CN115544362A - AI-based content recommendation system - Google Patents

AI-based content recommendation system Download PDF

Info

Publication number
CN115544362A
CN115544362A CN202211241342.3A CN202211241342A CN115544362A CN 115544362 A CN115544362 A CN 115544362A CN 202211241342 A CN202211241342 A CN 202211241342A CN 115544362 A CN115544362 A CN 115544362A
Authority
CN
China
Prior art keywords
information
control unit
central control
recommendation
recommendation information
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
CN202211241342.3A
Other languages
Chinese (zh)
Other versions
CN115544362B (en
Inventor
刘亚男
陈家峰
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.)
Readboy Education Technology Co Ltd
Original Assignee
Readboy Education 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 Readboy Education Technology Co Ltd filed Critical Readboy Education Technology Co Ltd
Priority to CN202211241342.3A priority Critical patent/CN115544362B/en
Publication of CN115544362A publication Critical patent/CN115544362A/en
Application granted granted Critical
Publication of CN115544362B publication Critical patent/CN115544362B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9536—Search customisation based on social or collaborative filtering
    • 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
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/01—Social networking
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • Primary Health Care (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Information Transfer Between Computers (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention discloses an AI-based content recommendation system, which comprises an intelligent acquisition unit, a recommendation unit and a recommendation unit, wherein the intelligent acquisition unit is used for acquiring chat information according to chat contents of a user; the analysis processing unit is used for analyzing and integrating the chat information to judge the chat information type and extracting key information; the central control unit is used for judging the chat content type of the user, whether the user has a requirement on the content in the picture information, whether the picture information is successfully matched with the recommendation information and how to specifically push the recommendation information according to the comparison result of the key parameters detected by the analysis processing unit and the preset standard; the cloud platform comprises a recommendation information base used for storing recommendation information and a key information repository used for storing key information; the system comprises a manual service unit and a content recommendation system, wherein the manual service unit is used for actively controlling the content recommendation system through a central control unit.

