A call center hotspot discovery method, device and electronic equipment

Abstract

The application relates to the field of data processing, in particular to a call center hotspot discovery method and device and electronic equipment, the method comprising the following steps: acquiring at least one new work order data, the new work order data being formed after a user establishes contact with a call center; clustering the at least one new work order data to obtain at least one new work order cluster; judging whether each new work order cluster meets preset clustering standards, and performing clustering optimization on the new work order cluster that does not meet the preset clustering standards until the optimization result meets the preset clustering standards; determining clustering boundary information corresponding to each new work order cluster that meets the preset clustering standards; and determining corresponding hotspot information according to the clustering boundary information corresponding to each new work order cluster, and feeding back each hotspot information. The application has the effect of improving the accuracy of hotspot identification.

Description

Technical Field

[0001] This application relates to the field of data processing, and in particular to a method, apparatus and electronic device for hotspot discovery in call centers. Background Technology

[0002] Call centers primarily handle telephone inquiries from businesses and users. Due to the large number of businesses and users they serve, they handle a significant amount of inquiries daily. By extracting key user information from this large volume of data and providing feedback and analysis based on this information, they can better meet the needs of most users.

[0003] Hot topics refer to public service news or information that is of great interest or popularity to the public, or public service news or issues that attract attention during a certain period. In related technologies, when identifying hot topics from a large amount of call data, the data is first clustered, and then the clustering results are used to directly identify hot topics. However, due to the large amount of call data, the clustering results are prone to bias, which affects the accuracy of identifying hot topics. Summary of the Invention

[0004] To improve the accuracy of identifying hotspot information, this application provides a method, apparatus, and electronic device for hotspot discovery in call centers.

[0005] Firstly, this application provides a method for discovering hotspots in call centers, employing the following technical solution:

[0006] A method for discovering hotspots in call centers includes:

[0007] Obtain at least one new work order data, which is generated after the user establishes contact with the call center;

[0008] Cluster the at least one new work order data to obtain at least one new work order cluster;

[0009] Determine whether each newly added work order cluster meets the preset clustering criteria, and optimize the clustering of newly added work order clusters that do not meet the preset clustering criteria until the optimization results meet the preset clustering criteria;

[0010] Determine the cluster boundary information corresponding to each new work order cluster that meets the preset clustering criteria;

[0011] Based on the cluster boundary information corresponding to each newly added work order cluster, the corresponding hotspot information is determined, and each hotspot information is fed back.

[0012] By adopting the above technical solution, after acquiring at least one new work order data, the data is clustered. This new work order data is generated after the user establishes communication with the call center. After clustering, at least one new work order cluster is obtained. Each new work order cluster is then judged to meet a preset clustering standard. Clusters that do not meet the preset standard are optimized until the optimization result meets the standard. The clustering space range corresponding to each new work order data that meets the preset standard is then determined. Finally, based on the clustering space range corresponding to each new work order cluster, hotspot information corresponding to each cluster is determined and fed back. By optimizing the clustering of new work order clusters that do not meet the preset standard, the accuracy of identifying hotspot information is improved.

[0013] In one possible implementation, clustering the at least one new work order data to obtain at least one new work order cluster includes:

[0014] Determine the new work order vector corresponding to each new work order data;

[0015] All new work order vectors are clustered according to the preset clustering association radius and the initial core vector to obtain at least one new work order cluster. The initial core vector is any new work order vector used to represent the reference vector when clustering all new work order vectors.

[0016] By adopting the above technical solution, the new work order vector corresponding to each new work order data is obtained, and then all new work order vectors are clustered according to the preset clustering association radius and the initial core vector, which facilitates the improvement of the accuracy of clustering new work order data.

[0017] In one possible implementation, the step of clustering all newly added work order vectors based on a preset clustering association radius and an initial core vector includes:

[0018] Based on the preset clustering association radius and the initial core vector, the core range corresponding to the initial core vector is determined;

[0019] Based on the core range and remaining vectors corresponding to the initial core vector, the associated vectors associated with the initial core vector are determined, and all the associated vectors are clustered with the initial core vector to form a new work order cluster. The remaining vectors are other new work order vectors besides at least one new work order vector corresponding to all clustered new work order clusters.

[0020] If the number of remaining vectors is not zero, then based on the remaining vectors, determine the next initial core vector, and iteratively execute the process of determining the associated vectors related to the initial core vector based on the core range corresponding to the initial core vector and the remaining vectors, and clustering all the associated vectors with the initial core vector to form a new work order cluster, until the number of remaining vectors is zero.

[0021] By adopting the above technical solution, the core range corresponding to the initial core vector is determined by preset cluster association radius and initial core vector. Then, based on the core range corresponding to the initial core vector and the remaining vectors determined by all newly added work order vectors other than at least one newly added work order vector corresponding to all clustered newly added work order clusters, the association vectors associated with the initial core vector are determined. Next, all association vectors and the initial core vector are clustered into a newly added work order cluster. Then, it is determined whether the number of other newly added work order vectors other than at least one newly added work order vector corresponding to all clustered newly added work order clusters is zero. If it is not zero, the next initial core vector is determined based on the remaining vectors, and the step of determining the association vector corresponding to the initial core vector is executed cyclically until the number of remaining vectors is zero, which improves the accuracy of determining the newly added work order clusters.

[0022] In one possible implementation, based on the core range and remaining vectors corresponding to any initial core vector, an associated vector is determined that is related to the initial core vector, and all the associated vectors are clustered with the initial core vector to form a new work order cluster, including:

[0023] From the remaining vectors, determine other new work order vectors that are located within the core range corresponding to the new work order cluster. The core range corresponding to the new work order cluster is the total range consisting of the core range corresponding to each associated vector and the core range corresponding to the initial core vector.

[0024] The newly added work order vector that meets the preset criteria is determined as the associated vector associated with the initial core vector, and any other newly added work order vector is clustered with the initial core vector to form a newly added work order cluster;

[0025] The process of determining other new work order vectors located within the core range corresponding to the new work order cluster from the remaining vectors is repeated until a preset condition is met. The preset condition includes any one of the following:

[0026] There are no other newly added work order vectors within the core range corresponding to the newly added work order cluster;

[0027] There are other newly added work order vectors within the core range corresponding to the newly added work order cluster, and none of the other newly added work order vectors meet the preset standard.

[0028] By adopting the above technical solution, after determining other new work order vectors within the core range corresponding to the new work order cluster from all new work order vectors that are not at least one new work order vector corresponding to all clustered new work order clusters, it is determined whether the other new work order vectors meet the preset criteria. Any other new work order vector that meets the preset criteria is determined as the associated vector with the initial core vector. Then, this other new work order vector is clustered with the initial core vector to form a new work order cluster. Finally, the steps of determining other new work order vectors are repeated until the conditions are met: there are no other new work order vectors within the core range corresponding to the new work order cluster, or there are other new work order vectors within the core range corresponding to the new work order cluster, and none of the corresponding other new work order vectors meet the preset criteria. This improves the accuracy of clustering new work order data vectors.

