CN113159855A - Live broadcast recommendation method - Google Patents

Live broadcast recommendation method Download PDF

Info

Publication number
CN113159855A
CN113159855A CN202110482943.2A CN202110482943A CN113159855A CN 113159855 A CN113159855 A CN 113159855A CN 202110482943 A CN202110482943 A CN 202110482943A CN 113159855 A CN113159855 A CN 113159855A
Authority
CN
China
Prior art keywords
live broadcast
user
recommendation
live
technology
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110482943.2A
Other languages
Chinese (zh)
Other versions
CN113159855B (en
Inventor
郑鑫
陈丽娜
霍胜军
吴正旭
韩大成
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Qingdao Mengdou Network Technology Co ltd
Original Assignee
Qingdao Mengdou Network Technology Co ltd
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 Qingdao Mengdou Network Technology Co ltd filed Critical Qingdao Mengdou Network Technology Co ltd
Priority to CN202110482943.2A priority Critical patent/CN113159855B/en
Publication of CN113159855A publication Critical patent/CN113159855A/en
Application granted granted Critical
Publication of CN113159855B publication Critical patent/CN113159855B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0251—Targeted advertisements
    • G06Q30/0269—Targeted advertisements based on user profile or attribute
    • G06Q30/0271—Personalized advertisement
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0277—Online advertisement

Landscapes

  • Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Engineering & Computer Science (AREA)
  • Accounting & Taxation (AREA)
  • Development Economics (AREA)
  • Finance (AREA)
  • Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

A live broadcast recommendation method is used for live broadcast recommendation for BCB e-commerce platform users and comprises the following steps: acquiring user information; calculating the recommendation index of each live broadcast to the user; calculating a recommendation index of the live broadcast similar user, and representing the matching degree of the similar user of the user and the live broadcast; calculating the recommendation degree of live broadcast, including the recommendation degree of the demand live broadcast and the recommendation degree of the technology-related live broadcast; generating a recommendation list, including a demand live broadcast recommendation list and a technology-related live broadcast recommendation list; and generating a user recommendation list, and generating the user recommendation list according to the technology direction and/or the demand direction as the direction in which the user is interested, and the technology-related live broadcast recommendation list and/or the demand live broadcast recommendation list. According to the method, the user can quickly and accurately acquire the effective information through recommendation, the interference of other information and invalid information is reduced, and the working efficiency and experience of the user are improved.