Description

AI-based content recommendation system
Technical Field
The invention relates to the field of Internet content recommendation, in particular to an AI-based content recommendation system.
Background
With the development of science and technology, the development of learning software on the internet is also leap forward, and a plurality of communication groups for communication between students and parents exist so as to better improve the learning performance of the students, and it is a hot spot that people pay attention to how to intelligently and automatically judge the needs of the students and the parents through the chat contents in the communication groups so as to better provide the students and the parents with contents more fitting the needs of the students and the parents on the learning software.
Chinese patent publication No. CN111506804A discloses a client content recommendation system and method based on user terminal behavior, including a data recording module, a user preference module, a resource acquisition module, and a content recommendation module, wherein the data recording module is configured to record data generated when a user accesses content; the user preference module is configured to determine user preferences based at least in part on the data recorded by the data recording module; the resource acquisition module is configured to acquire a content resource list from a server; the content recommendation module is configured to recommend one or more content in the content resource list based at least in part on the determined user preferences. Therefore, the client content recommendation system and method based on the user terminal behaviors have the following problems: the content recommendation is inaccurate because the user demand cannot be acquired through the usual chat content of the user.
Disclosure of Invention
Therefore, the invention provides an AI-based content recommendation system, which is used for solving the problem that content recommendation is inaccurate because the user requirements cannot be acquired through the usual chat content of the user in the prior art.
To achieve the above object, the present invention provides an AI-based content recommendation system, including:
the intelligent acquisition unit is arranged at the user side and used for acquiring chat information according to the chat content of the user receiving the authorization application;
the analysis processing unit is connected with the intelligent acquisition unit and is used for analyzing and integrating the chat information by identifying characters in the chat information so as to judge the type of the chat information and extract key information; the chat information types comprise text information, voice information and picture information; the key parameters comprise the proportion of the number of exercises in the picture information, the proportion of the number of knowledge points in the picture information, the number of required keywords in the text information and the number of required keywords in the voice information;
the central control unit is respectively connected with the intelligent acquisition unit, the analysis processing unit and the cloud platform and used for judging the chat content type of the user, whether the user has a demand for the content in the picture information, whether the picture information is successfully matched with the recommendation information and the specific mode of pushing the recommendation information according to the comparison result of the key parameters detected by the analysis processing unit and the preset standard;
the cloud platform comprises a recommendation information base used for storing recommendation information and a key information repository used for storing key information;
and the manual service unit is respectively connected with the central control unit and the cloud platform and is used for actively controlling the content recommendation system through the central control unit.
Furthermore, the analysis processing unit calculates the exercise quantity ratio P1 and the knowledge point quantity ratio P2 in the picture information when analyzing and integrating the chat information and transmits the calculation result to the central control unit, the central control unit judges the chat content type of the user according to the comparison result of the exercise quantity ratio P1 and the knowledge point quantity ratio P2,
if P1 is less than P2, the central control unit judges that the chat content type of the user is knowledge point learning;
if P1 is more than P2, the central control unit judges the chat content type of the user as a discussion question;
if P1= P2, the central control unit judges that the chat content type of the user can not be confirmed and transmits the judgment information to the manual service unit so as to enable the manual service unit to carry out manual judgment.
Further, the analysis processing unit detects the number N of required keywords in the text information and the voice information within a preset time range of the picture information sending time when the central control unit finishes judging the chat content type of the user and transmits the N to the central control unit, and the central control unit compares the N with a preset standard to judge whether the user has a requirement on the content in the picture information; the central control unit is provided with a first preset number Nz, wherein, 0 is more than Nz,
if N is less than Nz, the central control unit judges that the user does not have a requirement on the content in the picture information, does not need to match the picture information with the corresponding classification information of the recommendation information base, and transmits the judgment information to the manual service unit for manual judgment;
and if N is larger than or equal to Nz, the central control unit judges that the user has a demand on the content in the picture information, matches the picture information with the recommendation information of the recommendation information base, and transmits the picture information to a key information repository in the cloud platform for storage.
Further, when the central control unit matches the picture information with the recommendation information in the recommendation information base, the central control unit calculates the similarity between the picture information and each recommendation information and compares each similarity with a preset standard in sequence to judge whether the picture information is successfully matched with the recommendation information, wherein for the ith recommendation information in the recommendation information base, i =1,2,3.. N is the total number of the recommendation information in the recommendation information base, and the central control unit marks the similarity between the picture information and the recommendation information as Si; the central control unit is provided with a first preset similarity Sz1 and a second preset similarity Sz2, wherein Sz1 is more than 0 and Sz2 is more than 0,
if the Si is less than or equal to Sz1, the central control unit judges that the matching between the picture information and the recommendation information fails;
if Sz1 is larger than Si and smaller than or equal to Sz2, the central control unit judges that the picture information is successfully matched with the recommendation information, and records the recommendation information as second recommendation information;
if Sz2 is less than Si, the central control unit determines that the matching between the picture information and the recommendation information is successful, marks the recommendation information as first recommendation information, and pushes the first recommendation information to the user side.
Further, when the matching between the picture information and the recommendation information of the recommendation information base is completed, the central control unit counts the number Ns1 of the first recommendation information and the number Ns2 of the second recommendation information, and compares Ns1 with a preset standard respectively to determine how to push the recommendation information specifically; the central control unit is provided with a first recommended information quantity Nz1, wherein, 0 is more than Nz1,