[0029] In one possible implementation, determining whether each newly added work order cluster meets a preset clustering criterion, and performing clustering optimization on newly added work order clusters that do not meet the preset clustering criterion, includes:

[0030] Determine whether the clustering space range corresponding to each newly added work order cluster conforms to the standard space range corresponding to the preset clustering standard. The clustering space range is determined by at least one newly added work order vector in each newly added work order cluster.

[0031] If there are new work order clusters that do not meet the preset clustering criteria, the clustering anomaly type is determined based on the clustering space range corresponding to the new work order clusters that do not meet the preset clustering criteria and the standard space range corresponding to the preset clustering criteria.

[0032] If the clustering anomaly type is that the clustering space range corresponding to the newly added work order cluster is smaller than the standard space range, then the newly added work order clusters that do not meet the preset clustering standard will be merged.

[0033] If the clustering anomaly type is that the clustering space range corresponding to the newly added work order cluster is larger than the standard space range, then the newly added work order cluster that does not conform to the preset clustering standard will be refined.

[0034] By adopting the above technical solution, it is determined whether the clustering spatial range corresponding to each newly added work order cluster conforms to the preset clustering standard spatial range. The clustering spatial range corresponding to the newly added work order clusters that do not conform to the preset clustering standard spatial range is determined. Then, the clustering spatial range is compared with the clustering standard spatial range to determine the clustering anomaly type. The newly added work order clusters that do not conform to the preset clustering standard are optimized according to the clustering anomaly type. When the clustering anomaly type is that the clustering spatial range corresponding to the newly added work order cluster is larger than the standard spatial range, the corresponding newly added work order cluster is refined. When the clustering anomaly type is that the clustering spatial range corresponding to the newly added work order cluster is smaller than the standard spatial range, the corresponding newly added work order clusters are merged. By optimizing the clustering of newly added work order data that does not conform to the preset clustering standard, the accuracy of identifying corresponding hotspot information based on the clustering results is improved.

[0035] In one possible implementation, any new work order cluster that does not conform to the preset clustering criteria is merged, including:

[0036] Obtain the cluster boundary information of any new work order cluster that does not conform to the preset clustering criteria;

[0037] Based on the cluster boundary information, determine whether there are related clusters;

[0038] If they exist, the associated cluster will be merged with any new work order cluster that does not meet the preset clustering criteria.

[0039] By adopting the above technical solution, if there is a new work order cluster with a clustering space range smaller than the preset standard space range, the clustering boundary information is determined based on the new work order cluster, and then the existence of related clusters is determined based on the clustering boundary information. If the corresponding new work order cluster has related clusters, the clustering space ranges corresponding to the determined related clusters and the corresponding new work order clusters are merged to achieve the merging of the corresponding new work order clusters. Clustering and merging through related clusters can improve the accuracy of the clustering results.

[0040] In one possible implementation, determining the corresponding hotspot information based on the clustering boundary information corresponding to each newly added work order cluster, and feeding back each hotspot information, includes:

[0041] Based on the cluster boundary information corresponding to each new work order data, the corresponding hotspot information is determined by traversing the preset hotspot database.

[0042] If there are new work order data for which no corresponding known hotspot information is identified, then the corresponding new work order cluster will be marked with unknown hotspot information;

[0043] Feedback will be provided for all hotspot information corresponding to all newly added work order clusters.

[0044] By adopting the above technical solution, based on the cluster boundary information corresponding to each new work order data, the system traverses the preset hotspot database to determine the corresponding hotspot information. If there are new work order clusters for which no corresponding known hotspot information is identified, the corresponding new work order clusters are marked as unknown hotspot information. Finally, the hotspot information corresponding to all new work order clusters is fed back. By identifying known and unknown hotspot information, relevant staff can perform hotspot event analysis on all new work order data based on the feedback results.

[0045] In one possible implementation, prior to acquiring at least one new work order data, the method further includes:

[0046] Obtain at least one call data point;

[0047] Clean the data for each call data to obtain the corresponding initial new work order data;

[0048] Each initial new work order data is segmented into words to obtain the final new work order data.

[0049] By adopting the above technical solution, and by performing data cleaning and word segmentation on the acquired call data, it is easier to remove some meaningless words, modal particles, auxiliary words, and conjunctions from the call data, thereby improving the effectiveness of the call data.

[0050] Secondly, this application provides a call center hotspot discovery device, which adopts the following technical solution:

[0051] A call center hotspot discovery device, comprising:

[0052] The work order data acquisition module is used to acquire at least one new work order data, which is generated after the user establishes contact with the call center.

[0053] The data clustering module is used to cluster the at least one new work order data to obtain at least one new work order cluster.

[0054] The clustering optimization module is used to determine whether each newly added work order cluster meets the preset clustering criteria, and to perform clustering optimization on newly added work order clusters that do not meet the preset clustering criteria until the optimization result meets the preset clustering criteria.

[0055] The boundary information determination module is used to determine the cluster boundary information corresponding to each new work order cluster that meets the preset clustering criteria;

[0056] The hotspot information determination module is used to determine the corresponding hotspot information based on the cluster boundary information corresponding to each new work order cluster, and to feed back each hotspot information.

[0057] By adopting the above technical solution, after acquiring at least one new work order data, the data is clustered. This new work order data is generated after the user establishes communication with the call center. After clustering, at least one new work order cluster is obtained. Each new work order cluster is then judged to meet a preset clustering standard. Clusters that do not meet the preset standard are optimized until the optimization result meets the standard. The clustering space range corresponding to each new work order data that meets the preset standard is then determined. Finally, based on the clustering space range corresponding to each new work order cluster, hotspot information corresponding to each cluster is determined and fed back. By optimizing the clustering of new work order clusters that do not meet the preset standard, the accuracy of identifying hotspot information is improved.

[0058] Thirdly, this application provides an electronic device that adopts the following technical solution:

[0059] An electronic device comprising:

[0060] At least one processor;

[0061] Memory;

[0062] At least one application, wherein the at least one application is stored in memory and configured to be executed by at least one processor, the at least one application being configured to: perform the above-described method for call center hotspot discovery.

[0063] Fourthly, this application provides a computer-readable storage medium, which adopts the following technical solution:

[0064] A computer-readable storage medium includes: a computer program stored thereon that can be loaded by a processor and execute the above-described call center hotspot discovery method.

[0065] In summary, this application includes at least one of the following beneficial technical effects:

[0066] 1. After acquiring at least one new work order data, cluster the data. New work order data is generated after a user establishes communication with the call center. Clustering yields at least one new work order cluster. Determine if each cluster meets a preset clustering standard. Clusters that do not meet the standard are optimized until they satisfy the standard. Then, determine the clustering space range for each new work order data that meets the standard. Finally, based on the clustering space range, determine the hotspot information for each cluster and provide feedback. Optimizing the clustering of new work order clusters that do not meet the preset standard improves the accuracy of identifying hotspot information.

