穷尽性列表)包括以下:具有一个或多个布线的电连接部(电子装置),便携式计算机盘盒(磁装置),随机存取存储器(RAM),只读存储器(ROM),可擦除可编辑只读存储器(EPROM或闪速存储器),光纤装置,以及便携式只读存储器(CDROM)。另外,计算机可读存储介质甚至可以是可在其上打印程序的纸或其他合适的介质,因为可以例如通过对纸或其他介质进行光学扫描,接着进行编辑、解译或必要时以其他合 适方式进行处理来以电子方式获得程序,然后将其存储在计算机存储器中。
在本申请实施例中,计算机可读信号介质可以包括在基带中或者作为载波一部分传播的数据信号,其中承载了计算机可读的程序代码。这种传播的数据信号可以采用多种形式,包括但不限于电磁信号、光信号或上述的任意合适的组合。计算机可读的信号介质还可以是计算机可读存储介质以外的任何计算机可读介质,该计算机可读介质可以发送、传播或者传输用于指令执行系统、输入法或者器件使用或者与其结合使用的程序。计算机可读介质上包含的程序代码可以用任何适当的介质传输,包括但不限于:无线、电线、光缆、射频(Radio Frequency,RF)等等,或者上述的任意合适的组合。
应当理解,本申请的各部分可以用硬件、软件、固件或它们的组合来实现。在上述实施方式中,多个步骤或方法可以用存储在存储器中且由合适的指令执行系统执行的软件或固件来实现。例如,如果用硬件来实现,和在另一实施方式中一样,可用本领域公知的下列技术中的任一项或他们的组合来实现:具有用于对数据信号实现逻辑功能的逻辑门电路的离散逻辑电路,具有合适的组合逻辑门电路的专用集成电路,可编程门阵列(PGA),现场可编程门阵列(FPGA)等。
本技术领域的普通技术人员可以理解实现上述实施例方法携带的全部或部分步骤是可以通过程序来指令相关的硬件完成的程序,该程序可以存储于一种计算机可读存储介质中,该程序在执行时,包括方法实施例的步骤之一或其组合。
此外,在本申请各个实施例中的各功能单元可以集成在一个处理模块中,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模块中。上述集成的模块既可以采用硬件的形式实现,也可以采用软件功能模块的形式实现。集成的模块如果以软件功能模块的形式实现并作为独立的产品销售或使用时,也可以存储在一个计算机可读存储介质中。存储介质可以是只读存储器,磁盘或光盘等。
以上,仅为本申请的具体实施方式,但本申请的保护范围并不局限于此,任何熟悉本技术领域的技术人员在本申请揭露的技术范围内,可轻易想到其各种变化或替换,这些都应涵盖在本申请的保护范围之内。因此,本申请的保护范围应以权利要求的保护范围为准。

Claims (16)