if Ns1 is not more than Nz1, the central control unit judges that the number of the first recommendation information is less than a preset standard and compares Ns2 with (Nz 1-Ns 1) to judge whether any (Nz 1-Ns 1) second recommendation information is randomly selected for pushing, if Ns2 is not less than Nz1-Ns1, the central control unit judges that any (Nz 1-Ns 1) second recommendation information is randomly selected for pushing, if Ns2 is less than Nz1-Ns1, the central control unit judges that the number of the first recommendation information and the number of the second recommendation information do not accord with the preset standard and transmits the judgment information to the manual service unit for manual adjustment;
if Nz1 is less than Ns1, the central control unit judges that the quantity of the first recommendation information meets a preset standard and randomly selects the first recommendation information to be pushed to the user side.
Further, the central control unit calculates a difference value Δ N between the number N of the required keywords and a first preset number Nz when the determination of how to push the recommendation information is completed, and compares the Δ N with a preset standard to determine whether to adjust the push frequency F of the recommendation information, and sets that Δ N = N-Nz; the central control unit is provided with a first preset difference value delta N1, a second preset difference value delta N2, a standard pushing frequency Fz, a first adjusting coefficient alpha 1 and a second adjusting coefficient alpha 2, wherein, 0 is more than delta N1 and less than delta N2,0 is more than alpha 1 and less than alpha 2,0 is more than Fz,
if the delta N is less than or equal to the delta N1, the central control unit judges that the pushing frequency of the recommendation information does not need to be adjusted and set, and the pushing frequency F = Fz of the recommendation information;
if the number of the delta N1 is less than the number of the delta N and less than or equal to the number of the delta N2, the central control unit judges that the pushing frequency F of the recommendation information is adjusted by using alpha 1, the adjusted pushing frequency is recorded as F ', and F' = Fz multiplied by alpha 1 is set;
if Δ N2 is less than Δ N, the central control unit determines that α 2 is used to adjust the push frequency F of the recommendation information, and the adjusted push frequency is recorded as F ', and F' = Fz × α 2 is set.
Further, the intelligent acquisition unit detects the click rate Q of the user when pushing the recommendation information and transmits the click rate Q to the central control unit, and the central control unit compares the Q with a preset standard respectively to judge whether to adjust the display duration T of the recommendation information; the central control unit is provided with a first preset click rate Q1, a second preset click rate Q2, a third preset click rate Q3, a standard display time length Tz, a first time length adjusting coefficient beta 1 and a second time length adjusting coefficient beta 2, wherein Q1 is more than 0 and less than Q2 and less than Q3, beta 1 is more than 0 and less than beta 2, tz is more than 0,
if Q is less than or equal to Q1, the central control unit preliminarily judges that the user has no requirement on the recommendation information, sends a requirement questionnaire to the user side and transmits the judgment information to the manual service unit;
if Q1 is larger than Q and smaller than or equal to Q2, the central control unit judges that the display time length of the recommendation information does not need to be adjusted and sets the display time length of the recommendation information as T, and T = Tz;
if Q2 is more than or equal to Q3, the central control unit judges that the display time length T of the recommendation information is adjusted by using beta 1, the adjusted display time length is marked as T ', and T' = T multiplied by beta 1 is set;
if Q3 is less than Q, the central control unit judges that the display time length T of the recommendation information is adjusted by using beta 2, the adjusted display time length is marked as T ', and T' = T multiplied by beta 2 is set;
and the central control unit detects the click rate of the user again when the adjustment of the display duration of the recommended information is completed so as to judge whether to continue to adjust the display duration of the recommended information.
Further, the central control unit records the adjustment times M when detecting the click rate of the user again to judge whether to continue to adjust the display duration of the recommended information, and compares M and the adjusted display duration T' with a preset standard respectively to judge whether to allow the adjustment to be carried out; the central control unit is provided with a maximum adjusting time Mmax and a maximum display duration Tmax, wherein Mmax is more than 0, tmax is more than 0,
if M is greater than or equal to Max or N is greater than or equal to Nmax, the central control unit judges that the adjustment is not allowed to be carried out and transmits the judgment information to the manual service unit;
if M < Mmax and N < Nmax, the central control unit determines that this adjustment is allowed.
Furthermore, the cloud platform is in remote communication connection with the central control unit and the manual service unit.
Furthermore, the manual service unit is provided with an audio-visual display screen for receiving the judgment information of the central control unit.
Compared with the prior art, the method has the advantages that the central control unit classifies the chat content of the user, whether the user has a demand for the content in the picture information, whether the picture information is successfully matched with the recommendation information and specifically how to push the recommendation information through the comparison result of the key parameters in the chat information detected by the analysis processing unit and the preset standard, and the comprehensive analysis of the picture information, the text information and the voice information in the chat content improves the judgment efficiency of the method for the demand of the user.
Furthermore, the central control unit judges the chat content classification of the user according to the comparison result of the exercise quantity ratio and the knowledge point quantity ratio, so that the judgment range of the user requirement content is more targeted, the judgment speed of the user requirement is increased, and the judgment efficiency of the user requirement is improved.
Furthermore, the central control unit compares the number of the required keywords in the text information and the voice information with a preset standard to judge whether the user has a requirement on the content in the picture information, so that the problem of wrong pushing caused by inconsistency between the picture information and the requirement of the user is avoided, the accuracy and pertinence of information pushing are improved while the judging speed is ensured, and the judging efficiency of the invention for the requirement of the user is further improved.
Furthermore, the central control unit is provided with a first preset similarity and a second preset similarity, and the range division enables the central control unit to improve the matching speed of matching the picture information with the recommendation information of the recommendation information base, so that the judgment precision is ensured, the judgment speed is improved, and the judgment efficiency of the invention for the customer requirements is improved.
Furthermore, the central control unit records the matched recommendation information as the first recommendation information and the second recommendation information, and supplements the first recommendation information and the second recommendation information when the number of the first recommendation information does not meet the preset standard, so that the problem that the user requirements cannot be met due to the fact that the number of the recommendation information is too small is solved, and the judgment efficiency of the method for the client requirements is improved.
