CN115936904A - Finance and tax knowledge pushing method based on user behaviors - Google Patents
Finance and tax knowledge pushing method based on user behaviors Download PDFInfo
- Publication number
- CN115936904A CN115936904A CN202211302411.7A CN202211302411A CN115936904A CN 115936904 A CN115936904 A CN 115936904A CN 202211302411 A CN202211302411 A CN 202211302411A CN 115936904 A CN115936904 A CN 115936904A
- Authority
- CN
- China
- Prior art keywords
- user
- knowledge
- similarity
- users
- item set
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 26
- 230000006399 behavior Effects 0.000 title claims abstract description 12
- 238000004364 calculation method Methods 0.000 claims abstract description 12
- 230000029305 taxis Effects 0.000 claims description 3
- 238000004883 computer application Methods 0.000 abstract description 2
- 238000007726 management method Methods 0.000 description 9
- 230000008901 benefit Effects 0.000 description 5
- 230000008859 change Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 241000208125 Nicotiana Species 0.000 description 1
- 235000002637 Nicotiana tabacum Nutrition 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000007639 printing Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to the technical field of computer application, and particularly relates to a finance and tax knowledge pushing method based on user behaviors. According to the method, the similarity and the relevance calculation are formed by utilizing the relationship between the users and the relationship between the fiscal and tax knowledge, and the pushed knowledge is more suitable for the fiscal and tax learning condition by selecting the knowledge with higher similarity and relevance to push, so that the precise learning requirement of fiscal and tax staff is met.
Description
Technical Field
The invention belongs to the technical field of computer application, and particularly relates to a finance and tax knowledge pushing method based on user behaviors.
Background
Fiscal tax is the abbreviation for finance and tax. Generally including various specialties such as finance, tax, etc. The enterprise finance and tax management is a very important content of enterprise management and management. As market economics become more standard and sophisticated, tax management plays an increasingly significant role in enterprise competition. The financial and tax management of enterprises is to start from the method and measures of tax management emphatically, the learning of tax law knowledge is strengthened, so that the tax paying consciousness is further improved, and on the premise of mastering the theoretical knowledge of the tax law, the aims of reducing the enterprise operating cost, improving the financial management level and finally improving the enterprise competitiveness are fulfilled by some reasonable and legal tax avoiding methods such as tax discount, tax planning and the like.
Under the market economic condition, the market competition is fierce, and enterprises need to make good work on production, operation and financial management for the survival and development of the enterprises, so as to avoid risks and obtain the best economic benefit. The aerospace information ERP product carries out information transmission and element continuation with a national tax system, carries out business processing on various taxes (value-added tax, income tax, business tax, consumption tax, customs tax, export tax refund and the like) related to the enterprise operation process, can accurately account various tax due to declare tax payment, improves the working efficiency of tax staff, can also evaluate the enterprise accounts, tickets, operation, accounting and tax payment conditions, better helps the enterprise to correctly execute national tax policies, carries out integral operation planning and tax payment risk prevention, contributes to decision and contribution planning of enterprise management, and lays a solid foundation for benefit creation.
At present, the laws and regulations related to finance and tax in China are mainly divided into laws and relevant normative documents which are set by the general society of people and the general commission thereof, and administrative laws and relevant normative documents which are set by the state department, wherein the administrative laws and relevant normative documents which are set by the state department are divided into the following five basic types: the method comprises the following steps of firstly, basic system of tax (various tax types such as current value-added tax, consumption tax, business tax, vehicle purchase tax, land value-added tax, real estate tax, town land use tax, farmland occupation tax, tax liability tax, resource tax, ship tonnage tax, printing tax, city maintenance construction tax, tobacco leaf tax, customs tax and the like); second, law enforcement regulations or enforcement fine rules (personal income tax Law, enterprise income tax Law, vehicle and Ship tax Law, tax collection and administration Law, made by the national Community and its Commission, corresponding enforcement regulations or enforcement fine rules); thirdly, a non-basic system of tax revenue; fourthly, explanation is made on the specific regulations of tax administration laws and regulations; and fifthly, the normative documents approved by the state department and released by the department of the state department are viewed as the state department documents.