Description

Live broadcast recommendation method
Technical Field
The invention relates to the technical field of live broadcasting, in particular to a method for carrying out live broadcasting recommendation on a BCB electronic commerce platform user.
Background
In recent years, China B2B E-commerce has grown rapidly, even prizing trillion market-scale but long-term relatively traditionally closed fields, such as steel, machinery, chemical industry, raw materials, and the like. For the B2B model, it is a business-oriented and inter-business transaction model. Aiming at the live broadcast of a BCB electronic commerce platform, the live broadcast types of the current platform mainly include new technology live broadcast, technical requirement live broadcast and purchase requirement live broadcast (the purchase requirement live broadcast is simply called as requirement live broadcast, and the new technology live broadcast and the technical requirement live broadcast are collectively called as technology-related live broadcast). In order to enable a user to more accurately acquire effective information and reduce interference of other information and invalid information, a certain recommendation method is adopted for live broadcast recommendation of the user, so that the user can quickly acquire information required by the user from a large amount of information.
Disclosure of Invention
The purpose of the invention is: aiming at the problems described in the background art, the invention provides a method for live broadcast recommendation for a BCB e-commerce platform user, which can quickly and accurately provide live broadcast recommendation for the user.
In order to solve the problems, the technical scheme adopted by the invention is as follows:
a live broadcast recommendation method is used for live broadcast recommendation for BCB electronic commerce platform users and is characterized by comprising the following steps:
acquiring user information which comprises information of a user center, user history records and information of the user centers of other users consistent with the user company information;
step two, calculating a recommendation index of each live broadcast to the user, wherein the live broadcast comprises new technology live broadcast, technology demand live broadcast and purchase demand live broadcast; the purchase demand live broadcast is called demand live broadcast for short, and the new technology live broadcast and the technical demand live broadcast are called technical related live broadcast;
step three, calculating a recommendation index of the live broadcast similar user, and representing the matching degree of the similar user of the user and the live broadcast;
step four, calculating the recommendation degree of live broadcast, including the recommendation degree of the required live broadcast and the recommendation degree of the technology-related live broadcast;
generating a recommendation list which comprises a demand live broadcast recommendation list and a technology-related live broadcast recommendation list;
and step six, generating a user recommendation list, and generating the user recommendation list according to the technical direction and/or the demand direction of the user interested direction and the technical related live broadcast recommendation list and/or the demand live broadcast recommendation list.
Further, in the first step:
the information of the user center comprises information of user registration and later maintenance; specifically, the interested directions and corresponding labels of the user comprise technical related live broadcast category labels and live broadcast category labels required; a region of the user including a service region of a supplier identity and a delivery region of a buyer identity;
the user history records comprise live broadcast viewing records of the user and speech records in live broadcast;
the method comprises the steps of obtaining information of user centers of other users consistent with user company information, wherein the information comprises information of unified company user registration and later maintenance, specifically interested directions and corresponding labels of the users, and the information comprises technical related live broadcast categories and demand labels.
Further, in the second step, the calculation of the recommendation index requiring live broadcasting includes:
recommendation index for live broadcast of demand:
Figure BDA0003049918500000021
wherein f isiThe weight of the ith weight is preset, nf is 5, and the weight is adjusted appropriately according to the later use condition; initial setting: f. of1＝45，f2＝10，f3＝30，f4＝-0.5，f5＝10；
(1)F1E.g., {0, 0.125, 0.25, 0.5, 1}, which represents the matching degree of the interest index of the user and the label required for live broadcast, wherein the interest label required by the user and the label required for live broadcast are both final-level labels; the tags are divided into four stages, namely a first-stage tag, a second-stage tag, a third-stage tag and a last-stage tag;
if the end-level label of the user's demand interest matches the end-level label of the demand live broadcastIf matching is successful, F1Matching the first-level label, namely the third-level label, if the matching fails, 1;
if the third-level label of the interest of the user is matched with the third-level label of the live broadcast, and if the matching is successful, F1If the matching fails, continuing to match the upper-level label, namely the second-level label, if the matching fails, continuing to match the upper-level label;
if the secondary label of the user's interest is matched with the secondary label of the live broadcast, F1If the matching fails, continuing to match the upper-level label, namely the first-level label;
if the first-level label of the user's interest is matched with the first-level label of the live broadcast, F1F, matching failure, 0.1251＝0；
(2)F2E {0, 1}, which represents the matching degree of the area of the user and the delivery place requiring live broadcasting, wherein the specific service area and the delivery place are both specific to provinces, namely, the matching is successful on province addresses, namely, the delivery or transaction can be carried out by default;
if the area where the user is the buyer is the same as the delivery request place requiring live broadcasting, F2＝1；
If the user is the supplier's supply area containing the delivery place requiring live broadcast, F2＝1；
If the above condition is not satisfied, it indicates that the region of the user fails to match the delivery place requiring live broadcast, i.e., F2＝0；
(3)F3∈[0，1]The matching degree of the historical data labels of the user and the labels needing live broadcast is represented, and a label frequency set num for clicking the labels needing live broadcast and speaking in the needing live broadcast in days of the user is counted;
numlabelnumber of times of live broadcast for label as click label × 2
+ number of speech times of live broadcast with label as table
Figure BDA0003049918500000031
The num represents a calculation frequency set of historical clicks and speeches of all final-stage labels in a user's near days, and the number of elements in the num is the same as that of the final-stage labels; max (num) is the maximum value in the number of calculations set num, and when max (num) is 0, there is F for any desired live broadcast30, no calculation by the above formula is required; numlabelRepresenting the corresponding calculation times of the last-stage label table; f3Representing the matching degree of the required live broadcast under the historical data label of the user;
(4)F4≥0F4e [0, infinity)), representing the matching degree of the current time of the user and the required live broadcast time, wherein the current time refers to the latest refreshing time, the login time is used as the standard if the user does not refresh after logging in, and the latest refreshing time is used as the standard if the user has refreshing action after logging in; the live broadcast state is divided into:
if the live broadcast is in an on-going state, F4＝0；
If the live broadcast is in the state of waiting for live broadcast, F4Current time-required live start time |. Wherein F4Representing an hour difference between the two times, wherein the start time of the demand live broadcast is the start time of the demand live broadcast in a state to be live broadcast, which is a certain time point in the future and is not arrived temporarily;
if the live broadcast is required to be in the live broadcast ending state, F4The live broadcasting method comprises the steps that 1, current time-required live broadcasting end time | +2, wherein | the current time-required live broadcasting start time | represents an hour difference between two times, and the required live broadcasting end time is a certain time point of history of end time of required live broadcasting in a live broadcasting end state and occurs;
(5)F5the method comprises the following steps that (1) an element of {0, 0.125, 0.25, 0.5, 1}, which represents the matching degree of interest indexes of users of the same company with users and tags needing live broadcast, wherein the interest indexes of the users of the same company with the users are called enterprise tags, the enterprise interest tags and the tags needing live broadcast are final-stage tags, and the tags are divided into four stages, namely first-stage tags, second-stage tags, third-stage tags and final-stage tags;
if the enterprise interest last-level label is not live broadcastMatching the grade labels, if the matching is successful, F5Matching the first-level label, namely the third-level label, if the matching fails, 1;
f, if the enterprise interest third-level label is matched with the third-level label needing live broadcast and the matching is successful5If the matching fails, continuing to match the upper-level label, namely the second-level label, if the matching fails, continuing to match the upper-level label;
if the enterprise interest secondary label is matched with the secondary label required for live broadcast, F is carried out if the matching is successful5If the matching fails, continuing to match the upper-level label, namely the first-level label;
if the primary label of the enterprise interest label is matched with the primary label required for live broadcasting, F is carried out if the matching is successful5F, matching failure, 0.1255＝0。
Further, in the second step, the calculating of the recommendation index of the technology-related live broadcast includes:
recommendation index for technical-related live broadcasts:
Figure BDA0003049918500000041
wherein d isiThe weight of the ith weight is set in advance, nd is 5, and the weight is adjusted appropriately according to the later use condition; initial setting: d1＝45，d2＝10，d3＝30，d4＝-0.5，d5＝10
(1)D1E {0, 1}, representing the matching degree of the new technical interest index of the user and the tags of the technical related live broadcasts, and if the new technical interest index of the user contains the tags of the technical related live broadcasts, D11 is ═ 1; otherwise, if the new technical interest index of the user does not contain the tag of the technical related live broadcast, D1＝0；
(2)D2The element is {0.2, 0.4, 0.7, 1}, and the score of the stage where the new technology in the technology-related live broadcast is located is represented;
if the new technology in the technology-related live broadcast is in the mass production stage, D2＝1；
