Enterprise consultation management method and system based on cloud platform

Abstract

The invention relates to the technical field of consultation management systems, and particularly discloses an enterprise consultation management method and system based on a cloud platform, and the method comprises the steps: obtaining user information which comprises user identity information and user input information; acquiring historical data accessed by the user according to the user identity information, wherein the historical data is empty when the user is not a historical access user; performing keyword identification on the user input information and the historical data based on the cloud platform to obtain a keyword priority sequence; consultation recommendation is carried out on the user according to the keyword priority sequence, and consultation recommendation content is dynamically adjusted according to real-time browsing data of the user; according to the method and the device, the keyword priority sequence is obtained, so that the consultation requirement of the user can be better met in the process of analyzing the user requirement, and then content recommendation is adaptively performed, so that the recommended content better meets the user requirement.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of consultation management systems, in particular to an enterprise consultation management method and system based on a cloud platform. BACKGROUND

[0002] An enterprise consultation management system serves as a window for the outside world to understand enterprise business information, and plays an important role in enterprise development and promotion. With the rapid development of Internet technology and AI technology, the enterprise consultation management system is becoming more and more intelligent and automated. By introducing an existing AI model to analyze user demand information, corresponding content is provided according to the analysis result, thereby meeting different needs in the enterprise consultation process.

[0003] In the prior art, the process of analyzing user demand information is mainly through tagging. By processing user demand information, the corresponding keywords are obtained as tags, and the response content is recommended according to the tags, thereby being able to adaptively meet the user's demand to a certain extent.

[0004] For an enterprise with complex and extensive business, the existing tagging method has certain limitations in the application process. Because of the coincidence and intersection of many business categories, there is a certain error in the accuracy of the recommendation in the process of recommending to the user, thereby affecting the effect of enterprise consultation. Therefore, how to accurately obtain the user's demand and adaptively recommend the content is the fundamental problem to be solved by the application. SUMMARY

[0005] The purpose of the application is to provide an enterprise consultation management method and system based on a cloud platform, which solves the following technical problems: How to accurately obtain the user's demand and adaptively recommend the content.

[0006] The purpose of the application can be achieved by the following technical solutions: The enterprise consultation management method based on the cloud platform comprises the following steps: Obtaining user information, wherein the user information comprises user identity information and user input information; Obtaining historical data of user access according to the user identity information, wherein the historical data is empty when the user is not a historical access user; Performing keyword recognition on the user input information and the historical data based on the cloud platform to obtain a keyword priority sequence; Performing consultation recommendation on the user according to the keyword priority sequence, and dynamically adjusting the consultation recommendation content according to the user real-time browsing data.

[0007] Further, the process of obtaining the keyword priority sequence comprises the following steps: Performing AI recognition on the user input information through the cloud platform to obtain a first keyword group, wherein each keyword in the first keyword group has a corresponding criticality ratio, and the sum of the criticality ratios of all keywords in the first keyword group is 1; When the historical data is empty, the criticality ratio corresponding to each keyword is used as the adjusted criticality ratio, and the keywords are arranged in descending order according to the adjusted criticality ratio to obtain the keyword priority sequence; When the historical data is not empty, the content and browsing time of each visit in the historical data of user visits are obtained, and the keyword recognition of the content of each visit is performed on the cloud platform to obtain a second keyword group, and each keyword in the second keyword group has a corresponding criticality ratio, and the sum of the criticality ratios of all keywords in the second keyword group is 1; the keyword priority sequence is determined by the first keyword group, the second keyword group and the browsing time of each visit in the historical data.

[0008] Furthermore, when the historical data is not empty, the process of obtaining the keyword priority sequence further includes: Through the equation Calculate the adjusted criticality ratio of the i-th keyword , according to the adjusted criticality ratio Arrange the keywords from largest to smallest to obtain a keyword priority sequence; in, is the number of keywords in the second keyword group that overlap with the i-th keyword in the first keyword group, j∈[1, ], is the browsing time of the content corresponding to the jth keyword, for indivual The mean of is the criticality ratio of the ith keyword, and m is the number of keywords in the first keyword group.

[0009] Furthermore, the process of recommending consultations to users according to the keyword priority sequence includes: Pre-set recommended keywords and their corresponding recommendation ratios for the consulting recommendation content. The sum of the recommendation ratios of all recommended keywords for the same consulting recommendation content is 1. Through the equation Calculate the compliance coefficient G of each consultation recommendation content, and make consultation recommendations to users in descending order of the compliance coefficient G, where q is the number of keywords that overlap with the current consultation recommendation content and the keyword priority sequence, k∈[1,q], is the adjusted criticality ratio of the k-th keyword, is the recommendation ratio of the k-th keyword.

[0010] Further, the process of dynamically adjusting the consultation recommendation content according to the real-time browsing data of the user comprises: obtaining the real-time browsing data of the user, matching the real-time browsing data with the keyword priority sequence, when the judgment result is a high matching state, continuing to consult and recommend the user according to the keyword priority sequence, when the judgment result is a low matching state, performing similarity analysis on the browsing data corresponding to the low matching state, and determining the recommendation strategy according to the similarity analysis result.

[0011] Further, the process of matching degree judgment comprises: obtaining the coincidence coefficient G of the user browsing content and the keyword priority sequence, comparing the coincidence coefficient G with the preset threshold Gt, if the coincidence coefficient G of all user browsing contents is greater than or equal to Gt, the judgment result is a high matching state, otherwise, the judgment result is a low matching state.

[0012] Further, the process of similarity analysis comprises: obtaining the number u of user browsing contents, and obtaining the coincidence coefficient G of any two user browsing contents in all user browsing contents; obtaining the similarity coefficient s through the equation , wherein, is the number of any two user browsing content combinations in the u groups of user browsing contents, is the coincidence coefficient obtained in the xth combination; comparing the similarity coefficient s with the similarity threshold, when the similarity coefficient s is less than the similarity threshold, judging as high similarity, selecting similar consultation contents according to all user browsing contents to recommend the user, when the similarity coefficient s is greater than or equal to the similarity threshold, judging as low similarity, selecting the consultation contents not browsed by the user to recommend the user.

[0013] Further, in the high similarity state, the process of selecting similar consultation contents according to all user browsing contents to recommend the user comprises: calculating the browsing tendency key importance proportion of each keyword in the user browsing content through the equation , wherein, is the number of the yth keyword appearing in the keywords corresponding to the user browsing content, is the total browsing time of the yth keyword corresponding to the user browsing content, ts is the total browsing time of all user browsing contents, and the keywords are arranged in descending order according to the browsing tendency key importance proportion to obtain the bias keyword priority sequence; ​The coincidence coefficient G of the bias keyword priority sequence and each consultation recommendation content is calculated, and the user is consulted and recommended in real time according to the order from large to small of the coincidence coefficient G.

[0014] The enterprise consultation management system based on the cloud platform comprises a user information acquisition end, a user information analysis end, a cloud platform and a recommendation analysis unit. The user information acquisition end is used for acquiring user information, and the user information comprises user identity information and user input information. The user information analysis end is used for acquiring historical data of user access according to the user identity information, and the historical data is empty when the user is not a historical access user. The cloud platform is used for keyword recognition on the user input information and the historical data, and a keyword priority sequence is acquired. The recommendation analysis unit is used for consultation recommendation on the user according to the keyword priority sequence, and dynamically adjusts the consultation recommendation content according to real-time browsing data of the user.

[0015] The beneficial effects of the present application are as follows: (1) The keyword priority sequence is acquired at the same time of keyword recognition, and thus the user consultation demand can be more met in the process of analyzing the user demand, and the content recommendation can be adaptively performed. BRIEF DESCRIPTION OF DRAWINGS

[0016] The present application will be further described below in conjunction with the drawings.

[0017] Figure 1 It is a step flow chart of the enterprise consultation management method based on the cloud platform of the present application. Figure 2 It is a logic block diagram of the enterprise consultation management system based on the cloud platform of the present application. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings of the embodiments of the present application.

[0019] In one embodiment, the enterprise consultation management method based on the cloud platform is provided, please refer to Figure 1As shown, the method comprises: obtaining user information, the user information comprising user identity information and user input information; obtaining historical data of user access according to the user identity information, the historical data being empty when the user is not historically accessed by the user; identifying keywords based on the cloud platform, obtaining a keyword priority sequence; recommending consultation to the user according to the keyword priority sequence, and dynamically adjusting the consultation recommendation content according to the user real-time browsing data. The above scheme, compared with the prior art, obtains the keyword priority sequence while identifying the keywords, that is, each keyword has different weights, and when the keyword matches the user input information better, the weight is larger, and vice versa. Through the acquisition of the keyword priority sequence, the user's consultation demand can be better met in the process of analyzing the user's demand, and the content recommendation can be adaptively performed. The process of obtaining the keyword priority sequence comprises: first, the cloud platform identifies the user input information by AI to obtain a first keyword group. The cloud platform integrates a plurality of tools, wherein the AI identification model can directly use an existing intelligent model, and the content keywords can be obtained through the AI identification model by preset instructions and output forms. Each keyword in the first keyword group corresponds to a key proportion, and the sum of the key proportions of all keywords in the first keyword group is 1. The key proportion can be used to determine the weight of different keywords, and the accuracy of the recommended content can be improved by considering the weight in the analysis. In addition, when the historical data is empty, the key proportion corresponding to each keyword is taken as the adjusted key proportion, and the keywords are arranged in descending order of the adjusted key proportion to obtain the keyword priority sequence. When the historical data is not empty, the content and browsing time of each access in the historical data of user access are obtained, the cloud platform identifies the keywords of each access content to obtain a second keyword group, and each keyword in the second keyword group corresponds to a key proportion. The sum of the key proportions of all keywords in the second keyword group is 1. The keyword priority sequence is determined by the first keyword group, the second keyword group and the browsing time of each access in the historical data. Through the above process, when the user has access history record, the keyword priority sequence can be further adjusted and optimized according to the content and browsing time of each access in the user's historical data, so that the recommended content is more in line with the user's demand.

[0020] When the historical data is not empty, the process of obtaining the keyword priority sequence comprises: calculating the adjusted key proportion of the i-th keyword by the equation . , wherein, is the number of keywords in the second keyword group that coincide with the i-th keyword in the first keyword group, j∈[1, ], The browsing time length of the jth keyword corresponding to the access content, The average value of The average value of The average value of The average value of The process of obtaining the adjusted importance proportion of each keyword by comprehensively adjusting the coincidence of historical access data (keyword coincidence amount) and the attention of users to the content (user browsing time length), so that the adjusted importance proportion can better meet the actual needs of users, and the consultation content is recommended according to the adjusted importance proportion The keywords are arranged in descending order of the adjusted importance proportion, and a keyword priority sequence is obtained, and the consultation content is recommended according to the keyword priority sequence, which can improve the accuracy of recommendation and thus improve the probability of successful transaction; in addition, it should be noted that the numerator in the equation is the part adjusted by the ith keyword according to the keyword coincidence amount and the user browsing time length, and the denominator is the sum of the adjusted parts of all keywords, so that the adjusted importance proportion of all keywords is ensured The sum is still 1.

[0021] In addition, the process of recommending consultation to users according to the keyword priority sequence includes: setting the recommendation keywords and their corresponding recommendation proportion for the consultation recommendation content in advance, and the sum of the recommendation proportions of all recommendation keywords of the same consultation recommendation content is 1; since the recommendation content of the consultation management platform is limited, the process of setting the consultation recommendation content in advance can use the AI recognition model of the cloud platform, or it can be directly set by artificial, wherein the artificial setting method can ensure more accurate matching between the recommendation content and the user demand, and then the coincidence coefficient G of each consultation recommendation content is calculated by the equation , wherein q is the number of keywords that coincide with the keyword priority sequence in the current consultation recommendation content, k∈[1, q], The adjusted importance proportion of the kth keyword, The recommendation proportion of the kth keyword, so when the number of keywords that coincide with the keyword priority sequence in the recommendation keywords of the consultation recommendation content is more, the recommendation proportion is higher, and the importance proportion is higher, it means that the consultation recommendation content is more matched with the user's demand, and then the consultation is recommended to the user according to the size of the coincidence coefficient G, that is, the consultation is recommended to the user according to the descending order of the coincidence coefficient G, which can provide more accurate and matched consultation content according to the demand state of different access users, improve the user experience and the success rate of cooperation.

[0022] In one embodiment, a process for dynamically adjusting the consulting recommendation content according to the user's real-time browsing data is provided, which specifically includes: first obtaining the user's real-time browsing data, matching the real-time browsing data with the keyword priority sequence, the matching process mainly determines whether the content of interest has changed relative to the input information, so as to timely adjust the recommendation strategy, the specific process includes: first obtaining the coincidence coefficient G of the user's browsing content and the keyword priority sequence, comparing the coincidence coefficient G with the preset threshold Gt, the preset threshold Gt is set according to the test data, if the coincidence coefficient G of all user browsing content is greater than or equal to Gt, the judgment result is high matching state, otherwise, the judgment result is low matching state, therefore, when the judgment result is high matching state, continue to consult and recommend the user according to the keyword priority sequence, ensure that the recommended content is highly matched with the user's demand, when the judgment result is low matching state, according to the similarity analysis of the browsing data corresponding to the low matching state, determine the recommendation strategy according to the similarity analysis result, through the process of similarity analysis, when the similarity is high, it shows that the user's demand is focused on a certain direction, at this time, select similar consulting content according to all user browsing content to recommend to the user, ensure that the recommended content is matched with the user's demand, otherwise, when the similarity is low, it shows that the user's demand is not fixed, different directions of recommended content need to be selected, at this time, select the user's unvisited content for recommendation, which can improve the user's involvement and improve the user experience and the success rate of cooperation, the specific process of similarity analysis includes: first obtaining the number u of user browsing content, obtaining the coincidence coefficient G of any two user browsing content in all user browsing content; the similarity coefficient s is calculated by the equation , wherein, is the number of any two user browsing content combinations in the u user browsing content, is the coincidence coefficient obtained in the xth combination; therefore, when the value of the similarity coefficient s tends to 0, it shows that the similarity of all browsing data is higher, otherwise it shows that the similarity of the browsing data is lower, compare the similarity coefficient s with the similarity threshold, the similarity threshold is obtained according to the critical data in the test process, therefore, when the similarity coefficient s is less than the similarity threshold, it is judged as high similarity, select similar consulting content according to all user browsing content to recommend to the user, when the similarity coefficient s is greater than or equal to the similarity threshold, it is judged as low similarity, select the user's unvisited consulting content to recommend to the user, through the above process, the consulting recommendation content can be dynamically adjusted according to the user's real-time browsing data, thereby improving the user experience and the success rate of cooperation.

[0023] In addition, in the high similarity state, the process of selecting similar consulting content according to all user browsing content to recommend to the user includes: through the equation Calculate the browsing tendency and criticality ratio of each keyword in the user's browsing content ,in, is the number of times the yth keyword appears in the corresponding keyword of the content browsed by the user, is the total browsing time of the content browsed by the user corresponding to the yth keyword, and ts is the total browsing time of the content browsed by all users. The larger the value, the The larger the relative ts ratio is, the more important the keyword is, and the corresponding browsing tendency is the key ratio. The higher the browsing tendency, the more important the proportion Arrange the keywords in descending order to obtain a biased keyword priority sequence; calculate the compliance coefficient G between the biased keyword priority sequence and each consultation recommendation content, and make consultation recommendations to users in real time according to the compliance coefficient G in descending order, so that the recommended content is highly consistent with the content browsed by the user, satisfying the user's targeted consultation process.

[0024] In one embodiment, a cloud-based enterprise consulting management system is provided. Figure 2 As shown, it includes a user information collection end, a user information analysis end, a cloud platform and a recommendation analysis unit; the user information collection end is used to obtain user information, and the user information includes user identity information and user input information; the user information analysis end is used to obtain historical data of user visits based on user identity information, and the historical data is empty when the user is not a historical visitor; the cloud platform is used to perform keyword recognition on user input information and historical data to obtain a keyword priority sequence; the recommendation analysis unit is used to make consultation recommendations to users according to the keyword priority sequence, and dynamically adjust the consultation recommendation content according to the user's real-time browsing data.

[0025] The above is a detailed description of an embodiment of the present invention. However, the content described is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. The enterprise consulting management method based on the cloud platform is characterized by: The method comprises: Obtaining user information, including user identity information and user input information; Get the user's access history data based on the user's identity information. If the user is not a historical access user, the history data will be empty. Perform keyword recognition on user input information and historical data based on the cloud platform to obtain keyword priority sequence; Provide consultation recommendations to users according to keyword priority sequence, and dynamically adjust the consultation recommendation content based on users' real-time browsing data.

2. The enterprise consulting management method based on the cloud platform according to claim 1 is characterized in that: The process of obtaining the keyword priority sequence includes: Performing AI recognition on the user input information through the cloud platform to obtain a first keyword group, wherein each keyword in the first keyword group has a corresponding criticality ratio, and the sum of the criticality ratios of all keywords in the first keyword group is 1; When the historical data is empty, the criticality ratio corresponding to each keyword is used as the adjusted criticality ratio, and the keywords are arranged in descending order according to the adjusted criticality ratio to obtain the keyword priority sequence; When the historical data is not empty, the content and browsing time of each visit in the historical data of user visits are obtained, and the keyword recognition of the content of each visit is performed on the cloud platform to obtain a second keyword group, and each keyword in the second keyword group has a corresponding criticality ratio, and the sum of the criticality ratios of all keywords in the second keyword group is 1; the keyword priority sequence is determined by the first keyword group, the second keyword group and the browsing time of each visit in the historical data.

3. The enterprise consulting management method based on the cloud platform according to claim 2 is characterized in that: When the historical data is not empty, the process of obtaining the keyword priority sequence further includes: Through the equation Calculate the adjusted criticality ratio of the i-th keyword , according to the adjusted criticality ratio Arrange the keywords from largest to smallest to obtain a keyword priority sequence; in, is the number of keywords in the second keyword group that overlap with the i-th keyword in the first keyword group, j∈[1, ], is the browsing time of the content corresponding to the j-th keyword, for indivual The mean of is the criticality ratio of the ith keyword, and m is the number of keywords in the first keyword group.

4. The enterprise consulting management method based on the cloud platform according to claim 3 is characterized in that: The process of recommending consultations to users according to the keyword priority sequence includes: Pre-set recommended keywords and their corresponding recommendation ratios for the consulting recommendation content. The sum of the recommendation ratios of all recommended keywords for the same consulting recommendation content is 1. Through the equation Calculate the compliance coefficient G of each consultation recommendation content, and make consultation recommendations to users in descending order of the compliance coefficient G, where q is the number of keywords that overlap with the current consultation recommendation content and the keyword priority sequence, k∈[1,q], is the adjusted criticality ratio of the k-th keyword, is the recommendation ratio of the k-th keyword.

5. The enterprise consulting management method based on the cloud platform according to claim 4 is characterized in that: The process of dynamically adjusting consultation recommendations based on real-time user browsing data includes: Obtain the user's real-time browsing data and judge the matching degree between the real-time browsing data and the keyword priority sequence. When the judgment result is a high match state, continue to make consultation recommendations to the user according to the keyword priority sequence. When the judgment result is a low match state, perform similarity analysis based on the browsing data corresponding to the low match state, and determine the recommendation strategy based on the similarity analysis results.

6. The enterprise consulting management method based on the cloud platform according to claim 5 is characterized in that: The matching degree judgment process includes: Obtain the matching coefficient G between the user's browsing content and the keyword priority sequence, and compare the matching coefficient G with the preset threshold Gt. If the matching coefficient G of all user's browsing content is greater than or equal to Gt, the judgment result is a high match state, otherwise, the judgment result is a low match state.

7. The enterprise consulting management method based on a cloud platform according to claim 5, characterized in that: The similarity analysis process includes: Obtain the number u of user-browsed contents and obtain the matching coefficient G between any two user-browsed contents among all user-browsed contents; Through the equation Calculate the similarity coefficient s, where The number of combinations of content browsed by any two users in group u. is the compliance coefficient obtained in the xth combination; Compare the similarity coefficient s with the similarity threshold. When the similarity coefficient s is less than the similarity threshold, it is judged as high similarity, and similar consulting content is selected based on all the user's browsed content to recommend to the user. When the similarity coefficient s is greater than or equal to the similarity threshold, it is judged as low similarity, and consulting content that the user has not browsed is selected to recommend to the user.

8. The enterprise consulting management method based on the cloud platform according to claim 7 is characterized in that: In a high similarity state, the process of selecting similar consulting content from all user browsing content and recommending it to the user includes: Through the equation Calculate the browsing tendency and criticality ratio of each keyword in the user's browsing content ,in, is the number of times the yth keyword appears in the corresponding keyword of the content browsed by the user, is the total browsing time of the content corresponding to the yth keyword, ts is the total browsing time of all users browsing the content, and the proportion of key browsing tendency is Arrange the keywords from largest to smallest to obtain a biased keyword priority sequence; Calculate the matching coefficient G between the biased keyword priority sequence and each consultation recommendation content, and make consultation recommendations to users in real time according to the order of matching coefficient G from large to small.

9. The enterprise consulting management system based on the cloud platform is characterized by: The system adopts the enterprise consulting management method based on a cloud platform according to any one of claims 1 to 8, comprising a user information collection terminal, a user information analysis terminal, a cloud platform and a recommendation analysis unit; The user information collection end is used to obtain user information, which includes user identity information and user input information; the user information analysis end is used to obtain historical data of user visits based on user identity information, and the historical data is empty when the user is not a historical visitor; the cloud platform is used to perform keyword recognition on user input information and historical data to obtain a keyword priority sequence; the recommendation analysis unit is used to make consultation recommendations to users according to the keyword priority sequence, and dynamically adjust the consultation recommendation content based on the user's real-time browsing data.

Images