Due to market changes and various factors (such as policy changes caused by epidemic situations or other conditions), some tax policies are adjusted, and the financial staff or learning staff of the finance and tax as enterprises do not need to understand all laws and regulations related to finance and tax, and only need to learn according to the related properties of the enterprises and some policies of local governments, while the existing website or APP for learning finance and tax cannot accurately push according to the related needs of users, and some users hardly know what contents they need to learn, so how to accurately push is the difficulty of finance and tax learning at present.
Disclosure of Invention
Aiming at the technical problems in the finance and tax knowledge learning process, the invention provides the finance and tax knowledge pushing method based on the user behavior, which has the advantages of reasonable design, simple method and convenient operation and can realize accurate pushing according to the needs of the user.
In order to achieve the purpose, the invention adopts the technical scheme that the financial and tax knowledge pushing method based on the user behavior comprises the following steps:
a. firstly, obtaining basic attributes and reading records of a user, wherein the basic attributes comprise the industry of a unit where the user is located, the property of the unit where the user is located and the location of the unit where the user is located, and if the user is a person without a unit, the basic attributes are the expected work place of the user; the reading records are the learning duration of the user and the number of chapters and sections learned by the user;
b. b, searching similar users of the user according to the basic attributes and the reading records of the user obtained in the step a, and selecting users with similarity TOP-N;
c. then, taking each piece of knowledge related to the fiscal taxes as an item, forming an item set by all the knowledge learned by each user, constructing a firm database by all the knowledge learned by all the users, and calculating the association degree between each piece of knowledge and other knowledge;
d. according to the association degree between the knowledge learned by the TOP-N user selected in the step b and the knowledge obtained in the step c, mixed recommendation is realized for the user;
in the step b, the method for searching the similar user of the user comprises the following steps:
b1, firstly, determining similarity among basic attributes of users, wherein the similarity among the basic attributes comprises the similarity of industries of units where the users are located, the similarity of properties of the units where the users are located, and the similarity of places where the units where the users are located or places where the users expect to work, and the similarity of the industries of the units where the users are located is as follows:
that is, if the industries of the user m and the user n are consistent, S is i (m, n) is 1, if the industries of the user m and the user n are not consistent, S is i (m, n) is 0;
the similarity of the properties of the unit where the user is located is as follows:
that is, if the properties of the units where user m and user n are located are the same, S is n (m, n) is 1, if the industries of the user m and the user n are not consistent, S is n (m, n) is 0;
similarity of unit location where the user is located or expected working place of the user:
that is, if the locations of the units where the user m and the user n are located are the same or the expected work places are the same, S h (m, n) is 1, if the industries of the user m and the user n are not consistent, S is h (m, n) is 0;
b2, determining the similarity among the learning contents of the users according to the reading records of the users, wherein the calculation formula of the similarity among the learning contents of the users is as follows:
wherein i is learning content, i = (1, 2,3.. K), LT m,i And LT n,i Learning coefficients for the learning content i for the user m and the user n respectively,
b3, then according to the formula:
SIM comprehensive (m,n)＝λ(sim l (m,n))+(1-λ)(αS i (m,n)+βS n (m,n)+δS h (m,n))
and calculating the comprehensive similarity of the user m and the user n, wherein lambda is a weighting system of the comprehensive similarity between the learning contents of the user m and the user n between the user m and the user n, and alpha, beta and delta are the similarity of the industry of the unit where the user is located, the similarity of the property of the unit where the user is located and the weighting coefficient of the similarity of the place where the user is located or the place where the user expects to work respectively.