[0067] 2. Based on the cluster boundary information corresponding to each new work order data, the system traverses the preset hotspot database to determine the corresponding hotspot information. If there are new work order clusters for which no corresponding known hotspot information is identified, the corresponding new work order clusters are marked as unknown hotspot information. Finally, the hotspot information corresponding to all new work order clusters is fed back. By identifying known and unknown hotspot information, relevant staff can perform hotspot event analysis on all new work order data based on the feedback results. Attached Figure Description

[0068] Figure 1 This is a flowchart illustrating a call center hotspot discovery method according to an embodiment of this application;

[0069] Figure 2 This is a schematic diagram of the structure of a call center hotspot discovery device according to an embodiment of this application;

[0070] Figure 3 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0071] The following is in conjunction with the appendix Figure 1-3 This application will be described in further detail.

[0072] After reading this specification, those skilled in the art may make modifications to this embodiment without contributing any inventive step, but such modifications are protected by patent law as long as they fall within the scope of the claims of this application.

[0073] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0074] To improve the accuracy of hotspot identification for newly added work order data, in this embodiment, after acquiring at least one new work order data point, the data is clustered. The new work order data is generated after a user establishes communication with the call center. After clustering, at least one new work order cluster is obtained. Each new work order cluster is then judged to meet a preset clustering standard. Clusters that do not meet the preset standard are optimized until the optimization result meets the standard. The clustering space range corresponding to each new work order data point that meets the preset standard is then determined. Finally, based on the clustering space range corresponding to each new work order cluster, the hotspot information corresponding to each cluster is determined and fed back. By optimizing the clustering of new work order clusters that do not meet the preset standard, the accuracy of identifying hotspot information is improved.

[0075] Specifically, this application provides a call center hotspot discovery method, executed by an electronic device. This electronic device can be a server or a terminal device. The server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. The terminal device can be a smartphone, tablet, laptop, desktop computer, etc., but is not limited to these. The terminal device and the server can be directly or indirectly connected via wired or wireless communication, and this application does not impose any limitations on this connection.

[0076] refer to Figure 1 , Figure 1 This is a flowchart illustrating a call center hotspot discovery method according to an embodiment of this application. The method includes steps S110, S120, S130, S140, and S150, wherein:

[0077] Step S110: Obtain at least one new work order data, which is generated after the user establishes contact with the call center.

[0078] Specifically, the newly added work order data refers to work order data added within a preset period. The preset period can be one hour or one day, and is not specifically limited in this embodiment. It can be modified according to actual needs. In this embodiment, by default, each newly added work order data corresponds to a hotspot.

[0079] Furthermore, step S110 acquires at least one new work order data, which is preceded by steps S01 (not shown in the figures), S02 (not shown in the figures), and S03 (not shown in the figures), wherein:

[0080] Step S01: Obtain at least one call data.

[0081] Specifically, call data is the communication record between the user and the call center after communication is established. Call data can be formed by the call center operator manually entering the communication record, or it can be automatically generated through voice recognition.

[0082] Step S02: Clean up each call data to obtain the corresponding initial new work order data.

[0083] Specifically, data cleaning of call traffic data refers to a process of identifying and correcting identifiable errors in the data, facilitating the removal of meaningless data. Call traffic data cleaning can be performed using binning, clustering, or regression methods; this application does not impose specific limitations, as long as the data cleaning process is feasible.

[0084] Step S03: Perform word segmentation on each initial new work order data to obtain the final new work order data.

[0085] Specifically, word segmentation of the initial new work order data refers to identifying and removing modal particles, auxiliary words, and conjunctions from the initial new work order data to reduce the amount of data processing. When performing word segmentation on the initial new work order data, NLP (Natural Language Processing) algorithms can be used. The specific word segmentation method is not specifically limited in this embodiment, as long as it can effectively segment the initial new work order data.

[0086] In this embodiment of the application, by performing data cleaning and word segmentation on the acquired call data, it is convenient to remove some meaningless words, modal particles, auxiliary words and conjunctions in the call data, thereby improving the effectiveness of the call data.

[0087] Step S120: Cluster at least one new work order data to obtain at least one new work order cluster.

[0088] Specifically, all new work order data can be clustered using the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm. DBSCAN is a density-based clustering algorithm. Other methods for clustering new work order data include K-Means and Single-Pass clustering algorithms. In this embodiment, the specific method for clustering new work order data is not limited, as long as it can cluster all new work order data.

[0089] Step S130: Determine whether each newly added work order cluster meets the preset clustering criteria, and perform clustering optimization on newly added work order clusters that do not meet the preset clustering criteria until the optimization result meets the preset clustering criteria.

[0090] Specifically, the preset clustering criteria are determined based on historical clustering experience and are used to check the clustering results of each new work order data, i.e., each new work order cluster, thereby improving the accuracy of determining the clustering results. The preset clustering criteria can be modified, added, and deleted according to user needs. If a new work order cluster does not conform to the preset clustering criteria, at least one new work order data contained in the new work order cluster is re-clustered to form an optimized result. Then, it is determined whether the optimized result conforms to the preset clustering criteria. If it does, the clustering optimization stops; otherwise, the clustering optimization continues with at least one new work order data contained in the optimized result. The optimized result is the optimized new work order cluster formed after clustering optimization of at least one new work order data that does not conform to the preset clustering criteria.

[0091] If the maximum set of densely connected points is too large or too small during clustering, the corresponding new work order clusters may not meet the preset clustering criteria. For example, new work order data related to food delivery might be clustered into a food safety cluster. When the maximum set of densely connected points is too large, the new work order data corresponding to the new work order clusters that do not meet the preset clustering criteria can be optimized by refining the clustering. When the maximum set of densely connected points is too small, the new work order data corresponding to the new work order clusters that do not meet the preset clustering criteria can be optimized by merging the clustering.

[0092] When determining whether each newly added work order cluster meets the preset clustering criteria, each cluster can be fed back to the corresponding terminal device for manual evaluation by relevant staff. Since each cluster contains at least one new work order, to improve the speed at which staff can assess its compliance, the clusters can be displayed. Because each new work order contains a significant amount of content, the TextRank algorithm can be used to summarize the text of each cluster. By calculating the weight of each sentence within each cluster and ranking them, the central sentence of each cluster can be determined, facilitating reading and demonstrating good practicality.

[0093] Step S140: Determine the cluster boundary information corresponding to each new work order cluster that meets the preset clustering criteria.

[0094] Specifically, the clustering boundary information corresponding to the newly added work order cluster is determined by the boundary vectors in multiple newly added work order clusters that meet the preset clustering criteria. The boundary vectors in the newly added work order clusters are the newly added work order vectors corresponding to the boundary newly added work order data. Vectorization processing of each newly added work order data yields the corresponding newly added work order vector.