If the new technology in the technology-related live broadcast is in a new product test stageThen D is2＝0.7；
If the stage of the new technology in the technology-related live broadcast is an experimental demonstration stage, D2＝0.4；
If the stage of the new technology in the technology-related live broadcast is the scheme demonstration stage, D2＝0.2；
(3)D3∈[0，1]The matching degree of the historical data labels of the user and the labels of the technology-related live broadcasts is represented, and the times d _ num of clicking the labels of the technology-related live broadcasts and speaking in the technology-related live broadcasts in d _ days of the user are counted;
d_numlabelnumber of times 2 of related live broadcast of technique of clicking tag as cable
+ technology-related live broadcast speaking times with label of cable
Figure BDA0003049918500000051
D _ num represents a calculation frequency set of historical clicks and speeches of all new technical interest tags in a d _ days near by a user, and the number of elements in d _ num is the same as that of the new technical interest tags; max (D _ num) is the maximum value in the calculation number set D _ num, and when max (D _ num) is 0, there is D for any required live broadcast30, no calculation by the above formula is required; d _ numlabelRepresenting the corresponding calculation times of the interest label of the new technology, namely the table; d3Representing the matching degree of the related live broadcast of the technology under the historical data label of the user;
(4)D4e [0, + ∞), representing how well the user's current time matches the technology-related live time;
the state of the related live broadcast is divided into:
if the technology-related live broadcast is in an on-going state, D4＝0；
If the technology-related live broadcast is in a state of waiting for live broadcast,
D4current time-technology dependent live start time |; wherein D4Representing the difference in days between two times, technical relevanceThe live broadcast starting time is a technical related live broadcast starting time in a state of waiting for live broadcast, which is a certain time point in the future and is not yet arrived temporarily;
if the technology-related live broadcast is in a live broadcast end state,
D4current time-technology dependent live broadcast end time | + 2; wherein | the current time-the required live broadcast start time | represents a difference in days between two times, and the technology-related live broadcast end time is a certain time point in history, at which the technology-related live broadcast end time in the live broadcast end state has occurred;
(5)D5the element is {0, 1}, the matching degree of the new technical interest category index of the user of the same company with the user and the category label of the related live broadcast of the technology is represented, and the new technical interest category index of the user of the same company with the user is called an enterprise new technical label;
if the enterprise new technology interest label is matched with the label of the related technology live broadcast, and the matching is successful, D51, failure of matching then5＝0。
Further, the third step of calculating the recommendation index of the live broadcast similar user includes the following steps:
1. counting similar users with users, selecting n users with higher similarity, called similar users, temporarily setting n as 20, and properly adjusting the value of n according to the use condition in the later period; calculating the similarity between the current user and other users, wherein the first n other users with larger similarity values are used as the similarity users of the current user;
the similarity between two users is calculated as:
Figure BDA0003049918500000061
wherein, simL-uThe similarity between L and U is represented, the L and the U respectively represent tag sets of two users, and the tag sets comprise user new technology interest tag sets, required final level interest tag sets and tags corresponding to elements with element values larger than sn in a user calculation number set num; sThe value of n is temporarily set as sn-5, and is adjusted according to the use condition in the later period; l | represents the length of the tag set L, i.e. the number of elements in the tag set L; l | n |, U |, represents the length of the intersection of the tag set L and the tag set U, i.e., the number of elements in the intersection;
2. and counting the number of similar users of each live broadcast, which are checked in the last login and current latest refreshing time of the user. And calculates the TS value for it:
Figure BDA0003049918500000071
wherein u represents the number of people who click the live broadcast by similar users (no repeated calculation), and n represents the number of the similar users; the live broadcast without viewing has TS equal to 0, and does not need to be calculated by the above formula.
Further, the fourth step of calculating the recommendation degree of the live broadcast includes:
recommendation degree of demand-related live broadcast: RF ═ F + RF × TS
Recommendation degree of related live broadcast: RD + RD × TS
Wherein, rf, rd are used for controlling the important degree that live similar user's recommendation index occupied in live recommendation degree, and rf, rd are tentatively 7, and the later stage is adjusted according to the use thing appropriately.
Further, in the fifth step, a recommendation list is generated, which comprises a demand live broadcast recommendation list and a technology-related live broadcast recommendation list; the method comprises the following steps:
sequencing the demand live broadcasts according to the recommendation degrees of the demand live broadcasts from large to small to form a demand live broadcast recommendation list; and sequencing the related live broadcasts of the new technology according to the recommendation degrees of the related live broadcasts from large to small to form a related live broadcast recommendation list of the technology.
Further, in the sixth step, generating a user recommendation list includes:
(1) if the interested direction of the user is the technical direction, selecting the top tn live broadcast recommendation lists and the top qn live broadcast recommendation lists from the technology-related live broadcast recommendation lists, randomly displaying the demand live broadcast and the technology-related live broadcast without changing the front and back positions of the demand live broadcast recommendation lists and the technology-related live broadcast recommendation lists, and deleting the demand live broadcast recommendation lists and the technology-related live broadcast recommendation lists from the demand live broadcast list and the technology-related live broadcast recommendation lists; and analogizing in turn, namely (tn + qn) live broadcasts in the user recommendation list are taken as a group, wherein the group comprises tn technical-related live broadcasts and qn demand live broadcasts; the specific quantity is adjusted in the later use, and is temporarily set to be tn-5 and qn-1; when the demand live broadcast recommendation list or the technology-related live broadcast recommendation list is empty, directly adding the recommendation list which is not empty to the tail part of the user recommendation list;
(2) if the interested direction of the user is a demand direction, selecting the top tm from the technology-related live broadcast recommendation list and the top qm from the demand live broadcast recommendation list, and randomly displaying the demand live broadcast and the technology-related live broadcast without changing the front and back positions of the demand live broadcast and the technology-related live broadcast recommendation list respectively, and deleting the demand live broadcast and the technology-related live broadcast recommendation list; and analogizing in turn, namely taking (tm + qm) live broadcasts in the user recommendation list as a group, wherein the group comprises tm technology-related live broadcasts and qm demand live broadcasts, and the specific number is adjusted in the later period of use and is temporarily set to be tm equal to 10, and qm equal to 1; when the demand live broadcast recommendation list or the technology-related live broadcast recommendation list is empty, directly adding the recommendation list which is not empty to the tail part of the user recommendation list;
(3) if the interested direction of the user is a required direction and a technical direction, selecting the first t in a technical related live broadcast recommendation list and the first q in the required live broadcast recommendation list, randomly displaying the required live broadcast and the technical related live broadcast without changing the front and back positions of the required live broadcast and the technical related live broadcast recommendation list, and deleting the required live broadcast and the technical related live broadcast recommendation list; and analogizing in sequence, namely, the (t + q) live broadcasts in the user recommendation list are taken as a group, wherein t technical related live broadcasts and q required live broadcasts are included; the specific quantity is adjusted in the later use, and t is temporarily set to be 2, and q is temporarily set to be 3; and when the demand live broadcast recommendation list or the technology-related live broadcast recommendation list is empty, directly adding the recommendation list which is not empty to the tail part of the user recommendation list.
The technical scheme provided by the embodiment of the invention has the beneficial effects that at least: the live broadcast recommendation method has the beneficial effects that: the user can be enabled to:
1. and effective information is rapidly and accurately acquired.
2. And the interference of other information and invalid information is reduced.
3. The working efficiency and the experience of the user are improved.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
fig. 1 is a flowchart of a live broadcast recommendation method disclosed in an embodiment of the present invention.
DETAILED DESCRIPTION OF EMBODIMENT (S) OF INVENTION
Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
The embodiment of the invention provides a live broadcast recommendation method, which comprises the following six steps:
acquiring user information which comprises information of a user center, user history records and information of the user centers of other users consistent with the user company information;
step two, calculating a recommendation index of each live broadcast to the user, wherein the live broadcast comprises new technology live broadcast, technology demand live broadcast and purchase demand live broadcast; the purchase demand live broadcast is called demand live broadcast for short, and the new technology live broadcast and the technical demand live broadcast are called technical related live broadcast;
step three, calculating a recommendation index of the live broadcast similar user, and representing the matching degree of the similar user of the user and the live broadcast;
step four, calculating the recommendation degree of live broadcast, including the recommendation degree of the required live broadcast and the recommendation degree of the technology-related live broadcast;
generating a recommendation list which comprises a demand live broadcast recommendation list and a technology-related live broadcast recommendation list;