Furthermore, the central control unit is provided with a first preset difference value and a second preset difference value, and the range division enables the judgment speed of whether to adjust the pushing frequency of the recommendation information to be improved, so that the requirements of users are better met, and the judgment efficiency of the invention for the requirements of the users is improved while the judgment precision is ensured.
Drawings
Fig. 1 is a schematic structural diagram of an AI-based content recommendation system according to an embodiment of the present invention;
fig. 2 is a flowchart illustrating the central control unit determining chat content classification of the user according to a comparison result of the exercise quantity ratio and the knowledge point quantity ratio according to the embodiment of the present invention;
fig. 3 is a flowchart illustrating that the central control unit compares the number of required keywords in the text information and the voice information with a preset standard to determine whether the user has a requirement for the content in the picture information according to the embodiment of the present invention;
fig. 4 is a flowchart illustrating that the central control unit compares the similarity Si between the picture information and each piece of recommended information with a preset standard to determine whether the picture information is successfully matched with the recommended information;
fig. 5 is a flowchart illustrating how the central control unit compares the first recommendation information amount with a preset standard to determine how to specifically push the recommendation information according to the embodiment of the present invention.
Detailed Description
In order that the objects and advantages of the invention will be more clearly understood, the invention is further described below with reference to examples; it should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only for explaining the technical principle of the present invention, and do not limit the scope of the present invention.
It should be noted that in the description of the present invention, the terms of direction or positional relationship indicated by the terms "upper", "lower", "left", "right", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings, which are only for convenience of description, and do not indicate or imply that the device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and thus, should not be construed as limiting the present invention.
Furthermore, it should be noted that, in the description of the present invention, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood by those skilled in the art according to specific situations.
Please refer to fig. 1, which is a schematic structural diagram of an AI-based content recommendation system according to an embodiment of the present invention, the AI-based content recommendation system includes:
the intelligent acquisition unit is arranged at the user side and used for acquiring chat information according to the chat content of the user receiving the authorization application;
the analysis processing unit is connected with the intelligent acquisition unit and is used for analyzing and integrating the chat information by identifying characters in the chat information so as to judge the type of the chat information and extract key information; the chat information types comprise text information, voice information and picture information; the key parameters comprise the proportion of the number of exercises in the picture information, the proportion of the number of knowledge points in the picture information, the number of required keywords in the text information and the number of required keywords in the voice information;
the central control unit is respectively connected with the intelligent acquisition unit, the analysis processing unit and the cloud platform and used for judging the chat content type of the user, whether the user has a demand for the content in the picture information, whether the picture information is successfully matched with the recommendation information and the specific mode of pushing the recommendation information according to the comparison result of the key parameters detected by the analysis processing unit and the preset standard;
the cloud platform comprises a recommendation information base used for storing recommendation information and a key information repository used for storing key information;
and the manual service unit is respectively connected with the central control unit and the cloud platform and is used for actively controlling the content recommendation system through the central control unit.
Please refer to fig. 2, which is a flowchart illustrating the central control unit determining the chat content classification of the user according to the comparison result of the exercise quantity ratio and the knowledge point quantity ratio according to the embodiment of the present invention, when the analysis processing unit performs analysis and integration processing on the chat information, the analysis processing unit calculates the exercise quantity ratio P1 and the knowledge point quantity ratio P2 in the picture information and transmits the calculation result to the central control unit, the central control unit determines the chat content category of the user according to the comparison result of the exercise quantity ratio P1 and the knowledge point quantity ratio P2,
if P1 is less than P2, the central control unit judges that the chat content type of the user is knowledge point learning;
if P1 is more than P2, the central control unit judges the chat content type of the user as a discussion question;
if P1= P2, the central control unit judges that the chat content type of the user can not be confirmed and transmits the judgment information to the manual service unit so as to enable the manual service unit to carry out manual judgment.
Please refer to fig. 3, which is a flowchart illustrating that the central control unit compares the number of required keywords in the text information and the voice information with a preset standard to determine whether the user has a requirement for the content in the picture information according to an embodiment of the present invention, the analysis processing unit detects the number N of required keywords in the text information and the voice information within a preset time range of the sending time of the picture information and transmits N to the central control unit when the central control unit determines the chat content type of the user, and the central control unit compares N with the preset standard to determine whether the user has a requirement for the content in the picture information; the central control unit is provided with a first preset number Nz, wherein Nz =2,
if N is less than Nz, the central control unit judges that the user does not have requirements on the content in the picture information, does not need to match the picture information with corresponding classification information of a recommendation information base, and transmits the judgment information to the manual service unit for manual judgment;
and if N is larger than or equal to Nz, the central control unit judges that the user has a demand on the content in the picture information, matches the picture information with the recommendation information of the recommendation information base, and transmits the picture information to a key information repository in the cloud platform for storage.
Please refer to fig. 4, which is a flowchart illustrating that a central control unit compares similarity Si between picture information and each piece of recommendation information with a preset standard to determine whether the picture information is successfully matched with the recommendation information, wherein when the central control unit matches the picture information with the recommendation information in a recommendation information base, the central control unit calculates the similarity between the picture information and each piece of recommendation information and compares each similarity with the preset standard in sequence to determine whether the picture information is successfully matched with the recommendation information, i =1,2,3.. N is the total number of recommendation information in the recommendation information base for the ith piece of recommendation information in the recommendation information base, and the central control unit records the similarity between the picture information and the recommendation information as Si; the central control unit is provided with a first preset similarity Sz1 and a second preset similarity Sz2, wherein Sz1=50%, sz2=80%,