[0095] Step S150: Based on the cluster boundary information corresponding to each newly added work order cluster, determine the corresponding hotspot information and feed back each hotspot information.

[0096] Specifically, when determining the hotspot information corresponding to each newly added work order cluster, the clustering boundary information corresponding to each newly added work order cluster is compared with the hotspot clustering boundary information corresponding to each hotspot information in the preset hotspot information table. The preset hotspot information table stores multiple known hotspot information, and each known hotspot information includes the name of the known hotspot and the corresponding hotspot clustering boundary information.

[0097] In this embodiment, after acquiring at least one new work order data, the at least one new work order data is clustered. The new work order data is formed after the user establishes communication with the call center. After clustering, at least one new work order cluster is obtained. It is determined whether each new work order cluster meets the preset clustering criteria. The new work order clusters that do not meet the preset clustering criteria are clustered and optimized until the optimization result meets the preset clustering criteria. Then, the clustering space range corresponding to each new work order data that meets the preset clustering criteria is determined. Finally, based on the clustering space range corresponding to each new work order cluster, the hotspot information corresponding to each new work order cluster is determined and the hotspot information is fed back. By optimizing the clustering of new work order clusters that do not meet the preset clustering criteria, the accuracy of identifying hotspot information can be improved.

[0098] Furthermore, in step S120, at least one newly added work order data is clustered to obtain at least one newly added work order cluster, which may specifically include steps S1201 (not shown in the figure) and S1202 (not shown in the figure), wherein:

[0099] Step S1201: Determine the new work order vector corresponding to each new work order data.

[0100] Specifically, vectorizing each new work order data can yield a corresponding new work order vector. This can be achieved by inputting each new work order data into the Doc2Vec document vector model and outputting the corresponding new work order vector. This process involves establishing a high-dimensional coordinate system based on the text information corresponding to each new work order data, and then forming the corresponding new work order vector by considering the frequency of each word segmentation in the text and the relationship between the word segmentation context.

[0101] Doc2Vec is an unsupervised algorithm that can obtain vector representations of newly added work order data. During vectorization, it considers the contextual relevance between words in the new work order data and demonstrates good language model representation performance after training with a large amount of sample data. Work order data is core business data in call centers, and most call centers retain a large amount of historical work order data for initializing document vector model training. Furthermore, Doc2Vec is an unsupervised algorithm, requiring no manual annotation or optimization. In this embodiment, a Doc2Vec model is constructed using a large amount of historical work order data and used to vectorize new work order data. This ensures that each new work order data is vectorized according to the same standard, and that the output new work order vectors after vectorization are located in a unified coordinate system.

[0102] Step S1202: Cluster all new work order vectors according to the preset clustering association radius and the initial core vector to obtain at least one new work order cluster. The initial core vector is any new work order vector, which is used to represent the reference vector when clustering all new work order vectors.

[0103] Specifically, the initial core vector is randomly selected from multiple newly added work order vectors. The preset clustering association radius can be modified as needed, and is not specifically limited in this embodiment. When clustering multiple newly added work order vectors, the initial core vector is used as the clustering benchmark, and the preset clustering association radius is used as the radius to determine the clustering neighborhood corresponding to the initial core vector. Then, the clustering of multiple newly added work order vectors is completed based on the determined clustering neighborhood. At least one newly added work order vector is clustered into a single newly added work order cluster.

[0104] In this embodiment of the application, by obtaining the new work order vector corresponding to each new work order data, and then clustering all new work order vectors according to the preset clustering association radius and the initial core vector, it is easier to improve the accuracy of clustering new work order data.

[0105] Furthermore, in step S1202, all newly added work order vectors are clustered according to the preset clustering association radius and the initial core vector. Specifically, this may include steps S1202a (not shown in the attached figure), S1202b (not shown in the attached figure), and S1202c (not shown in the attached figure), wherein:

[0106] Step S1202a: Based on the preset clustering association radius and the initial core vector, determine the core range corresponding to the initial core vector.

[0107] Specifically, the core range corresponding to the initial core vector is a circle with the coordinates corresponding to the initial core vector as the center and the preset clustering association radius as the radius.

[0108] Step S1202b: Based on the core range and remaining vectors corresponding to the initial core vector, determine the associated vectors related to the initial core vector, and cluster all the associated vectors with the initial core vector to form a new work order cluster.

[0109] The remaining vectors are the new work order vectors other than at least one new work order vector corresponding to all clustered new work order clusters.

[0110] Specifically, the initial core vector is determined as the initial cluster, meaning that each initial cluster contains only one core vector. Clustering is then completed by updating the initial cluster with the association vectors associated with the initial core vector. From all newly added work order vectors, those that have already undergone clustering are removed, and the remaining newly added work order vectors are the residual vectors.

[0111] Step S1202c: If the number of remaining vectors is not zero, determine the next initial core vector based on the remaining vectors, and repeat the process until the number of remaining vectors is zero. Based on the core range corresponding to the initial core vector and the remaining vectors, determine the associated vectors related to the initial core vector, and cluster all the associated vectors with the initial core vector to form a new work order cluster.

[0112] Specifically, when the number of remaining vectors is not zero, a random one is selected from the remaining new work order vectors as the next initial core vector, and step S1202b is executed repeatedly to cluster the remaining vectors until all remaining vectors have been clustered and the loop stops.

[0113] In this embodiment, the core range corresponding to the initial core vector is determined by preset clustering association radius and initial core vector. Then, based on the core range corresponding to the initial core vector and the remaining vectors determined by all newly added work order vectors other than at least one newly added work order vector corresponding to all clustered newly added work order clusters, the association vectors associated with the initial core vector are determined. Next, all association vectors and the initial core vector are clustered into a newly added work order cluster. Then, it is determined whether the number of other newly added work order vectors other than at least one newly added work order vector corresponding to all clustered newly added work order clusters is zero. If it is not zero, the next initial core vector is determined based on the remaining vectors, and the step of determining the association vector corresponding to the initial core vector is executed cyclically until the number of remaining vectors is zero, which improves the accuracy of determining the newly added work order clusters.

[0114] Furthermore, in step S1202b, based on the core range and remaining vectors corresponding to any initial core vector, the associated vectors are determined, and all associated vectors are clustered with any initial core vector to form a new work order cluster. This can specifically include steps Sa (not shown in the attached figure), Sb (not shown in the attached figure), and Sc (not shown in the attached figure), wherein:

[0115] Step Sa: Determine other new work order vectors located within the core range corresponding to the new work order cluster from the remaining vectors. The core range corresponding to the new work order cluster is the total range consisting of the core range corresponding to each associated vector and the core range corresponding to the initial core vector.

[0116] Specifically, other newly added vectors within the core range corresponding to the newly added work order cluster may or may not be associated vectors of the initial core vector. The newly added work order cluster contains both the initial core vector and associated vectors related to it. Therefore, the core range corresponding to the newly added work order cluster is the core range corresponding to the initial core vector plus the core range corresponding to the associated vectors related to the initial core vector.

[0117] Step Sb: Determine the newly added work order vector that meets the preset criteria as the associated vector with the initial core vector, and cluster any other newly added work order vector with the initial core vector to form a newly added work order cluster.

[0118] Specifically, the preset criterion for determining whether other newly added work order vectors within the core range corresponding to a newly added work order are related vectors of the initial core vector is to determine whether the number of coordinate points corresponding to the newly added work order vectors within a clustering neighborhood with the other newly added work order vectors within the core range corresponding to any newly added work order cluster as the center and the preset clustering association radius as the radius is not less than the preset minimum number of sample points. Other newly added work order vectors whose number of coordinate points corresponding to the newly added work order vectors within any clustering neighborhood is greater than or equal to the minimum number of sample points are determined as related vectors of the initial core vector.

[0119] When determining the initial core vector from multiple new work order vectors, the multiple new work order vectors can be numbered, and then a random number selector can be used to determine a random number. The new work order vector corresponding to the random number is then determined as the initial core vector.

[0120] Step Sc: Iteratively execute the process of determining other new work order vectors within the core range corresponding to the new work order cluster from the remaining vectors until a preset condition is met. The preset condition includes any one of the following:

[0121] There are no other newly added work order vectors within the core scope corresponding to the newly added work order cluster;

[0122] There are other newly added work order vectors within the core scope corresponding to the newly added work order cluster, and none of the other newly added work order vectors meet the preset standards.

[0123] Specifically, if there are no other new work order vectors within the core range corresponding to the newly added work order cluster, it means that there are no other new work order vectors besides at least one new work order vector corresponding to the clustered new work order cluster; if there are other new work order vectors within the core range corresponding to the newly added work order cluster, the number of coordinate points corresponding to the new work order vector within the core range corresponding to any other new work order vector is less than the preset minimum number of sample points.

[0124] In this embodiment, after determining other new work order vectors within the core range corresponding to the new work order cluster from all new work order vectors that are not at least one new work order vector corresponding to all clustered new work order clusters, it is determined whether the other new work order vectors meet the preset criteria. Any other new work order vector that meets the preset criteria is determined as an associated vector associated with the initial core vector. Then, this other new work order vector is clustered with the initial core vector to form a new work order cluster. Finally, the step of determining other new work order vectors is repeated until the conditions are met: there are no other new work order vectors within the core range corresponding to the new work order cluster, or there are other new work order vectors within the core range corresponding to the new work order cluster, and none of the corresponding other new work order vectors meet the preset criteria. This improves the accuracy of clustering new work order data vectors.

[0125] Furthermore, in step S130, it is determined whether each newly added work order cluster meets the preset clustering criteria, and the newly added work order clusters that do not meet the preset clustering criteria are subjected to clustering optimization. Specifically, this may include steps S1301 (not shown in the figure), S1302 (not shown in the figure), S1303 (not shown in the figure), and S1304 (not shown in the figure), wherein:

[0126] Step S1301: Determine whether the clustering space range corresponding to each newly added work order cluster conforms to the standard space range corresponding to the preset clustering standard.

[0127] The clustering space range is determined by at least one new work order vector in each new work order cluster.

[0128] Specifically, the clustering space range corresponding to each newly added work order cluster is composed of multiple boundary newly added work order vectors in each newly added work order cluster, and the standard space range is composed of multiple boundary standard work order vectors in the preset clustering standard.

[0129] Step S1302: If there are new work order clusters that do not meet the preset clustering criteria, then determine the clustering anomaly type based on the clustering space range corresponding to the new work order clusters that do not meet the preset clustering criteria and the standard space range corresponding to the preset clustering criteria.

[0130] Specifically, clustering anomalies include clustering spatial ranges that are smaller than the standard spatial range for any newly added work order cluster and clustering spatial ranges that are larger than the standard spatial range for any newly added work order cluster.

[0131] Step S1303: If the clustering anomaly type is that the clustering space range corresponding to any newly added work order cluster is smaller than the standard space range, then the newly added work order clusters that do not meet the preset clustering criteria will be merged.

[0132] Step S1304: If the clustering anomaly type is that the clustering space range corresponding to any newly added work order cluster is larger than the standard space range, then the newly added work order clusters that do not meet the preset clustering criteria will be refined.

[0133] Specifically, if the maximum set of points that can be reached by density during the clustering process is too large or too small, that is, if the clustering space range corresponding to the new work order cluster is larger than the standard space range or the new work order cluster is smaller than the standard space unit, the clustering results may not conform to business experience. For example, new work order data related to food delivery may be clustered into the food safety cluster.

[0134] When refining new work order clusters that do not meet the preset clustering criteria, the preset clustering association radius and the preset minimum number of sample points can be reduced, thereby reducing the maximum set of densely connected points during the clustering process. This refines new work order clusters whose clustering space range is larger than the standard space range. When merging new work order clusters that do not meet the preset clustering criteria, the preset clustering association radius and the preset minimum number of sample points can be increased, thereby increasing the maximum set of densely connected points during the clustering process. This merges new work order clusters whose clustering space range is smaller than the standard space range. Alternatively, the clustering space ranges corresponding to two new work order clusters can be merged to merge new work order clusters whose clustering space range is smaller than the standard space range.

[0135] In this embodiment, it is determined whether the clustering space range corresponding to each newly added work order cluster conforms to the preset clustering standard space range. The clustering space range corresponding to the newly added work order clusters that do not conform to the preset clustering standard space range is determined. Then, the clustering space range is compared with the clustering standard space range to determine the clustering anomaly type. The newly added work order clusters that do not conform to the preset clustering standard are optimized according to the clustering anomaly type. When the clustering anomaly type is that the clustering space range corresponding to the newly added work order cluster is larger than the standard space range, the corresponding newly added work order cluster is refined. When the clustering anomaly type is that the clustering space range corresponding to the newly added work order cluster is smaller than the standard space range, the corresponding newly added work order clusters are merged. By optimizing the clustering of newly added work order data that does not conform to the preset clustering standard, the accuracy of identifying corresponding hotspot information based on the clustering results is improved.

[0136] Furthermore, step S1303 involves merging any newly added work order clusters, which may specifically include steps S1303a (not shown in the figures), S1303b (not shown in the figures), and S1303c (not shown in the figures), wherein:

[0137] Step S1301a: Obtain the cluster boundary information of any newly added work order cluster that does not conform to the preset clustering criteria.

[0138] Specifically, the clustering space is composed of the value ranges of multiple new work order vectors in each dimension. The clustering space can be represented as [Amin: Amax, Bmin: Bmax, Cmin: Cmax, ..., Nmin: Nmax], where A, B, C...N represent each dimension, Amin represents the minimum value in dimension A, and Amax represents the maximum value in dimension A. The clustering boundary information of the new work order cluster is composed of the new work order vectors located at the cluster boundary of at least one new work order vector corresponding to the new work order cluster.

[0139] Step S1303b: Determine whether there are related clusters based on the cluster boundary information.

[0140] Step S1303c: If there is an associated cluster, merge the associated cluster with any new work order cluster that does not meet the preset clustering criteria.

[0141] Specifically, the associated cluster is another newly added work order cluster whose clustering space range is within a preset range corresponding to the newly added work order cluster. The preset range can be the straight-line distance between the target core vector coordinates in the two newly added work order clusters. The specific preset range can be modified according to user needs, and is not specifically limited in this embodiment.

[0142] When the straight-line distance between the target core vector coordinates in two newly added work order clusters is less than a preset range, it indicates that the two newly added work order clusters are related clusters. The clustering space ranges corresponding to the two newly added work order clusters are merged to achieve the merging of related clusters. The clustering space ranges corresponding to the two related clusters may be adjacent or separated.

[0143] In this embodiment of the application, if the clustering space range corresponding to any newly added work order cluster is smaller than the preset standard space range, the clustering boundary information is determined according to the corresponding newly added work order cluster, and then the existence of related clusters is determined according to the clustering boundary information. If the corresponding newly added work order cluster has related clusters, the clustering space ranges corresponding to the determined related clusters and the corresponding newly added work order clusters are merged to merge the corresponding newly added work order clusters. Clustering and merging through related clusters can improve the accuracy of clustering results.

[0144] Furthermore, in step S150, based on the cluster boundary information corresponding to each newly added work order cluster, the corresponding hotspot information is determined, and each hotspot information is fed back. Specifically, this includes steps S1501 (not shown in the attached figure), S1502 (not shown in the attached figure), and S1503 (not shown in the attached figure), wherein:

[0145] Step S1501: Based on the cluster boundary information corresponding to each newly added work order cluster, traverse the preset hotspot database to determine the corresponding hotspot information.

[0146] Step S1502: If there are new work order clusters for which no corresponding known hotspot information has been identified, then the corresponding new work order clusters are marked as unknown hotspot information.

[0147] Specifically, the pre-defined hotspot database stores multiple known hotspot information entries, each including a known hotspot name and corresponding hotspot cluster boundary information. When determining the hotspot information corresponding to each new work order cluster, the cluster boundary information corresponding to each new work order cluster is matched with the hotspot cluster boundary information corresponding to known hotspots. For any new work order cluster, if the cluster boundary information corresponding to any new work order cluster falls within the range of the hotspot cluster boundary information corresponding to any known hotspot, then the new work order cluster is named based on that known hotspot information.

[0148] If, after traversing the preset hotspot database, there are newly added work order clusters that have not been named as hotspots, then the corresponding newly added work order clusters will be named unknown hotspots, and the clustering boundary information of the newly added work order clusters corresponding to the unknown hotspots will be saved.

[0149] Step S1503: Feed back all hotspot information corresponding to all newly added work order clusters.

[0150] Specifically, when providing feedback on trending topics, both known and unknown trending topics are reported. By sending information about unknown trending topics to the corresponding terminal devices, relevant staff can identify them. After receiving confirmation information from the corresponding terminal devices, the preset trending topic database is updated based on this confirmation information.

[0151] In this embodiment, based on the cluster boundary information corresponding to each new work order data, the system traverses the preset hotspot database to determine the corresponding hotspot information. If there are new work order clusters for which no corresponding known hotspot information is identified, the corresponding new work order clusters are marked as unknown hotspot information. Finally, the hotspot information corresponding to all new work order clusters is fed back. By identifying known and unknown hotspot information, relevant staff can perform hotspot event analysis on all new work order data based on the feedback results.

[0152] The above embodiments describe a method for discovering hotspots in a call center from the perspective of the process flow. The following embodiments describe a device for discovering hotspots in a call center from the perspective of a virtual module or virtual unit. For details, please refer to the following embodiments.

[0153] This application provides a device for discovering call center hotspots, such as... Figure 2 As shown, the device may specifically include a work order data acquisition module 210, a data clustering module 220, a clustering optimization module 230, a boundary information determination module 240, and a hotspot information determination module 250, wherein:

[0154] The work order data acquisition module 210 is used to acquire at least one new work order data, which is generated after the user establishes contact with the call center.

[0155] The data clustering module 220 is used to cluster at least one new work order data to obtain at least one new work order cluster.

[0156] The clustering optimization module 230 is used to determine whether each newly added work order cluster meets the preset clustering criteria, and to perform clustering optimization on newly added work order clusters that do not meet the preset clustering criteria until the optimization result meets the preset clustering criteria.

[0157] The boundary information determination module 240 is used to determine the clustering boundary information corresponding to each new work order cluster that meets the preset clustering criteria.

[0158] The hotspot information determination module 250 is used to determine the corresponding hotspot information based on the cluster boundary information corresponding to each newly added work order cluster, and to feed back each hotspot information.

[0159] In one possible implementation, the data clustering module 220 includes:

[0160] The "Determine New Work Order Vector" submodule is used to determine the new work order vector corresponding to each new work order data.

[0161] The clustering submodule is used to cluster all new work order vectors according to the preset clustering association radius and the initial core vector to obtain at least one new work order cluster. The initial core vector is any new work order vector, which is used to represent the reference vector when clustering all new work order vectors.

[0162] In one possible implementation, the clustering submodule includes:

[0163] The core range unit is used to determine the core range corresponding to the initial core vector based on the preset clustering association radius and the initial core vector.

[0164] The associated vector unit is determined based on the core range and remaining vectors corresponding to the initial core vector. The associated vectors are then clustered with the initial core vector to form a new work order cluster. The remaining vectors are the new work order vectors other than at least one new work order vector corresponding to all clustered new work order clusters.

[0165] The process involves iteratively determining the associated vector unit. When the number of remaining vectors is not zero, the process determines the next initial core vector based on the remaining vectors. The process iteratively executes the determination of associated vectors related to the initial core vector based on the core range corresponding to the initial core vector and the remaining vectors. All associated vectors are then clustered with the initial core vector to form a new work order cluster, until the number of remaining vectors is zero.

[0166] In one possible implementation, determining the associated vector unit includes:

[0167] Determine other newly added work order vector sub-units to identify other newly added work order vectors located within the core range corresponding to the newly added work order cluster from the remaining vectors. The core range corresponding to the newly added work order cluster is the total range consisting of the core range corresponding to each associated vector and the core range corresponding to the initial core vector.

[0168] The association clustering subunit is used to determine the newly added work order vector that meets the preset criteria as the association vector associated with the initial core vector, and to cluster any other newly added work order vector with the initial core vector to form a new work order cluster;

[0169] The iterative determination sub-unit is used to iteratively execute the determination of other new work order vectors located within the core range corresponding to the new work order cluster from the remaining vectors until a preset condition is met. The preset condition includes any one of the following:

[0170] There are no other newly added work order vectors within the core scope corresponding to the newly added work order cluster;

[0171] There are other newly added work order vectors within the core scope corresponding to the newly added work order cluster, and none of the other newly added work order vectors meet the preset standards.

[0172] In one possible implementation, the clustering optimization module 230 includes:

[0173] The judgment submodule is used to determine whether the clustering space range corresponding to each newly added work order cluster conforms to the standard space range corresponding to the preset clustering standard. The clustering space range is determined by at least one newly added work order vector in each newly added work order cluster.

[0174] The anomaly type determination submodule is used to determine the clustering anomaly type when there are new work order clusters that do not conform to the preset clustering criteria, based on the clustering space range corresponding to the new work order clusters that do not conform to the preset clustering criteria and the standard space range corresponding to the preset clustering criteria.

[0175] The cluster merging submodule is used to merge new work order clusters that do not meet the preset clustering criteria when the clustering space range corresponding to any new work order cluster is smaller than the standard space range.

[0176] The clustering refinement submodule is used to refine new work order clusters that do not meet the preset clustering criteria when the clustering space range corresponding to any new work order cluster is larger than the standard space range due to the clustering anomaly type.

[0177] In one possible implementation, the clustering merging submodule includes:

[0178] The boundary information acquisition unit is used to acquire the cluster boundary information of any new work order cluster that does not conform to the preset clustering criteria;

[0179] The association determination unit is used to determine whether there are associated clusters based on the cluster boundary information;

[0180] The cluster merging unit is used to merge the associated clusters with any new work order cluster that does not meet the preset clustering criteria, if associated clusters exist.

[0181] In one possible implementation, the hotspot information determination module 250 includes:

[0182] The hotspot information determination submodule is used to traverse the preset hotspot database based on the cluster boundary information corresponding to each new work order data to determine the corresponding hotspot information.

[0183] The hotspot marking submodule is used to mark the corresponding new work order clusters with unknown hotspot information when there is new work order data that has not been identified with corresponding known hotspot information.

[0184] The Hotspot Feedback submodule is used to provide feedback on all hotspot information corresponding to all newly added work order clusters.

[0185] In one possible implementation, the device further includes:

[0186] The call data acquisition module is used to acquire at least one call data point;

[0187] The module for determining the initial new work order data is used to clean up each call data to obtain the corresponding initial new work order data.

[0188] The module for determining new work order data is used to perform word segmentation on each initial new work order data to obtain the final new work order data.

[0189] This application provides an electronic device, such as... Figure 3 As shown, Figure 3 The illustrated electronic device 300 includes a processor 301 and a memory 303. The processor 301 and the memory 303 are connected, for example, via a bus 302. Optionally, the electronic device 300 may also include a transceiver 304. It should be noted that in practical applications, the transceiver 304 is not limited to one type, and the structure of this electronic device 300 does not constitute a limitation on the embodiments of this application.

[0190] Processor 301 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 301 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0191] Bus 302 may include a pathway for transmitting information between the aforementioned components. Bus 302 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 302 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0192] The memory 303 may be a ROM (Read Only Memory) or other type of static storage device capable of storing static information and instructions, RAM (Random Access Memory) or other type of dynamic storage device capable of storing information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0193] The memory 303 is used to store application code that executes the solution of this application, and its execution is controlled by the processor 301. The processor 301 is used to execute the application code stored in the memory 303 to implement the content shown in the embodiments of this application.

[0194] Electronic devices include, but are not limited to: mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), and in-vehicle terminals (such as in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Servers can also be included. Figure 3 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0195] This application provides a computer-readable storage medium storing a computer program that, when run on a computer, enables the computer to execute the corresponding content in the aforementioned method embodiments. Compared with related technologies, in this application embodiment, after obtaining at least one new work order data, the at least one new work order data is clustered. The new work order data is formed after a user establishes communication with a call center. After clustering, at least one new work order cluster is obtained. It is determined whether each new work order cluster meets a preset clustering standard. Clustering optimization is performed on new work order clusters that do not meet the preset clustering standard until the optimization result meets the preset clustering standard. Then, the clustering space range corresponding to each new work order data that meets the preset clustering standard is determined. Finally, based on the clustering space range corresponding to each new work order cluster, hotspot information corresponding to each new work order cluster is determined, and the hotspot information is fed back. By optimizing the clustering of new work order clusters that do not meet the preset clustering standard, the accuracy of identifying hotspot information is improved.

[0196] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0197] The above description is only a partial embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for discovering hotspots in call centers, characterized in that, include: Obtain at least one new work order data, which is generated after the user establishes contact with the call center; Cluster the at least one new work order data to obtain at least one new work order cluster; Determine whether each newly added work order cluster meets the preset clustering criteria, and optimize the clustering of newly added work order clusters that do not meet the preset clustering criteria until the optimization results meet the preset clustering criteria; Determine the cluster boundary information corresponding to each new work order cluster that meets the preset clustering criteria. The cluster boundary information is determined by the boundary vectors in multiple new work order clusters that meet the preset clustering criteria. The boundary vector is the new work order vector corresponding to the boundary new work order data. Based on the cluster boundary information corresponding to each newly added work order cluster, the corresponding hotspot information is determined, and each hotspot information is fed back. The step of clustering the at least one new work order data to obtain at least one new work order cluster includes: Determine the new work order vector corresponding to each new work order data; All new work order vectors are clustered according to the preset clustering association radius and the initial core vector to obtain at least one new work order cluster. The initial core vector is any new work order vector, which is used to characterize the reference vector when clustering all new work order vectors. The step of clustering all newly added work order vectors according to a preset clustering association radius and an initial core vector includes: Based on the preset clustering association radius and the initial core vector, the core range corresponding to the initial core vector is determined; Based on the core range and remaining vectors corresponding to the initial core vector, the associated vectors associated with the initial core vector are determined, and all the associated vectors are clustered with the initial core vector to form a new work order cluster. The remaining vectors are other new work order vectors besides at least one new work order vector corresponding to all clustered new work order clusters. If the number of remaining vectors is not zero, then based on the remaining vectors, determine the next initial core vector, and iteratively execute the process of determining the associated vectors related to the initial core vector based on the core range corresponding to the initial core vector and the remaining vectors, and clustering all the associated vectors with the initial core vector to form a new work order cluster, until the number of remaining vectors is zero. Specifically, based on the core range and remaining vectors corresponding to any initial core vector, associated vectors are determined that are related to the initial core vector. All associated vectors are then clustered with the initial core vector to form a new work order cluster, including: From the remaining vectors, determine other new work order vectors that are located within the core range corresponding to the new work order cluster. The core range corresponding to the new work order cluster is the total range consisting of the core range corresponding to each associated vector and the core range corresponding to the initial core vector. The newly added work order vector that meets the preset criteria is determined as the associated vector associated with the initial core vector, and any other newly added work order vector is clustered with the initial core vector to form a new work order cluster; The process of determining other new work order vectors located within the core range corresponding to the new work order cluster from the remaining vectors is repeated until a preset condition is met. The preset condition includes any one of the following: There are no other newly added work order vectors within the core range corresponding to the newly added work order cluster; There are other new work order vectors within the core range corresponding to the newly added work order cluster, and none of the other new work order vectors meet the preset standard. The step of determining whether each newly added work order cluster meets the preset clustering criteria, and optimizing the clustering of newly added work order clusters that do not meet the preset clustering criteria, includes: Determine whether the clustering space range corresponding to each newly added work order cluster conforms to the standard space range corresponding to the preset clustering standard. The clustering space range is determined by at least one newly added work order vector in each newly added work order cluster. If there are new work order clusters that do not meet the preset clustering criteria, the clustering anomaly type is determined based on the clustering space range corresponding to the new work order clusters that do not meet the preset clustering criteria and the standard space range corresponding to the preset clustering criteria. If the clustering anomaly type is that the clustering space range corresponding to the newly added work order cluster is smaller than the standard space range, then the newly added work order clusters that do not meet the preset clustering standard will be merged. If the clustering anomaly type is that the clustering space range corresponding to the newly added work order cluster is larger than the standard space range, then the newly added work order cluster that does not meet the preset clustering standard will be refined. The step of determining the corresponding hotspot information based on the clustering boundary information corresponding to each newly added work order cluster, and feeding back each hotspot information, includes: Based on the cluster boundary information corresponding to each new work order data, the corresponding hotspot information is determined by traversing the preset hotspot database. If there are new work order data for which no corresponding known hotspot information is identified, then the corresponding new work order cluster will be marked with unknown hotspot information; Feedback will be provided for all hotspot information corresponding to all newly added work order clusters.

2. The call center hotspot discovery method according to claim 1, characterized in that, Merge any new work order cluster that does not meet the preset clustering criteria, including: Obtain the cluster boundary information of any new work order cluster that does not conform to the preset clustering criteria; Based on the cluster boundary information, determine whether there are related clusters; If they exist, the associated cluster will be merged with any new work order cluster that does not meet the preset clustering criteria.

3. The method for discovering hotspots in a call center according to claim 1, characterized in that, Before obtaining at least one new work order data, the process also includes: Obtain at least one call data point; Clean the data for each call data to obtain the corresponding initial new work order data; Each initial new work order data is segmented into words to obtain the final new work order data.

4. A call center hotspot discovery device, characterized in that, include: The work order data acquisition module is used to acquire at least one new work order data, which is generated after the user establishes contact with the call center. The data clustering module is used to cluster the at least one new work order data to obtain at least one new work order cluster. The clustering optimization module is used to determine whether each newly added work order cluster meets the preset clustering criteria, and to perform clustering optimization on newly added work order clusters that do not meet the preset clustering criteria until the optimization result meets the preset clustering criteria. The boundary information determination module is used to determine the cluster boundary information corresponding to each new work order cluster that meets the preset clustering criteria. The cluster boundary information is determined by the boundary vectors in multiple new work order clusters that meet the preset clustering criteria. The boundary vector is the new work order vector corresponding to the boundary new work order data. The hotspot information determination module is used to determine the corresponding hotspot information based on the cluster boundary information corresponding to each new work order cluster, and to feed back each hotspot information. The data clustering module includes: The "Determine New Work Order Vector" submodule is used to determine the new work order vector corresponding to each new work order data. The clustering submodule is used to cluster all new work order vectors according to a preset clustering association radius and an initial core vector to obtain at least one new work order cluster. The initial core vector is any new work order vector used to characterize the reference vector when clustering all new work order vectors. The clustering submodule includes: The core range unit is used to determine the core range corresponding to the initial core vector based on the preset clustering association radius and the initial core vector; The associated vector unit is used to determine the associated vectors associated with the initial core vector based on the core range and the remaining vectors corresponding to the initial core vector, and to cluster all the associated vectors with the initial core vector to form a new work order cluster. The remaining vectors are other new work order vectors other than at least one new work order vector corresponding to all the clustered new work order clusters. The associated vector unit is determined in a loop. If the number of remaining vectors is not zero, the next initial core vector is determined based on the remaining vectors. The process of determining the associated vectors associated with the initial core vector based on the core range and remaining vectors is repeated in a loop, and all the associated vectors are clustered with the initial core vector to form a new work order cluster, until the number of remaining vectors is zero. Determining the associated vector unit includes: Determine other newly added work order vector sub-units to determine other newly added work order vectors located within the core range corresponding to the newly added work order cluster from the remaining vectors. The core range corresponding to the newly added work order cluster is the total range composed of the core range corresponding to each associated vector and the core range corresponding to the initial core vector. The association clustering subunit is used to determine the newly added work order vector that meets the preset criteria as the association vector associated with the initial core vector, and to cluster any other newly added work order vector with the initial core vector to form a newly added work order cluster; The iterative determination subunit is used to iteratively execute the process of determining other new work order vectors located within the core range corresponding to the new work order cluster from the remaining vectors until a preset condition is met. The preset condition includes any one of the following: There are no other newly added work order vectors within the core range corresponding to the newly added work order cluster; There are other new work order vectors within the core range corresponding to the newly added work order cluster, and none of the other new work order vectors meet the preset standard. The clustering optimization module includes: The judgment submodule is used to determine whether the clustering space range corresponding to each newly added work order cluster conforms to the standard space range corresponding to the preset clustering standard. The clustering space range is determined by at least one newly added work order vector in each newly added work order cluster. The anomaly type determination submodule is used to determine the clustering anomaly type based on the clustering space range corresponding to the new work order cluster that does not conform to the preset clustering standard and the standard space range corresponding to the preset clustering standard if there is a new work order cluster that does not conform to the preset clustering standard. The clustering and merging submodule is used to merge the new work order clusters that do not meet the preset clustering criteria if the clustering space range corresponding to the new work order cluster is smaller than the standard space range. The clustering refinement submodule is used to refine the new work order cluster that does not meet the preset clustering standard if the clustering space range corresponding to the new work order cluster is larger than the standard space range. The hotspot information identification module includes: The hotspot information determination submodule is used to traverse the preset hotspot database based on the cluster boundary information corresponding to each new work order data to determine the corresponding hotspot information. The hotspot marking submodule is used to mark the corresponding new work order clusters as unknown hotspot information if there are new work order data for which no corresponding known hotspot information is identified. The Hotspot Feedback submodule is used to provide feedback on all hotspot information corresponding to all newly added work order clusters.

5. An electronic device, characterized in that, The electronic device includes: At least one processor; Memory; At least one application, wherein the at least one application is stored in memory and configured to be executed by at least one processor, the at least one application being configured to: perform the call center hotspot discovery method of any one of claims 1-3.

Images