and step six, generating a user recommendation list, and generating the user recommendation list according to the technical direction and/or the demand direction of the user interested direction and the technical related live broadcast recommendation list and/or the demand live broadcast recommendation list.
The steps are described in detail as follows:
step one, acquiring user information which comprises information of a user center, user history records and information of the user centers of other users consistent with the user company information.
The information of the user center comprises information of user registration and later maintenance. Specifically, the interested directions and corresponding labels of the user comprise technical related live broadcast category labels and live broadcast category labels required; the user's area, including the supplier-identified service area and the buyer-identified delivery area.
And the user history record comprises a live viewing record of the user and a speech record in the live broadcast.
And acquiring the information of the user centers of other users consistent with the user company information, wherein the information comprises the information of unified company user registration and later maintenance. Specifically, the interest direction and the corresponding tag of the user include the category and demand tag of the related live broadcast of the technology.
And step two, calculating the recommendation index of each live broadcast to the user. Live broadcast comprises new technology live broadcast, technical requirement live broadcast and purchase requirement live broadcast; the purchase demand live broadcast is called demand live broadcast for short, and the new technology live broadcast and the technical demand live broadcast are called technical related live broadcast together. The calculation of the recommendation index of demand live broadcast and the calculation of the recommendation index of technology-related live broadcast (new technology live broadcast and technology demand live broadcast) are as follows:
1. recommendation index for live broadcast of demand:
Figure BDA0003049918500000101
wherein f isiThe i-th weight is set in advance, and nf is 5, which can be adjusted as appropriate according to the later usage.
Initial setting: f. of1＝45，f2＝10，f3＝30，f4＝-0.5，f5＝10。
(1)F1The element belongs to {0, 0.125, 0.25, 0.5, 1}, and the matching degree of the interest indexes of the user and the tags needing live broadcasting is represented, and the tags needing interest and live broadcasting of the user are final-level tags. The tags are divided into four levels, namely a first-level tag, a second-level tag, a third-level tag and a fourth-level tag (a final-level tag). The label with the hierarchical relationship comprises an electronic component (a first-level label), a chip (a second-level label), a logic chip (a third-level label) and a counter (a last-level label).
F, if the final-stage label of the demand interest of the user is matched with the final-stage label of the demand live broadcast, if the matching is successful1Matching the first-level label, namely the third-level label, if the matching fails, 1;
if the third-level label of the interest of the user is matched with the third-level label of the live broadcast, and if the matching is successful, F1If the matching fails, continuing to match the upper-level label, namely the second-level label, if the matching fails, continuing to match the upper-level label;
if the secondary label of the user's interest is matched with the secondary label of the live broadcast, F1If the matching fails, continuing to match the upper-level label, namely the first-level label;
if the first-level label of the user's interest is matched with the first-level label of the live broadcast, F1F, matching failure, 0.1251＝0。
Such as: the demand interest tag of the user is a trigger, and the demand live tag is a counter. The hierarchical relationship of the flip-flops and counters is as follows:
failure of matching First stage Second stage Three-stage Four stages
Demand live broadcast Electronic component Chip and method for manufacturing the same Logic chip Counter with a memory
Interest tag Electronic component Chip and method for manufacturing the same Logic chip Flip-flop
F1 0 0.125 0.25 0.5 1
Matching the interest label of the user and the label needing live broadcasting successfully on the third-level label, wherein F is the moment1＝0.5。
(2)F2E {0, 1}, which represents the matching degree of the user's area and the delivery place requiring live broadcasting, and the specific service area and the delivery place are both specific to province, namely, the matching is successful on the province address, namely, the delivery or transaction is acquiescent.
If the area where the user is the buyer is the same as the delivery request place requiring live broadcasting, F2＝1；
If the user is the supplier's supply area containing the delivery place requiring live broadcast, F2＝1；
If the above condition is not satisfied, it indicates that the region of the user fails to match the delivery place requiring live broadcast, i.e., F2＝0。
(3)F3∈[0，1]And the matching degree of the historical data label of the user and the label needing live broadcasting is shown. And counting a tag frequency set num for clicking the live demand broadcast and speaking in the live demand broadcast in a days of nearly days of the user.
numlabelNumber of times of live broadcast for label as click label × 2
+ number of speech times of live broadcast with label as table
Figure BDA0003049918500000111
The num represents a calculation frequency set of historical clicks and speeches of all final-stage labels in a user's near days, and the number of elements in the num is the same as that of the final-stage labels; max (num) is the number of countsThe maximum value of num, when max (num) is 0, F is available for any desired live broadcast30, no calculation by the above formula is required; numlabelRepresenting the corresponding calculation times of the last-stage label table; f3Indicating the matching degree of the demand live broadcast under the historical data label of the user.
(4)F4E [0, + ∞)) representing how well the user's current time matches the desired live time. The current time, refers to the most recent refresh time. If the refresh action exists after the login, the latest refresh time is taken as the standard.
The live broadcast state is divided into:
if the live broadcast is in an on-going state, F4＝0；
If the live broadcast is in the state of waiting for live broadcast, F4Current time-required live start time |. Wherein F4Representing an hour difference between the two times, wherein the start time of the demand live broadcast is the start time of the demand live broadcast in a state to be live broadcast, which is a certain time point in the future and is not arrived temporarily;
if the live broadcast is required to be in the live broadcast ending state, F4Current time-required live end time | + 2. Wherein | current time-required live broadcast start time | represents an hour difference between two times, and the required live broadcast end time is a time point in the history of the end time of the required live broadcast in the live broadcast end state, which has already occurred.
(5)F5And the element is e {0, 0.125, 0.25, 0.5, 1}, and represents the matching degree of the interest indexes of the users of the same company with the user and the tags needing live broadcasting, and the interest indexes of the users of the same company with the user are called enterprise tags. The enterprise interest tags and the tags requiring live broadcasting are final-stage tags. The tags are divided into four levels, namely a first-level tag, a second-level tag, a third-level tag and a fourth-level tag (a final-level tag). The label with the hierarchical relationship comprises an electronic component (a first-level label), a chip (a second-level label), a logic chip (a third-level label) and a counter (a last-level label).
If the enterprise interest final label and the demand live broadcast final labelMatching labels, if matching is successful, F5Matching the first-level label, namely the third-level label, if the matching fails, 1;
f, if the enterprise interest third-level label is matched with the third-level label needing live broadcast and the matching is successful5If the matching fails, continuing to match the upper-level label, namely the second-level label, if the matching fails, continuing to match the upper-level label;
if the enterprise interest secondary label is matched with the secondary label required for live broadcast, F is carried out if the matching is successful5If the matching fails, continuing to match the upper-level label, namely the first-level label;
if the primary label of the enterprise interest label is matched with the primary label required for live broadcasting, F is carried out if the matching is successful5F, matching failure, 0.1255＝0。
2. Recommendation index for technical-related live broadcasts:
Figure BDA0003049918500000121
wherein d isiThe i-th weight is set in advance, and nd is 5, which can be adjusted appropriately according to the later use condition.
Initial setting: d1＝45，d2＝10，d3＝30，d4＝-0.5，d5＝10
(1)D1And e {0, 1}, which represents the matching degree of the new technical interest index of the user and the tags of the technical related live broadcasts. If the new technical interest index of the user contains a tag of the technical related live broadcast, D11 is ═ 1; otherwise, if the new technical interest index of the user does not contain the tag of the technical related live broadcast, D1＝0。
The types of the new technical labels are as follows:
Figure BDA0003049918500000131
(2)D2e {0.2, 0.4, 0.7, 1}, which represents the score of the stage where the new technology in the technology-related live broadcast is located.
The following shows the stage situation where the new technology in the technology-dependent live broadcast is located.
Figure BDA0003049918500000132
If the new technology in the technology-related live broadcast is in the mass production stage, D2＝1；
If the stage of the new technology in the technology-related live broadcast is a new product test stage, D2＝0.7；
If the stage of the new technology in the technology-related live broadcast is an experimental demonstration stage, D2＝0.4；
If the stage of the new technology in the technology-related live broadcast is the scheme demonstration stage, D2＝0.2。
(3)D3∈[0，1]Indicating how well the user's historical data tags match the tags of the technology-related live broadcast. And counting the times d _ num that the user clicks tags of the technical related live broadcast and the speech in the technical related live broadcast in a day of d _ days.
d_numlabelNumber of times 2 of related live broadcast of technique of clicking tag as cable
+ technology-related live broadcast speaking times with label of cable
Figure BDA0003049918500000133
D _ num represents a calculation frequency set of historical clicks and speeches of all new technical interest tags in a d _ days near by a user, and the number of elements in d _ num is the same as that of the new technical interest tags; max (D _ num) is the maximum value in the calculation number set D _ num, and when max (D _ num) is 0, there is D for any required live broadcast30, no calculation by the above formula is required; d _ numlabelRepresenting the corresponding calculation times of the interest label of the new technology, namely the table; d3Indicating the degree of match of the technology-related live under the user's historical data tags.