Preferably, in the step c, the method for calculating the association degree between each piece of knowledge and other pieces of knowledge includes:
c1, scanning a transaction database to obtain the support degree count of each knowledge, and deleting the knowledge with the support degree count smaller than the minimum support degree to obtain a frequent 1 item set;
c2, generating a candidate 2 item set according to the obtained frequent 1 item set in a self-connection mode, then deleting the knowledge that the number of support degrees in the candidate 2 item set is smaller than the minimum support degree to obtain a frequent 2 item set, and repeating the steps in a layer-by-layer iteration mode to generate a candidate K item set according to the frequent K-1 item set in a self-connection mode;
c3, calculating the confidence coefficient between each knowledge and the associated knowledge in the frequent item set for the frequent item set obtained in the step, and deleting the knowledge with the confidence coefficient smaller than the minimum confidence coefficient to obtain strong associated knowledge;
and c4, determining the association degree of the knowledge according to the confidence degree of the knowledge.
Preferably, in the step c4, the calculation formula of the association degree between the knowledge is as follows:
wherein con (x, y) is the proportion of the x item set containing knowledge and the y item set containing knowledge, and con (y, x) is the proportion of the y item set containing knowledge and the x item set containing knowledge.
Preferably, in the step c, the knowledge having the smallest number of support degrees is deleted while the knowledge having the smallest number of support degrees is deleted in the candidate k item sets.
Compared with the prior art, the invention has the advantages and positive effects that,
1. the invention provides a finance and tax knowledge pushing method based on user behaviors, which utilizes the relationship between users and the relationship between finance and tax knowledge to form similarity and association calculation, and selects knowledge with higher similarity and association to push, so that the pushed knowledge is more in line with the finance and tax learning condition, and the accurate learning requirement of finance and tax staff is met.
Detailed Description
In order that the above objects, features and advantages of the present invention can be more clearly understood, the present invention will be further described with reference to the following examples. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflict.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those specifically described herein, and thus the present invention is not limited to the specific embodiments of the present disclosure.
Embodiment 1, this embodiment provides a finance and tax knowledge pushing method based on user behaviors
Firstly, basic attributes and reading records of a user are obtained, data which needs to be filled in when the user registers the basic attributes of the user, and the user needs to accurately fill in the basic attributes to ensure the pushing accuracy.
The basic attributes comprise the industry of the unit where the user is located, the property of the unit where the user is located and the location of the unit where the user is located, and the industry of the unit where the user is located is determined according to 'national economy industry classification' (GB/T4754-2017) 2019 modified version, because the applicable financial and tax policies of different industries are different, and therefore, the recommendation needs to be carried out according to the located industry.
The nature of the unit where the user is located is mainly to determine the nature of the enterprise, because different types of enterprises adapt different policies, such as small-scale enterprises and regular enterprises, individual payers, joint venture enterprises and sole proprietor enterprises, and different types of enterprises adapt different policies.
In addition, local governments may also set policies according to their own needs, so that the policies are adjusted according to locations, in this embodiment, locations are filled in by taking areas as units, and particularly, the policy preference is different from that of other areas when the existing high and new areas or economic development areas are set up, so that the accuracy of the policies can be ensured by taking areas as units.
If the user is a person without a unit, the basic attribute is a place where the user desires to work, and a general person who does not work may record only the place where the user desires to work, and similarly, the basic attribute is also in units of areas, because the basic attribute is only a part of the weight, and therefore, only the place where the user desires to work may be recorded.
The reading records are the learning duration of the user and the number of chapters learned by the user, because the training course of the law or policy, which is generally pushed by finance and tax learning, includes a plurality of chapters, and some policies may be about 1 course or one article, for this reason, the pushing of the article takes the duration of most students as the total learning duration.
Then, according to the basic attributes and reading records of the user obtained in step a, searching for a similar user of the user, and selecting a user with a similarity TOP-N, wherein the determination of the first few users can be determined according to a test condition, in this embodiment, selecting a user with a similarity TOP 10.