  1. 一种推荐方法,其特征在于,包括:
    确定候选对象的影响因子的权重和变化量;
    获取所述影响因子的期望值和实际值;
    根据所述影响因子的期望值、实际值、变化量和权重,计算所述影响因子对所述候选对象的影响程度;以及
    根据所述影响因子的影响程度和权重,计算所述候选对象的推荐程度。
  2. 如权利要求1所述的方法,其特征在于,所述影响因子的类型为数值类型,所述变化量包括上限变化量和下限变化量,以及,
    所述根据所述影响因子的期望值、实际值、变化量和权重,计算所述影响因子对所述候选对象的影响程度,包括:
    计算所述影响因子的期望值与所述下限变化量的差值,得到所述影响因子的下限值;
    计算所述影响因子的期望值与所述上限变化量的差值,得到所述影响因子的上限值;
    确定所述影响因子的实际值落在所述下限值与所述上限值之间;以及
    根据所述影响因子的期望值、实际值、上限变化量、下限变化量和权重,计算所述影响因子对所述候选对象的影响程度。
  3. 如权利要求2所述的方法,其特征在于,根据所述影响因子的期望值、实际值、上限变化量、下限变化量和权重,计算所述影响因子对所述候选对象的影响程度,包括:
    采用式1和式2,计算所述影响因子对所述候选对象的影响程度,
    如果t>s,则
    Figure PCTCN2019095560-appb-100001
    如果t<s,则
    Figure PCTCN2019095560-appb-100002
    其中,t为所述影响因子的实际值,s为所述影响因子的期望值,sl为所述影响因子对所述候选对象的影响程度,div为所述影响因子的实际值与期望值的差值的绝对值;up为所述影响因子的上限变化量;down为所述影响因子的下限变化量,q为所述影响因子的权重。
  4. 如权利要求2所述的方法,其特征在于,所述方法还包括:
    确定所述影响因子的实际值落在所述下限值与所述上限值之外,并且确定所述影响因子对所述候选对象的影响程度为零。
  5. 如权利要求1所述的方法,其特征在于,根据所述影响因子的影响程度和权重,计算所述候选对象的推荐程度,包括:
    计算所述影响因子的影响程度的总和与权重的总和的比值,得到所述候选对象的推荐程度。
  6. 如权利要求1所述的方法,其特征在于,所述影响因子的类型为非数值类型,以及,
    所述根据所述影响因子的期望值、实际值、变化量和权重,计算所述影响因子对所述候选对象的影响程度,包括:
    将所述影响因子的期望值和实际值确定为所述影响因子的期望条件和实际条件,并确定所述影响因子的期望条件与实际条件是否相同;
    在确定所述影响因子的期望条件与实际条件相同的情况下,确定所述影响因子对所述候选对象的影响程度为100%;以及
    在确定所述影响因子的期望条件与实际条件不相同的情况下,确定所述影响因子对所述候选对象的影响程度为零。
  7. 如权利要求1至6任一项所述的方法,其特征在于,所述方法还包括:
    获取多个候选对象,按照所述多个候选对象的推荐程度大小,并且对所述多个候选对象进行排序,得到排序结果;以及
    将所述排序结果发送给用户终端。
  8. 一种推荐装置,其特征在于,包括:
    确定模块,用于确定候选对象的影响因子的权重和变化量;
    获取模块,用于获取所述影响因子的期望值和实际值;
    第一计算模块,用于根据所述影响因子的期望值、实际值、变化量和权重,计算所述影响因子对所述候选对象的影响程度;以及
    第二计算模块,用于根据所述影响因子的影响程度和权重,计算所述候选对象的推荐程度。
  9. 如权利要求8所述的装置,其特征在于,所述影响因子的类型为数值类型,所述变化量包括上限变化量和下限变化量,以及所述第一计算模块包括:
    下限计算单元,用于计算所述影响因子的期望值与所述下限变化量的差值,得到所述影响因子的下限值;
    上限计算单元,用于计算所述影响因子的期望值与所述上限变化量的差值,得到所述影响因子的上限值;
    第一确定单元,用于确定所述影响因子的实际值落在所述下限值与所述上限值之间;以及
    第一计算单元,用于根据所述影响因子的期望值、实际值、上限变化量、下限变化量和权重,计算所述影响因子对所述候选对象的影响程度。
  10. 如权利要求9所述的装置,其特征在于,所述第一计算单元进一步用于:
    采用式1和式2,计算所述影响因子对所述候选对象的影响程度,
    如果t>s,则
    Figure PCTCN2019095560-appb-100003
    如果t<s,则
    Figure PCTCN2019095560-appb-100004
    其中,t为所述影响因子的实际值,s为所述影响因子的期望值,sl为所述影响因子对所述候选对象的影响程度,div为所述影响因子的实际值与期望值的差值的绝对值;up为所述影响因子的上限变化量;down为所述影响因子的下限变化量,q为所述影响因子的权重。
  11. 如权利要求9所述的装置,其特征在于,所述第一计算模块还包括:
    第二计算单元,用于确定所述影响因子的实际值落在所述下限值与所述上限值之外,并且确定所述影响因子对所述候选对象的影响程度为零。
  12. 如权利要求8所述的装置,其特征在于,所述第二计算模块进一步用于:
    计算所述影响因子的影响程度的总和与权重的总和的比值,得到所述候选对象的推荐程度。
  13. 如权利要求8所述的装置,其特征在于,所述影响因子的类型为非数值类型,以及,所述第一计算模块包括:
    获取单元,用于将所述影响因子的期望值和实际值确定为所述影响因子的期望条件和实际条件,并确定所述影响因子的期望条件与实际条件是否相同;
    第二确定单元,用于在确定所述影响因子的期望条件与实际条件相同的情况下,确定所述影响因子对所述候选对象的影响程度为100%;以及;
    第三确定单元,用于在确定所述影响因子的期望条件与实际条件不相同的情况下,确定所述影响因子对所述候选对象的影响程度为零。
  14. 如权利要求8至13任一项所述的装置,其特征在于,所述装置还包括:
    排序模块,用于获取多个候选对象,按照所述多个候选对象的推荐程度大小,并且对所述多个候选对象进行排序,得到排序结果;以及
    推送模块,用于将所述排序结果发送给用户终端。
  15. 一种实现推荐的终端设备,其特征在于,所述终端设备包括:
    一个或多个处理器;
    存储装置,用于存储一个或多个程序;
    当所述一个或多个程序被所述一个或多个处理器执行时,使得所述一个或多个处理器实现如权利要求1-7中任一所述的推荐方法。
  16. 一种计算机可读存储介质,其存储有计算机程序,其特征在于,该程序被处理器执行时实现如权利要求1-7中任一所述的推荐方法。
PCT/CN2019/095560 2018-08-28 2019-07-11 推荐方法、装置、存储介质和终端设备 WO2020042776A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
US16/957,260 US20210365776A1 (en) 2018-08-28 2019-07-11 Recommendation method, device, storage medium, and terminal apparatus
EP19855724.1A EP3678067A4 (en) 2018-08-28 2019-07-11 RECOMMENDED PROCEDURE AND DEVICE, STORAGE MEDIUM AND TERMINAL DEVICE
KR1020207019637A KR102402918B1 (ko) 2018-08-28 2019-07-11 추천 방법, 장치, 저장 매체 및 단말 기기
JP2020534513A JP7150228B2 (ja) 2018-08-28 2019-07-11 推薦方法、装置、記憶媒体、端末設備、及びプログラム

