Intelligent pushing method for information consultation service

Abstract

The invention discloses an intelligent pushing method for information consultation service, and relates to the technical field of service pushing. The method comprises the steps that search information of a user is collected, and target consultation information of the user is extracted according to the search information; searching whether associated content of the target consultation information exists on the network, and if the associated content of the target consultation information exists, collecting personal information of the user; service information of the information consultation service is collected, and whether the information consultation service is pushed or not is judged by combining personal information of the user; if the information consultation service is pushed, a pushing scheme is set according to the personal information of the user, and the information consultation service is pushed according to the pushing scheme; and if the associated content of the target consultation information does not exist, collecting personal information of the user, generating a pushing scheme according to the personal information, and pushing the information consultation service according to the pushing scheme. According to the invention, the accuracy of intelligent pushing for information consultation service is improved.

Description

Technical Field

[0001] This application relates to the technical field of service push, and particularly to an intelligent push method for information consultation services. Background Art

[0002] Information consultation service is an information service activity in which an information service department, according to the needs put forward by users, based on professional knowledge, practical experience and innovation ability, fully develops and utilizes information resources, and uses scientific methods and modern technical means to provide users with suggestions, solutions, strategies, plans or measures, etc. to solve problems. The intelligent push of information consultation service saves users' time and energy and provides convenience for users' daily life. Existing information consultation services select the information consultation services to be pushed and the push time according to user portraits, and for whether to push information consultation services, it only depends on the user's search situation. If the user searches, it is judged that the user needs to receive the push of information consultation services, and then it is directly pushed. This single judgment method makes the push result inaccurate, because in addition to information consultation services, users can also obtain content through other means and do not necessarily need information consultation services. Pushing information consultation services based on a single judgment basis is likely to cause user annoyance and bring inconvenience to users. Summary of the Invention

[0003] The purpose of the present invention is to provide an intelligent push method for information consultation services to solve the problems put forward in the above background art.

[0004] An intelligent push method for information consultation services provided by this application adopts the following technical solutions: Collect the search information of the user, and extract the target consultation information of the user according to the search information; Search on the network to find whether there is associated content of the target consultation information. If there is associated content of the target consultation information, collect the user's personal information; Collect the service information of the information consultation service, and combine the user's personal information to judge whether to push the information consultation service; If the information consultation service is to be pushed, set a push plan according to the user's personal information, and push the information consultation service according to the push plan; If there is no associated content of the target consultation information, collect the user's personal information, generate a push plan according to the personal information, and push the information consultation service according to the push plan.

[0005] Preferably, the step of collecting the service information of the information consultation service and combining the user's personal information to judge whether to push the information consultation service is specifically: Analyze the user's retrieval ability according to the user's personal information, obtain the retrieval difficulty of the target consultation information, and combine the retrieval ability and the retrieval difficulty to obtain a retrieval adaptation value; The information consultation service corresponding to the matching target consultation information is recorded as the corresponding service, and the service information of the corresponding service is collected; The adaptation value of the corresponding service is obtained by evaluating according to the service information of the corresponding service, and the maximum adaptation value of the corresponding service is selected as the service adaptation value; Compare the retrieval adaptation value with the service adaptation value. If the retrieval adaptation value is not less than the service adaptation value, it is determined not to push the information consultation service; If the retrieval adaptation value is less than the service adaptation value, it is determined to push the information consultation service.

[0006] Preferably, the step of analyzing the user's retrieval ability based on the user's personal information, obtaining the retrieval difficulty of the target consultation information, and combining the retrieval ability and the retrieval difficulty to obtain the retrieval adaptation value is specifically as follows: Extract the historical retrieval information according to the user's personal information, and count the historical retrieval platforms used by the user according to the historical retrieval information; Divide the historical retrieval information to obtain historical retrieval data, and obtain the user's basic retrieval value according to the analysis of the historical retrieval data; Obtain the discrimination ability value of the user's discrimination information according to the evaluation of the historical retrieval data, and combine the basic retrieval value to obtain the user's retrieval ability value; Obtain the parameter data of the target consultation information, and obtain the retrieval difficulty value of the target consultation information according to the evaluation of the parameter data; Set the proportionality coefficients of the retrieval ability value and the retrieval difficulty value respectively, and calculate the retrieval adaptation value of the target consultation information according to the proportionality coefficients.

[0007] Preferably, the step of dividing the historical retrieval information to obtain historical retrieval data and obtaining the user's basic retrieval value according to the analysis of the historical retrieval data is specifically as follows: Divide the historical retrieval information according to the retrieval time to obtain the historical retrieval data of multiple retrievals; Extract the final retrieval content retrieved by the user from the historical retrieval data, and record the keywords of the final retrieval content as the final keywords; Obtain the keywords of the user's initial retrieval and record them as the retrieval keywords, and compare to obtain the keyword similarity between the final keywords and the retrieval keywords; Statistically calculate the average value of the keyword similarities of multiple retrievals, and record it as the retrieval accuracy; Statistically calculate the average number of historical retrieval platforms used by the user each time according to the historical retrieval data, and record it as the retrieval quantity; Statistically calculate the number of information sources of the historical retrieval platforms, and record it as the number of information sources. Combine the retrieval accuracy and the retrieval quantity to obtain the basic retrieval value.