if Si is less than or equal to Sz1, the central control unit judges that the matching of the picture information and the recommendation information fails;
if Sz1 is larger than Si and is smaller than or equal to Sz2, the central control unit judges that the picture information is successfully matched with the recommendation information, and records the recommendation information as second recommendation information;
if Sz2 is less than Si, the central control unit determines that the matching between the picture information and the recommendation information is successful, records the recommendation information as first recommendation information, and pushes the first recommendation information to the user side.
Please refer to fig. 5, which is a flowchart illustrating how the central control unit compares the first recommended information quantity with a preset standard respectively to determine how to specifically push the recommended information according to the embodiment of the present invention, wherein when the central control unit completes matching between the picture information and the recommended information in the recommended information base, the central control unit counts the first recommended information quantity Ns1 and the second recommended information quantity Ns2, and compares Ns1 with the preset standard respectively to determine how to specifically push the recommended information; the central control unit is provided with a first recommended information quantity Nz1, wherein Nz1=8,
if Ns1 is not more than Nz1, the central control unit judges that the quantity of the first recommendation information is less than a preset standard and compares Ns2 with (Nz 1-Ns 1) to judge whether any (Nz 1-Ns 1) second recommendation information is randomly selected for pushing, if Ns2 is not less than Nz1-Ns1, the central control unit judges that any (Nz 1-Ns 1) second recommendation information is randomly selected for pushing, and if Ns2 is less than Nz1-Ns1, the central control unit judges that the quantity of the first recommendation information and the quantity of the second recommendation information do not accord with the preset standard and transmits the judgment information to the manual service unit for manual adjustment;
if Nz1 is less than Ns1, the central control unit judges that the quantity of the first recommendation information meets a preset standard and randomly selects the first recommendation information to be pushed to the user side.
As shown in fig. 1 to 5, when the determination of how to push the recommendation information is completed, the central control unit calculates a difference Δ N between the number N of the required keywords and the first preset number Nz, and compares Δ N with a preset standard to determine whether to adjust the pushing frequency F of the recommendation information, where Δ N = N-Nz; the central control unit is provided with a first preset difference Δ N1, a second preset difference Δ N2, a standard push frequency Fz, a first adjustment coefficient α 1 and a second adjustment coefficient α 2, wherein Δ N1=2, Δ N2=5, α 1=2, α 2=3, fz =2 times/h,
if the delta N is less than or equal to the delta N1, the central control unit judges that the pushing frequency of the recommendation information does not need to be adjusted and set, and the pushing frequency F = Fz of the recommendation information;
if the number of the delta N1 is less than the number of the delta N and less than or equal to the number of the delta N2, the central control unit judges that the pushing frequency F of the recommendation information is adjusted by using alpha 1, the adjusted pushing frequency is recorded as F ', and F' = Fz multiplied by alpha 1 is set;
if Δ N2 is less than Δ N, the central control unit determines that α 2 is used to adjust the push frequency F of the recommendation information, and the adjusted push frequency is recorded as F ', and F' = Fz × α 2 is set.
Specifically, the intelligent acquisition unit detects a user click rate Q when pushing recommendation information and transmits the user click rate Q to the central control unit, and the central control unit compares Q with a preset standard respectively to judge whether to adjust the display duration T of the recommendation information; the central control unit is provided with a first preset click rate Q1, a second preset click rate Q2, a third preset click rate Q3, a standard display time length Tz, a first time length adjusting coefficient beta 1 and a second time length adjusting coefficient beta 2, wherein Q1=20%, Q2=50%, Q3=80%, beta 1=1.6, beta 2=1.8, tz =2min,
if Q is less than or equal to Q1, the central control unit preliminarily judges that the user has no requirement on the recommendation information, sends a requirement questionnaire to the user side and transmits the judgment information to the manual service unit;
if Q1 is larger than Q and smaller than or equal to Q2, the central control unit judges that the display time length of the recommendation information does not need to be adjusted and sets the display time length of the recommendation information as T, and T = Tz;
if Q2 is more than Q and less than or equal to Q3, the central control unit judges that the display time length T of the recommendation information is adjusted by using beta 1, the adjusted display time length is marked as T ', and T' = T multiplied by beta 1 is set;
if Q3 is less than Q, the central control unit judges that the display time length T of the recommendation information is adjusted by using beta 2, the adjusted display time length is marked as T ', and T' = T multiplied by beta 2 is set;
and the central control unit detects the click rate of the user again when the adjustment of the display duration of the recommended information is finished so as to judge whether to continue adjusting the display duration of the recommended information.
Specifically, the central control unit records the adjustment times M when detecting the click rate of the user again to determine whether to continue adjusting the display duration of the recommended information, and compares M and the adjusted display duration T' with a preset standard respectively to determine whether to allow the adjustment; the central control unit is provided with the maximum adjusting times Mmax and the maximum display duration Tmax, wherein, mmax is more than 0, tmax is more than 0,
if M is greater than or equal to Max or N is greater than or equal to Nmax, the central control unit judges that the adjustment is not allowed to be carried out and transmits the judgment information to the manual service unit;
if M < Mmax and N < Nmax, the central control unit determines that this adjustment is allowed.
Specifically, the cloud platform is in remote communication connection with the central control unit and the manual service unit.
Specifically, the manual service unit is provided with a video display screen for receiving the judgment information of the central control unit.
Specifically, the intelligent acquisition unit sends an authorization application to the user before the intelligent acquisition unit acquires the chat information, and if the user accepts the authorization application, the intelligent acquisition unit acquires the chat information according to the chat content of the user receiving the authorization application; if the user refuses the authorization application, the intelligent acquisition unit transmits the information of the user refusing the authorization application to the manual service unit through the central control unit.
Specifically, the analysis processing unit is provided with a voice recognition module for recognizing key parameters in voice information in the chat content; the analysis processing unit is provided with a character recognition module for recognizing key parameters in character information in the chat content; the analysis processing unit is provided with a picture identification module for identifying key parameters in the text information in the chat content; the analysis processing unit has an intelligent screenshot function.