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810988166.7A CN109165742A (zh) 2018-08-28 2018-08-28 推荐方法、装置、存储介质和终端设备
CN201810988166.7 2018-08-28

Publications (1)

Publication Number Publication Date
WO2020042776A1 true WO2020042776A1 (zh) 2020-03-05

Family

ID=64893100

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/095560 WO2020042776A1 (zh) 2018-08-28 2019-07-11 推荐方法、装置、存储介质和终端设备

Country Status (6)

Country Link
US (1) US20210365776A1 (zh)
EP (1) EP3678067A4 (zh)
JP (1) JP7150228B2 (zh)
KR (1) KR102402918B1 (zh)
CN (1) CN109165742A (zh)
WO (1) WO2020042776A1 (zh)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109165742A (zh) * 2018-08-28 2019-01-08 北京百度网讯科技有限公司 推荐方法、装置、存储介质和终端设备
CN113379436A (zh) * 2020-03-09 2021-09-10 阿里巴巴集团控股有限公司 信息处理方法、装置、计算设备及介质
CN113807550B (zh) * 2021-03-01 2023-11-07 北京京东振世信息技术有限公司 一种会议室预定方法、系统、装置及其存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105761106A (zh) * 2016-02-07 2016-07-13 徐璇炫 一种基于用户设置进行对象排序的方法和装置
US20160239765A1 (en) * 2015-02-17 2016-08-18 Amadeus S.A.S. Personalized ranking for search results of a travel-related database query
CN107368907A (zh) * 2017-07-25 2017-11-21 携程计算机技术(上海)有限公司 酒店信息展示方法及装置、电子设备、存储介质
CN108268559A (zh) * 2017-01-04 2018-07-10 阿里巴巴集团控股有限公司 基于票务搜索的信息提供方法和装置
CN109165742A (zh) * 2018-08-28 2019-01-08 北京百度网讯科技有限公司 推荐方法、装置、存储介质和终端设备