(4)D4E [0, + ∞)) represents the user's currentThe degree of match of the previous time with the technology-dependent live time.
The state of the related live broadcast is divided into:
if the technology-related live broadcast is in an on-going state, D4＝0；
If the related live broadcast is in a state of waiting for live broadcast, D4Current time-technology dependent live start time |. Wherein D4Representing the day difference between the two times, wherein the technology-related live broadcast starting time is the starting time of the technology-related live broadcast in a state of waiting for live broadcast, is a certain time point in the future and temporarily does not arrive;
if the technology-related live broadcast is in a live broadcast end state,
D4current time-technology dependent live end time | + 2.
Where | current time-required live broadcast start time | represents the difference in days between two times, and the technology-related live broadcast end time is a point in time when the end time of the technology-related live broadcast in the live broadcast end state is historical and has occurred.
(5)D5And e {0, 1}, which represents the matching degree of the new technical interest category index of the user in the same company with the user and the category label of the technical related live broadcast, wherein the new technical interest category index of the user in the same company with the user is called an enterprise new technical label.
If the enterprise new technology interest label is matched with the label of the related technology live broadcast, and the matching is successful, D51, failure of matching then5＝0；
And step three, calculating the recommendation index TS of the live broadcast similar user. TS is more than or equal to 0 and less than or equal to 1, and represents the matching degree of similar users of the users and the live broadcast.
1. Counting similar users with users, selecting n users with higher similarity, called similar users, and temporarily setting n to 20, wherein the value of n can be adjusted appropriately according to the use condition in the later period. And calculating the similarity between the current user and other users, wherein the first n other users with larger similarity values are used as the similarity users of the current user.
The similarity between two users is calculated as:
Figure BDA0003049918500000141
wherein, simL-uAnd the similarity between L and U is represented, and L and U respectively represent the tag sets of two users, wherein the tag sets comprise the tag sets of interest of the new technology of the users, the required tag sets of interest of the final stage and tags corresponding to elements with element values larger than sn in the num of the calculation times of the users. And the value of sn is temporarily set as sn-5, and is adjusted according to the use condition in the later period. L | represents the length of the tag set L, i.e. the number of elements in the tag set L; l | n |, U |, represents the length of the intersection of the tag set L and the tag set U, i.e., the number of elements in the intersection.
2. And counting the number of similar users of each live broadcast, which are checked in the last login and current latest refreshing time of the user. And calculates the TS value for it:
Figure BDA0003049918500000151
wherein u represents the number of people similar users click on the live broadcast (no repeated calculation), and n represents the number of similar users. The live broadcast without viewing has TS equal to 0, and does not need to be calculated by the above formula.
Step four, calculating the recommendation degree of live broadcast, including the recommendation degree of the required live broadcast and the recommendation degree of the technology-related live broadcast;
recommendation degree of demand-related live broadcast: RF ═ F + RF × TS
Recommendation degree of related live broadcast: RD + RD × TS
Wherein, rf, rd are used for controlling the important degree that live similar user's recommendation index occupied in live recommendation degree, and rf, rd are tentatively 7, and the later stage can carry out appropriate adjustment according to the use thing.
Generating a recommendation list which comprises a demand live broadcast recommendation list and a technology-related live broadcast recommendation list;
sequencing the demand live broadcasts according to the recommendation degrees of the demand live broadcasts from large to small to form a demand live broadcast recommendation list; and sequencing the technical related live broadcasts according to the recommendation degrees of the technical related live broadcasts from large to small to form a technical related live broadcast recommendation list.
And step six, generating a user recommendation list, and generating the user recommendation list according to the technical direction and/or the demand direction of the user interested direction and the technical related live broadcast recommendation list and/or the demand live broadcast recommendation list.
(1) If the interested direction of the user is the technical direction, the top tn live broadcast recommendation lists and the top qn live broadcast recommendation lists in the technology correlation are selected, the front positions and the back positions of the required live broadcast recommendation lists and the technology correlation live broadcast recommendation lists in the requirement live broadcast recommendation lists and the technology correlation live broadcast recommendation lists are randomly displayed but are not changed, and the required live broadcast recommendation lists and the technology correlation live broadcast recommendation lists are deleted. And the rest can be done, that is, (tn + qn) live broadcasts in the user recommendation list are grouped, wherein the (tn) technical related live broadcasts and qn demand live broadcasts are included (the specific number can be adjusted when the system is used later, and is temporarily set to tn-5, and qn-1). When the demand live broadcast recommendation list or the technology-related live broadcast recommendation list is empty, directly adding the recommendation list which is not empty to the tail part of the user recommendation list;
(2) if the interested direction of the user is the required direction, selecting the top tm from the technical related live broadcast recommendation list and the top qm from the required live broadcast recommendation list, and randomly displaying the required live broadcast and the technical related live broadcast without changing the front and back positions of the required live broadcast recommendation list and the technical related live broadcast recommendation list respectively, and deleting the required live broadcast recommendation list and the technical related live broadcast recommendation list. And analogizing in turn, that is, the (tm + qm) live broadcasts in the user recommendation list are grouped into a group, which includes tm technology-related live broadcasts and qm demand live broadcasts (the specific number can be adjusted when being used later, and is temporarily set to tm being 10, and qm being 1). When the demand live broadcast recommendation list or the technology-related live broadcast recommendation list is empty, directly adding the recommendation list which is not empty to the tail part of the user recommendation list;
(3) if the interested direction of the user is a required direction and a technical direction, selecting the first t in the technical related live broadcast recommendation list and the first q in the required live broadcast recommendation list, and randomly displaying the required live broadcast and the technical related live broadcast without changing the front and back positions of the required live broadcast and the technical related live broadcast recommendation list respectively, and deleting the required live broadcast and the technical related live broadcast recommendation list. And analogizing in turn, that is, the (t + q) live broadcasts in the user recommendation list are grouped into a group, wherein t technical-related live broadcasts and q demand live broadcasts are included (the specific number can be adjusted when the system is used later, and is temporarily set to t-2 and q-3). And when the demand live broadcast recommendation list or the technology-related live broadcast recommendation list is empty, directly adding the recommendation list which is not empty to the tail part of the user recommendation list.
It should be understood that the specific order or hierarchy of steps in the processes disclosed is an example of exemplary approaches. Based upon design preferences, it is understood that the specific order or hierarchy of steps in the processes may be rearranged without departing from the scope of the present disclosure. The accompanying method claims present elements of the various steps in a sample order, and are not intended to be limited to the specific order or hierarchy presented.
In the foregoing detailed description, various features are grouped together in a single embodiment for the purpose of streamlining the disclosure. This method of disclosure is not to be interpreted as reflecting an intention that the claimed embodiments of the subject matter require more features than are expressly recited in each claim. Rather, as the following claims reflect, invention lies in less than all features of a single disclosed embodiment. Thus, the following claims are hereby expressly incorporated into the detailed description, with each claim standing on its own as a separate preferred embodiment of the invention.
Those of skill would further appreciate that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present disclosure.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor such the processor can read information from, and write information to, the storage medium. Of course, the storage medium may also be integral to the processor. The processor and the storage medium may reside in an ASIC. The ASIC may reside in a user terminal. Of course, the processor and the storage medium may reside as discrete components in a user terminal.
For a software implementation, the techniques described herein may be implemented with modules (e.g., procedures, functions, and so on) that perform the functions described herein. The software codes may be stored in memory units and executed by processors. The memory unit may be implemented within the processor or external to the processor, in which case it can be communicatively coupled to the processor via various means as is known in the art.
What has been described above includes examples of one or more embodiments. It is, of course, not possible to describe every conceivable combination of components or methodologies for purposes of describing the aforementioned embodiments, but one of ordinary skill in the art may recognize that many further combinations and permutations of various embodiments are possible. Accordingly, the embodiments described herein are intended to embrace all such alterations, modifications and variations that fall within the scope of the appended claims. Furthermore, to the extent that the term "includes" is used in either the detailed description or the claims, such term is intended to be inclusive in a manner similar to the term "comprising" as "comprising" is interpreted when employed as a transitional word in a claim. Furthermore, any use of the term "or" in the specification of the claims is intended to mean a "non-exclusive or".