The method for searching the similar user of the user comprises the following steps:
firstly, determining similarity among basic attributes of users, wherein the similarity among the basic attributes comprises the similarity of industries of units where the users are located, the similarity of properties of the units where the users are located, and the similarity of places where the users are located or places where the users expect to work, and the similarity of the industries of the units where the users are located is as follows:
that is, if the industries of the user m and the user n are consistent, S is i (m, n) is 1, if the industries of the user m and the user n are not consistent, S is i And (m and n) are 0, and the difference of financial and tax policies of different industries is just introduced, so that the similarity of the industries needs to be considered.
The similarity of the properties of the unit where the user is located is as follows:
that is, if the properties of the unit where the user m and the user n are located are consistent, S n (m, n) is 1, if the industries of the user m and the user n are not consistent, S is n (m, n) is 0.
Similarity of unit location where the user is located or expected working place of the user:
that is, if the locations of the units where the user m and the user n are located or the expected work places are consistent, S h (m, n) is 1, if the industries of the user m and the user n are inconsistent, S h (m, n) is 0.
In this way, the similarity between the user basic attributes is determined.
And then determining the similarity between the learning contents of the users according to the reading records of the users, wherein the calculation formula of the similarity between the learning contents of the users is as follows:
wherein i is learning content, i = (1, 2,3.. K),LT m,i and LT n,i The learning coefficients of the user m and the user n for the learning content i are determined based on a law or an article, and in a simple manner, the "vehicle and vessel tax law" includes thirteen laws, the videos of the thirteen laws are generally a set of learning videos, and for this reason, the learning contents are classified as the learning contents of the law, and the interpretation videos or the article of the article are also classified as the learning contents individually.
The companies for the calculation of the learning coefficients are as follows:
in this way, mainly considering whether the user learns one learning content or is in an open state only, after learning for a period, the learning content is found to be possibly irrelevant to the work thereof, and the learning content is abandoned, so that the maximum value of the learning coefficient is 1, and if the learning duration of the user x for the learning content i exceeds the total learning duration of the learning content i, the learning duration of the user x for the learning content i is equal to the total learning duration of the learning content i.
Finally, then according to the formula:
SIM comprehensive (m,n)＝λ(sim l (m,n))+(1-λ)(αS i (m,n)+βS n (m,n)+δS h (m,n))
and calculating the comprehensive similarity of the user m and the user n, wherein lambda is a weighting system of the comprehensive similarity between the learning contents of the user m and the user n between the user m and the user n, and alpha, beta and delta are the similarity of the industry of the unit where the user is located, the similarity of the property of the unit where the user is located and the weighting coefficient of the similarity of the place where the user is located or the place where the user expects to work respectively. Thus, the similarity between users is calculated, the user most similar to the user is found, and then the TOP 10 is selected according to the selected TOP-N.
The knowledge that the user with the highest similarity may learn is very extensive, and then the correlation strength between the knowledge and the knowledge needs to be calculated to make a recommendation more accurate.
Therefore, each knowledge related to the finance and tax is taken as one item, in the embodiment, each knowledge refers to a specific legal item, and each legal item is taken as one knowledge, for example, the third first rule (one) of the tax Law of the vehicles and ships in the people's republic of China states that fishing boats for culture belong to the tax free vehicles and ships, and the second rule that vehicles and ships special for military forces and armed police force belong to the tax free vehicles and ships, namely two pieces of knowledge, which are different from the learning contents.
All knowledge learned by each user forms an item set, and an item set may contain a plurality of items, and if an item set contains K items, the item set becomes a K item set. Each transaction is a set of items and has a unique identifier. All the knowledge learned by the users constructs a firm database, i.e. the firm database is a collection of affairs. Meanwhile, two concepts, namely support degree counting and minimum support degree, need to be popularized. Suppose there are six knowledge of knowledge A, B, C, D, E, F, and there are 4 users at the same time, users are X, Y, Z, Q. Wherein, the knowledge learned by user X is A, B, C, the knowledge learned by user Y is B, C, D, E, F, the knowledge learned by user Z is at least A, D, E, F, the knowledge learned by user Q is C, E, F, then the support degree count for knowledge A is 2, which appears in user (item set) X and user Z, the support degree count for knowledge B is 2, which appears in user X and user Y, the support degree count for knowledge C is 3, which appears in user X, user Y and user Q, so as to determine the support degree technique, and the minimum support degree is the number determined according to the test result.
After the above contents are determined, the association degree between each piece of knowledge and other pieces of knowledge needs to be calculated, and the specific method is as follows:
firstly, scanning a transaction database to obtain the support degree count of each knowledge, deleting the knowledge with the support degree count smaller than the minimum support degree to obtain a frequent 1 item set, and according to the just-formed upright column, setting the support degree count of the knowledge A to be 2, setting the support degree technology of the knowledge B to be 2, setting the support degree count of the knowledge C to be 3, setting the support degree count of the knowledge D to be 2, setting the support degree count of the knowledge E to be 3, setting the minimum support degree count to be 3, and setting the frequent 1 item set to be the knowledge C, the knowledge E and the knowledge F.
And generating a candidate 2 item set according to the obtained frequent 1 item set by self-connection, then deleting the knowledge with the support degree smaller than the minimum support degree in the candidate 2 item set to obtain the frequent 2 item set, wherein the candidate 2 items are the item set containing two kinds of knowledge, for example, if a user containing knowledge A and knowledge B only has a user X, the count is 1, and if a user containing knowledge E and knowledge F has a user Y, a user Z and a user Q, the count is 3, and similarly, deleting the knowledge with the support degree smaller than the minimum support degree in the candidate 2 item set. And analogizing, generating a candidate K item set by self-connection according to the frequent K-1 item set in a layer-by-layer iteration mode, and thus finishing the determination of the frequent item set. Since an item set with high confidence is recommended when actually recommended, for this reason, the knowledge with the minimum support degree is deleted while the knowledge with the minimum support degree is deleted in the candidate k item set. The purpose of reducing screening can be achieved by setting in this way, and after all, each screening needs to traverse the whole transaction database, so that the deletion is minimum and the calculation amount can be reduced.
And for the frequent item set obtained in the steps, calculating the confidence coefficient between each knowledge in the frequent item set and the associated knowledge thereof, and deleting the knowledge with the confidence coefficient smaller than the minimum confidence coefficient to obtain the strongly associated knowledge. Confidence is the confidence of one knowledge relative to the other, and
that is, con (x, y) is a ratio of the x-term set containing knowledge and the y-term set, and the minimum confidence is also determined by simulation. Thus, knowledge above the minimum confidence level results in an association between knowledge that is strongly associated knowledge.
Then, the degree of association between the knowledge is determined according to the degree of confidence between the knowledge. Due to confidence
And confidence ≧>
The results of (a) are completely different, and for this reason, in order to effectively determine the degree of association between two pieces of knowledge, in the present embodiment, a calculation formula of the degree of association between the pieces of knowledge is also provided as:
wherein con (x, y) is the proportion of the knowledge x item set containing knowledge y, and con (y, x) is the proportion of the knowledge y item set containing knowledge x.
Thus, the knowledge learned by the user based on TOP-N and the knowledge learned by the user at present recommend knowledge with high association degree, i.e. form a mixed recommendation. That is, assuming that the user has learned knowledge a, and has calculated knowledge x, y, z from the degree of association between the knowledge, only x is recommended among users of TOP-N, and if all the knowledge x is learned, pushing is performed based on the degree of association of x, y, z. And if the new user does not have any learning record, recommending according to the knowledge with the most learning content of the TOP-N user. Of course, the recommendation may also be performed by adopting a weight calculation method, where the calculation formula is:
roc＝τ.SIM comprehenave (m,n)+(1-τ)cor(x,y)。
through the arrangement, accurate pushing for the new user and the old user is effectively achieved, the learning of the users for inaccurate finance and tax knowledge is reduced, and the learning efficiency is improved.
The above description is only a preferred embodiment of the present invention, and not intended to limit the present invention in other forms, and any person skilled in the art may apply the above modifications or changes to the equivalent embodiments with equivalent changes, without departing from the technical spirit of the present invention, and any simple modification, equivalent change and change made to the above embodiments according to the technical spirit of the present invention still belong to the protection scope of the technical spirit of the present invention.
Claims (4)
1. A finance and tax knowledge pushing method based on user behaviors is characterized by comprising the following steps:
a. firstly, obtaining basic attributes and reading records of a user, wherein the basic attributes comprise the industry of a unit where the user is located, the property of the unit where the user is located and the location of the unit where the user is located, and if the user is a person without a unit, the basic attributes are the expected work place of the user; the reading records are the learning duration of the user and the number of chapters and sections learned by the user;
b. b, searching similar users of the user according to the basic attributes and the reading records of the user obtained in the step a, and selecting users with similarity TOP-N;
c. then, taking each piece of knowledge related to the fiscal taxes as an item, forming an item set by all the knowledge learned by each user, constructing a firm database by all the knowledge learned by all the users, and calculating the association degree between each piece of knowledge and other knowledge;
d. according to the association degree between the knowledge learned by the TOP-N user selected in the step b and the knowledge obtained in the step c, mixed recommendation is realized for the user;
in the step b, the method for searching the similar user of the user comprises the following steps:
b1, firstly, determining similarity among basic attributes of users, wherein the similarity among the basic attributes comprises the similarity of industries of units where the users are located, the similarity of properties of the units where the users are located, and the similarity of places where the units where the users are located or places where the users expect to work, and the similarity of the industries of the units where the users are located is as follows:
that is, if the industries of the user m and the user n are consistent, S is i (m, n) is 1, if the industries of the user m and the user n are not consistent, S is i (m, n) is 0;
the similarity of the properties of the unit where the user is located is as follows:
that is, if the properties of the units where user m and user n are located are the same, S is n (m, n) is 1, if the industries of the user m and the user n are not consistent, S is n (m, n) is 0;
similarity of unit location where the user is located or expected working place of the user:
that is, if the locations of the units where the user m and the user n are located or the expected work places are consistent, S h (m, n) is 1, if the industries of the user m and the user n are not consistent, S is h (m, n) is 0;
b2, determining the similarity between the learning contents of the users according to the reading records of the users, wherein the calculation formula of the similarity between the learning contents of the users is as follows:
wherein i is learning content, i = (1, 2,3.. K), LT m,i And LT n,i The learning coefficients of the user m and the user n for the learning content i are respectively:
b3, then according to the formula:
SIM comprehensive (m,n)＝λ(sim l (m,n))+(1-λ)(αS i (m,n)+βS n (m,n)+δS h (m,n))
and calculating the comprehensive similarity of the user m and the user n, wherein lambda is a weighting system of the comprehensive similarity between the learning contents of the user m and the user n between the user m and the user n, and alpha, beta and delta are the similarity of the industry of the unit where the user is located, the similarity of the property of the unit where the user is located and the weighting coefficient of the similarity of the place where the user is located or the place where the user expects to work respectively.
2. The method for pushing fiscal knowledge based on user's behavior as claimed in claim 1, wherein in the step c, the method for calculating the degree of association between each piece of knowledge and other pieces of knowledge is:
c1, scanning a transaction database to obtain the support degree count of each knowledge, and deleting the knowledge with the support degree count smaller than the minimum support degree to obtain a frequent 1 item set;
c2, generating a candidate 2 item set according to the obtained frequent 1 item set in a self-connection mode, then deleting the knowledge that the number of support degrees in the candidate 2 item set is smaller than the minimum support degree to obtain a frequent 2 item set, and repeating the steps in a layer-by-layer iteration mode to generate a candidate K item set according to the frequent K-1 item set in a self-connection mode;
c3, calculating the confidence coefficient between each knowledge and the associated knowledge in the frequent item set for the frequent item set obtained in the step, and deleting the knowledge with the confidence coefficient smaller than the minimum confidence coefficient to obtain strong associated knowledge;
and c4, determining the association degree of the knowledge according to the confidence degree of the knowledge.
3. The method for pushing fiscal knowledge based on user's behavior as claimed in claim 2, wherein in the step c4, the calculation formula of the correlation degree between knowledge is:
wherein con (x, y) is the proportion of the knowledge x item set containing knowledge y, and con (y, x) is the proportion of the knowledge y item set containing knowledge x.
4. The property and tax knowledge push method based on user behavior according to claim 3, wherein in the step c, the knowledge with the lowest support degree is deleted at the same time of deleting the knowledge with the support degree smaller than the minimum support degree in the candidate k item set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211302411.7A CN115936904A (en) | 2022-10-24 | 2022-10-24 | Finance and tax knowledge pushing method based on user behaviors |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211302411.7A CN115936904A (en) | 2022-10-24 | 2022-10-24 | Finance and tax knowledge pushing method based on user behaviors |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115936904A true CN115936904A (en) | 2023-04-07 |
Family
ID=86554861
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211302411.7A Pending CN115936904A (en) | 2022-10-24 | 2022-10-24 | Finance and tax knowledge pushing method based on user behaviors |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115936904A (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663047A (en) * | 2012-03-29 | 2012-09-12 | 中国科学院计算技术研究所 | Method and device for mining social relationship during mobile reading |
| CN107943946A (en) * | 2017-11-24 | 2018-04-20 | 重庆科技学院 | Relevance method for digging between test item bank knowledge point based on Apriori algorithm |
| CN108628967A (en) * | 2018-04-23 | 2018-10-09 | 西安交通大学 | A kind of e-learning group partition method generating network similarity based on study |
| CN109241405A (en) * | 2018-08-13 | 2019-01-18 | 华中师范大学 | A kind of associated education resource collaborative filtering recommending method of knowledge based and system |
| CN109829110A (en) * | 2019-01-29 | 2019-05-31 | 四川长虹电器股份有限公司 | A kind of personalized recommendation method of learning materials |
| CN111581529A (en) * | 2020-05-07 | 2020-08-25 | 之江实验室 | Course recommendation method and device combining student fitness and course collocation degree |
| CN112733035A (en) * | 2021-01-21 | 2021-04-30 | 唐亮 | Knowledge point recommendation method and device based on knowledge graph, storage medium and electronic device |
| CN114025006A (en) * | 2021-10-20 | 2022-02-08 | 浪潮软件科技有限公司 | Method and system for accurately pushing tax-related knowledge, electronic equipment and storage medium |
| CN114385901A (en) * | 2020-10-22 | 2022-04-22 | 北京大学 | Method and system for recommending dissociative abnormal image attention of system |
| CN115168744A (en) * | 2022-08-09 | 2022-10-11 | 舟山广播电视总台 | Radio and television technology knowledge recommendation method based on user portrait and knowledge graph |
-
2022
- 2022-10-24 CN CN202211302411.7A patent/CN115936904A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663047A (en) * | 2012-03-29 | 2012-09-12 | 中国科学院计算技术研究所 | Method and device for mining social relationship during mobile reading |
| CN107943946A (en) * | 2017-11-24 | 2018-04-20 | 重庆科技学院 | Relevance method for digging between test item bank knowledge point based on Apriori algorithm |
| CN108628967A (en) * | 2018-04-23 | 2018-10-09 | 西安交通大学 | A kind of e-learning group partition method generating network similarity based on study |
| CN109241405A (en) * | 2018-08-13 | 2019-01-18 | 华中师范大学 | A kind of associated education resource collaborative filtering recommending method of knowledge based and system |
| CN109829110A (en) * | 2019-01-29 | 2019-05-31 | 四川长虹电器股份有限公司 | A kind of personalized recommendation method of learning materials |
| CN111581529A (en) * | 2020-05-07 | 2020-08-25 | 之江实验室 | Course recommendation method and device combining student fitness and course collocation degree |
| CN114385901A (en) * | 2020-10-22 | 2022-04-22 | 北京大学 | Method and system for recommending dissociative abnormal image attention of system |
| CN112733035A (en) * | 2021-01-21 | 2021-04-30 | 唐亮 | Knowledge point recommendation method and device based on knowledge graph, storage medium and electronic device |
| CN114025006A (en) * | 2021-10-20 | 2022-02-08 | 浪潮软件科技有限公司 | Method and system for accurately pushing tax-related knowledge, electronic equipment and storage medium |
| CN115168744A (en) * | 2022-08-09 | 2022-10-11 | 舟山广播电视总台 | Radio and television technology knowledge recommendation method based on user portrait and knowledge graph |
Non-Patent Citations (1)
| Title |
|---|
| 高鹏;: "基于关联模式的导向型工艺知识发现", 制造业自动化, no. 10, 25 May 2013 (2013-05-25) * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Nigrini | Forensic analytics: Methods and techniques for forensic accounting investigations | |
| Ahituv | A systematic approach toward assessing the value of an information system | |
| Harvey et al. | Databases in historical research | |
| Liao et al. | Intellectual capital disclosure and accounting standards | |
| Hedin et al. | Overview of the TREC 2009 Legal Track. | |
| Iaryczower et al. | More than politics: Ability and ideology in the British Appellate Committee | |
| Yao et al. | Soft or hard information? A trade-off selection of environmental disclosures by way of peer imitation and geographical distance | |
| Puspasari et al. | Implementation K-means clustering method in job recommendation system | |
| Özar et al. | Micro and small enterprise growth in Turkey: Under the shadow of financial crisis | |
| Ezenagu | Boom or bust, extractives are no longer saviours: the need for robust tax regimes in Gulf countries | |
| McGuire et al. | Penalties for industrial accidents: The impact of the Deepwater Horizon accident on BP’s reputation and stock market returns | |
| Berry et al. | The distribution of income in Colombia: an overview | |
| Young | Mergers and Aquisitions: Planning and Action | |
| Murphy et al. | Resources allocated to tackling the tax gap: a comparative EU study | |
| Chen et al. | Bagging or boosting? Empirical evidence from financial statement fraud detection | |
| Mohammadi et al. | Presenting a Model for Financial Reporting Fraud Detection using Genetic Algorithm | |
| Miller | Which Supreme Court Cases Influenced Recent Supreme Court IP Decisions? A Citation Study | |
| Lin et al. | The application of decision tree and artificial neural network to income tax audit: the examples of profit-seeking enterprise income tax and individual income tax in Taiwan | |
| CN115936904A (en) | Finance and tax knowledge pushing method based on user behaviors | |
| Cowling et al. | The hazards of delivering a public loan guarantee scheme: An analysis of borrower and lender characteristics | |
| Bripi et al. | Services trade and credit frictions: Evidence with matched bank–firm data | |
| Chimonaki et al. | Perspectives in fraud theories–A systematic review approach | |
| Vandaele | The economics of crime: an econometric investigation of auto theft in the United States. | |
| Li | Study on the Personal Credit Risk Evaluation based on Improved Fuzzy AHP Comprehensive Evaluation Method | |
| Fang | Credit Risk Evaluation Model of Small-Micro Enterprises for Rural Commercial Bank Based on XGBoost and Random Forest |
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 |