Family Cites Families (26)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006243950A (ja) 2005-03-01 2006-09-14 Canon System Solutions Inc 情報処理装置及びその制御方法、プログラム
US8065175B1 (en) * 2007-01-30 2011-11-22 Intuit Inc. System and method for relational scheduling of people and/or resources
JP5277619B2 (ja) 2007-11-26 2013-08-28 富士通株式会社 情報処理装置及び訪問スケジュール調整システム
CN102523247B (zh) * 2011-11-24 2014-09-24 合肥工业大学 一种基于多属性匹配的云服务推荐方法及装置
JP5676062B2 (ja) 2012-10-31 2015-02-25 三菱電機株式会社 カーレンタル支援装置
CN104636339B (zh) * 2013-11-06 2018-11-16 华为技术有限公司 一种会议室预定的方法和装置
US20170351978A1 (en) * 2013-11-06 2017-12-07 Swift Travel Services, Llc Dynamic recommendation platform with artificial intelligence
WO2015081212A1 (en) * 2013-11-27 2015-06-04 Er Express, Llc Emergency-room reservation system and method
JP2015176221A (ja) 2014-03-13 2015-10-05 カシオ計算機株式会社 登山情報検索装置、登山情報検索方法、およびプログラム
JP6309362B2 (ja) * 2014-06-19 2018-04-11 株式会社イトーキ 会議室予約システム
CN104967679B (zh) * 2015-06-09 2017-07-21 深圳市腾讯计算机系统有限公司 信息推荐系统、方法及装置
KR101760181B1 (ko) * 2015-07-17 2017-07-31 김지은 비딩을 이용한 호텔 예약 방법
JP2017078985A (ja) 2015-10-21 2017-04-27 株式会社Fasコンサルティング M&a候補表示方法及びm&a候補表示装置並びにスコアリング方法及びスコアリング装置
JP6621307B2 (ja) * 2015-11-12 2019-12-18 アズビル株式会社 予約システム
US20170141967A1 (en) * 2015-11-16 2017-05-18 International Business Machines Corporation Resource forecasting for enterprise applications
CN106940703B (zh) * 2016-01-04 2020-09-11 腾讯科技(北京)有限公司 推送信息粗选排序方法及装置
CN105678587B (zh) * 2016-01-12 2020-11-24 腾讯科技(深圳)有限公司 一种推荐特征确定方法、信息推荐方法及装置
CN105677881B (zh) * 2016-01-12 2020-07-03 腾讯科技(深圳)有限公司 一种信息推荐方法、装置及服务器
JP6834152B2 (ja) * 2016-03-14 2021-02-24 株式会社リコー 会議室予約装置、会議室予約方法、及びプログラム
WO2018018201A1 (zh) * 2016-07-24 2018-02-01 严映军 一种酒店推荐时的信息展示方法以及订房系统
JP2018063469A (ja) 2016-10-11 2018-04-19 オムロン株式会社 マッチングシステム、マッチング方法及びプログラム
CN106503822A (zh) 2016-10-27 2017-03-15 深圳市家创易科技有限公司 一种智能预约匹配方法及系统
JP2018121212A (ja) * 2017-01-25 2018-08-02 京セラドキュメントソリューションズ株式会社 予約管理装置及び予約システム
CN108334539B (zh) * 2017-12-15 2021-10-15 努比亚技术有限公司 对象推荐方法、移动终端及计算机可读存储介质
CN108055334A (zh) * 2017-12-21 2018-05-18 武汉虹信技术服务有限责任公司 一种智能化会议室管理系统及方法
US20190311416A1 (en) * 2018-04-10 2019-10-10 International Business Machines Corporation Trend identification and modification recommendations based on influencer media content analysis