Claims (8)

1. A live broadcast recommendation method is used for live broadcast recommendation for BCB electronic commerce platform users and is characterized by comprising the following steps:
acquiring user information which comprises information of a user center, user history records and information of the user centers of other users consistent with the user company information;
step two, calculating a recommendation index of each live broadcast to the user, wherein the live broadcast comprises new technology live broadcast, technology demand live broadcast and purchase demand live broadcast; the purchase demand live broadcast is called demand live broadcast for short, and the new technology live broadcast and the technical demand live broadcast are called technical related live broadcast;
step three, calculating a recommendation index of the live broadcast similar user, and representing the matching degree of the similar user of the user and the live broadcast;
step four, calculating the recommendation degree of live broadcast, including the recommendation degree of the required live broadcast and the recommendation degree of the technology-related live broadcast;
generating a recommendation list which comprises a demand live broadcast recommendation list and a technology-related live broadcast recommendation list;
and step six, generating a user recommendation list, and generating the user recommendation list according to the technical direction and/or the demand direction of the user interested direction and the technical related live broadcast recommendation list and/or the demand live broadcast recommendation list.
2. The live recommendation method of claim 1, wherein in step one:
the information of the user center comprises information of user registration and later maintenance; specifically, the interested directions and corresponding labels of the user comprise technical related live broadcast category labels and live broadcast category labels required; a region of the user including a service region of a supplier identity and a delivery region of a buyer identity;
the user history records comprise live broadcast viewing records of the user and speech records in live broadcast;
the method comprises the steps of obtaining information of user centers of other users consistent with user company information, wherein the information comprises information of unified company user registration and later maintenance, specifically interested directions and corresponding labels of the users, and the information comprises technical related live broadcast categories and demand labels.
3. The live recommendation method of claim 1, wherein in the second step: the recommendation index for demand live broadcast is calculated as follows:
recommendation index for live broadcast of demand:
Figure FDA0003049918490000021
wherein f isiThe weight of the ith weight is preset, nf is 5, and the weight is adjusted appropriately according to the later use condition; initial setting: f. of1＝45，f2＝10，f3＝30，f4＝-0.5，f5＝10；
(1)F1E.g., {0, 0.125, 0.25, 0.5, 1}, which represents the matching degree of the interest index of the user and the label required for live broadcast, wherein the interest label required by the user and the label required for live broadcast are both final-level labels; the tags are divided into four stages, namely a first-stage tag, a second-stage tag, a third-stage tag and a last-stage tag;
f, if the final-stage label of the demand interest of the user is matched with the final-stage label of the demand live broadcast, if the matching is successful1Matching the first-level label, namely the third-level label, if the matching fails, 1;
if the third-level label of the interest of the user is matched with the third-level label of the live broadcast, and if the matching is successful, F1If the matching fails, continuing to match the upper-level label, namely the second-level label, if the matching fails, continuing to match the upper-level label;
if the secondary label of the user's interest is matched with the secondary label of the live broadcast, F1If matching fails, the matching process continues to the upper label, i.e. the first label, which is 0.25Matching;
if the first-level label of the user's interest is matched with the first-level label of the live broadcast, F1F, matching failure, 0.1251＝0；
(2)F2E {0, 1}, which represents the matching degree of the area of the user and the delivery place requiring live broadcasting, wherein the specific service area and the delivery place are both specific to provinces, namely, the matching is successful on province addresses, namely, the delivery or transaction can be carried out by default;
if the area where the user is the buyer is the same as the delivery request place requiring live broadcasting, F2＝1；
If the user is the supplier's supply area containing the delivery place requiring live broadcast, F2＝1；
If the above condition is not satisfied, it indicates that the region of the user fails to match the delivery place requiring live broadcast, i.e., F2＝0；
(3)F3∈[0，1]The matching degree of the historical data labels of the user and the labels needing live broadcast is represented, and a label frequency set num for clicking the labels needing live broadcast and speaking in the needing live broadcast in days of the user is counted;
numlabelnumber of times of live broadcasting with tab as clicked × 2+ number of times of speech with tab as table as clicked
Figure FDA0003049918490000031
The num represents a calculation frequency set of historical clicks and speeches of all final-stage labels in a user's near days, and the number of elements in the num is the same as that of the final-stage labels; max (num) is the maximum value in the number of calculations set num, and when max (num) is 0, there is F for any desired live broadcast30, no calculation by the above formula is required; numlabelRepresenting the corresponding calculation times of the last-stage label table; f3Representing the matching degree of the required live broadcast under the historical data label of the user;
(4)F4e [0, + ∞)) representing the user's current time and desired live timeThe matching degree, namely the current time refers to the latest refreshing time, the login time is taken as the standard when the user does not refresh after logging in, and the latest refreshing time is taken as the standard when the user has refreshing action after logging in; the live broadcast state is divided into:
if the live broadcast is in an on-going state, F4＝0；
If the live broadcast is in the state of waiting for live broadcast, F4Current time-required live start time |. Wherein F4Representing an hour difference between the two times, wherein the start time of the demand live broadcast is the start time of the demand live broadcast in a state to be live broadcast, which is a certain time point in the future and is not arrived temporarily;
if the live broadcast is required to be in the live broadcast ending state, F4The live broadcasting method comprises the steps that 1, current time-required live broadcasting end time | +2, wherein | the current time-required live broadcasting start time | represents an hour difference between two times, and the required live broadcasting end time is a certain time point of history of end time of required live broadcasting in a live broadcasting end state and occurs;
(5)F5the method comprises the following steps that (1) an element of {0, 0.125, 0.25, 0.5, 1}, which represents the matching degree of interest indexes of users of the same company with users and tags needing live broadcast, wherein the interest indexes of the users of the same company with the users are called enterprise tags, the enterprise interest tags and the tags needing live broadcast are final-stage tags, and the tags are divided into four stages, namely first-stage tags, second-stage tags, third-stage tags and final-stage tags;
f, if the enterprise interest final label is matched with the final label needing live broadcast and the matching is successful5Matching the first-level label, namely the third-level label, if the matching fails, 1;
f, if the enterprise interest third-level label is matched with the third-level label needing live broadcast and the matching is successful5If the matching fails, continuing to match the upper-level label, namely the second-level label, if the matching fails, continuing to match the upper-level label;
if the enterprise interest secondary label is matched with the secondary label required for live broadcast, F is carried out if the matching is successful5If the matching fails, continuing to match the upper-level label, namely the first-level label;
if the first level label of the enterprise interest labelMatching the label with a first-level label required for live broadcasting, and F if the matching is successful5F, matching failure, 0.1255＝0。
4. The live recommendation method of claim 3, wherein in the second step: the recommendation index for technology-dependent live broadcast is calculated as follows:
recommendation index for technical-related live broadcasts:
Figure FDA0003049918490000041
wherein d isiThe weight of the ith weight is set in advance, nd is 5, and the weight is adjusted appropriately according to the later use condition; initial setting: d1＝45，d2＝10，d3＝30，d4＝-0.5，d5＝10
(1)D1E {0, 1}, representing the matching degree of the new technical interest index of the user and the tags of the technical related live broadcasts, and if the new technical interest index of the user contains the tags of the technical related live broadcasts, D11 is ═ 1; otherwise, if the new technical interest index of the user does not contain the tag of the technical related live broadcast, D1＝0；
(2)D2The element is {0.2, 0.4, 0.7, 1}, and the score of the stage where the new technology in the technology-related live broadcast is located is represented;
if the new technology in the technology-related live broadcast is in the mass production stage, D2＝1；
If the stage of the new technology in the technology-related live broadcast is a new product test stage, D2＝0.7；