Specifically, the analysis processing unit is provided with a preset time range D, set, D =3min.
Example 1
In this embodiment, when analyzing and integrating chat information, the analysis processing unit calculates a proportion of exercise numbers P1 and a proportion of knowledge points P2 in picture information, P1=20%, and P2=80%, where P1 is less than P2, the central control unit determines that chat content of a user is a knowledge book, in this embodiment, the analysis processing unit detects that there are N =6 required keywords in text information and voice information within a preset time range of the picture information sending time, and at this time, nz is less than N, the central control unit determines that the user has a requirement for content in the picture information, matches the picture information with recommendation information in a recommendation information repository, and transmits the picture information to a key information repository in the cloud platform for storage, in this embodiment, the central control unit calculates a similarity between the picture information and the recommendation information as S2, S2=85%, at this time, sz2 is less than S2, the central control unit determines that the picture information is successfully matched with the recommendation information, records the recommendation information as first recommendation information and pushes the first recommendation information to a client, ns2= 10, and at this time, the central control unit selects Ns2 is less than Ns1, and pushes the recommendation information to a recommendation number of a recommendation information, and selects a recommendation number of a recommendation information in this embodiment, and pushes Ns1 to a user terminal, and selects a recommendation information item. In this embodiment, Δ N = N-Nz =6-2=4, in this case, Δ N1 < [ Δ N ] < [ Δ N2 ], the central control unit determines that the push frequency F of the recommendation information is adjusted using α 1, and the adjusted push frequency is set as F ', where F' =2 × 2=4 times/h. In this embodiment, when the intelligent obtaining unit detects that the click rate Q =80% of the user when pushing the recommendation information, and Q3= Q at this time, the central control unit determines to adjust the display duration T of the recommendation information by using β 2, where the adjusted display duration is denoted as T ', and T' =2 × 1.6=3.2min is set.
Example 2
In this embodiment, the analysis processing unit calculates, when performing analysis and integration processing on chat information, a question number ratio P1 and a knowledge point number ratio P2 in picture information, where P1=80% and P2=20%, where P2 is less than P1, the central control unit determines that chat content of a user is a question document type, in this embodiment, the analysis processing unit detects that there is no requirement for content in picture information by the user, and N =1 is a requirement keyword number in text information and voice information within a preset time range at the time of sending the picture information, where N is less than Nz, the central control unit determines that there is no requirement for content in the picture information by the user, and does not need to match the picture information with corresponding classification information of a recommendation information base, and transmits determination information to the manual service unit for manual determination.
So far, the technical solutions of the present invention have been described in connection with the preferred embodiments shown in the drawings, but it is easily understood by those skilled in the art that the scope of the present invention is obviously not limited to these specific embodiments. Equivalent changes or substitutions of related technical features can be made by those skilled in the art without departing from the principle of the invention, and the technical scheme after the changes or substitutions can be within the protection scope of the invention.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention; various modifications and alterations to this invention will become apparent to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1. An AI-based content recommendation system, comprising:
the intelligent acquisition unit is arranged at the user side and used for acquiring chat information according to the chat content of the user receiving the authorization application;
the analysis processing unit is connected with the intelligent acquisition unit and is used for analyzing and integrating the chat information by identifying characters in the chat information so as to judge the type of the chat information and extract key information; the chat information type comprises character information, voice information and picture information; the key parameters comprise the proportion of the number of exercises in the picture information, the proportion of the number of knowledge points in the picture information, the number of required keywords in the text information and the number of required keywords in the voice information;
the central control unit is respectively connected with the intelligent acquisition unit, the analysis processing unit and the cloud platform and used for judging the chat content type of the user, whether the user has a demand for the content in the picture information, whether the picture information is successfully matched with the recommendation information and the specific mode of pushing the recommendation information according to the comparison result of the key parameters detected by the analysis processing unit and the preset standard;
the cloud platform comprises a recommendation information base used for storing recommendation information and a key information repository used for storing key information;
and the manual service unit is respectively connected with the central control unit and the cloud platform and is used for actively controlling the content recommendation system through the central control unit.
2. The AI-based content recommendation system according to claim 1, wherein the analysis processing unit calculates a problem number ratio P1 and a knowledge point number ratio P2 in picture information and transmits calculation results to the central control unit when analyzing and integrating chat information, the central control unit determines a chat content type of the user based on a comparison result of the problem number ratio P1 and the knowledge point number ratio P2,
if P1 is less than P2, the central control unit judges that the chat content type of the user is knowledge point learning;
if P1 is more than P2, the central control unit judges the chat content type of the user as a discussion question;
if P1= P2, the central control unit judges that the chat content type of the user can not be confirmed and transmits the judgment information to the manual service unit so as to enable the manual service unit to carry out manual judgment.
3. The AI-based content recommendation system of claim 2, wherein the analysis processing unit detects the number N of required keywords in the text information and the voice information within a preset time range of the sending time of the picture information when the central control unit completes the determination of the chat content type of the user and transmits N to the central control unit, and the central control unit compares N with a preset standard to determine whether the user has a need for the content in the picture information; the central control unit is provided with a first preset number Nz, wherein, nz is more than 0,
if N is less than Nz, the central control unit judges that the user does not have a requirement on the content in the picture information, does not need to match the picture information with the corresponding classification information of the recommendation information base, and transmits the judgment information to the manual service unit for manual judgment;
and if N is larger than or equal to Nz, the central control unit judges that the user has a demand on the content in the picture information, matches the picture information with the recommendation information of the recommendation information base, and transmits the picture information to a key information repository in the cloud platform for storage.