[0008] Preferably, the step of obtaining the retrieval ability value of the user by combining the discrimination ability value of the user identification information evaluated according to the historical retrieval data with the basic retrieval value is specifically as follows: Extract the retrieval requirement content of the user from the historical retrieval data, and divide the retrieval requirement content into subjective content and objective content; Extract the historical retrieval data of the subjective content and denote it as the subjective retrieval data, and extract the average retrieval times of the user in the subjective retrieval data; Extract the average amount of information of the user reading the retrieval requirement content in the subjective retrieval data, and extract the average change times of the retrieval expressions in the same retrieval in the subjective retrieval data; Combine the average retrieval times, the average amount of information and the average change times to calculate the subjective discrimination value of the user; Extract the historical retrieval data of the objective content and denote it as the objective retrieval data, analyze the objective retrieval data to obtain the objective discrimination value, and combine the subjective discrimination value to obtain the discrimination ability value of the user; Set the proportionality coefficients of the discrimination ability value and the basic retrieval value respectively, and calculate the retrieval ability value of the user according to the proportionality coefficients.

[0009] Preferably, the step of extracting the historical retrieval data of the objective content and denoting it as the objective retrieval data, and analyzing the objective retrieval data to obtain the objective discrimination value is specifically as follows: Judge whether there is a clear answer to the retrieval requirement content of the user through retrieval, denote the objective retrieval data with a clear answer as the clear retrieval data, and denote the objective retrieval data without a clear answer as the non-clear retrieval data; Extract the correct answer content in the clear retrieval data, and count the average frequency of the user retrieving the correct answer content as the first discrimination value; Extract the average correlation degree between the adjusted retrieval expression of the user and the previous retrieval content in the non-clear retrieval data as the second discrimination value; Calculate the average value of the first discrimination value and the second discrimination value to obtain the objective discrimination value.

[0010] Preferably, the step of obtaining the parameter data of the target consultation information and evaluating the retrieval difficulty value of the target consultation information according to the parameter data is specifically as follows: Obtain the keywords of the target consultation information and denote them as the target keywords, and search for the information containing the target keywords in the historical retrieval platform and denote it as the target information; Count the amount of information of the target information and count the number of types of the target information; Set the proportionality coefficients of the amount of information and the number of types of the target information respectively, and calculate the retrieval difficulty value of the target consultation information according to the proportionality coefficients.

[0011] Preferably, the step of evaluating the adaptation value of the corresponding service according to the service information of the corresponding service and selecting the maximum adaptation value of the corresponding service as the service adaptation value is specifically as follows: Search for the information type corresponding to the target consultation information, and determine whether the user understands the information type according to the user's personal information; If the user does not understand the information type, extract the average understanding duration of the information type as the service duration; If the user understands the information type, extract the average consultation duration of the corresponding service as the service duration; Extract the average queuing duration of the corresponding service, and extract the average satisfaction degree feedback of the corresponding service; Calculate the adaptation value of the corresponding service by combining the service duration, the average queuing duration, and the average satisfaction degree, and select the maximum adaptation value of the corresponding service as the service adaptation value.

[0012] Preferably, the step of, if the user understands the information type, extracting the average consultation duration of the corresponding service as the service duration is specifically as follows: Obtain the historical consultation data of the corresponding service, set the duration time period, and respectively count the average consultation times of the same user within different duration time periods; Select the duration time period with the largest average consultation times of the same user as the understanding duration period; Calculate the average consultation duration of the user within the understanding duration period as the service duration of the corresponding service.

[0013] Preferably, the step of, if the information consultation service is to be pushed, setting a push plan according to the user's personal information and pushing the information consultation service according to the push plan is specifically as follows: Extract the platform usage rate of the user according to the user's personal information, and select the platform with the highest platform usage rate as the push platform; Extract the online rule of the user according to the user's personal information, and generate the push time according to the online rule; Form a push plan according to the push time and the push platform, and push the information consultation service with the highest service adaptation value according to the push plan.

[0014] In summary, the present application includes at least one of the following beneficial technical effects: 1. Collect the user's search information to confirm the user's target consultation information. Evaluate the retrieval adaptation value of the user's retrieval of the target consultation information based on the user's personal information, and evaluate the service adaptation value of the target consultation information based on the service information. By comparing the magnitudes of the retrieval adaptation value and the service adaptation value, determine whether to push information consultation services to the user. Determining whether to push information consultation services according to the user's own retrieval ability is beneficial to reducing the interference caused by the push of information consultation services to the user. It not only reduces the workload of information consultation service personnel, but also provides the user with a more adaptable information acquisition method, improving the accuracy of intelligent push for information consultation services.

[0015] 2. Confirm the user's basic retrieval value based on the retrieval accuracy and retrieval quantity in the user's historical retrieval data. Then distinguish between subjective retrieval and objective retrieval, and evaluate the discrimination values of the user under different types of retrieval respectively. Comprehensively obtain the user's discrimination ability value. Thus, evaluate the user's retrieval ability based on the basic retrieval value and the discrimination ability value, and combine with the retrieval difficulty value obtained through the amount of information and the number of types of target information to confirm the retrieval adaptation value of the target consultation information. By evaluating the adaptation degree between the target consultation information and the retrieval, unnecessary consultations can be reduced, improving the quality and speed of information consultation services.

[0016] 3. According to the judgment result of whether the user understands the information type corresponding to the target consultation information, obtain different service durations respectively, and then calculate the adaptation value of the corresponding service by comprehensively averaging the queuing duration and the average satisfaction degree. Evaluate the adaptation value of the information consultation service according to the actual situation, which is not only beneficial to judging whether to push the information consultation service, but also can select the best evaluation information consultation service during the push, improving the accuracy of intelligent push for information consultation services and the adaptation degree between the intelligent push of information consultation services and the user. Description of the Drawings

[0017] Figure 1 It is a schematic diagram of the specific steps of an embodiment of an intelligent push method for information consultation services of the present invention. Detailed Embodiment

[0018] The following combines the embodiments and Figure 1 Further detailed description of the present invention is provided, but the embodiments of the present invention are not limited thereto.

[0019] The present invention discloses an intelligent push method for information consultation services, which specifically includes the following steps: Step S1, collect the user's search information, and extract the user's target consultation information according to the search information.

[0020] The target consultation information corresponding to the search information can be found in the historical search information by searching for keywords. For example, machine learning algorithms (such as support vector machines, naive Bayes, decision trees, etc.) can be used to classify or cluster the search information, so as to extract the user's target consultation information.

[0021] Step S2, search on the network to find whether there is associated content of the target consultation information. If there is associated content of the target consultation information, collect the user's personal information.

[0022] It is possible to perform a preliminary search by using a search engine to input keywords related to the target consultation information, so as to confirm whether there is associated content of the target consultation information.

[0023] Step S3, collect the service information of the information consultation service, and combine the user's personal information to judge whether to push the information consultation service.

[0024] Step S4, if the information consultation service is to be pushed, set the push plan according to the user's personal information, and push the information consultation service according to the push plan.

[0025] Step S5, if there is no associated content of the target consultation information, collect the user's personal information, generate a push plan according to the personal information, and push the information consultation service according to the push plan.

[0026] If there is no associated content of the target consultation information on the Internet, the user cannot obtain the corresponding information through retrieval. At this time, it is necessary to obtain the required information through the consultation service, so the consultation information service is pushed.

[0027] In actual application, in the intelligent push of the existing information consultation service, usually a suitable information consultation service is selected for push according to the user's retrieval, without considering whether the user needs the information consultation service. Some simple content can be easily retrieved on the Internet. For such content, if it is blindly pushed, on the one hand, it will deepen the user's aversion to the information consultation service, and on the other hand, it will also interfere with the user's retrieval. For example, the user accidentally touches the consultation page and other situations. At the same time, if the user selects the information consultation service for such services, it will also increase the workload of the service personnel, resulting in a worse experience for the services of other people with real needs. For example, the user wants to find out how to connect the A mobile phone to the Bluetooth headset. There is a lot of such content on the Internet, and the corresponding connection steps can be obtained easily. At this time, if the consultation service of the mobile phone application is pushed, the user still needs to wait in line, which is far less convenient than retrieving by himself. And when the server accesses the user and answers for the user, it also needs to wait for the user to end the service, increasing the service volume, and correspondingly reducing the service experience. So at this time, there is no need to push the consultation service, and it is more suitable for the user to obtain the required information through retrieval.

[0028] The steps of collecting service information of the information consultation service and determining whether to push the information consultation service in combination with the user's personal information are as follows: Step S31, analyze the user's retrieval ability based on the user's personal information, obtain the retrieval difficulty of the target consultation information, and obtain a retrieval adaptation value by combining the retrieval ability and the retrieval difficulty.

[0029] Step S32, match the information consultation service corresponding to the target consultation information and record it as the corresponding service, and collect the service information of the corresponding service.

[0030] The target consultation information has its corresponding information consultation service. For example, if the user wants to consult questions related to refrigerators, then the corresponding service is the consultation service of platforms such as selling refrigerators and repairing refrigerators. And if the user has a specific refrigerator brand, the scope of the corresponding service is further narrowed.

[0031] Step S33, evaluate the adaptation value of the corresponding service based on the service information of the corresponding service, and select the maximum adaptation value of the corresponding service as the service adaptation value.

[0032] Step S34, compare the retrieval adaptation value and the service adaptation value. If the retrieval adaptation value is not less than the service adaptation value, it is determined not to push the information consultation service.

[0033] Step S35, if the retrieval adaptation value is less than the service adaptation value, it is determined to push the information consultation service.

[0034] In actual application, in addition to obtaining information through the information consultation service, users can also obtain the required information through personal retrieval. Due to the development of the modern network, a lot of information can be found on the network. Therefore, sometimes retrieval is more convenient and faster than consulting to obtain information. Evaluate the retrieval adaptation value and the service adaptation value of the target consultation information, compare the two values after normalization processing, and compare the sizes, so as to select a more suitable information acquisition method for the user. For example, when the retrieval adaptation value is not less than the service adaptation value, then the user can obtain information through his own retrieval, which is faster and more convenient than obtaining the required content through consultation. At this time, there is no need to push the information consultation service, which can reduce the user's boredom with the push on the one hand and improve the convenience of the user to obtain the required content on the other hand.

[0035] The steps of analyzing the user's retrieval ability based on the user's personal information, obtaining the retrieval difficulty of the target consultation information, and obtaining a retrieval adaptation value by combining the retrieval ability and the retrieval difficulty are as follows: Step S311, extract the historical retrieval information according to the user's personal information, and count the historical retrieval platforms used by the user according to the historical retrieval information.

[0036] Step S312: Divide the historical retrieval information to obtain historical retrieval data, and analyze the historical retrieval data to obtain the user's basic retrieval value.

[0037] Step S313: Evaluate the discrimination ability value of the user's discrimination information based on the historical retrieval data, and combine it with the basic retrieval value to obtain the user's retrieval ability value.

[0038] Step S314: Obtain the parameter data of the target consultation information, and evaluate the retrieval difficulty value of the target consultation information based on the parameter data.

[0039] Step S315: Set the proportionality coefficients of the retrieval ability value and the retrieval difficulty value respectively, and calculate the retrieval adaptation value of the target consultation information according to the proportionality coefficients.

[0040] In practical applications, the user's own retrieval ability is one of the key factors affecting the retrieval results. Users with strong retrieval ability can more accurately understand their information needs, select appropriate retrieval tools, and proficiently use various retrieval techniques to improve the accuracy and efficiency of retrieval. In addition, users with strong retrieval ability are also better at screening and evaluating retrieval results and can quickly find the most valuable information from them. Therefore, the improvement of the user's retrieval ability helps to obtain more accurate and comprehensive retrieval results. At the same time, the retrieval difficulty of information also has an important impact on the retrieval results. When information resources are scattered, disordered, and frequently updated, it is difficult for users to quickly locate the required information, resulting in low retrieval efficiency. In addition, if there are copyright restrictions or intellectual property issues in information resources, it may also bring additional obstacles to information retrieval. These factors will increase the retrieval difficulty and make it difficult for users to obtain satisfactory retrieval results. When the retrieval ability value is larger, it is easier for users to quickly retrieve the accurate required content, so the retrieval adaptation value is larger. When the retrieval difficulty is greater, it is more difficult for users to accurately and quickly find the required content, and the retrieval adaptation value is smaller. The larger the retrieval adaptation value, the more convenient it is for users to obtain information through retrieval in this scenario.

[0041] The steps of dividing the historical retrieval information to obtain historical retrieval data and analyzing the historical retrieval data to obtain the user's basic retrieval value are specifically as follows: Step S3121: Divide the historical retrieval information according to the retrieval time to obtain historical retrieval data of multiple retrievals.

[0042] Among the historical retrieval information, there is data of multiple retrievals. The continuous and uninterrupted retrieval information is recorded as one piece of historical retrieval data.

[0043] Step S3122: Extract the final retrieval content retrieved by the user from the historical retrieval data, and record the keywords of the final retrieval content as the final keywords.

[0044] Step S3123: Obtain the keywords initially retrieved by the user, denoted as the retrieval keywords, and compare to obtain the keyword similarity between the final keywords and the retrieval keywords.

[0045] Step S3124: Statistically calculate the average value of the keyword similarities for multiple retrievals, denoted as the retrieval accuracy.

[0046] During the retrieval process, users generally stop retrieving after obtaining the answers they want. Therefore, the final retrieval content refers to the content that the user views last in each retrieval. Extract keywords from the final retrieval content and compare them with the user's retrieval keywords. The larger the average value of the keyword similarity, the more accurate the user's retrieval expression. For example, if the user's retrieval keywords are Beijing, Guangdong, and price, and the keywords in the final retrieval content are Beijing, Shantou, high-speed rail, and price, this indicates that the user wants to know the high-speed rail price from Beijing to Shantou, Guangdong. The keyword expression description is not precise enough, so the retrieval accuracy is not very high. According to the user's retrieval keywords, a lot of irrelevant information will appear, increasing the difficulty for the user to obtain the required information.

[0047] Step S3125: Statistically calculate the average number of historical retrieval platforms used by the user in each retrieval based on historical retrieval data, denoted as the retrieval quantity.

[0048] Step S3126: Statistically calculate the number of information sources of the historical retrieval platforms, denoted as the number of information sources, and combine the retrieval accuracy and the retrieval quantity to obtain the basic retrieval value.

[0049] In practical applications, set the proportionality coefficients for the number of information sources, retrieval accuracy, and retrieval quantity respectively, and calculate the basic retrieval value according to the proportionality coefficients. The higher the user's retrieval accuracy, the stronger the user's retrieval ability, so the basic retrieval value is larger. When the number of platforms used by the user in each retrieval is larger, it means that the user's retrieval scope is wider, and the user is more proficient in using multiple retrieval platforms, which also reflects that the user's retrieval ability is stronger. And the larger the number of information sources, the wider the user's retrieval scope, which is more conducive to the user to find the required information, so the basic retrieval value is larger. For example, User A has a small number of information sources and fails to find Information A. While User B has a large number of information sources and retrieves Information A. Therefore, a larger number of information sources is conducive to retrieving some less numerous information and improving the user's retrieval ability.

[0050] The steps to obtain the user's retrieval ability value by evaluating the user's information discrimination ability value based on historical retrieval data and combining the basic retrieval value are specifically as follows: Step S3131: Extract the user's retrieval requirement content from historical retrieval data, and classify the retrieval requirement content into subjective content and objective content.

[0051] When users search, the search content can be divided into subjective and objective. For example, searching for personal evaluation or impressions of a movie is subjective content. Searching for specific specifications, performance parameters or price information of a product is objective content. We distinguish between subjective content and objective content based on whether the search content is mixed with subjective judgment or opinion.

[0052] Step S3132: extract historical search data of subjective content and record it as subjective search data, and extract the average search times of users in the subjective search data.

[0053] Step S3133, extracting the average amount of information of the user's reading retrieval requirement content in the subjective retrieval data, and extracting the average number of changes of the retrieval expression in the same retrieval in the subjective retrieval data.

[0054] Step S3134, combining the average number of retrievals, the average amount of information and the average number of changes, calculates the user's subjective discrimination value.

[0055] The proportional coefficients of the average number of retrievals, the average amount of information, and the average number of changes are set respectively, and the user's subjective discrimination value is calculated based on the proportional coefficients.

[0056] Step S3135, extract the historical search data of objective content and record it as objective search data, analyze the objective search data to obtain the objective discrimination value, and combine it with the subjective discrimination value to obtain the user's discrimination ability value.

[0057] The proportional coefficients of the objective discrimination value and the subjective discrimination value are respectively set, and the user's discrimination ability value is calculated according to the proportional coefficients.

[0058] Step S3136, respectively set the proportional coefficients of the identification capability value and the basic search value, and calculate the user's search capability value based on the proportional coefficients.

[0059] In actual use, when a user wants to retrieve the required content, on the one hand, it comes from the user's retrieval ability, and on the other hand, it comes from the user's discrimination ability. The greater the basic retrieval value and the discrimination ability value, the stronger the user's retrieval ability. For example, if a user wants to retrieve whether a certain brand of rice cooker has a reservation function and the retrieved results say yes for some and no for others, the user needs to have sufficient discrimination ability to confirm the accurate information. Moreover, the discrimination ability for subjective content and objective content is different. Since subjective content comes from human subjectivity, it is evaluated separately. In the retrieval of subjective content, some are real feelings while some are malicious evaluations. The user mainly needs to distinguish whether the subjective content is a malicious evaluation. The more times the user retrieves and the more the average amount of information in the retrieved required content, it means the more content the user views, and the more likely the user can find the data they want. And the more times the average change of the retrieval expression in the same retrieval, it means the user may retrieve from multiple perspectives, and the obtained information content is more comprehensive, which is more conducive to the user to screen out the sincere subjective content. For example, when a user searches for movie reviews of Movie A and sees a large number of positive reviews and decides to watch Movie A. However, if the large number of positive reviews are from paid posters and the user doesn't see a small number of negative reviews, it means the user's retrieval is not comprehensive enough, so the obtained information is relatively one-sided and the discrimination ability is poor.

[0060] The historical retrieval data of objective content is recorded as objective retrieval data. The steps to analyze the objective retrieval data to obtain the objective discrimination value are specifically as follows: Step S31351: Determine whether there is a clear answer to the user's retrieval required content through retrieval. Record the objective retrieval data with a clear answer as clear retrieval data, and record the objective retrieval data without a clear answer as non-clear retrieval data.

[0061] Use a search engine for retrieval and observe the quantity and quality of the search results. If there are numerous search results and they contain clear answers from authoritative sources, it is determined that there is a clear answer to the user's retrieval requirement. If the search results are few, the content is vague or there are disputes, it is determined that it is difficult to find a clear answer. For example, there is a clear answer to what is 1 + 1. And the cause of abdominal pain is also objective retrieval content, but since the conditions given are insufficient and there are many involved factors, no clear answer can be given.

[0062] Step S31352: Extract the correct answer content from the clear retrieval data and count the average frequency of the user retrieving the correct answer content as the first discrimination value.

[0063] If there is a clear answer in the retrieved data, the average frequency of the user retrieving the correct answer content is used as the first discrimination value. For example, for a certain math problem with an accurate answer, but there is a lot of information on the Internet. If a netizen gives an answer through a post question, the answer may be correct or wrong. Even for target consultation information with a clear and accurate answer, there is likely to be a lot of wrong information on the Internet. Therefore, the higher the average frequency of the user retrieving the correct answer content, the larger the first discrimination value.

[0064] Step S31353, extract the average correlation degree between the retrieved expression adjusted by the user in the non-clear retrieved data and the previous retrieved content as the second discrimination value.

[0065] For some non-clear retrieved content, the user needs to adjust it by himself to obtain more accurate information content. For example, when the user wants to know the cause of abdominal pain, the more conditions are given, the more accurate the answer will be. When the user adjusts the retrieved expression according to the previous retrieved content, the second discrimination value will be larger. For example, if the previous retrieved content states that abdominal pain may be due to eating cold food, then the user adjusts the retrieved expression to "the cause of abdominal pain without eating cold food". The higher the average correlation degree between the retrieved expression adjusted by the user and the previous retrieved content, the more the user can adjust the retrieved expression according to the retrieval situation to reduce the interference of wrong information, reflecting that the user has a stronger ability to distinguish information.

[0066] Step S31354, calculate the average value of the first discrimination value and the second discrimination value to obtain the objective discrimination value.

[0067] In actual application, for distinguishing objective information, on the one hand, it is necessary to reduce the acquisition of wrong information, and on the other hand, it is necessary to think about the information to improve the accuracy of information acquisition. The user can adjust the retrieved expression according to the retrieved content obtained, indicating that the user thinks about the retrieved content and readjusts the expression to exclude the wrong information that has appeared, indicating that the user has a stronger ability to distinguish objective content.

[0068] The steps to obtain the parameter data of the target consultation information and evaluate the retrieval difficulty value of the target consultation information according to the parameter data are specifically as follows: Step S3141, obtain the keywords of the target consultation information and record them as target keywords, and search for the information containing the target keywords in the historical retrieval platform and record it as target information.

[0069] Step S3142, count the amount of information of the target information and count the number of types of the target information.

[0070] Step S3143, respectively set the proportionality coefficients of the amount of information and the number of types of the target information, and calculate the retrieval difficulty value of the target consultation information according to the proportionality coefficients.

[0071] In actual use, whether users can retrieve accurate required content is also related to the retrieval difficulty. For example, if there is no relevant information about content A to be retrieved in the retrieval platforms used by the users in the past, even if the users have strong retrieval capabilities, it is difficult to retrieve accurate information. Among them, the less the target information volume is, the more difficult it is for users to find the corresponding information content, indicating that the retrieval difficulty is greater. And the more types of target information there are, it means that it is more difficult for users to screen out accurate information, so the retrieval difficulty is even greater. The number of types of target information refers to the number of different contents. For example, when a user retrieves the answer to a multiple-choice question and retrieves that some people choose A and some choose B, then the number of types is 2. If there is also an option of C, then the number of types is 3.

[0072] The steps of obtaining the adaptation value of the corresponding service according to the service information of the corresponding service and selecting the maximum adaptation value of the corresponding service as the service adaptation value are specifically as follows: Step S331: Find the information type corresponding to the target consultation information, and judge whether the user understands the information type according to the user's personal information.

[0073] The target consultation information has a corresponding information type. For example, if consulting about the bank deposit rate, it belongs to the financial information type, and it can be judged whether there is contact with the corresponding information type according to the user's educational background information, work information, etc.

[0074] Step S332: If the user does not understand the information type, extract the average understanding duration of the information type as the service duration.

[0075] If the user does not understand the information type, then count the existing preliminary entry duration, estimated understanding duration, etc. as the service duration, or the average consultation duration of customers who do not understand this information type extracted from historical consultation information can also be used as the service duration. Because the user does not understand the information type, the service staff needs to further explain the basic information of this type of information, which will increase the service duration. For example, if the user does not know what the bank deposit rate is, the service staff needs to explain the meaning of the bank deposit rate, and use the average understanding duration of understanding the bank deposit rate as the service duration.

[0076] Step S333: If the user understands the information type, extract the average consultation duration of the corresponding service as the service duration.

[0077] Step S334: Extract the average queuing duration of the corresponding service and extract the average satisfaction degree of the corresponding service feedback.

[0078] Step S335: Calculate the adaptation value of the corresponding service by combining the service duration, average queuing duration and average satisfaction degree, and select the maximum adaptation value of the corresponding service as the service adaptation value.

[0079] In actual application, the proportionality coefficients of service duration, average queuing duration, and average satisfaction are set respectively, and the adaptation value of the corresponding service is calculated according to the proportionality coefficients. When the service duration is longer, the average queuing duration is longer, and the average satisfaction is lower, the adaptation value is lower. For example, if the average queuing duration is 1 hour, it is obvious that the convenience of user consultation service is reduced, so the adaptation degree with users is lower.

[0080] If the user knows the information type, the step of extracting the average consultation duration of the corresponding service as the service duration is as follows: Step S3331: Obtain the historical consultation data of the corresponding service, set the duration time period, and respectively count the average consultation times of the same user within different duration time periods.

[0081] The consultation durations of different users are different. For example, some users consult for ten minutes, and some users consult for 30 minutes. The duration time period can be set. For example, 0 - 10 minutes is a duration time period, and 10 - 20 minutes is a duration time period.

[0082] Step S3332: Select the duration time period with the largest average consultation times of the same user as the understanding duration period.

[0083] For example, User A consults 3 times, and User B consults 5 times, and the durations are all within 0 - 10 minutes. Then within the duration time period of 0 - 10 minutes, the average consultation duration of the same user is 4 times. User C consults 1 time, and User D consults 2 times, and the durations are all within 10 - 20 minutes. Then within the duration time period of 10 - 20 minutes, the average consultation duration of the same user is 1.5 times.

[0084] Step S3333: Calculate the average consultation duration of users within the understanding duration period as the service duration of the corresponding service.

[0085] In actual application, when the number of times a user consults is more, it is considered that the user is more familiar with the information type, and the corresponding average consultation duration is used as the service duration of the user who is familiar with this type. For example, within 0 - 10 minutes, the average consultation times of the same user are the largest. The consultation durations of User A are 4 minutes, 6 minutes, and 3 minutes respectively, and the consultation durations of User B are 3 minutes, 5 minutes, 4 minutes, and 3 minutes respectively. Then the average consultation duration each time is 4 minutes, and 4 minutes is used as the service duration of the corresponding service.

[0086] If information consultation service is to be pushed, the step of setting the push plan according to the user's personal information and pushing the information consultation service according to the push plan is as follows: Step S41: Extract the platform usage rate of the user according to the user's personal information, and select the platform with the highest platform usage rate as the push platform.

[0087] Step S42: Extract the user's online pattern based on the user's personal information, and generate a push time according to the online pattern.

[0088] Step S43: Form a push plan based on the push time and the push platform, and push the information consultation service with the highest service adaptation value according to the push plan.

[0089] In actual application, obtain the platform most frequently used by the user as the push platform, set the push time according to the user's online pattern on this platform, and increase the probability that the user sees the push. For example, if the user most frequently uses Platform A and is generally online after 8 pm, and push the information consultation service with the highest service adaptation degree on Platform A after point B, it is beneficial for the user to see the push information, and the service adaptation degree is higher, which is beneficial for the user to obtain a better consultation experience.

[0090] The above are all the preferred embodiments of this application. The protection scope of this application is not limited by this. Therefore, all equivalent changes made according to the structure, shape, and principle of this application should be covered within the protection scope of this application.

Claims

1. An intelligent push method for information consulting services, characterized in that: The following steps are involved: Collect the user's search information, and extract the user's target consulting information based on the search information; Searching the Internet for any content related to the target consulting information, and if so, collecting the user's personal information; Collect service information of information consulting services and determine whether to push information consulting services based on user personal information; If information consulting services are pushed, a push plan will be set based on the user's personal information, and the information consulting services will be pushed according to the push plan; If there is no relevant content of the target consulting information, the user's personal information is collected, a push plan is generated based on the personal information, and the information consulting service is pushed according to the push plan.

2. The intelligent push method for information consulting service according to claim 1, characterized in that: The steps of collecting the service information of the information consulting service and determining whether to push the information consulting service in combination with the user's personal information are specifically as follows: The user's retrieval ability is obtained based on the user's personal information analysis, the retrieval difficulty of the target consulting information is obtained, and the retrieval adaptation value is obtained by combining the retrieval ability and the retrieval difficulty; The information consulting service corresponding to the matched target consulting information is recorded as the corresponding service, and the service information of the corresponding service is collected; Evaluate the adaptation value of the corresponding service according to the service information of the corresponding service, and select the largest adaptation value of the corresponding service as the service adaptation value; Comparing the search adaptation value with the service adaptation value, if the search adaptation value is not less than the service adaptation value, it is determined that the information consulting service will not be pushed; If the retrieval adaptation value is less than the service adaptation value, it is determined that the information consulting service is pushed.

3. The intelligent push method for information consulting service according to claim 2, characterized in that: The steps of obtaining the user's retrieval ability according to the user's personal information analysis, obtaining the retrieval difficulty of the target consulting information, and obtaining the retrieval adaptation value by combining the retrieval ability and the retrieval difficulty are specifically as follows: Extract historical search information based on user personal information, and count historical search platforms used by users based on the historical search information; Divide the historical search information to obtain historical search data, and obtain the user's basic search value based on the historical search data analysis; The identification capability value of the user identification information is obtained based on the historical search data evaluation, and combined with the basic search value, the user's search capability value is obtained; Obtaining parameter data of target consulting information, and obtaining a retrieval difficulty value of the target consulting information based on the parameter data evaluation; The proportional coefficients of the retrieval capability value and the retrieval difficulty value are set respectively, and the retrieval adaptation value of the target consulting information is calculated according to the proportional coefficients.

4. The intelligent push method for information consulting service according to claim 3, characterized in that: The steps of dividing the historical search information to obtain historical search data and obtaining the user's basic search value according to the historical search data analysis are specifically as follows: Divide the historical search information according to the search time to obtain historical search data of multiple searches; Extract the final search content of the user's search from the historical search data, and record the keywords of the final search content as the final keywords; Obtain the keyword that the user initially searches for as the search keyword, and compare and obtain the keyword similarity between the final keyword and the search keyword; The average value of keyword similarities of multiple searches is calculated and recorded as search accuracy; According to the historical search data, the average number of historical search platforms used by the user for each search is recorded as the number of searches; The number of information sources of the historical retrieval platform is recorded as the number of information sources, and the basic retrieval value is obtained by combining the retrieval accuracy and the number of retrievals.

5. The intelligent push method for information consulting service according to claim 4, characterized in that: The step of obtaining the identification capability value of the user identification information according to the historical search data evaluation and combining it with the basic search value to obtain the user's search capability value is specifically as follows: Extract the user's search demand content based on historical search data, and divide the search demand content into subjective content and objective content; The historical search data of the extracted subjective content is recorded as subjective search data, and the average search times of users in the subjective search data are extracted; Extract the average amount of information that users read in the search demand content from the subjective search data, and extract the average number of changes in the search expression in the same search from the subjective search data; The user's subjective discrimination value is calculated by combining the average number of retrievals, the average amount of information, and the average number of changes; The historical search data of objective content is recorded as objective search data, and the objective search data is analyzed to obtain the objective discrimination value, which is combined with the subjective discrimination value to obtain the user's discrimination ability value; The proportional coefficients of the identification ability value and the basic search value are set respectively, and the user's search ability value is calculated according to the proportional coefficients.

6. The intelligent push method for information consulting service according to claim 5, characterized in that: The historical search data for extracting objective content is recorded as objective search data, and the step of analyzing the objective search data to obtain the objective resolution value is specifically as follows: Determine whether the user's search demand content has a clear answer through the search, record the objective search data with a clear answer as clear search data, and record the objective search data without a clear answer as non-clear search data; Extract the correct answer content from the explicit search data, and count the average frequency of users retrieving the correct answer content as the first resolution value; Extracting the average relevance between the search expression adjusted by the user and the previous search content in the non-explicit search data as the second resolution value; The average of the first resolution value and the second resolution value is calculated to obtain the objective resolution value.

7. The intelligent push method for information consulting service according to claim 3, characterized in that: The step of obtaining parameter data of the target consulting information and evaluating the retrieval difficulty value of the target consulting information according to the parameter data is specifically as follows: The keywords for obtaining the target consulting information are recorded as the target keywords, and the information containing the target keywords found in the historical search platform is recorded as the target information; Count the amount of target information and the number of types of target information; The proportional coefficients of the information volume and the number of types of the target information are set respectively, and the retrieval difficulty value of the target consulting information is calculated according to the proportional coefficients.

8. The intelligent push method for information consulting service according to claim 2, characterized in that: The step of obtaining the adaptation value of the corresponding service according to the service information of the corresponding service and selecting the largest adaptation value of the corresponding service as the service adaptation value is specifically as follows: Find the information type corresponding to the target consultation information, and determine whether the user understands the information type based on the user's personal information; If the user does not understand the information type, the average understanding time of the information type is extracted as the service time; If the user knows the information type, the average consultation time of the corresponding service is extracted as the service time; Extract the average queuing time of the corresponding service and the average satisfaction of the corresponding service feedback; The adaptation value of the corresponding service is calculated by combining the service time, average queuing time and average satisfaction, and the largest adaptation value of the corresponding service is selected as the service adaptation value.

9. The intelligent push method for information consulting service according to claim 8, characterized in that: If the user knows the information type, the step of extracting the average consultation duration of the corresponding service as the service duration is specifically as follows: Get the historical consultation data of the corresponding service, set the time period, and count the average number of consultations for the same user in different time periods; Select the time period with the largest average number of consultations for the same user as the learning time period; Calculate the average consultation duration of users within the understanding time period as the service duration of the corresponding service.

10. The intelligent push method for information consulting service according to claim 2, characterized in that: If the information consulting service is pushed, a push plan is set according to the user's personal information, and the steps of pushing the information consulting service according to the push plan are specifically as follows: Extract the user's platform usage rate based on the user's personal information, and select the platform with the highest platform usage rate as the push platform; Extract the user's online pattern based on the user's personal information, and generate the push time based on the online pattern; A push plan is formulated based on the push time and push platform, and the information consulting service with the highest service adaptation value is pushed according to the push plan.

Images