If the stage of the new technology in the technology-related live broadcast is an experimental demonstration stage, D2＝0.4；
If the stage of the new technology in the technology-related live broadcast is the scheme demonstration stage, D2＝0.2；
(3)D3∈[0，1]Showing the matching degree of the historical data label of the user and the label of the technology-related live broadcast, counting the click on the technology-related live broadcast and the technology-related live broadcast in the near d _ days of the userThe number of times d _ num of the tag of the middle utterance;
d_numlabelnumber of times of click on technical-related live broadcast with label of label 2+ number of talk times of technical-related live broadcast with label of cable
Figure FDA0003049918490000042
D _ num represents a calculation frequency set of historical clicks and speeches of all new technical interest tags in a d _ days near by a user, and the number of elements in d _ num is the same as that of the new technical interest tags; max (D _ num) is the maximum value in the calculation number set D _ num, and when max (D _ num) is 0, there is D for any required live broadcast30, no calculation by the above formula is required; d _ numlabelRepresenting the corresponding calculation times of the interest label of the new technology, namely the table; d3Representing the matching degree of the related live broadcast of the technology under the historical data label of the user;
(4)D4e [0, + ∞), representing how well the user's current time matches the technology-related live time;
the state of the related live broadcast is divided into:
if the technology-related live broadcast is in an on-going state, D4＝0；
If the technology-related live broadcast is in a state of waiting for live broadcast,
D4current time-technology dependent live start time |; wherein D4Representing the day difference between the two times, wherein the technology-related live broadcast starting time is the starting time of the technology-related live broadcast in a state of waiting for live broadcast, is a certain time point in the future and temporarily does not arrive;
if the technology-related live broadcast is in a live broadcast end state,
D4current time-technology dependent live broadcast end time | + 2; wherein | the current time-the required live broadcast start time | represents a difference in days between two times, and the technology-related live broadcast end time is a certain time point in history, at which the technology-related live broadcast end time in the live broadcast end state has occurred;
(5)D5the element is {0, 1}, the matching degree of the new technical interest category index of the user of the same company with the user and the category label of the related live broadcast of the technology is represented, and the new technical interest category index of the user of the same company with the user is called an enterprise new technical label;
if the enterprise new technology interest label is matched with the label of the related technology live broadcast, and the matching is successful, D51, failure of matching then5＝0。
5. The live broadcasting recommendation method of claim 1, wherein the third step of calculating the recommendation index of the live broadcasting similar user comprises the following steps:
1. counting similar users with users, selecting n users with higher similarity, called similar users, temporarily setting n as 20, and properly adjusting the value of n according to the use condition in the later period; calculating the similarity between the current user and other users, wherein the first n other users with larger similarity values are used as the similarity users of the current user;
the similarity between two users is calculated as:
Figure FDA0003049918490000061
wherein, simL-uThe similarity between L and U is represented, the L and the U respectively represent tag sets of two users, and the tag sets comprise user new technology interest tag sets, required final level interest tag sets and tags corresponding to elements with element values larger than sn in a user calculation number set num; the value of sn is temporarily set as sn-5, and is adjusted according to the use condition in the later period; l | represents the length of the tag set L, i.e. the number of elements in the tag set L; l | n |, U |, represents the length of the intersection of the tag set L and the tag set U, i.e., the number of elements in the intersection;
2. and counting the number of similar users of each live broadcast, which are checked in the last login and current latest refreshing time of the user. And calculates the TS value for it:
Figure FDA0003049918490000062
wherein u represents the number of people who click the live broadcast by similar users (no repeated calculation), and n represents the number of the similar users; the live broadcast without viewing has TS equal to 0, and does not need to be calculated by the above formula.
6. The live broadcast recommendation method according to claim 1, wherein the fourth step of calculating the live broadcast recommendation degree includes:
recommendation degree of demand-related live broadcast: RF ═ F + RF × TS
Recommendation degree of related live broadcast: RD + RD × TS
Wherein, rf, rd are used for controlling the important degree that live similar user's recommendation index occupied in live recommendation degree, and rf, rd are tentatively 7, and the later stage is adjusted according to the use thing appropriately.
7. The live recommendation method of claim 1, wherein step five, generating recommendation lists, including a demand live recommendation list and a technology-dependent live recommendation list; the method comprises the following steps:
sequencing the demand live broadcasts according to the recommendation degrees of the demand live broadcasts from large to small to form a demand live broadcast recommendation list; and sequencing the technical related live broadcasts according to the recommendation degrees of the technical related live broadcasts from large to small to form a technical related live broadcast recommendation list.
8. The live recommendation method of claim 1, wherein the sixth step of generating a user recommendation list comprises:
(1) if the interested direction of the user is the technical direction, selecting the top tn live broadcast recommendation lists and the top qn live broadcast recommendation lists from the technology-related live broadcast recommendation lists, randomly displaying the demand live broadcast and the technology-related live broadcast without changing the front and back positions of the demand live broadcast recommendation lists and the technology-related live broadcast recommendation lists, and deleting the demand live broadcast recommendation lists and the technology-related live broadcast recommendation lists from the demand live broadcast list and the technology-related live broadcast recommendation lists; and analogizing in turn, namely (tn + qn) live broadcasts in the user recommendation list are taken as a group, wherein the group comprises tn technical-related live broadcasts and qn demand live broadcasts; the specific quantity is adjusted in the later use, and is temporarily set to be tn-5 and qn-1; when the demand live broadcast recommendation list or the technology-related live broadcast recommendation list is empty, directly adding the recommendation list which is not empty to the tail part of the user recommendation list;
(2) if the interested direction of the user is a demand direction, selecting the top tm from the technology-related live broadcast recommendation list and the top qm from the demand live broadcast recommendation list, and randomly displaying the demand live broadcast and the technology-related live broadcast without changing the front and back positions of the demand live broadcast and the technology-related live broadcast recommendation list respectively, and deleting the demand live broadcast and the technology-related live broadcast recommendation list; and analogizing in turn, namely taking (tm + qm) live broadcasts in the user recommendation list as a group, wherein the group comprises tm technology-related live broadcasts and qm demand live broadcasts, and the specific number is adjusted in the later period of use and is temporarily set to be tm equal to 10, and qm equal to 1; when the demand live broadcast recommendation list or the technology-related live broadcast recommendation list is empty, directly adding the recommendation list which is not empty to the tail part of the user recommendation list;
(3) if the interested direction of the user is a required direction and a technical direction, selecting the first t in a technical related live broadcast recommendation list and the first q in the required live broadcast recommendation list, randomly displaying the required live broadcast and the technical related live broadcast without changing the front and back positions of the required live broadcast and the technical related live broadcast recommendation list, and deleting the required live broadcast and the technical related live broadcast recommendation list; and analogizing in sequence, namely, the (t + q) live broadcasts in the user recommendation list are taken as a group, wherein t technical related live broadcasts and q required live broadcasts are included; the specific quantity is adjusted in the later use, and t is temporarily set to be 2, and q is temporarily set to be 3; and when the demand live broadcast recommendation list or the technology-related live broadcast recommendation list is empty, directly adding the recommendation list which is not empty to the tail part of the user recommendation list.
CN202110482943.2A 2021-04-30 2021-04-30 Live broadcast recommendation method Active CN113159855B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110482943.2A CN113159855B (en) 2021-04-30 2021-04-30 Live broadcast recommendation method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110482943.2A CN113159855B (en) 2021-04-30 2021-04-30 Live broadcast recommendation method