4. The AI-based content recommendation system according to claim 3, wherein when the central control unit matches the picture information with the recommendation information in the recommendation information base, the central control unit calculates the similarity between the picture information and each recommendation information and compares the similarity with a preset standard in sequence to determine whether the picture information is successfully matched with the recommendation information, wherein for the ith recommendation information in the recommendation information base, i =1,2,3.. N, where n is the total number of recommendation information in the recommendation information base, and the central control unit marks the similarity between the picture information and the recommendation information as Si; the central control unit is provided with a first preset similarity Sz1 and a second preset similarity Sz2, wherein Sz1 is more than 0 and Sz2 is more than 0,
if Si is less than or equal to Sz1, the central control unit judges that the matching of the picture information and the recommendation information fails;
if Sz1 is larger than Si and smaller than or equal to Sz2, the central control unit judges that the picture information is successfully matched with the recommendation information, and records the recommendation information as second recommendation information;
if Sz2 is less than Si, the central control unit determines that the matching between the picture information and the recommendation information is successful, records the recommendation information as first recommendation information, and pushes the first recommendation information to the user side.
5. The AI-based content recommendation system according to claim 4, wherein the central control unit counts the number Ns1 of the first recommendation information and the number Ns2 of the second recommendation information when the matching of the picture information and the recommendation information of the recommendation information base is completed, and compares Ns1 with a preset standard respectively to determine how to specifically push the recommendation information; the central control unit is provided with a first recommended information quantity Nz1, wherein, 0 is more than Nz1,
if Ns1 is not more than Nz1, the central control unit judges that the quantity of the first recommendation information is less than a preset standard and compares Ns2 with (Nz 1-Ns 1) to judge whether any (Nz 1-Ns 1) second recommendation information is randomly selected for pushing, if Ns2 is not less than Nz1-Ns1, the central control unit judges that any (Nz 1-Ns 1) second recommendation information is randomly selected for pushing, and if Ns2 is less than Nz1-Ns1, the central control unit judges that the quantity of the first recommendation information and the quantity of the second recommendation information do not accord with the preset standard and transmits the judgment information to the manual service unit for manual adjustment;
if Nz1 is less than Ns1, the central control unit judges that the quantity of the first recommendation information meets a preset standard and randomly selects the first recommendation information to be pushed to the user side.
6. The AI-based content recommendation system according to claim 5, wherein the central control unit calculates a difference Δ N between the number N of required keywords and a first preset number Nz when the determination of how to push the recommendation information is completed and compares Δ N with a preset criterion to determine whether to adjust the push frequency F of the recommendation information, setting Δ N = N-Nz; the central control unit is provided with a first preset difference value delta N1, a second preset difference value delta N2, a standard pushing frequency Fz, a first adjusting coefficient alpha 1 and a second adjusting coefficient alpha 2, wherein, 0 is more than delta N1 and less than delta N2,0 is more than alpha 1 and less than alpha 2,0 is more than Fz,
if the delta N is less than or equal to the delta N1, the central control unit judges that the pushing frequency of the recommendation information does not need to be adjusted and set, and the pushing frequency F = Fz of the recommendation information;
if the delta N1 is less than the delta N and less than or equal to the delta N2, the central control unit judges that the pushing frequency F of the recommendation information is adjusted by using alpha 1, the adjusted pushing frequency is recorded as F ', and F' = Fz multiplied by alpha 1 is set;
if Δ N2 < [ delta ] N, the central control unit determines that α 2 is used to adjust the push frequency F of the recommendation information, and the adjusted push frequency is denoted as F ', and F' = Fz × α 2 is set.
7. The AI-based content recommendation system of claim 6 wherein the intelligent acquisition unit detects a user click-through rate Q when pushing recommendation information and transmits it to the central control unit, which compares Q with preset criteria to determine whether to adjust the display duration T of the recommendation information; the central control unit is provided with a first preset click rate Q1, a second preset click rate Q2, a third preset click rate Q3, a standard display time length Tz, a first time length adjusting coefficient beta 1 and a second time length adjusting coefficient beta 2, wherein Q1 is more than 0 and less than Q2 and less than Q3, beta 1 is more than 0 and less than beta 2, tz is more than 0,
if Q is less than or equal to Q1, the central control unit preliminarily judges that the user has no requirement on the recommendation information, sends a requirement questionnaire to the user side and transmits the judgment information to the manual service unit;
if Q1 is more than Q and less than or equal to Q2, the central control unit judges that the display time length of the recommendation information does not need to be adjusted and set, the display time length of the recommendation information is T, and T = Tz;
if Q2 is more than or equal to Q3, the central control unit judges that the display time length T of the recommendation information is adjusted by using beta 1, the adjusted display time length is marked as T ', and T' = T multiplied by beta 1 is set;
if Q3 is less than Q, the central control unit judges that the display time length T of the recommendation information is adjusted by using beta 2, the adjusted display time length is marked as T ', and T' = T multiplied by beta 2 is set;
and the central control unit detects the click rate of the user again when the adjustment of the display duration of the recommended information is finished so as to judge whether to continue adjusting the display duration of the recommended information.
8. The AI-based content recommendation system of claim 7, wherein the central control unit records the adjustment times M when re-detecting the user click rate to determine whether to continue adjusting the display duration of the recommendation information, and compares M and the adjusted display duration T' with preset criteria, respectively, to determine whether to allow the adjustment; the central control unit is provided with a maximum adjusting time Mmax and a maximum display duration Tmax, wherein Mmax is more than 0, tmax is more than 0,
if M is greater than or equal to Max or N is greater than or equal to Nmax, the central control unit judges that the adjustment is not allowed to be carried out and transmits the judgment information to the manual service unit;
if M < Mmax and N < Nmax, the central control unit determines that this adjustment is allowed.
9. The AI-based content recommendation system of claim 8, wherein the cloud platform is in telecommunication connection with the central control unit and the human service unit.
10. The AI-based content recommendation system of claim 9, wherein the manual service unit is configured with an audio/video display screen for receiving the determination information from the central control unit.
CN202211241342.3A 2022-10-11 2022-10-11 Content recommendation system based on AI Active CN115544362B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211241342.3A CN115544362B (en) 2022-10-11 2022-10-11 Content recommendation system based on AI