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160239765A1 (en) * 2015-02-17 2016-08-18 Amadeus S.A.S. Personalized ranking for search results of a travel-related database query
CN105761106A (zh) * 2016-02-07 2016-07-13 徐璇炫 一种基于用户设置进行对象排序的方法和装置
CN108268559A (zh) * 2017-01-04 2018-07-10 阿里巴巴集团控股有限公司 基于票务搜索的信息提供方法和装置
CN107368907A (zh) * 2017-07-25 2017-11-21 携程计算机技术(上海)有限公司 酒店信息展示方法及装置、电子设备、存储介质
CN109165742A (zh) * 2018-08-28 2019-01-08 北京百度网讯科技有限公司 推荐方法、装置、存储介质和终端设备

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3678067A4

Also Published As

Publication number Publication date
KR102402918B1 (ko) 2022-05-30
EP3678067A1 (en) 2020-07-08
KR20200098586A (ko) 2020-08-20
US20210365776A1 (en) 2021-11-25
JP7150228B2 (ja) 2022-10-11
CN109165742A (zh) 2019-01-08
EP3678067A4 (en) 2021-04-28
JP2021508116A (ja) 2021-02-25

Similar Documents

Publication Publication Date Title
WO2020042776A1 (zh) 推荐方法、装置、存储介质和终端设备
CN108021691A (zh) 答案查找方法、客服机器人以及计算机可读存储介质
TW201820165A (zh) 用於雲端巨量資料運算架構之伺服器及其雲端運算資源最佳化方法
CN111047563B (zh) 一种应用于医学超声图像的神经网络构建方法
WO2021097962A1 (zh) 一种异构芯片的任务处理方法、任务处理装置及电子设备
WO2019062405A1 (zh) 应用程序的处理方法、装置、存储介质及电子设备
CN108665148B (zh) 一种电子资源质量评价方法、装置和存储介质
WO2022028147A1 (zh) 图像分类模型训练方法、装置、计算机设备及存储介质
CN111158828A (zh) 应用程序app的用户界面确定方法及装置、存储介质
WO2023093375A1 (zh) 一种计算资源获取方法、装置、电子设备和存储介质
WO2021098265A1 (zh) 缺失信息预测方法、装置、计算机设备及存储介质
CN113392974A (zh) 模型训练方法、装置、电子设备以及存储介质
WO2023207487A1 (zh) 一种电路走线确定方法及相关设备
CN110909040B (zh) 一种业务投放辅助方法、装置及电子设备
CN109657840A (zh) 决策树生成方法、装置、计算机可读存储介质及电子设备
CN113204614B (zh) 模型训练方法、优化训练数据集的方法及其装置
WO2020107264A1 (zh) 神经网络架构搜索的方法与装置
CN113592590A (zh) 一种用户画像生成方法及装置
CN108961071B (zh) 自动预测组合业务收益的方法及终端设备
CN111382793A (zh) 一种特征提取方法、装置和存储介质
Morente-Molinera et al. Using clustering methods to deal with high number of alternatives on group decision making
CN109086995A (zh) 应用程序下载途径的评价方法、装置、设备及存储介质
WO2018040561A1 (zh) 数据处理方法、装置及系统
US20060167835A1 (en) Optimization-based media allocation
CN114332529A (zh) 图像分类模型的训练方法、装置、电子设备及存储介质

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2019855724

Country of ref document: EP

Effective date: 20200402

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19855724

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2020534513

Country of ref document: JP

Kind code of ref document: A

ENP Entry into the national phase

Ref document number: 20207019637

Country of ref document: KR

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE