阅读说明：本技术 一种自动推荐客户服务策略的系统和方法 (System and method for automatically recommending customer service strategy ) 是由 龙丹 许睿 刘佳 许蕾 禹汪宏 于 2021-06-10 设计创作，主要内容包括：本发明涉及客户服务技术领域,且公开了一种自动推荐客户服务策略的系统和方法,包括处理器,处理器的输入端与输入单元的输出端电连接,处理器的输出端与输出模块的输入端电连接,处理器与推荐客户服务单元双向电连接,该自动推荐客户服务策略的方法,通过分析单元、联网检索单元和生成单元的配合使用,且生成单元能够对网络上自动推荐服务使用点击率,及时转化为数据完成分析反馈至中央处理器,方便工作人员了解对该系统在使用中存在的不足情况,通过垃圾信息模块的设置,能够对检测出的垃圾信息通过反馈模块传送至中央处理器,在得到最终检索信息时,能够得到人性化的解决方法,更提高了该系统在使用中的检测效率。(The invention relates to the technical field of customer service, and discloses a system and a method for automatically recommending customer service strategies, which comprises a processor, wherein the input end of the processor is electrically connected with the output end of an input unit, the output end of the processor is electrically connected with the input end of an output module, and the processor is electrically connected with a recommended customer service unit in a bidirectional way, the method for automatically recommending the customer service strategies is realized by the cooperation of an analysis unit, a network searching unit and a generation unit, the generation unit can use click rate for automatically recommending services on a network, timely converts the click rate into data, completes analysis and feeds back the data to a central processing unit, and is convenient for workers to know the defects existing in the use of the system, the detected junk information can be transmitted to the central processing unit through a feedback module through the arrangement of a junk information module, and a humanized solution can be obtained when final searching information is obtained, the detection efficiency of the system in use is improved.)

1. A system for automatically recommending customer service policies, characterized by: the system comprises a processor, wherein the processor is connected with a recommended client service unit, an analysis unit, a command generation unit and a junk information processing unit, and a retrieval unit is connected with the analysis unit, the generation unit and the junk information processing unit;

the command generation unit disassembles the command task issued by analysis and matches the command generation work of the instruction set based on the service keywords input by the retrieval unit;

the analysis unit is used for converting the click rate of the automatic recommendation service on the network into data in time and feeding the data back to the processor;

the garbage information processing unit comprises a feedback module and a deletion screening unit, the deletion screening unit carries out secondary cleaning on the recommended service information retrieved by the retrieval unit and transmits the detected garbage information to the processor through the feedback module;

the recommendation client service unit comprises an intelligent recommendation unit and a service information display unit, wherein the intelligent recommendation unit analyzes the most interesting service of the client by adopting a K-means algorithm according to the historical ordering condition of the client.

2. The system for automatically recommending customer service policies according to claim 1, wherein: the process of performing the secondary cleaning is as follows:

the deleting and screening unit records and stores a plurality of pieces of recommended information every time, analyzes a piece of recommended repeated information which is successfully recommended and recommended for more than 3 times, defines the information as junk information if no clicking action is performed on the recommended information, deletes the screening unit, and records the information in a junk record database; the next time the above recommendation appears, the direct deletion is not analyzed.

3. The system for automatically recommending customer service policies according to claim 1, wherein: the process of analyzing the services of most interest to the customer is as follows:

step (1) selecting characteristic factors of ordering under user interest, wherein the characteristic factors comprise a user name, account information, a user portrait characteristic label and ordering information of customer service, and preprocessing data:

1.1 historical data of the service condition of placing the order, and if the month in which the data is analyzed is N, extracting the times of placing the order of the specific user in N-4 month, N-3 month, N-2 month and N-1 month, namely the previous 4 months;

1.2 weighting historical order-placing amount data, if the month in which the data is analyzed is N, extracting the sum of the amounts of orders placed by specific users in N-4 month, N-3 month, N-2 month and N-1 month, namely the previous 4 months, calculating the sum, judging the demand degree of ordering by each service, and weighting according to the demand degree;

discretizing the sample, and dividing into 4 grades according to the size, wherein the number of each grade is recorded as a group;

and (3) processing data by adopting a K-means algorithm.

4. The system for automatically recommending customer service policies of claim 3, wherein: in the step (3), the data processing by adopting the K-means algorithm is as follows:

from the resulting data set, the given data was analyzed: the algorithm develops clustering algorithm mining analysis aiming at a given sample set Q:

3.1 randomly selecting k points from the sample set Q within the Q data range as respective centers of k groups;

3.2 respectively calculating the dissimilarity degree of the points in the sample set Q to the center of the k groups, and classifying the dissimilarity degrees into the group with the lowest dissimilarity degree; the dissimilarity degree between two points is measured by an Euclidean distance formula, and the dissimilarity degree is measured by the following steps:

d＝sqrt((X₁-X₂)²+(Y₁-Y₂)²；

d is the Euclidean distance between two points T0(x1, y2) and T1(x2, y2), each point having two dimensions, namely a characteristic factor; the smaller the Euclidean distance is, the smaller the dissimilarity degree is;

3.3 according to the clustering result, recalculating the respective centers of the k groups by taking the arithmetic mean of the respective dimensions of all points in the groups;

3.4 clustering all the points in the sample set Q again according to the new centers;

3.5 repeating the step 3.4 until the clustering result is not changed;

3.6 outputting the user interest clustering objects according to k groups;

3.7 steps 3.1-3.6 are performed using k-8, k-4, k-2, respectively;

calculating a result through the algorithm, and finally obtaining the user interest evaluation of the whole sample set, wherein the evaluation is used as intelligent recommendation to recommend services with times and money similar to the algorithm result in the existing services;

the method comprises the steps that the name of a person user and account information of characteristic factors are used as indexes for algorithm analysis, SQL sentences are adopted for the indexes to call and analyze historical data of target data; the user portrait feature label is used for recommending in a further limited range according to the gender, age and the like of a user in intelligent recommendation as much as possible, so that the recommendation deviation is avoided, and client service ordering information is core analysis data of the intelligent recommendation algorithm.

5. A method for automatically recommending customer service policies, characterized by: the method comprises the following steps:

inputting keywords of related services;

resolving the command task issued by analysis based on the service keywords input by the retrieval unit, and matching the command generation work of the command set;

the click rate of the automatic recommendation service on the network is converted into data in time and fed back to the processor;

and performing secondary cleaning on the recommended service information retrieved by the retrieval unit, and transmitting the detected spam information to the processor through the feedback module.

6. The system of claim 5, wherein: the automatic recommendation service proceeds as follows:

selecting characteristic factors of ordering of user interest, wherein the characteristic factors comprise a user name, account information, a user portrait characteristic label and ordering information of customer service, and preprocessing data:

historical data of the service condition of placing the order, and if the month in which the data are analyzed is N, extracting the times of placing the order of the specific user in the (N-4) month, (N-3) month, (N-2) month, (N-1) month, namely the previous 4 months;

weighting historical ordering amount data, if the month in which the data is analyzed is N, extracting the sum of the amounts of ordering of specific users in N-4 month, N-3 month, N-2 month and N-1 month, namely the previous 4 months, calculating the sum, judging the demand degree of ordering of each service, and weighting according to the demand degree;

then carrying out discretization treatment on the sample, and dividing 4 grades according to the size, wherein the quantity of each grade is recorded as a group; the data was processed using the K-means algorithm.

7. The system for automatically recommending customer service policies of claim 6, wherein: in the step (3), the data processing by adopting the K-means algorithm is as follows:

(1) from the resulting data set, the given data was analyzed: the algorithm develops clustering algorithm mining analysis aiming at a given sample set Q:

(2) randomly selecting k points from a sample set Q within a Q data range as respective centers of k groups;

(3) respectively calculating the dissimilarity degree from the point in the sample set Q to the center of the k groups, and classifying the dissimilarity degree into the group with the lowest dissimilarity degree; the dissimilarity degree between two points is measured by an Euclidean distance formula, and the dissimilarity degree is measured by the following steps:

a＝sqrt((X₁-X₂)²+(Y₁-Y₂)²；

d is the Euclidean distance between two points T0(x1, y2) and T1(x2, y2), each point having two dimensions, namely a characteristic factor; the smaller the Euclidean distance is, the smaller the dissimilarity degree is;

according to the clustering result, re-calculating the respective centers of the k groups by taking the arithmetic mean of the respective dimensions of all points in the groups;

(4) re-clustering all points in the sample set Q according to the new center;

(5) repeating the step 3.4 until the clustering result is not changed;

(6) outputting the user interest clustering objects according to k groups;

(7) steps 3.1-3.6 are performed using k-8, k-4, k-2, respectively;

calculating a result through the algorithm, and finally obtaining the user interest evaluation of the whole sample set, wherein the evaluation is used as intelligent recommendation to recommend services with times and money similar to the algorithm result in the existing services;

the method comprises the steps that the name of a person user and account information of characteristic factors are used as indexes for algorithm analysis, SQL sentences are adopted for the indexes to call and analyze historical data of target data; the user portrait feature label is used for recommending in a further limited range according to the gender, age and the like of a user in intelligent recommendation as much as possible, so that the recommendation deviation is avoided, and client service ordering information is core analysis data of the intelligent recommendation algorithm.

8. The system for automatically recommending customer service policies of claim 6, wherein:

the deleting and screening unit records and stores a plurality of pieces of recommended information every time, analyzes a piece of recommended repeated information which is successfully recommended and recommended for more than 3 times, defines the information as junk information if no clicking action is performed on the recommended information, deletes the screening unit, and records the information in a junk record database; the next time the above recommendation appears, the direct deletion is not analyzed.

9. An electronic device comprising a memory, a processor, and a computer program that is executable on the memory and on the processor, wherein: the processor, when executing the computer program, performs the steps of the method of any of the preceding claims 5 to 8.

10. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that: the computer program, when executed by a processor, implementing the steps of the method as claimed in any one of claims 5 to 8.

Technical Field

The invention relates to the technical field of customer service, in particular to a system and a method for automatically recommending a customer service strategy.

Background

A web site is one of the most important channels for a business to attract potential customers, sell products or services, and communicate with customers, and for customer service operations, various information, including billing information, promotional information, and technical support, may be provided to customers through a customer service web site, customers browse the customer service site for useful information, and may spend some time answering questions to guide customers on the customer service web site.

However, even after a lot of time is spent on exploring the customer service website, the customers may not find the information they are looking for, and therefore, the arrangement work of the strategy information by the staff is still needed, and an attempt may be made to contact the customer service agent or representative to solve their problems, so that there are disadvantages that the staff is not easy to make data analysis and understanding on the viewing records of different service information on the website for the customers using the website, so as to optimize the strategy requirement on the services required by the customers, and the website targeted matching information also has the condition of outsourcing junk information, which is easy to cause confusion about the service information after screening, is not beneficial to the convenient arrangement work of the information in the later period, how to enable the customers to quickly and efficiently obtain the effective information, and fully mine the value in obtaining the effective information, providing more efficient and valuable information push service for users becomes another problem to be urgently researched and solved.

Disclosure of Invention

In view of the deficiencies of the prior art, the present invention provides a system and method for automatically recommending customer service policies. In order to achieve the purpose, the invention provides the following technical scheme:

a system for automatically recommending customer service policies, characterized by: the system comprises a processor, wherein the processor is connected with a recommended client service unit, an analysis unit, a command generation unit and a junk information processing unit, and a retrieval unit is connected with the analysis unit, the generation unit and the junk information processing unit;

the command generation unit disassembles the command task issued by analysis and matches the command generation work of the instruction set based on the service keywords input by the retrieval unit;

the analysis unit is used for converting the click rate of the automatic recommendation service on the network into data in time and feeding the data back to the processor;

the garbage information processing unit comprises a feedback module and a deletion screening unit, the deletion screening unit carries out secondary cleaning on the recommended service information retrieved by the retrieval unit and transmits the detected garbage information to the processor through the feedback module;

the recommendation client service unit comprises an intelligent recommendation unit and a service information display unit, wherein the intelligent recommendation unit analyzes the most interesting service of the client by adopting a K-means algorithm according to the historical ordering condition of the client.

Further, the process of performing the secondary cleaning is as follows:

the deleting and screening unit records and stores a plurality of pieces of recommended information every time, analyzes a piece of recommended repeated information which is successfully recommended and recommended for more than 3 times, defines the information as junk information if no clicking action is performed on the recommended information, deletes the screening unit, and records the information in a junk record database; the next time the above recommendation appears, the direct deletion is not analyzed.

That is, after receiving data, firstly analyzing whether the record is in a garbage record database, directly deleting the record, then processing the record by looking at the data which is repeatedly recommended for more than 3 times, and finally, if only one record is left, recommending the record, and if a plurality of records are left, extracting the record with the best user interest degree for recommendation, and deleting the other records.

Further, the process of analyzing the services of most interest to the customer is as follows:

step (1) selecting characteristic factors of ordering under user interest, wherein the characteristic factors comprise a user name, account information, a user portrait characteristic label and ordering information of customer service, and preprocessing data:

1.1 historical data of the service condition of placing the order, and if the month in which the data is analyzed is N, extracting the times of placing the order of the specific user in N-4 month, N-3 month, N-2 month and N-1 month, namely the previous 4 months;

1.2 weighting historical order-placing amount data, if the month in which the data is analyzed is N, extracting the sum of the amounts of orders placed by specific users in N-4 month, N-3 month, N-2 month and N-1 month, namely the previous 4 months, calculating the sum, judging the demand degree of ordering by each service, and weighting according to the demand degree;

discretizing the sample, and dividing into 4 grades according to the size, wherein the number of each grade is recorded as a group;

and (3) processing data by adopting a K-means algorithm.

Further, in the step (3), the data processing by using the K-means algorithm is as follows:

from the resulting data set, the given data was analyzed: the algorithm develops clustering algorithm mining analysis aiming at a given sample set Q:

3.1 randomly selecting k points from the sample set Q within the Q data range as respective centers of k groups;

3.2 respectively calculating the dissimilarity degree of the points in the sample set Q to the center of the k groups, and classifying the dissimilarity degrees into the group with the lowest dissimilarity degree; the dissimilarity degree between two points is measured by an Euclidean distance formula, and the dissimilarity degree is measured by the following steps:

d＝sqrt((X₁-V₂)²+(Y₁-Y₂)²；

d is the Euclidean distance between two points T0(x1, y2) and T1(x2, y2), each point having two dimensions, namely a characteristic factor; the smaller the Euclidean distance is, the smaller the dissimilarity degree is;

3.3 according to the clustering result, recalculating the respective centers of the k groups by taking the arithmetic mean of the respective dimensions of all points in the groups;

3.4 clustering all the points in the sample set Q again according to the new centers;

3.5 repeating the step 3.4 until the clustering result is not changed;

3.6 outputting the user interest clustering objects according to k groups;

3.7 steps 3.1-3.6 are performed using k-8, k-4, k-2, respectively;

calculating a result through the algorithm, and finally obtaining the user interest evaluation of the whole sample set, wherein the evaluation is used as intelligent recommendation to recommend services with times and money similar to the algorithm result in the existing services;

the method comprises the steps that the name of a person user and account information of characteristic factors are used as indexes for algorithm analysis, SQL sentences are adopted for the indexes to call and analyze historical data of target data; the user portrait feature label is used for recommending in a further limited range according to the gender, age and the like of a user in intelligent recommendation as much as possible, so that the recommendation deviation is avoided, and client service ordering information is core analysis data of the intelligent recommendation algorithm.

The invention also relates to a method for automatically recommending the customer service strategy, which is carried out as follows:

inputting keywords of related services;

resolving the command task issued by analysis based on the service keywords input by the retrieval unit, and matching the command generation work of the command set;

the click rate of the automatic recommendation service on the network is converted into data in time and fed back to the processor;

and performing secondary cleaning on the recommended service information retrieved by the retrieval unit, and transmitting the detected spam information to the processor through the feedback module.

Further, the automatic recommendation service proceeds as follows:

selecting characteristic factors of ordering of user interest, wherein the characteristic factors comprise a user name, account information, a user portrait characteristic label and ordering information of customer service, and preprocessing data:

historical data of the service condition of placing the order, and if the month in which the data are analyzed is N, extracting the times of placing the order of the specific user in the (N-4) month, (N-3) month, (N-2) month, (N-1) month, namely the previous 4 months;

weighting historical ordering amount data, if the month in which the data is analyzed is N, extracting the sum of the amounts of ordering of specific users in N-4 month, N-3 month, N-2 month and N-1 month, namely the previous 4 months, calculating the sum, judging the demand degree of ordering of each service, and weighting according to the demand degree;

then carrying out discretization treatment on the sample, and dividing 4 grades according to the size, wherein the quantity of each grade is recorded as a group; the data was processed using the K-means algorithm.

Further, in the step (3), the data processing by using the K-means algorithm is as follows:

(1) from the resulting data set, the given data was analyzed: the algorithm develops clustering algorithm mining analysis aiming at a given sample set Q:

(2) randomly selecting k points from a sample set Q within a Q data range as respective centers of k groups;

(3) respectively calculating the dissimilarity degree from the point in the sample set Q to the center of the k groups, and classifying the dissimilarity degree into the group with the lowest dissimilarity degree; the dissimilarity degree between two points is measured by an Euclidean distance formula, and the dissimilarity degree is measured by the following steps:

d＝sqrt((X₁-X₂)²+(Y₁-Y₂)²；

d is the Euclidean distance between two points T0(x1, y2) and T1(x2, y2), each point having two dimensions, namely a characteristic factor; the smaller the Euclidean distance is, the smaller the dissimilarity degree is;

according to the clustering result, re-calculating the respective centers of the k groups by taking the arithmetic mean of the respective dimensions of all points in the groups;

(4) re-clustering all points in the sample set Q according to the new center;

(5) repeating the step 3.4 until the clustering result is not changed;

(6) outputting the user interest clustering objects according to k groups;

(7) steps 3.1-3.6 are performed using k-8, k-4, k-2, respectively;

calculating a result through the algorithm, and finally obtaining the user interest evaluation of the whole sample set, wherein the evaluation is used as intelligent recommendation to recommend services with times and money similar to the algorithm result in the existing services;

the method comprises the steps that the name of a person user and account information of characteristic factors are used as indexes for algorithm analysis, SQL sentences are adopted for the indexes to call and analyze historical data of target data; the user portrait feature label is used for recommending in a further limited range according to the gender, age and the like of a user in intelligent recommendation as much as possible, so that the recommendation deviation is avoided, and client service ordering information is core analysis data of the intelligent recommendation algorithm.

Further, the deleting and screening unit records and stores a plurality of pieces of recommended information each time, analyzes a piece of recommended repeated information which is successfully recommended and recommended for more than 3 times, defines the recommended repeated information as junk information if no clicking action is performed on the recommended information, deletes the screening unit, and records the deleted screening unit in a junk record database; the next time the above recommendation appears, the direct deletion is not analyzed.

An electronic device comprising a memory, a processor, and a computer program that is executable on the memory and on the processor, wherein: the processor realizes the steps of the above method when executing the computer program.

A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that: the computer program, when executed by a processor, performs the steps of the above-described method.

Compared with the prior art, the invention has the following beneficial effects:

1. according to the invention, through the matching use of the analysis unit, the networking retrieval unit and the generation unit, a client inputs keywords of related services on a website, so that the analysis unit can provide service information with higher matching degree for the client, the manual retrieval condition of the client or a worker for the service information is improved, the automatic recommendation effect of the system for the client is improved, the generation unit can use click rate for the automatic recommendation service on the network, the click rate is converted into data in time to complete analysis, and the data is fed back to the central processing unit, so that the worker can conveniently know the shortage condition of the system in use.

2. According to the invention, through the arrangement of the junk information module, the feedback module and the deleting and screening unit in the junk information module, the recommended service information searched out through networking can be cleaned for the second time, the detected junk information can be transmitted to the central processing unit through the feedback module, and when the information of the recommended service is automatically searched, the selection of the information content is selected by a worker, so that the worker can obtain a humanized solution when obtaining the final searched information, and the detection efficiency of the system in use is further improved.

Drawings

FIG. 1 is a block diagram of the system of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

As shown in fig. 1, the system for automatically recommending a customer service policy of the present embodiment includes a central processing unit, and an input end of the central processing unit is electrically connected to an output end of an input unit.

The input unit can include the keyboard, infrared camera and mouse etc, the output of keyboard is connected with central processing unit's input electricity, infrared camera's output is connected with central processing unit's input electricity, the output of mouse is connected with central processing unit's input electricity, central processing unit's output is connected with output module's input electricity, central processing unit is connected with recommendation customer service unit two-way electricity, recommendation customer service unit includes intelligent recommendation module and service information display module, central processing unit is connected with the two-way electricity of command generation unit, the command generation unit disassembles the order task of analysis assignment, the order of matching instruction set generates work.

The output end of the command generation unit is electrically connected with the input end of the networking retrieval unit, the output end of the networking retrieval unit is connected with the input end of the command generation unit, and the output end of the networking retrieval unit is connected with the input end of the spam module.

The analysis unit comprises a computer history comparison module, a data analysis module and a reference value module. The history comparison module compares the recommended information with the history library and checks whether the record is recommended or not; the data analysis module analyzes whether the recommendation information has click action or not; and the reference value module analyzes the recommendation times of the recommendation data to determine whether the recommendation times exceed 3 times. The analysis unit cooperates with the deletion screening unit.

The output end of the computer history comparison module is electrically connected with the input end of the central processing unit, the output end of the data analysis module is electrically connected with the input end of the central processing unit, the output end of the reference value module is electrically connected with the input end of the central processing unit, the garbage information module is electrically connected with the central processing unit in a bidirectional mode, the garbage information module comprises a feedback module and a deletion screening unit, the feedback module is electrically connected with the central processing unit in a bidirectional mode, the deletion screening unit is electrically connected with the central processing unit in a bidirectional mode, the output end of the garbage information module is electrically connected with the input end of the transmission module, the transmission module is electrically connected with the safety firewall in a bidirectional mode, and the safety firewall is electrically connected with the storage unit in a bidirectional mode.

The storage unit comprises a classification processing module, an encryption conversion module and a decryption conversion module. The classification processing module classifies the data to be stored, classifies and stores the data; the encryption conversion module encrypts data recommended to be sent to the webpage end; and simultaneously, the encrypted information returned by the webpage is decrypted through the decryption conversion module.

The storage unit is in bidirectional signal connection with the information extraction unit, the information extraction unit comprises a data port, the storage unit is in bidirectional electrical connection with the central processing unit, the central processing unit is in bidirectional electrical connection with the communication module, the input end of the communication module is in signal connection with the output end of the base station, the output end of the communication module is in signal connection with the input end of the base station, the base station is in bidirectional signal connection with the user service center, the user service center comprises a mobile intelligent terminal, the user service center is in bidirectional signal connection with the mobile intelligent terminal, and the communication module is in bidirectional signal connection with the user terminal unit.

The recommendation client service unit comprises an intelligent recommendation module and a service information display module, the central processing unit is electrically connected with the analysis unit in a bidirectional mode, the output end of the analysis unit is electrically connected with the input end of the networking retrieval unit, the output end of the networking retrieval unit is connected with the input end of the generation unit, the output end of the networking retrieval unit is connected with the input end of the spam information module, the generation unit comprises a computer history comparison module, a data analysis module and a reference value module, the output end of the computer history comparison module is electrically connected with the input end of the central processing unit, the output end of the data analysis module is electrically connected with the input end of the central processing unit, the output end of the reference value module is electrically connected with the input end of the central processing unit, the spam information module is electrically connected with the central processing unit in a bidirectional mode, the spam information module comprises a feedback module and a deletion screening unit, the feedback module is in bidirectional electric connection with the central processing unit, the deleting and screening unit is in bidirectional electric connection with the central processing unit, the output end of the garbage information module is in electric connection with the input end of the transmission module, the transmission module is in bidirectional electric connection with the security firewall, the security firewall is in bidirectional electric connection with the storage unit, the central processing unit is in bidirectional electric connection with the communication module, and the communication module is in bidirectional signal connection with the user terminal unit.

The networked retrieval unit includes an information screening processor and an information tagging module. The information screening processor processes the selected user interest ordering feature factors, wherein the feature factors comprise user names, account information, user portrait feature tags and customer service ordering information, and data are preprocessed; the information marking module marks characteristic factors required by analysis, the characteristic factors are used as indexes of algorithm analysis, SQL sentences are adopted for the indexes to call historical data of target data, analysis is carried out, the output end of the analysis unit is electrically connected with the input end of the information screening processor, and the output end of the analysis unit is electrically connected with the input end of the information marking module.

The input unit comprises a keyboard, an infrared camera and a mouse, wherein the output end of the keyboard is electrically connected with the input end of the central processing unit, the output end of the infrared camera is electrically connected with the input end of the central processing unit, and the output end of the mouse is electrically connected with the input end of the central processing unit.

The input end of the communication module is in signal connection with the output end of the base station, the output end of the communication module is in signal connection with the input end of the base station, and the base station is in bidirectional signal connection with the user service center.

The storage unit comprises a classification processing module, an encryption conversion module and a decryption conversion module, the storage unit is in bidirectional signal connection with the information extraction unit, the information extraction unit comprises a data port, and the storage unit is in bidirectional electrical connection with the central processing unit.

The user service center comprises a mobile intelligent terminal, and the user service center is in bidirectional signal connection with the mobile intelligent terminal. The analysis unit provides service information with a high matching degree based on the service keyword input by the search unit.

The generation unit is used for converting the click rate of the automatic recommendation service on the network into data in time and feeding the data back to the processor;

the garbage information processing unit comprises a feedback module and a deletion screening unit, the deletion screening unit carries out secondary cleaning on the recommended service information retrieved by the retrieval unit and transmits the detected garbage information to the processor through the feedback module; the deleting and screening unit records and stores a plurality of pieces of recommended information every time, analyzes a piece of recommended repeated information which is successfully recommended and recommended for more than 3 times, defines the information as junk information if no clicking action is performed on the recommended information, deletes the screening unit, and records the information in a junk record database; the next time the above recommendation appears, the direct deletion is not analyzed.

That is, after receiving data, firstly analyzing whether the record is in a garbage record database, directly deleting the record, then processing the record by looking at the data which is repeatedly recommended for more than 3 times, and finally, if only one record is left, recommending the record, and if a plurality of records are left, extracting the record with the best user interest degree for recommendation, and deleting the other records.

The recommendation client service unit comprises an intelligent recommendation unit and a service information display unit, wherein the intelligent recommendation unit analyzes the most interesting service of the client by adopting a K-means algorithm according to the historical ordering condition of the client.

The process of analyzing the services of most interest to the customer is as follows:

step (1) selecting characteristic factors of ordering under user interest, wherein the characteristic factors comprise a user name, account information, a user portrait characteristic label and ordering information of customer service, and preprocessing data:

1.1 historical data of the service condition of the order, and if the month in which the data are analyzed is N, extracting the times of the order of the specific user in the (N-4) month, the (N-3) month, the (N-2) month, the (N-1) month, namely the previous 4 months; the factor mainly reflects the ordering frequency of the service of the object and is in direct proportion to the interest and the required condition of the user;

1.2 weighting historical order-placing amount data, if the month in which the data is analyzed is N, extracting the sum of the amounts of orders placed by specific users in N-4 month, N-3 month, N-2 month and N-1 month, namely the previous 4 months, calculating the sum, judging the demand degree of ordering by each service, and weighting according to the demand degree; the factors mainly reflect the historical actual single conditions of the service object;

discretizing the sample, and dividing into 4 grades according to the size, wherein the number of each grade is recorded as a group;

and (3) processing data by adopting a K-means algorithm.

From the resulting data set, the given data was analyzed: the algorithm develops clustering algorithm mining analysis aiming at a given sample set Q:

3.1 randomly selecting k points from the sample set Q within the Q data range as respective centers of k groups;

3.2 respectively calculating the dissimilarity degree of the points in the sample set Q to the center of the k groups, and classifying the dissimilarity degrees into the group with the lowest dissimilarity degree; the dissimilarity degree between two points is measured by an Euclidean distance formula, and the dissimilarity degree is measured by the following steps:

d＝sqrt((X₁-X₂)²+(Y₁-Y₂)²；

d is the Euclidean distance between two points T0(x1, y2) and T1(x2, y2), each point having two dimensions, namely a characteristic factor; the smaller the Euclidean distance is, the smaller the dissimilarity degree is;

3.3 according to the clustering result, recalculating the respective centers of the k groups by taking the arithmetic mean of the respective dimensions of all points in the groups;

3.4 clustering all the points in the sample set Q again according to the new centers;

3.5 repeating the step 3.4 until the clustering result is not changed;

3.6 outputting the user interest clustering objects according to k groups;

3.7 steps 3.1-3.6 are performed using k-8, k-4, k-2, respectively;

through the calculation result of the algorithm, the user interest evaluation of the whole sample set is finally obtained, namely the order placing amount and the order placing frequency of a size (numerical value) according with the degree are the most interesting and most required by the client, and the order placing amount and the order placing frequency are used as intelligent recommendation to recommend services with the frequency and the amount close to the algorithm result in the existing services;

the method comprises the steps that the name of a person user and account information of characteristic factors are used as indexes for algorithm analysis, SQL sentences are adopted for the indexes to call and analyze historical data of target data; the user portrait feature label is used for recommending in a further limited range according to the gender, age and the like of a user in intelligent recommendation as much as possible, so that the recommendation deviation is avoided, and client service ordering information is core analysis data of the intelligent recommendation algorithm.

In summary, the analysis unit, the network retrieval unit and the generation unit are used cooperatively, so that a client inputs keywords of related services on a website, the analysis unit can provide service information with high matching degree for the service information, the manual retrieval condition of the client or a worker for the service information is improved, the automatic recommendation effect of the system for the client is improved, the click rate of the automatic recommendation service on the network can be used by the generation unit, the click rate is timely converted into data to complete analysis, and the data are fed back to the central processing unit, so that the worker can conveniently know the shortage condition of the system in use.

According to the invention, through the arrangement of the junk information module, the feedback module and the deleting and screening unit in the junk information module, the recommended service information searched out through networking can be cleaned for the second time, the detected junk information can be transmitted to the central processing unit through the feedback module, and when the information of the recommended service is automatically searched, the selection of the information content is selected by a worker, so that the worker can obtain a humanized solution when obtaining the final searched information, and the detection efficiency of the system in use is further improved.

The related modules involved in the system are all hardware system modules or functional modules combining computer software programs or protocols with hardware in the prior art, and the computer software programs or the protocols involved in the functional modules are all known in the technology of persons skilled in the art, and are not improvements of the system; the improvement of the system is the interaction relation or the connection relation among all the modules, namely the integral structure of the system is improved, so as to solve the corresponding technical problems to be solved by the system.

Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.