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211241342.3A CN115544362B (en) 2022-10-11 2022-10-11 Content recommendation system based on AI

Publications (2)

Publication Number Publication Date
CN115544362A true CN115544362A (en) 2022-12-30
CN115544362B CN115544362B (en) 2023-06-13

Family

ID=84733353

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211241342.3A Active CN115544362B (en) 2022-10-11 2022-10-11 Content recommendation system based on AI

Country Status (1)

Country Link
CN (1) CN115544362B (en)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105243143A (en) * 2015-10-14 2016-01-13 湖南大学 Recommendation method and system based on instant voice content detection
CN106709049A (en) * 2017-01-05 2017-05-24 胡开标 Phonetic character key word identifying and searching system
CN108874975A (en) * 2018-06-08 2018-11-23 Oppo（重庆）智能科技有限公司 Search for content recommendation method, device, terminal device and storage medium
CN110633413A (en) * 2019-08-26 2019-12-31 浙江大搜车软件技术有限公司 Label recommendation method and device, computer equipment and storage medium
CN111475714A (en) * 2020-03-17 2020-07-31 北京声智科技有限公司 Information recommendation method, device, equipment and medium
CN111506804A (en) * 2020-03-18 2020-08-07 上海大犀角信息科技有限公司 Client content recommendation system and method based on user terminal behaviors
CN111741369A (en) * 2020-07-10 2020-10-02 安徽芯智科技有限公司 Smart television set top box based on voice recognition
CN113641781A (en) * 2020-04-27 2021-11-12 中科国力(镇江)智能技术有限公司 Semantic sharing platform system
CN114896501A (en) * 2022-05-19 2022-08-12 文楚霞 Cloud computing and block chain based service recommendation method and cloud computing system
CN115080862A (en) * 2022-07-20 2022-09-20 广州市保伦电子有限公司 Conference recommendation system based on recommendation algorithm

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105243143A (en) * 2015-10-14 2016-01-13 湖南大学 Recommendation method and system based on instant voice content detection
CN106709049A (en) * 2017-01-05 2017-05-24 胡开标 Phonetic character key word identifying and searching system
CN108874975A (en) * 2018-06-08 2018-11-23 Oppo（重庆）智能科技有限公司 Search for content recommendation method, device, terminal device and storage medium
CN110633413A (en) * 2019-08-26 2019-12-31 浙江大搜车软件技术有限公司 Label recommendation method and device, computer equipment and storage medium
CN111475714A (en) * 2020-03-17 2020-07-31 北京声智科技有限公司 Information recommendation method, device, equipment and medium
CN111506804A (en) * 2020-03-18 2020-08-07 上海大犀角信息科技有限公司 Client content recommendation system and method based on user terminal behaviors
CN113641781A (en) * 2020-04-27 2021-11-12 中科国力(镇江)智能技术有限公司 Semantic sharing platform system
CN111741369A (en) * 2020-07-10 2020-10-02 安徽芯智科技有限公司 Smart television set top box based on voice recognition
CN114896501A (en) * 2022-05-19 2022-08-12 文楚霞 Cloud computing and block chain based service recommendation method and cloud computing system
CN115080862A (en) * 2022-07-20 2022-09-20 广州市保伦电子有限公司 Conference recommendation system based on recommendation algorithm

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
LANGCAI CAO; HONGWEI LI; RONGBIAO XIE; JINRONG ZHU;: "A Text Detection Algorithm for Image of Student Exercises Based on CTPN and Enhanced YOLOv3", IEEE ACCESS, pages 176924 - 176934 *
李方方;马昊宇;: "基于聊天机器人的智能导购系统", 福建电脑, no. 02, pages 31 - 32 *
胡辉: "基于知识图谱的个性化习题推荐研究", 中国优秀硕士学位论文全文数据库, pages 127 - 23 *

Also Published As

Publication number Publication date
CN115544362B (en) 2023-06-13

Similar Documents

Publication Publication Date Title
EP3996070A1 (en) Curriculum optimisation method, apparatus, and system
US9495385B2 (en) Mixed media reality recognition using multiple specialized indexes
US8676810B2 (en) Multiple index mixed media reality recognition using unequal priority indexes
US8965145B2 (en) Mixed media reality recognition using multiple specialized indexes
US20170289619A1 (en) Method for positioning video, terminal apparatus and cloud server
RU2571593C2 (en) Scanning of social network data
CN110072152B (en) Method and apparatus for identifying and presenting internet-accessible content
US20090070302A1 (en) Mixed Media Reality Recognition Using Multiple Specialized Indexes
KR20180122926A (en) Method for providing learning service and apparatus thereof
US20080294655A1 (en) Method and apparatus for generation of a user profile
US20090031336A1 (en) Group preference control system
KR101899921B1 (en) Online korean language learning system and the method
US20180262878A1 (en) Account pushing method and apparatus, and computer storage medium
CN111107444B (en) User comment generation method, electronic device and storage medium
US20040128691A1 (en) Video browsing system, distribution server and browse client
CN110889034A (en) Data analysis method and data analysis system
US8473586B2 (en) Method for determining aesthetic preferences to define a style guide and transforming a presentation based thereon
CN117992674A (en) Intelligent course recommendation method for online learning platform
JP5356286B2 (en) Question recommendation device, method and program
KR102007504B1 (en) System and method for providing customized contents information per user
CN110427620B (en) Service quality optimization management system based on community system
CN113626638A (en) Short video recommendation processing method and device, intelligent terminal and storage medium
CN115544362A (en) AI-based content recommendation system
JP2015005130A (en) Document posting device, method and program
CN113127720A (en) Hot word searching determination method and 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