CN116521986A - Test question recommending method, device, equipment and storage medium based on behavior sequence - Google Patents

Test question recommending method, device, equipment and storage medium based on behavior sequence Download PDF

Info

Publication number
CN116521986A
CN116521986A CN202310376666.6A CN202310376666A CN116521986A CN 116521986 A CN116521986 A CN 116521986A CN 202310376666 A CN202310376666 A CN 202310376666A CN 116521986 A CN116521986 A CN 116521986A
Authority
CN
China
Prior art keywords
behavior
similarity
sequence
user
state
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.)
Pending
Application number
CN202310376666.6A
Other languages
Chinese (zh)
Inventor
杨华
雷清锋
邓太勇
曾山
周康
沈浩
林荡
彭程武
张海峰
省深洋
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.)
Wuhan Polytechnic University
Original Assignee
Wuhan Polytechnic University
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 Wuhan Polytechnic University filed Critical Wuhan Polytechnic University
Priority to CN202310376666.6A priority Critical patent/CN116521986A/en
Publication of CN116521986A publication Critical patent/CN116521986A/en
Pending legal-status Critical Current

Links

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
    • G06F16/958—Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
    • 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/20—Education
    • G06Q50/205—Education administration or guidance
    • 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

Abstract

The invention belongs to the technical field of Internet application, and discloses a test question recommending method, device and equipment based on a behavior sequence and a storage medium. The method comprises the following steps: determining the similarity of the behavior sequences according to the behavior sequences of the target user and the candidate user; determining user similarity according to the behavior sequence similarity; according to the preset selection quantity and the user similarity, determining similar users of the target user from candidate users, and acquiring error questions of the similar users in a preset time range as candidate recommendation test questions; acquiring association weight data between historical wrong questions of a target user and candidate recommended test questions, and determining a test question recommendation strategy according to the association weight data and a behavior sequence of the target user; and determining target recommended test questions in the candidate recommended test questions according to the test question recommendation strategy, and recommending the target recommended test questions to the target user. Through the mode, the user characteristics are fully mined, personalized test question recommendation is provided, the characteristic change is dynamically updated, and the flexibility is improved.

Description

Test question recommending method, device, equipment and storage medium based on behavior sequence
Technical Field
The invention relates to the technical field of internet application, in particular to a test question recommending method, device and equipment based on a behavior sequence and a storage medium.
Background
Nowadays, online themes become social trends, and the application is very wide, for example: a series of competition activities lead to the market of a large number of on-line question training systems, so that users can conveniently start question making at any time and any place. However, these training systems have many limitations and limitations, the user portrait is not constant, the capability level of the user can be changed in the process of making questions, most online test question systems are not deep enough to mine the user portrait, and the recommendation of test questions cannot be flexibly adapted or has very strong hysteresis.
The foregoing is provided merely for the purpose of facilitating understanding of the technical solutions of the present invention and is not intended to represent an admission that the foregoing is prior art.
Disclosure of Invention
The invention mainly aims to provide a test question recommending method, device, equipment and storage medium based on a behavior sequence, and aims to solve the technical problems that most online test question systems in the prior art are not deep enough in mining of user images, and test question recommending cannot be flexibly adapted or has strong hysteresis.
In order to achieve the above purpose, the invention provides a test question recommending method based on a behavior sequence, which comprises the following steps:
acquiring a behavior sequence of a target user and a behavior sequence of a candidate user, and determining the similarity of the behavior sequences between the target user and the candidate user according to the behavior sequence of the target user and the behavior sequence of the candidate user;
determining the user similarity between the target user and the candidate user according to the behavior sequence similarity;
according to the preset selection quantity and the user similarity, determining similar users of the target user from the candidate users, and acquiring error questions of the similar users in a preset time range as candidate recommendation test questions;
acquiring association weight data between the historical wrong questions of the target user and the candidate recommended test questions, and determining the test question recommendation strategy according to the association weight data and the behavior sequence of the target user;
and determining a target recommended test question in the candidate recommended test questions according to the test question recommendation strategy, and recommending the target recommended test question to the target user.
Optionally, the behavior sequence includes a plurality of state string data arranged according to a preset time sequence, the state string data includes access data and operation data corresponding to the access data, and determining, according to the behavior sequence of the target user and the behavior sequence of the candidate user, a similarity of the behavior sequences between the target user and the candidate user includes:
Sequentially connecting state string data in the behavior sequence of the target user to obtain the behavior state sequence of the target user;
sequentially connecting state string data in the candidate user behavior sequence to obtain the candidate user behavior state sequence;
determining the behavior state sequence similarity, the behavior state transition similarity and the behavior state value similarity between the target user and the candidate user according to the behavior state sequence of the target user and the behavior state sequence of the candidate user;
and determining the behavior sequence similarity between the target user and the candidate user according to the behavior state sequence similarity, the behavior state transition similarity and the behavior state value similarity.
Optionally, the determining the behavior state sequence similarity, the behavior state transition similarity and the behavior state value similarity between the target user and the candidate user according to the behavior state sequence of the target user and the behavior state sequence of the candidate user includes:
determining a maximum public state subsequence according to the behavior state sequence of the target user and the behavior state sequence of the candidate user, and determining the length of the maximum public state subsequence corresponding to the maximum public state subsequence;
Determining the quantity of common state transitions, the quantity of state transitions and the quantity of state strings and the quantity of common state strings according to the behavior state sequence of the target user and the behavior state sequence of the candidate user;
determining the behavior state sequence similarity according to the maximum common state subsequence length, the correspondence between the state strings and the quantity and the behavior state sequence similarity, the maximum common state subsequence length and the state strings and the quantity;
determining the behavior state transition similarity according to the common state transition quantity, the correspondence between the state transition sum quantity and the behavior state transition similarity, the common state transition quantity and the state transition sum quantity;
and determining the behavior state value similarity according to the state strings and the quantity, the corresponding relation between the public state string quantity and the behavior state value similarity, the state strings and the quantity and the public state string quantity.
Optionally, the determining the behavior sequence similarity between the target user and the candidate user according to the behavior state sequence similarity, the behavior state transition similarity and the behavior state value similarity includes:
Acquiring the corresponding relation among the behavior state sequence similarity, the behavior state transition similarity, the behavior state value similarity and the behavior sequence similarity;
determining coefficient data according to the behavior state sequence similarity, the behavior state transition similarity, the correspondence between the behavior state value similarity and the behavior sequence similarity and a preset coefficient range;
and determining the behavior sequence similarity between the target user and the candidate user according to the behavior state sequence similarity, the behavior state transition similarity, the correspondence between the behavior state value similarity and the behavior sequence similarity and the coefficient data.
Optionally, the determining the user similarity between the target user and the candidate user according to the behavior sequence similarity includes:
acquiring the generation sequence time and the action time interval of the target user and the generation sequence time and the action time interval of the candidate user;
distributing corresponding time weight data for the behavior sequence of the target user according to the time of the generated sequence of the target user;
Distributing corresponding time weight data for the candidate state sequence according to the generation sequence time and the behavior time interval of the candidate user;
acquiring the corresponding relation among the time weight data, the behavior sequence similarity and the user similarity;
and determining the user similarity between the target user and the candidate user according to the time weight data, the corresponding relation between the behavior sequence similarity and the user similarity, the time weight data of the target user, the time weight data of the candidate user and the behavior sequence similarity.
Optionally, the acquiring the association weight data between the historical wrong questions of the target user and the candidate recommended test questions includes:
acquiring the outer product of the historical wrong questions of the target user and the candidate recommended test questions;
splicing the outer product with the history wrong questions and the candidate recommended test questions to obtain spliced data;
inputting the spliced data into a preset activation network to obtain the association weight data between the historical wrong questions and the candidate recommended test questions, wherein the preset activation network is composed of preset activation functions.
Optionally, the determining the test question recommendation policy according to the association weight data and the behavior sequence of the target user includes:
acquiring a corresponding relation between the association weight data and the behavior sequence of the target user;
and determining the test question recommendation strategy according to the corresponding relation between the association weight data and the behavior sequence of the target user, the association weight data and the behavior sequence of the target user.
In addition, in order to achieve the above object, the present invention also provides a test question recommending apparatus based on a behavior sequence, the test question recommending apparatus based on the behavior sequence includes:
the acquisition module is used for acquiring the behavior sequence of the target user and the behavior sequence of the candidate user, and determining the similarity of the behavior sequences between the target user and the candidate user according to the behavior sequence of the target user and the behavior sequence of the candidate user;
the acquisition module is further used for determining user similarity between the target user and the candidate user according to the behavior sequence similarity;
the candidate module is used for determining similar users of the target user from the candidate users according to the preset selection quantity and the user similarity, and acquiring the misquestions of the similar users in a preset time range as candidate recommendation test questions;
The recommendation module is used for acquiring association weight data between the historical wrong questions of the target user and the candidate recommendation test questions and determining the test question recommendation strategy according to the association weight data and the behavior sequence of the target user;
and the recommendation module is also used for determining a target recommendation test question in the candidate recommendation test questions according to the test question recommendation strategy and recommending the target recommendation test question to the target user.
In addition, in order to achieve the above object, the present invention also provides a test question recommending apparatus based on a behavior sequence, the test question recommending apparatus based on the behavior sequence comprising: the system comprises a memory, a processor and a test question recommending program based on a behavior sequence, wherein the test question recommending program is stored on the memory and can run on the processor, and the test question recommending program based on the behavior sequence is configured to realize the steps of the test question recommending method based on the behavior sequence.
In addition, in order to achieve the above object, the present invention also proposes a storage medium having stored thereon a test question recommending program based on a behavior sequence, which when executed by a processor, implements the steps of the test question recommending method based on a behavior sequence as described above.
According to the method, a behavior sequence of a target user and a behavior sequence of a candidate user are obtained, the similarity of the behavior sequence between the target user and the candidate user is determined according to the behavior sequence of the target user and the behavior sequence of the candidate user, the user similarity between the target user and the candidate user is determined according to the similarity of the behavior sequence, the similar user of the target user is determined in the candidate user according to the preset selection quantity and the user similarity, a wrong question of the similar user in a preset time range is obtained as a candidate recommendation test question, association weight data between the historical wrong question of the target user and the candidate recommendation test question is obtained, a test question recommendation strategy is determined according to the association weight data and the behavior sequence of the target user, a target recommendation test question is determined in the candidate recommendation test question according to the test question recommendation strategy, and the target recommendation test question is recommended to the target user. Compared with the fact that most online test question systems have insufficient mining of user images, the test questions have strong hysteresis, the method and the device can fully mine user characteristics, find similar users of the users according to the historical behavior sequences of the users, provide candidate recommendation test questions according to the historical wrong questions of the similar users, and meanwhile utilize an attention mechanism to examine the relevance between the candidate recommendation test questions and the historical wrong questions of the users, so that recommendation accuracy is improved.
Drawings
FIG. 1 is a schematic structural diagram of a test question recommending device based on a behavior sequence in a hardware running environment according to an embodiment of the present invention;
FIG. 2 is a flowchart of a first embodiment of a test question recommending method based on a behavior sequence according to the present invention;
FIG. 3 is a schematic overall flow chart of an embodiment of a test question recommending method based on a behavior sequence;
FIG. 4 is a flowchart of a second embodiment of the method for recommending test questions based on a behavior sequence according to the present invention;
FIG. 5 is a flowchart of a third embodiment of a test question recommending method based on a behavior sequence according to the present invention;
FIG. 6 is a schematic diagram of an activation unit of an embodiment of a test question recommending method based on a behavior sequence according to the present invention;
fig. 7 is a block diagram of a first embodiment of the test question recommending apparatus based on a behavior sequence.
The achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Referring to fig. 1, fig. 1 is a schematic diagram of a test question recommending apparatus based on a behavior sequence in a hardware running environment according to an embodiment of the present invention.
As shown in fig. 1, the test question recommending apparatus based on the behavior sequence may include: a processor 1001, such as a central processing unit (Central Processing Unit, CPU), a communication bus 1002, a user interface 1003, a network interface 1004, a memory 1005. Wherein the communication bus 1002 is used to enable connected communication between these components. The user interface 1003 may include a Display, an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may further include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a Wireless interface (e.g., a Wireless-Fidelity (Wi-Fi) interface). The Memory 1005 may be a high-speed random access Memory (Random Access Memory, RAM) Memory or a stable nonvolatile Memory (NVM), such as a disk Memory. The memory 1005 may also optionally be a storage device separate from the processor 1001 described above.
It will be appreciated by those skilled in the art that the structure shown in fig. 1 does not constitute a limitation of the behavior sequence based question recommending apparatus, and may include more or less components than illustrated, or may combine certain components, or may be a different arrangement of components.
As shown in fig. 1, an operating system, a network communication module, a user interface module, and a question recommending program based on a behavior sequence may be included in the memory 1005 as one storage medium.
In the test question recommending device based on the behavior sequence shown in fig. 1, the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 in the test question recommending device based on the behavior sequence can be arranged in the test question recommending device based on the behavior sequence, and the test question recommending device based on the behavior sequence calls the test question recommending program based on the behavior sequence stored in the memory 1005 through the processor 1001 and executes the test question recommending method based on the behavior sequence provided by the embodiment of the invention.
The embodiment of the invention provides a test question recommending method based on a behavior sequence, and referring to fig. 2, fig. 2 is a flow diagram of a first embodiment of the test question recommending method based on the behavior sequence.
In this embodiment, the test question recommendation method based on the behavior sequence includes the following steps:
step S10: and acquiring a behavior sequence of a target user and a behavior sequence of a candidate user, and determining the similarity of the behavior sequences between the target user and the candidate user according to the behavior sequence of the target user and the behavior sequence of the candidate user.
It should be noted that, the execution body of the embodiment is an online question making system, in which a question recommending program based on a behavior sequence is provided, and the online question making system generally includes a plurality of functional modules, for example: the system comprises a question making module, a wrong question review module, a checking and counting module and the like, and the recommendation of the test questions is realized through a test question recommendation program based on a behavior sequence.
It can be understood that the target user refers to a user who needs to recommend questions for the target user, the candidate users refer to other users except the target user, and may be all other users, or may be other users that may be similar users after preliminary screening, where the similar users refer to users with higher similarity to the target user.
It should be understood that the behavior sequence includes a plurality of state string data arranged according to a preset time sequence, where the state string data refers to a state string of a user, that is, behavior of the user, including access data and operation data corresponding to the access data, where the access data refers to an access situation of the user to a function module in the online question making system, the operation data refers to an operation situation of the user in the function module, and is generally expressed by (, yi) to represent an ith state string si, where zi represents the function module accessed by the user by the ith state string, yi represents the access of the user by the ith state string, and the operation data refers to the access situation of the user by the ith state string si And corresponding operation when i functional modules are adopted. The preset time sequence refers to a time sequence, the state strings of the users are generally arranged according to the time sequence, a finite set { (z 1, y 1), (z 2, y 2) … (zn, yn) } is obtained, and is not less than 2, namely a behavior sequence S, wherein the behavior sequence in the embodiment comprises a behavior sequence of the target user and a behavior sequence of the candidate user, and S can be used respectively i And S is j To represent. The behavior sequence similarity refers to the similarity of behavior sequences between two users, in this embodiment, the similarity of behavior sequences between a target user and a candidate user is the similarity of behavior sequences between the target user and the candidate user, and is used to determine the similarity between the target user and the candidate user, sim (S i ,S j ) To represent.
In the specific implementation, firstly, according to the access condition and the operation condition of a user to a functional module, a behavior sequence of a target user and a behavior sequence of a candidate user are respectively determined, the obtained behavior sequences are used for calculating the similarity of the behavior sequences between the target user and the candidate user.
Step S20: and determining the user similarity between the target user and the candidate user according to the behavior sequence similarity.
Further, the step S20 includes: and obtaining the corresponding relation among the time weight data, the behavior sequence similarity and the user similarity, and determining the user similarity among the target user and the candidate user according to the corresponding relation among the time weight data, the behavior sequence similarity and the user similarity, the time weight data of the target user, the time weight data of the candidate user and the behavior sequence similarity.
It should be noted that, the generated sequence time refers to a time from the first generation of the behavior sequence to the current generated user sequence, and the behavior time interval refers to an interval between behaviors in the behavior sequence, where the generated sequence time in this embodiment includes a generated sequence time of the target user and a generated sequence time of the candidate user, and the behavior time interval includes a behavior time interval of the target user and a behavior time interval of the candidate user.
It can be understood that, as the time varies, various feature information of the user is dynamically changed, so that when the similarity between the users is calculated by the similarity of the behavior sequences, a time weight coefficient is also required to be introduced, a larger weight coefficient is given to the recent behavior sequence, the reference value of the behavior sequence with a longer time interval is smaller, and a smaller weight coefficient is allocated, where the time weight data is the weight coefficient allocated to the behavior sequence in the embodiment, and includes the time weight data of the target user and the time weight data of the candidate user, and the time weight data can be allocated according to the following calculation relation:
wherein W (A, S) i ) Time weight data representing the user a,representing the time, L, from the first generation of user sequence by user A to the current generation of user sequence A Time interval representing the behavior of user a in the behavior sequence, S i Representing the behavior sequence of user A, a is a real number between 0 and 1 (excluding 0 and 1), according to L A And->The measurement was performed. In the present embodiment, the target user a is assigned time weight data W (a, S i ) Time weight data W (B, S) are assigned to candidate users B j )。
It should be understood that the user similarity refers to the similarity between two users, and in this embodiment, the user similarity is the similarity between the target user and the candidate user. The corresponding relation among the time weight data, the behavior sequence similarity and the user similarity refers to a calculation relational expression of the user similarity according to the corresponding relation among the time weight data, the behavior sequence similarity and the user similarity, and the calculation relational expression is as follows:
wherein sim (A, B) represents the user similarity between the target user A and the candidate user B, W (A, S) i ) Time weight data is assigned to target user a, W (B, S j ) Time weight data, sim (S) i ,S j ) Representing the similarity of the behavior sequences between the target user A and the candidate user B, S A And S is equal to B Respectively representing the behavior sequences of the target user A and the candidate user B. Substituting the obtained time weight data of the target user, the time weight data of the candidate user and the behavior sequence similarity into the calculation relation of the user similarity, and obtaining the user similarity between the target user and the candidate user.
In the specific implementation, a time weight coefficient is firstly introduced for the behavior sequence of the target user and the behavior sequence of the candidate user, and then the user similarity between the target user and the candidate user is calculated according to the similarity between the allocated time weight coefficient and the previously obtained behavior sequence.
Step S30: and determining similar users of the target user from the candidate users according to the preset selection quantity and the user similarity, and acquiring error questions of the similar users in a preset time range as candidate recommendation test questions.
It should be noted that, the similar users are candidate users with higher similarity to the target user, the preset selection number refers to the selection number of the similar users, usually 5-10 or other values may be set, and the selection number may be determined according to actual situations, which is not limited in this embodiment. The preset time range refers to a preset test question selection range, and is usually a recent time, for example: the time period of 10 days and one week can be adjusted according to actual requirements, and the embodiment is not limited to this. The candidate recommended test questions refer to error questions in the near term of similar users, are candidate test questions, have the possibility of being recommended, and can be selected from the candidate recommended test questions for recommendation.
In specific implementation, similar users of the target user are found, error questions in the recent times of the similar users are used as candidate recommendation error questions, and a candidate recommendation test question set is generated and used for determining follow-up recommendation test questions.
Step S40: and acquiring association weight data between the historical wrong questions of the target user and the candidate recommended test questions, and determining the test question recommendation strategy according to the association weight data and the behavior sequence of the target user.
It can be understood that the historical wrong question of the target user refers to a wrong question of the target user in the near future, the association weight data between the historical wrong question of the target user and the candidate recommended test question refers to the association degree between the historical wrong question and the candidate recommended test question, and the greater the association degree is, the more similar the candidate recommended test question is to the wrong question of the target user, and the target user has a high probability of not doing the question. The test question recommendation strategy refers to a recommendation mode of test questions, in this embodiment, the recommended test questions are determined according to the association degree between the historical wrong questions and the candidate recommended test questions, and generally, the recommendation with a large association degree is given to the target user.
It should be understood that, in this embodiment, an activation unit is provided, where the activation unit has two inputs, one is a fault question in the historical behavior sequence of the target user, and the other is a candidate recommended test question, and the associated weight data between the historical fault question of the target user and the candidate recommended test question may be output.
Step S50: and determining a target recommended test question in the candidate recommended test questions according to the test question recommendation strategy, and recommending the target recommended test question to the target user.
In a specific implementation, the target recommended test questions refer to test questions finally recommended to the target user, and in the embodiment, follow-up recommended test questions with a larger degree of relevance to the historical wrong questions of the target user are recommended to the target user, so that the user can be helped to quickly find out weak points to perform leak detection and repair.
As shown in the overall flow chart of FIG. 3, the similarity of the user behavior sequences is calculated according to the behavior sequences of the target users, so as to determine similar users of the target users, the recent wrong questions of the similar users are used as candidate recommended questions, the target users do missing questions and the candidate recommended questions are used as input together to obtain a score, the relevance between the target users and the candidate recommended questions is measured, and the higher the relevance between the selected questions and the historical wrong questions is, the higher the probability that the user does wrong questions is, so that the target users can be recommended.
In this embodiment, a behavior sequence of a target user and a behavior sequence of a candidate user are obtained, a behavior sequence similarity between the target user and the candidate user is determined according to the behavior sequence of the target user and the behavior sequence of the candidate user, a user similarity between the target user and the candidate user is determined according to the behavior sequence similarity, a similar user of the target user is determined in the candidate user according to a preset selection number and the user similarity, a wrong question of the similar user in a preset time range is obtained as a candidate recommendation test question, association weight data between the historical wrong question of the target user and the candidate recommendation test question is obtained, a test question recommendation strategy is determined according to the association weight data and the behavior sequence of the target user, a target recommendation test question is determined in the candidate recommendation test questions according to the test question recommendation strategy, and the target recommendation test question is recommended to the target user. According to the method and the device for recommending the user, the user characteristics can be fully mined, similar users of the users can be found according to the historical behavior sequences of the users, candidate recommendation test questions are provided according to the historical wrong questions of the similar users, meanwhile, the relevance between the candidate recommendation test questions and the historical wrong questions of the users is inspected by using an attention mechanism, the recommendation accuracy is improved, in addition, the recent historical behavior sequences of the users can be used for dynamically displaying the characteristic changes of the users, so that the user portraits are dynamically updated, obvious hysteresis does not occur, and flexible personalized recommendation is realized.
Referring to fig. 4, fig. 4 is a flow chart of a second embodiment of a test question recommending method based on a behavior sequence according to the present invention.
Based on the above embodiment, the step S10 includes:
step S101: and sequentially connecting the state string data in the behavior sequence of the target user to obtain the behavior state sequence of the target user, and sequentially connecting the state string data in the behavior sequence of the candidate user to obtain the behavior state sequence of the candidate user.
It should be noted that, the behavior state sequence is a string obtained by sequentially concatenating all state strings based on the behavior sequence, for example: the behavior sequence is { s1, s2, …, sn }, and when the behavior sequence is not less than 2, the behavior state sequence is s1s2 … sn.
In a specific implementation, the state strings in the behavior sequences of the target user and the candidate user are respectively connected with the behavior state sequence B in turn i And B j 。
Step S102: and determining the similarity of the behavior state sequence, the behavior state transition similarity and the behavior state value similarity between the target user and the candidate user according to the behavior state sequence of the target user and the behavior state sequence of the candidate user.
Further, the step S102 includes:
And determining a maximum public state subsequence according to the behavior state sequence of the target user and the behavior state sequence of the candidate user, and determining the length of the maximum public state subsequence corresponding to the maximum public state subsequence.
It can be understood that the similarity of the two behavior state sequences in the occurrence sequence, namely, the state sequence similarity, in order to calculate the state sequence similarity, the largest common state subsequence, namely, the largest common state subsequence in the behavior state sequence of the target user and the behavior state sequence of the candidate user, needs to be calculated first, and the length of the largest common state subsequence refers to the sequence length of the largest common state subsequence.
And determining the common state transition quantity, the state transition sum quantity, the state string sum quantity and the common state string quantity according to the behavior state sequence of the target user and the behavior state sequence of the candidate user.
It should be understood that the process of changing the state string of the user is referred to as state transition, for example, the transition from si to sj is the state transition of the user, and the state transition of the user can indirectly reflect the transition trend of the attention of the user, and the similarity of the user can be determined by performing a judgment analysis on the similar state transitions.
It should be noted that the number of common state transitions refers to a sum of numbers showing that two behavior state sequences have the same number of times in the occurrence of state transitions, the number of state transitions refers to a sum of numbers of state transitions of the two behavior state sequences, the number of common state strings refers to a number of common state strings included in the two behavior state sequences, and the number of state strings refers to a sum of numbers of state strings showing the two behavior state sequences.
And determining the behavior state sequence similarity according to the maximum common state subsequence length, the corresponding relation between the state strings and the quantity and the behavior state sequence similarity, the maximum common state subsequence length and the state strings and the quantity.
It is understood that the correspondence between the maximum common state subsequence length, the state strings and the number and the behavior state order similarity refers to a calculation relation of state order similarity, as follows:
in the formula, sim_seq (B i ,B j ) Representing behavior state sequence B of target user i Behavior state sequence B with candidate user j Similarity of state sequences between [ len [ comm (B) i ,B j )]Representation B i And B is connected with j Maximum common state subsequence length between com (B) i ,B j ) Representation B i And B is connected with j Maximum common state subsequence between, |b i ∪B j I represents B i And B is connected with j State strings and numbers in between. Substituting the maximum common state subsequence length, state strings and numbers into the calculation relation of the state sequence similarity, and calculating the behavior state sequence similarity.
And determining the behavior state transition similarity according to the common state transition quantity, the correspondence between the state transition sum quantity and the behavior state transition similarity, the common state transition quantity and the state transition sum quantity.
It should be understood that the state transition similarity refers to the similarity of state transitions in two behavior state sequences, and the correspondence between the number of common state transitions, the state transitions and the number and the behavior state transition similarity refers to a calculation relational expression of state transition similarity, as shown in the following:
in the formula, sim_trans (B i ,B j ) Representing behavior state sequence B of target user i Behavior state sequence B with candidate user j Similarity of state transitions between them, ω+δ representing B i And B is connected with j State transitions and amounts in between, Representation B i And B is connected with j The number of common state transitions between. Substituting the common state transition quantity and the state transition sum into the calculation relation of the state transition similarity to obtain the state transition similarity.
And determining the behavior state value similarity according to the state strings and the quantity, the corresponding relation between the public state string quantity and the behavior state value similarity, the state strings and the quantity and the public state string quantity.
It should be noted that, the behavior state value similarity refers to the similarity of state values in two behavior state sequences, and the correspondence between the state strings and the number, the number of common state strings, and the behavior state value similarity refers to a calculation relational expression of the behavior state value similarity, as follows:
in the formula, sim_value (B i ,B j ) Representing behavior state sequence B of target user i Behavior state sequence B with candidate user j Similarity of behavior state values between, |B i ∪B j I represents B i And B is connected with j State string and number between, |b i ∩B j I represents B i And B is connected with j Number of common state strings in between. Substituting the state strings and the number of the public state strings into the calculation relational expression of the behavior state value similarity to obtain the behavior state value similarity.
In specific implementation, the maximum common state subsequence, the number of common state transitions, the number of state transitions and states, the number of state strings and states, and the number of common state strings are first determined, so that the behavior state sequence similarity, the behavior state transition similarity, and the behavior state value similarity between the target user and the candidate user are calculated.
Step S103: and determining the behavior sequence similarity between the target user and the candidate user according to the behavior state sequence similarity, the behavior state transition similarity and the behavior state value similarity.
Further, the step S103 includes: and acquiring the corresponding relation among the behavior state sequence similarity, the behavior state transition similarity, the behavior state value similarity and the behavior sequence similarity, determining coefficient data according to the corresponding relation among the behavior state sequence similarity, the behavior state transition similarity, the behavior state value similarity and the behavior sequence similarity and a preset coefficient range, and determining the behavior sequence similarity between the target user and the candidate user according to the behavior state sequence similarity, the behavior state transition similarity, the corresponding relation among the behavior state value similarity and the behavior sequence similarity and the coefficient data.
It may be understood that the correspondence between the behavior state order similarity, the behavior state transition similarity, the behavior state value similarity and the behavior sequence similarity refers to a calculation relation of behavior sequence similarity, as follows:
sim(B i ，B j )＝α×sim_seq(B i ，B j )+β×sim_trans(B i ，B j )+γ×sim_value(B i ，B j )
wherein sim (B) i ，B j ) Representing behavior state sequence B of target user i Behavior state sequence B with candidate user j Similarity of behavioural sequences between, sim_seq (B i ，B j ) Representation B i And B is connected with j Similarity of state sequences between im_trans (B i ，B j ) Representation B i And B is connected with j Similarity of state transitions between each other, sim_value (B i ，B j ) Representation B i And B is connected with j The similarity of the behavior state values among the two is the coefficient data, the alpha, beta and gamma are the coefficient data, the preset coefficient range refers to the value constraint condition of the alpha, beta and gamma, in the embodiment, the sum of the alpha, beta and gamma is equal to 1 and is equal to or greater than 0, the optimal values of the alpha, beta and gamma are obtained by continuous correction in the process of calculating the similarity of the behavior sequence, and the calculated values are in direct proportion to the similarity among users. Substituting the obtained coefficient data, the behavior state sequence similarity, the behavior state transition similarity and the behavior state value similarity into the calculation relational expression of the behavior sequence similarity to obtain the behavior sequence similarity.
In this embodiment, the state string data in the behavior sequence of the target user is sequentially connected to obtain the behavior state sequence of the target user, the state string data in the behavior sequence of the candidate user is sequentially connected to obtain the behavior state sequence of the candidate user, the behavior state sequence similarity, the behavior state transition similarity and the behavior state value similarity between the target user and the candidate user are determined according to the behavior state sequence of the target user and the behavior state sequence of the candidate user, and the behavior sequence similarity between the target user and the candidate user is determined according to the behavior state sequence similarity, the behavior state transition similarity and the behavior state value similarity. According to the method and the device for recommending the user, the characteristics of the user can be fully mined, the similar users of the user can be found according to the historical behavior sequence of the user, the recommending accuracy is improved, in addition, the characteristic change of the user can be dynamically displayed by using the recent historical behavior sequence of the user, the portrait of the user is dynamically updated, obvious hysteresis cannot occur, and flexible personalized recommending is achieved.
Referring to fig. 5, fig. 5 is a schematic flow chart of a third embodiment of a test question recommending method based on a behavior sequence according to the present invention.
Based on the above embodiment, the step S40 includes:
step S401: and acquiring an outer product of the historical wrong questions of the target user and the candidate recommended test questions, and splicing the outer product with the historical wrong questions and the candidate recommended test questions to obtain spliced data.
It should be noted that, the spliced data refers to data obtained by splicing an outer product with a history wrong question and a candidate recommended test question.
It can be understood that the historical wrong questions and the candidate recommended questions in this embodiment may be regarded as related data corresponding to the historical wrong questions and the candidate recommended questions.
Step S402: and inputting the spliced data into a preset activation network to obtain the association weight data between the historical wrong questions and the candidate recommended test questions.
It should be understood that the preset activation network refers to a multi-layer network used in the activation unit, and the preset activation network is formed by preset activation functions, where the preset activation functions refer to preset activation functions, and the preset activation functions used in this embodiment are PRelu (Parametric Rectified Linear Unit)/Dice (Dice Activation Function) and Linear (Linear activation function) respectively, and may also be adjusted according to actual requirements, which is not limited in this embodiment.
Step S403: and acquiring a corresponding relation between the association weight data and the behavior sequence of the target user, and determining the test question recommendation strategy according to the corresponding relation between the association weight data and the behavior sequence of the target user, the association weight data and the behavior sequence of the target user.
It should be noted that, the correspondence between the association weight data and the behavior sequence of the target user refers to an expression of the candidate recommended test question, which may be obtained by multiplying the obtained association weight data by a historical error question, as follows:
in the formula, v U (A) Expression of target user U for candidate recommended test question A, { e 1 ,e 2 ,…,e H -representing the behavior sequence of the target user U, a (e) j ,v A ) And representing the associated weight corresponding to the first historical wrong question.
In the specific implementation, the association weight data and the behavior sequence of the target user are substituted into the expression of the candidate recommended test questions, so that the expression of the target user for the candidate recommended test questions can be obtained, and the test question recommendation strategy is determined.
As shown in FIG. 6, the activating unit has two inputs, one is a wrong question in the target user behavior sequence, the other is a candidate recommended question, the two are spliced with the outer product between them, then the two are input into a multi-layer network, a weight corresponding to the history wrong question is learned, and finally the obtained weight is multiplied by the history wrong, so that the expression of the target user for the candidate recommended question is obtained.
In specific implementation, by acquiring an outer product of a historical wrong question of a target user and a candidate recommended test question, splicing the outer product with the historical wrong question and the candidate recommended test question to obtain spliced data, inputting the spliced data into a preset activation network to obtain association weight data between the historical wrong question and the candidate recommended test question, acquiring a corresponding relation between the association weight data and a behavior sequence of the target user, and determining a test question recommendation strategy according to the corresponding relation between the association weight data and the behavior sequence of the target user, the association weight data and the behavior sequence of the target user. According to the method and the device for searching the candidate recommendation test questions, the attention mechanism is utilized to examine the relevance between the candidate recommendation test questions and the historical wrong questions of the user, if the relevance is strong, the candidate recommendation test questions are similar to the historical wrong questions of the target user, the user has high probability of not doing the questions, the user can be helped to find weak points quickly to detect leakage and repair defects, and the recommendation accuracy is improved.
In addition, the embodiment of the invention also provides a storage medium, wherein the storage medium is stored with a test question recommending program based on the behavior sequence, and the test question recommending program based on the behavior sequence realizes the steps of the test question recommending method based on the behavior sequence when being executed by a processor.
Referring to fig. 7, fig. 7 is a block diagram illustrating a first embodiment of a test question recommending apparatus based on a behavior sequence according to the present invention.
As shown in fig. 7, a test question recommending apparatus based on a behavior sequence according to an embodiment of the present invention includes:
the obtaining module 10 is configured to obtain a behavior sequence of a target user and a behavior sequence of a candidate user, and determine a similarity of the behavior sequences between the target user and the candidate user according to the behavior sequence of the target user and the behavior sequence of the candidate user.
The obtaining module 10 is further configured to determine a user similarity between the target user and the candidate user according to the behavior sequence similarity.
And the candidate module 20 is configured to determine similar users of the target user among the candidate users according to a preset selection number and the user similarity, and obtain misquestions of the similar users within a preset time range as candidate recommended test questions.
And the recommendation module 30 is used for acquiring the association weight data between the historical wrong questions of the target user and the candidate recommended test questions, and determining the test question recommendation strategy according to the association weight data and the behavior sequence of the target user.
The recommendation module 30 is further configured to determine a target recommended test question from the candidate recommended test questions according to the test question recommendation policy, and recommend the target recommended test question to the target user.
In this embodiment, a behavior sequence of a target user and a behavior sequence of a candidate user are obtained, a behavior sequence similarity between the target user and the candidate user is determined according to the behavior sequence of the target user and the behavior sequence of the candidate user, a user similarity between the target user and the candidate user is determined according to the behavior sequence similarity, a similar user of the target user is determined in the candidate user according to a preset selection number and the user similarity, a wrong question of the similar user in a preset time range is obtained as a candidate recommendation test question, association weight data between the historical wrong question of the target user and the candidate recommendation test question is obtained, a test question recommendation strategy is determined according to the association weight data and the behavior sequence of the target user, a target recommendation test question is determined in the candidate recommendation test questions according to the test question recommendation strategy, and the target recommendation test question is recommended to the target user. According to the method and the device for recommending the user, the user characteristics can be fully mined, similar users of the users can be found according to the historical behavior sequences of the users, candidate recommendation test questions are provided according to the historical wrong questions of the similar users, meanwhile, the relevance between the candidate recommendation test questions and the historical wrong questions of the users is inspected by using an attention mechanism, the recommendation accuracy is improved, in addition, the recent historical behavior sequences of the users can be used for dynamically displaying the characteristic changes of the users, so that the user portraits are dynamically updated, obvious hysteresis does not occur, and flexible personalized recommendation is realized.
In an embodiment, the behavior sequence includes a plurality of state string data arranged according to a preset time sequence, where the state string data includes access data and operation data corresponding to the access data, and the obtaining module 10 is further configured to sequentially connect the state string data in the behavior sequence of the target user to obtain a behavior state sequence of the target user;
sequentially connecting state string data in the candidate user behavior sequence to obtain the candidate user behavior state sequence;
determining the behavior state sequence similarity, the behavior state transition similarity and the behavior state value similarity between the target user and the candidate user according to the behavior state sequence of the target user and the behavior state sequence of the candidate user;
and determining the behavior sequence similarity between the target user and the candidate user according to the behavior state sequence similarity, the behavior state transition similarity and the behavior state value similarity.
In an embodiment, the obtaining module 10 is further configured to determine a maximum common state subsequence according to the behavior state sequence of the target user and the behavior state sequence of the candidate user, and determine a maximum common state subsequence length corresponding to the maximum common state subsequence;
Determining the quantity of common state transitions, the quantity of state transitions and the quantity of state strings and the quantity of common state strings according to the behavior state sequence of the target user and the behavior state sequence of the candidate user;
determining the behavior state sequence similarity according to the maximum common state subsequence length, the correspondence between the state strings and the quantity and the behavior state sequence similarity, the maximum common state subsequence length and the state strings and the quantity;
determining the behavior state transition similarity according to the common state transition quantity, the correspondence between the state transition sum quantity and the behavior state transition similarity, the common state transition quantity and the state transition sum quantity;
and determining the behavior state value similarity according to the state strings and the quantity, the corresponding relation between the public state string quantity and the behavior state value similarity, the state strings and the quantity and the public state string quantity.
In an embodiment, the obtaining module 10 is further configured to obtain a correspondence between the behavior state sequence similarity, the behavior state transition similarity, the behavior state value similarity, and the behavior sequence similarity;
Determining coefficient data according to the behavior state sequence similarity, the behavior state transition similarity, the correspondence between the behavior state value similarity and the behavior sequence similarity and a preset coefficient range;
and determining the behavior sequence similarity between the target user and the candidate user according to the behavior state sequence similarity, the behavior state transition similarity, the correspondence between the behavior state value similarity and the behavior sequence similarity and the coefficient data.
In an embodiment, the obtaining module 10 is further configured to obtain a generated sequence time and a behavior time interval of the target user and a generated sequence time and a behavior time interval of the candidate user;
distributing corresponding time weight data for the behavior sequence of the target user according to the time of the generated sequence of the target user;
distributing corresponding time weight data for the candidate state sequence according to the generation sequence time and the behavior time interval of the candidate user;
acquiring the corresponding relation among the time weight data, the behavior sequence similarity and the user similarity;
And determining the user similarity between the target user and the candidate user according to the time weight data, the corresponding relation between the behavior sequence similarity and the user similarity, the time weight data of the target user, the time weight data of the candidate user and the behavior sequence similarity.
In an embodiment, the recommendation module 30 is further configured to obtain an outer product of the historical error questions of the target user and the candidate recommended test questions;
splicing the outer product with the history wrong questions and the candidate recommended test questions to obtain spliced data;
inputting the spliced data into a preset activation network to obtain the association weight data between the historical wrong questions and the candidate recommended test questions, wherein the preset activation network is composed of preset activation functions.
In an embodiment, the recommendation module 30 is further configured to obtain a correspondence between the association weight data and the behavior sequence of the target user;
and determining the test question recommendation strategy according to the corresponding relation between the association weight data and the behavior sequence of the target user, the association weight data and the behavior sequence of the target user.
It should be understood that the foregoing is illustrative only and is not limiting, and that in specific applications, those skilled in the art may set the invention as desired, and the invention is not limited thereto.
It should be noted that the above-described working procedure is merely illustrative, and does not limit the scope of the present invention, and in practical application, a person skilled in the art may select part or all of them according to actual needs to achieve the purpose of the embodiment, which is not limited herein.
In addition, technical details which are not described in detail in the present embodiment can be referred to the test question recommendation method based on the behavior sequence provided in any embodiment of the present invention, and are not described herein again.
Furthermore, it should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The foregoing embodiment numbers of the present invention are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. Read Only Memory)/RAM, magnetic disk, optical disk) and including several instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method according to the embodiments of the present invention.
The foregoing description is only of the preferred embodiments of the present invention, and is not intended to limit the scope of the invention, but rather is intended to cover any equivalents of the structures or equivalent processes disclosed herein or in the alternative, which may be employed directly or indirectly in other related arts.

Claims (10)

1. The test question recommending method based on the behavior sequence is characterized by comprising the following steps of:
acquiring a behavior sequence of a target user and a behavior sequence of a candidate user, and determining the similarity of the behavior sequences between the target user and the candidate user according to the behavior sequence of the target user and the behavior sequence of the candidate user;
determining the user similarity between the target user and the candidate user according to the behavior sequence similarity;
according to the preset selection quantity and the user similarity, determining similar users of the target user from the candidate users, and acquiring error questions of the similar users in a preset time range as candidate recommendation test questions;
acquiring association weight data between the historical wrong questions of the target user and the candidate recommended test questions, and determining the test question recommendation strategy according to the association weight data and the behavior sequence of the target user;
and determining a target recommended test question in the candidate recommended test questions according to the test question recommendation strategy, and recommending the target recommended test question to the target user.
2. The method of claim 1, wherein the behavior sequence includes a plurality of state string data arranged according to a preset time sequence, the state string data includes access data and operation data corresponding to the access data, and determining the similarity of the behavior sequences between the target user and the candidate user according to the behavior sequence of the target user and the behavior sequence of the candidate user includes:
Sequentially connecting state string data in the behavior sequence of the target user to obtain the behavior state sequence of the target user;
sequentially connecting state string data in the candidate user behavior sequence to obtain the candidate user behavior state sequence;
determining the behavior state sequence similarity, the behavior state transition similarity and the behavior state value similarity between the target user and the candidate user according to the behavior state sequence of the target user and the behavior state sequence of the candidate user;
and determining the behavior sequence similarity between the target user and the candidate user according to the behavior state sequence similarity, the behavior state transition similarity and the behavior state value similarity.
3. The method of claim 2, wherein the determining the behavioral state order similarity, the behavioral state transition similarity, and the behavioral state value similarity between the target user and the candidate user from the behavioral state sequence of the target user and the behavioral state sequence of the candidate user comprises:
determining a maximum public state subsequence according to the behavior state sequence of the target user and the behavior state sequence of the candidate user, and determining the length of the maximum public state subsequence corresponding to the maximum public state subsequence;
Determining the quantity of common state transitions, the quantity of state transitions and the quantity of state strings and the quantity of common state strings according to the behavior state sequence of the target user and the behavior state sequence of the candidate user;
determining the behavior state sequence similarity according to the maximum common state subsequence length, the correspondence between the state strings and the quantity and the behavior state sequence similarity, the maximum common state subsequence length and the state strings and the quantity;
determining the behavior state transition similarity according to the common state transition quantity, the correspondence between the state transition sum quantity and the behavior state transition similarity, the common state transition quantity and the state transition sum quantity;
and determining the behavior state value similarity according to the state strings and the quantity, the corresponding relation between the public state string quantity and the behavior state value similarity, the state strings and the quantity and the public state string quantity.
4. The method of claim 2, wherein the determining the behavioral sequence similarity between the target user and the candidate user based on the behavioral state order similarity, the behavioral state transition similarity, and the behavioral state value similarity comprises:
Acquiring the corresponding relation among the behavior state sequence similarity, the behavior state transition similarity, the behavior state value similarity and the behavior sequence similarity;
determining coefficient data according to the behavior state sequence similarity, the behavior state transition similarity, the correspondence between the behavior state value similarity and the behavior sequence similarity and a preset coefficient range;
and determining the behavior sequence similarity between the target user and the candidate user according to the behavior state sequence similarity, the behavior state transition similarity, the correspondence between the behavior state value similarity and the behavior sequence similarity and the coefficient data.
5. The method of claim 1, wherein the determining the user similarity between the target user and the candidate user based on the behavior sequence similarity comprises:
acquiring the generation sequence time and the action time interval of the target user and the generation sequence time and the action time interval of the candidate user;
distributing corresponding time weight data for the behavior sequence of the target user according to the time of the generated sequence of the target user;
Distributing corresponding time weight data for the candidate state sequence according to the generation sequence time and the behavior time interval of the candidate user;
acquiring the corresponding relation among the time weight data, the behavior sequence similarity and the user similarity;
and determining the user similarity between the target user and the candidate user according to the time weight data, the corresponding relation between the behavior sequence similarity and the user similarity, the time weight data of the target user, the time weight data of the candidate user and the behavior sequence similarity.
6. The method of claim 1, wherein the obtaining of the association weight data between the historical mistopic of the target user and the candidate recommended test topic comprises:
acquiring the outer product of the historical wrong questions of the target user and the candidate recommended test questions;
splicing the outer product with the history wrong questions and the candidate recommended test questions to obtain spliced data;
inputting the spliced data into a preset activation network to obtain the association weight data between the historical wrong questions and the candidate recommended test questions, wherein the preset activation network is composed of preset activation functions.
7. The method of claim 6, wherein the determining the topic recommendation policy based on the association weight data and the behavior sequence of the target user comprises:
acquiring a corresponding relation between the association weight data and the behavior sequence of the target user;
and determining the test question recommendation strategy according to the corresponding relation between the association weight data and the behavior sequence of the target user, the association weight data and the behavior sequence of the target user.
8. The test question recommending device based on the behavior sequence is characterized by comprising:
the acquisition module is used for acquiring the behavior sequence of the target user and the behavior sequence of the candidate user, and determining the similarity of the behavior sequences between the target user and the candidate user according to the behavior sequence of the target user and the behavior sequence of the candidate user;
the acquisition module is further used for determining user similarity between the target user and the candidate user according to the behavior sequence similarity;
the candidate module is used for determining similar users of the target user from the candidate users according to the preset selection quantity and the user similarity, and acquiring the misquestions of the similar users in a preset time range as candidate recommendation test questions;
The recommendation module is used for acquiring association weight data between the historical wrong questions of the target user and the candidate recommendation test questions and determining the test question recommendation strategy according to the association weight data and the behavior sequence of the target user;
and the recommendation module is also used for determining a target recommendation test question in the candidate recommendation test questions according to the test question recommendation strategy and recommending the target recommendation test question to the target user.
9. A test question recommending apparatus based on a behavior sequence, the apparatus comprising: a memory, a processor and a behavior sequence based question recommendation program stored on the memory and executable on the processor, the behavior sequence based question recommendation program being configured to implement the steps of the behavior sequence based question recommendation method of any one of claims 1 to 7.
10. A storage medium, wherein a test question recommending program based on a behavior sequence is stored on the storage medium, and the test question recommending program based on the behavior sequence realizes the steps of the test question recommending method based on the behavior sequence according to any one of claims 1 to 7 when being executed by a processor.
CN202310376666.6A 2023-04-10 2023-04-10 Test question recommending method, device, equipment and storage medium based on behavior sequence Pending CN116521986A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310376666.6A CN116521986A (en) 2023-04-10 2023-04-10 Test question recommending method, device, equipment and storage medium based on behavior sequence

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310376666.6A CN116521986A (en) 2023-04-10 2023-04-10 Test question recommending method, device, equipment and storage medium based on behavior sequence

Publications (1)

Publication Number Publication Date
CN116521986A true CN116521986A (en) 2023-08-01

Family

ID=87402173

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310376666.6A Pending CN116521986A (en) 2023-04-10 2023-04-10 Test question recommending method, device, equipment and storage medium based on behavior sequence

Country Status (1)

Country Link
CN (1) CN116521986A (en)

Similar Documents

Publication Publication Date Title
CN110322169B (en) Task issuing method and device
CN112906384B (en) BERT model-based data processing method, BERT model-based data processing device, BERT model-based data processing equipment and readable storage medium
US20130097264A1 (en) System and method for optimizing response handling time and customer satisfaction scores
US20020152114A1 (en) System and method for updating an intranet portal
CN115203496A (en) Project intelligent prediction and evaluation method and system based on big data and readable storage medium
US20060230352A1 (en) Contextual error messaging
CN113742069A (en) Capacity prediction method and device based on artificial intelligence and storage medium
CN116521986A (en) Test question recommending method, device, equipment and storage medium based on behavior sequence
CN111062626A (en) Capability level evaluation method, device, equipment and storage medium
US20220327450A1 (en) Method for increasing or decreasing number of workers and inspectors in crowdsourcing-based project for creating artificial intelligence learning data
CN110825960A (en) Learning content recommendation method and device
CN113159537B (en) Assessment method and device for new technical project of power grid and computer equipment
CN114048137A (en) Application testing method and device and electronic equipment
JP6961148B1 (en) Information processing system and information processing method
CN114693116A (en) Method and device for detecting code review validity and electronic equipment
CN113297371A (en) Method, device and equipment for generating recommended question library and storage medium
CN114492905A (en) Customer appeal rate prediction method and device based on multi-model fusion and computer equipment
US20230360552A1 (en) Method, apparatus, device and medium for information updating
CN116664016B (en) Screening method and device of ESG (electronic service guide) sub-topics, electronic equipment and readable storage medium
CN113448872B (en) Test execution method, device, equipment and storage medium
CN114637921B (en) Item recommendation method, device and equipment based on modeling accidental uncertainty
CN114548523B (en) User viewing information prediction method, device, equipment and storage medium
JP6833235B1 (en) Information processing system and information processing method
CN116483983B (en) Method and related equipment for generating emotion change quantity of virtual character
JP7465622B2 (en) How to match employers with job seekers

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