Publications (2)

Publication Number Publication Date
CN113159855A true CN113159855A (en) 2021-07-23
CN113159855B CN113159855B (en) 2023-01-13

Family

ID=76872856

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110482943.2A Active CN113159855B (en) 2021-04-30 2021-04-30 Live broadcast recommendation method

Country Status (1)

Country Link
CN (1) CN113159855B (en)

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104504149A (en) * 2015-01-08 2015-04-08 中国联合网络通信集团有限公司 Application recommendation method and device
CN106126669A (en) * 2016-06-28 2016-11-16 北京邮电大学 User collaborative based on label filters content recommendation method and device
CN106294800A (en) * 2016-08-16 2017-01-04 武汉斗鱼网络科技有限公司 Method and system recommended by direct broadcasting room based on weighting k neighbour scoring
CN108271076A (en) * 2017-01-03 2018-07-10 武汉斗鱼网络科技有限公司 A kind of method and device for recommending direct broadcasting room
CN109218769A (en) * 2018-09-30 2019-01-15 武汉斗鱼网络科技有限公司 A kind of recommended method and relevant device of direct broadcasting room
CN109413461A (en) * 2018-09-30 2019-03-01 武汉斗鱼网络科技有限公司 A kind of recommended method and relevant device of direct broadcasting room
CN109543111A (en) * 2018-11-28 2019-03-29 广州虎牙信息科技有限公司 Recommendation information screening technique, device, storage medium and server
CN110730385A (en) * 2018-07-16 2020-01-24 武汉斗鱼网络科技有限公司 Live broadcast room recommendation method and device, server and storage medium
CN110896488A (en) * 2018-08-23 2020-03-20 武汉斗鱼网络科技有限公司 Recommendation method for live broadcast room and related equipment

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104504149A (en) * 2015-01-08 2015-04-08 中国联合网络通信集团有限公司 Application recommendation method and device
CN106126669A (en) * 2016-06-28 2016-11-16 北京邮电大学 User collaborative based on label filters content recommendation method and device
CN106294800A (en) * 2016-08-16 2017-01-04 武汉斗鱼网络科技有限公司 Method and system recommended by direct broadcasting room based on weighting k neighbour scoring
CN108271076A (en) * 2017-01-03 2018-07-10 武汉斗鱼网络科技有限公司 A kind of method and device for recommending direct broadcasting room
CN110730385A (en) * 2018-07-16 2020-01-24 武汉斗鱼网络科技有限公司 Live broadcast room recommendation method and device, server and storage medium
CN110896488A (en) * 2018-08-23 2020-03-20 武汉斗鱼网络科技有限公司 Recommendation method for live broadcast room and related equipment
CN109218769A (en) * 2018-09-30 2019-01-15 武汉斗鱼网络科技有限公司 A kind of recommended method and relevant device of direct broadcasting room
CN109413461A (en) * 2018-09-30 2019-03-01 武汉斗鱼网络科技有限公司 A kind of recommended method and relevant device of direct broadcasting room
CN109543111A (en) * 2018-11-28 2019-03-29 广州虎牙信息科技有限公司 Recommendation information screening technique, device, storage medium and server

Also Published As

Publication number Publication date
CN113159855B (en) 2023-01-13

Similar Documents

Publication Publication Date Title
EP3471374B1 (en) Method and device for identifying type of geographic location at where user is located
US7281001B2 (en) Data quality system
US20170364506A1 (en) System and method for generating phrase based categories of interactions
CN110428322A (en) A kind of adaptation method and device of business datum
CN107832291B (en) Man-machine cooperation customer service method, electronic device and storage medium
CN108681970A (en) Finance product method for pushing, system and computer storage media based on big data
CN104461863A (en) Service system testing method, device and system
CN109871861B (en) System and method for providing coding for target data
CN109299227B (en) Information query method and device based on voice recognition
CN107133289A (en) A kind of method and apparatus of determination commercial circle
CN110515999A (en) General record processing method, device, electronic equipment and storage medium
US20060293915A1 (en) Method for optimizing accuracy of real estate valuations using automated valuation models
CN112988963A (en) User intention prediction method, device, equipment and medium based on multi-process node
CN110675249A (en) Matching method, device, server and storage medium for network lending
CN110147286B (en) Custom information sending method, device and system based on message queue and recording medium
Lagos et al. International tourism and poverty alleviation: cross-country evidence using panel quantile fixed effects approach
US20240143605A1 (en) System And Method For Improved State Identification And Prediction In Computerized Queries
CN113159855B (en) Live broadcast recommendation method
CN112270350B (en) Method, apparatus, device and storage medium for portraying organization
US20200364537A1 (en) Systems and methods for training and executing a recurrent neural network to determine resolutions
CN117151794A (en) Advertisement task recommendation method and related device
CN107563809A (en) Advertisement placement method and device
CN116362601A (en) Work efficiency evaluation method, apparatus, device, medium and program product
CN112397147B (en) Gene sequencing delivery method and system for realizing collinear production based on intelligent scheduling technology
CN111507366B (en) Training method of recommendation probability model, intelligent completion method and related device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant