Modeling and visual analysis method for correlation of multi-dimensional questions in questionnaire

Abstract

The invention relates to the technical field of park management, and discloses a multi-dimensional question correlation modeling and visual analysis method for questionnaires, which comprises the following steps: receiving questionnaire data, operation index data and service event data, combining the questionnaire data and the operation index data into a time slice sequence according to a unified time step length, and setting an event interval, the method comprises the steps that two types of data are aligned at the same time aperture, linkage fragments are generated in an event interval according to questionnaire change types and index change types, anchor point pair sets are determined by the linkage fragments which are continuous in the interval and interrupted outside the interval, irrelevant linkage is suppressed, questionnaire questions before an event and operation indexes after the event are selected through the anchor point pair sets, and the questionnaire questions and the operation indexes are selected through the selected questionnaire questions. A service problem path is generated according to time sequence, cross-source linkage response delay is reduced, finally a questionnaire view, an index view and an event interval are obtained, a visual access track is recorded according to the path sequence, and the traceability of the service problem path is achieved.

Description

Technical Field

[0001] This invention relates to the field of park management technology, and more specifically, to a method for modeling and visualizing the correlation of multi-dimensional items in a questionnaire. Background Technology

[0002] In smart park and enterprise service operation scenarios, platforms typically collect multi-dimensional feedback data from park enterprises and end users regarding service quality, facility usage, operation, and maintenance through online questionnaires. This questionnaire data, along with business data such as park operation indicators, equipment operation records, and work order information, is integrated into a unified data analysis and visualization platform. Under the constraints of limited computing power and storage resources and the need to support multi-terminal interaction, existing technologies mostly adopt dimensional aggregation methods based on statistical summarization, inter-item relationship estimation methods based on correlation measurement, and result stability control methods based on time smoothing and filtering. They provide an operational view through fixed report displays and dashboard linkage. Under the conditions of relatively stable questionnaire template structure, limited number of parks, and relatively smooth changes in operational indicators, the above methods can support basic satisfaction analysis and simple operational trend judgment to a certain extent.

[0003] With the increase in the number and types of smart parks and the diversification of questionnaire sources, the frequent adjustments to the dimensions and structure of questionnaire questions lead to a continuously changing mapping relationship between questionnaire questions and park operation indicators, which is one type of objective factor. Simultaneously, the high frequency of updates to service event data and operation indicators during park operation, coupled with differences in the arrival time and refresh rhythm of multi-source data, results in a continuous increase in visualization linkage paths, which is another type of objective factor. When these factors work together, the abstract methods based on existing statistical aggregation correlation measurements and fixed linkage rules are prone to alignment deviations and linkage delays when constructing the alignment relationship between questionnaire questions and operation indicators and driving multi-source visualization linkage. This leads to incomplete and unstable visualization paths for service problem localization. Therefore, the technical problem to be solved is ensuring the traceability of service problem localization paths while improving the consistency between questionnaire question relevance and park operation indicator alignment, and reducing the response delay of multi-source data visualization linkage.

[0004] In view of this, the present invention proposes a method for multi-dimensional item relevance modeling and visualization analysis in questionnaires to solve the above problems. Summary of the Invention

[0005] To overcome the aforementioned shortcomings of existing technologies, this invention provides a method for multi-dimensional item relevance modeling and visualization analysis in questionnaires.

[0006] To achieve the above objectives, the present invention provides the following technical solution: Firstly, a method for multi-dimensional item relevance modeling and visualization analysis in questionnaires is provided, including: Receive questionnaire data, operational indicator data, and service event data. Combine the questionnaire data and operational indicator data into a time slice sequence with a unified time step. Set the event interval in the time slice sequence around the time of each service event, covering the time slices before and after it. The questionnaire change type is classified according to the change in responses in adjacent time slices, and the indicator change type is classified according to the change in values ​​in adjacent time slices. Linkage segments are generated in the event interval according to the correspondence between the questionnaire change type and the indicator change type. The anchor point pair set between the questionnaire questions and the operational indicators is determined by the case that the linkage segments are continuous in the event interval and interrupted outside the event interval. Within the set of anchor points in each event interval, select questionnaire items in the time slice before the event and operational indicators in the time slice after the event, and connect the selected questionnaire items and operational indicators in chronological order to generate a service question path. In the service issue path, for each path, obtain the questionnaire view and indicator view on the path, as well as the event interval corresponding to the path, and record the visual access trajectory according to the order of the questionnaire view and indicator view in the service issue path.

[0007] In some embodiments, setting an event interval in the time slice sequence around the occurrence time of each service event, covering the time slices before and after, includes: From the questionnaire data, the number of responses to questionnaire questions corresponding to service event types within each time segment in the park dimension is counted, a questionnaire statistics table is generated, and the time segments in the questionnaire statistics table that contain responses to questionnaire questions are grouped into a questionnaire segment set. From the operational indicator data, the numerical status of operational indicators corresponding to service event types within each time slice is statistically analyzed at the park level, an indicator status table is generated, and the time slices in which the numerical status of operational indicators in the indicator status table changes are combined into an indicator slice set. For each service event record in the service event data, with the time slice where the service event occurred as the center, a search is performed simultaneously forward and backward along the questionnaire slice set and the indicator slice set in the time slice sequence. Time slices that continuously belong to the questionnaire slice set or the indicator slice set are formed into time slice segments, and the time range corresponding to the time slice segment is determined as the event interval of the service event record in the time slice sequence.

[0008] In some embodiments, setting an event interval in the time slice sequence around the occurrence time of each service event, covering the time slices before and after, further includes: At the park level, the event intervals corresponding to each service event record are summarized, and a service attribution table is established. The service attribution table is indexed by time slices and records a list of service event record identifiers containing that time slice. In the service attribution table, for time slices that have multiple service event records, based on the service event type corresponding to the time slice, time slices that are adjacent and have all the same service event type are grouped into the same type of segment, and time slices that are adjacent and do not all have the same service event type are grouped into different types of segment. For each service event record, within its corresponding event interval, the time slice in the same type of segment is defined as the primary impact time slice of the service event record, and the time slice in the different type of segment is defined as the secondary impact time slice of the service event record. The time range covered by the primary impact time slice and the secondary impact time slice is used as the updated event interval of the service event record in the time slice sequence.

[0009] In some embodiments, a linked segment is generated within the event interval according to the correspondence between questionnaire change types and indicator change types, including: In the park's operation cloud platform, service dimension tags are configured for questionnaire items, operation dimension tags are configured for operation indicators, and a dimension mapping table is established to record the pairing relationship between service dimension tags and operation dimension tags. For each questionnaire item in the time slice sequence, a questionnaire change table is generated based on the comparison of the number of responses to each option recorded in the questionnaire statistics table in adjacent time slices. In the questionnaire change table, records where the number of responses to high-scoring options decreases and the number of responses to medium- or low-scoring options increases are classified as questionnaire change types, and the corresponding service dimension tags are recorded. For each operational indicator in the time slice sequence, the operational status of each time slice is determined based on the position of the operational indicator value in the preset operational segment within the adjacent time slices. An indicator change table is generated, and records in the indicator change table that have changed from normal operation status to critical operation status or from critical operation status to alarm operation status are classified as indicator change types, and the corresponding operational dimension labels are registered. For each questionnaire question and each operational indicator combination within each event interval, the service dimension label and the operational dimension label are confirmed to be able to be paired in the dimension mapping table. Within the adjacent time slices covered by the event interval, when there are questionnaire change type records in the questionnaire change table and indicator change type records in the indicator change table, the time range corresponding to the time slice sequence formed by the consecutive arrangement of the corresponding adjacent time slices is determined as the linkage segment.

[0010] In some embodiments, the set of anchor point pairs between questionnaire items and operational metrics is determined based on the condition that the linked segments are continuous within an event interval but interrupted outside the event interval, including: For each questionnaire question and each operational indicator combination, summarize the corresponding linkage segments of the questionnaire question and the operational indicator in all event intervals, and establish a linkage position table. The linkage position table records the service event record identifier to which each linkage segment belongs and the position information of the time slice covered by the linkage segment in the event interval, whether it is in the early stage, middle stage or late stage of the event. In the linked location table, for each questionnaire question and each operational indicator combination, find linked segments with the same service event type and location information belonging to the same location category, group these linked segments into linked groups, and record the event interval identifiers involved in the linked groups. Remove all time slices covered by event intervals from the time slice sequence. Within the time slices after removal, check whether there are records of combinations of questionnaire change type and indicator change type corresponding to the linkage group based on the questionnaire change table and indicator change table. If there are no combination records within this range, register the corresponding questionnaire question and operation indicator combination as anchor point pairs, and form an anchor point pair set with all the registered anchor point pairs.

[0011] In some embodiments, within the set of anchor point pairs in each event interval, questionnaire items in the pre-event time slice and operational metrics in the post-event time slice are selected, and the selected questionnaire items and operational metrics are connected in chronological order to generate a service question path, including: For each set of anchor point pairs in each event interval, an anchor point table is established according to the questionnaire question identifier, the operation indicator identifier, and the time slice identifier. In the anchor point table, each anchor point record is marked with its time slice relative to the time of the service event, indicating whether it is in the pre-event or post-event stage. In the anchor table, anchor records with the location category of the pre-event segment are grouped into a pre-questionnaire set, and anchor records with the location category of the post-event segment are grouped into a post-indicator set. The anchor records in the pre-questionnaire set and the post-indicator set are arranged in time slice order to obtain the pre-anchor sequence and the post-anchor sequence. For each event interval, questionnaire item identifiers are extracted in the time slice order from the previous anchor sequence, and operational indicator identifiers are extracted in the time slice order from the subsequent anchor sequence. The questionnaire item identifier sequence and the operational indicator identifier sequence are connected in chronological order to form a path node sequence, and the cross-event path node sequence is used as the updated service issue path.

[0012] In some embodiments, the path for generating service questions by sequentially connecting selected questionnaire items and operational metrics in chronological order further includes: For each service issue path corresponding to each event interval, record the path number, the sequence of questionnaire questions in the path, and the sequence of operational indicators according to the event interval identifier, and generate a path record table; Based on the service event type, service issue paths with the same service event type are grouped into the same path group. Within the same path group, paths with the same starting questionnaire question identifier are searched, and the starting part of these paths is extracted as a prefix segment. Paths with the same ending operational indicator identifier are searched, and the ending part of these paths is extracted as a suffix segment. Prefix and suffix segments are connected according to the order of the corresponding event intervals in the time slice sequence and the connection relationship of path nodes in the same path group to form a cross-event path node sequence, and the cross-event path node sequence is used as the service problem path.

[0013] In some embodiments, within a service issue path, for each path, the questionnaire view and indicator view on the path, as well as the event interval corresponding to the path, are obtained, and the visualized access trajectory is recorded according to the order of the questionnaire view and indicator view in the service issue path, including: For each service issue path, the preset view configuration table is queried based on the questionnaire question identifier and the operation indicator identifier. The questionnaire question identifier is mapped to the questionnaire view identifier, and the operation indicator identifier is mapped to the indicator view identifier, generating a path view table indexed by the order of the path nodes. In the path view table, each view identifier is associated with a service question path, an event interval identifier, and its position within the event interval (before, during, or after the event). The questionnaire view identifiers and indicator view identifiers are then arranged consecutively according to the path node order to form a view segment set. For each service issue path, the view identifiers are read sequentially from the path nodes in the view segment set, and the sequence of view identifiers is associated with the corresponding event interval identifiers and recorded as a visual access trajectory.

[0014] In some embodiments, recording the visualized access trajectory according to the order of the questionnaire view and the indicator view in the service issue path further includes: Extract path identifiers, view identifiers, and interaction types from the front-end access records of the park's operation cloud platform to generate an operation log table indexed by access order. Interaction types include view opening, view switching, and view returning. Associate the path identifier and view identifier in the operation log table with the path node order in the path view table, filter out operation records whose access order includes the path node order, and generate a view operation table indexed by the path node order. For each service issue path, the view identifier and its interaction type are read sequentially from the path node in the view operation table. The view identifier sequence and interaction type sequence are combined with the event interval identifier and recorded as a visual access trajectory.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention receives questionnaire data, operational indicator data, and service event data. It combines the questionnaire data and operational indicator data into a time-slice sequence with a unified time step. Within this sequence, event intervals are set around the occurrence time of each service event, aligning questionnaire questions and park operational indicators under the same timeframe. Subsequently, questionnaire change types are categorized based on changes in responses within adjacent time slices, and indicator change types are categorized based on changes in numerical values ​​within adjacent time slices. Linkage segments are generated within the event intervals based on the correspondence between these two types of changes. Anchor point pairs are precisely defined as those where the linkage segments are continuous within the event interval and interrupted outside the event interval. This allows for the precise definition of the event interval's location within the event interval. Within the domain, the alignment and consistency between questionnaire questions and park operation indicators are strengthened, and irrelevant linkages are suppressed. Within the set of anchor points in each event interval, questionnaire questions in the time slice before the event and operation indicators in the time slice after the event are selected and connected in chronological order to generate a service issue path. This ensures that changes related to service events are tightly linked and reduces response delays in cross-source linkages. Finally, for each path, the questionnaire view and indicator view on the path, as well as the corresponding event interval, are obtained. The visual access trajectory is recorded according to the order of the questionnaire view and indicator view in the service issue path, so that the service issue location path remains traceable at the visualization level. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a method for multi-dimensional item relevance modeling and visualization analysis in a questionnaire, as described in this invention. Figure 2 This is a schematic diagram of a method for modeling and visualizing the correlation of multi-dimensional items in a questionnaire, as described in this invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains. The terms “first,” “second,” and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as “comprising” or “including” mean that the element or object preceding the term covers the element or object listed after the term and its equivalents, without excluding other elements or objects. Terms such as “connection” or “linked” are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect.

[0019] In the technical solution of this invention, the user information (including but not limited to user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, and displayed data) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data are all carried out in accordance with relevant laws, regulations, and standards, and necessary confidentiality measures have been taken. They do not violate public order and good morals, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0020] Example 1 Please see Figure 1 As shown in the figure, this embodiment discloses a method for multi-dimensional item relevance modeling and visualization analysis in questionnaires, including: S10: Receive questionnaire data, operational indicator data, and service event data. Combine the questionnaire data and operational indicator data into a time slice sequence with a unified time step. Set the event interval in the time slice sequence around the time of each service event, covering the time slices before and after it. In this embodiment, receiving questionnaire data, operational indicator data, and service event data and combining them into a time-slice sequence with a unified time step is to align data from different sources on the same time axis. This ensures that each time slice simultaneously carries the questionnaire response status and operational indicator value status within that time period. Based on this, several time slices are extended forward and backward around the occurrence time of each service event, forming an event interval. This ensures that each service event record corresponds to a time range containing contextual information before and after the event, thereby providing a unified and traceable time framework for subsequent analysis of the linkage between questionnaire change types and indicator change types, determination of anchor point pairs, and generation of service issue paths within the event interval.

[0021] For example, assuming a uniform time step of 1 hour, the following time-slice sequence can be generated to analyze the operation of a certain park on a certain day. 08:00, Summary of questionnaire responses and status of operational metrics; 09:00, Summary of questionnaire responses and status of operational metrics; 10:00 AM: Summary of questionnaire responses and operational metrics status; When a service event occurs at 09:15, the time slot containing 09:00 can be taken as the time slot where the service event occurred. According to preset rules, it can be extended forward to the 08:00 time slot and backward to the 10:00 time slot. The three time slots of 08:00, 09:00 and 10:00 constitute the event interval of the service event. Such an event interval includes both the changes in user perception and facility operation before the event and the recovery or deterioration of operational indicators after the event. This is conducive to a detailed analysis of the dynamic relationship between questionnaire questions and operational indicators in subsequent steps.

[0022] In the time-slice sequence, set the event interval around the occurrence time of each service event, covering the time slices before and after it, including: From the questionnaire data, the number of responses to questionnaire questions corresponding to service event types within each time segment in the park dimension is counted, a questionnaire statistics table is generated, and the time segments in the questionnaire statistics table that contain responses to questionnaire questions are grouped into a questionnaire segment set. From the operational indicator data, the numerical status of operational indicators corresponding to service event types within each time slice is statistically analyzed at the park level, an indicator status table is generated, and the time slices in which the numerical status of operational indicators in the indicator status table changes are combined into an indicator slice set. For each service event record in the service event data, with the time slice where the service event occurred as the center, a search is performed simultaneously forward and backward along the questionnaire slice set and the indicator slice set in the time slice sequence. Time slices that continuously belong to the questionnaire slice set or the indicator slice set are formed into time slice segments, and the time range corresponding to the time slice segment is determined as the event interval of the service event record in the time slice sequence.

[0023] It should be noted that generating a questionnaire statistics table and extracting a questionnaire slice set from the questionnaire data in the time slice sequence refers to counting the responses to questionnaire questions related to a certain type of service event within each time slice at the park level. The number of responses to each question or option in each time slice is recorded in the questionnaire statistics table, and time slices with at least one question response record are selected to form a questionnaire slice set. This can identify which time slices have user subjective feedback related to this type of service event at the time dimension, providing a candidate set for subsequent searches of time slices related to user perception around the occurrence time of service events.

[0024] It is understandable that generating an indicator status table from operational indicator data and extracting indicator slice sets refers to statistically analyzing the status of operational indicator values ​​corresponding to service event types within each time slice at the park level. For example, based on preset operational intervals, indicator values ​​are divided into normal, tense, or alarm states. The operational status of each time slice is recorded in the indicator status table, and time slices whose operational status changes relative to the previous time slice are selected to form indicator slice sets. In this way, objective operational changes related to service event types can be captured at the time dimension, so that the construction of subsequent event intervals not only considers the subjective perception changes reflected in the questionnaire slice sets, but also the objective operational changes reflected in the indicator slice sets.

[0025] For example, for a service incident type called "air conditioning cooling complaint" in a certain park, the number of responses to questionnaire questions related to air conditioning user experience can be counted in three time slots: 08:00, 09:00, and 10:00, resulting in a questionnaire statistics table excerpt as follows: At 08:00, Question A has been answered 10 times; At 09:00, Question A has been answered 25 times; At 10:00, Question A has been answered 5 times; Simultaneously, based on the operating indicators of the air conditioning equipment within the same time slot, the operating status is recorded as follows: At 08:00, the air outlet temperature was normal. At 09:00, the air outlet temperature is slightly high; At 10:00, the air outlet temperature returned to normal. Therefore, the time slices 08:00, 09:00, and 10:00 where questionnaire responses exist can be grouped into a questionnaire slice set, and the time slices 09:00 and 10:00 where operational status changes can be grouped into an indicator slice set. Based on this, for a service event record that occurs at 09:15, the system searches forward and backward along both the questionnaire slice set and the indicator slice set, with the 09:00 time slice as the center. The time slices that continuously belong to either the questionnaire slice set or the indicator slice set between 08:00 and 10:00 are formed into a time slice segment. The time range covered by this time slice segment is determined as the event interval of the service event in the time slice sequence, thus ensuring that the event interval covers both the time period when user subjective complaints are concentrated and the time period when operational indicators are abnormal and recovering.

[0026] Setting an event interval in the time-slice sequence that covers the time slices before and after each service event occurrence time also includes: At the park level, the event intervals corresponding to each service event record are summarized, and a service attribution table is established. The service attribution table is indexed by time slices and records a list of service event record identifiers containing that time slice. In the service attribution table, for time slices that have multiple service event records, based on the service event type corresponding to the time slice, time slices that are adjacent and have all the same service event type are grouped into the same type of segment, and time slices that are adjacent and do not all have the same service event type are grouped into different types of segment. For each service event record, within its corresponding event interval, the time slice in the same type of segment is defined as the primary impact time slice of the service event record, and the time slice in the different type of segment is defined as the secondary impact time slice of the service event record. The time range covered by the primary impact time slice and the secondary impact time slice is used as the updated event interval of the service event record in the time slice sequence.

[0027] It should be noted that summarizing the event intervals corresponding to each service event record at the park level and establishing a service attribution table means that, after determining the initial event interval for each service event, each time slice in the time slice sequence is used as an index to compile a list of service event record identifiers that contain that time slice in all event intervals. The correspondence between the time slice and the service event record identifiers it contains is recorded in the service attribution table. In this way, those skilled in the art can clearly know whether multiple service events overlap within the same time slice and the types and identifiers of these service events, providing data support for subsequent division of similar and dissimilar segments and further refinement of event intervals.

[0028] Understandably, in the service attribution table, time slices with multiple service event records are categorized into similar and dissimilar segments based on the service event type corresponding to the time slice. This is to distinguish between time segments in the time slice sequence that are continuously dominated by the same type of service event and time segments that are influenced by a mixture of multiple service event types. Similar segments consist of adjacent time slices with all service event types being the same, reflecting the concentrated outbreak or sustained impact of a certain type of service problem within a continuous period. Dissimilar segments consist of adjacent time slices with service event types that are not all the same, reflecting the complex scenario where multiple types of service events are intertwined and superimposed in time. This categorization allows different time slices to be assigned different influencing roles when updating event intervals later.

[0029] For example, we can assume that in a certain time slice sequence, the 08:00 time slice only contains the service event type "air conditioning cooling complaint", the 09:00 time slice contains both the service event types "air conditioning cooling complaint" and "elevator congestion complaint", and the 10:00 time slice only contains the service event type "elevator congestion complaint". According to the service attribution table, 08:00 can be regarded as a time slice occupied solely by air conditioning-related service events, 09:00 as a time slice where air conditioning and elevator-related service events coexist, and 10:00 as a time slice occupied solely by elevator-related service events. The 08:00 time slices that are adjacent in time and have all the same service event type can form a segment of the same type. The adjacent 08:00 and 09:00, and 09:00 and 10:00, which do not have all the same service event type, can form segments of different types.

[0030] Based on this, for a specific air conditioning service event record, the time slice within its event interval that contains only air conditioning events can be defined as the primary impact time slice, and the time slice that contains both air conditioning events and other types of events can be defined as the secondary impact time slice. The time range covered by both the primary and secondary impact time slices is then used as the updated event interval. This ensures that the event interval of the service event highlights both the time period in which it has a dominant influence and the time period in which it is affected by other events, thereby providing more refined time-layered information for subsequently constructing service problem paths and visualizing access trajectories.

[0031] S20: Classify questionnaire change types according to changes in responses in adjacent time slices, classify indicator change types according to changes in values ​​in adjacent time slices, generate linkage segments in the event interval according to the correspondence between questionnaire change types and indicator change types, and determine the set of anchor point pairs between questionnaire items and operational indicators based on the situation that linkage segments are continuous within the event interval and interrupted outside the event interval. In this embodiment, the questionnaire change type is classified according to the change in responses in adjacent time slices, and the indicator change type is classified according to the change in values ​​in adjacent time slices. This is to abstract the original questionnaire response count and operational indicator values, which are originally difficult to compare directly, into a set of alignable discrete change events. This allows those skilled in the art to identify questionnaire change type "decline in satisfaction" and indicator change type "deterioration in operational status" in a unified time slice sequence. Based on this, linkage segments are generated within the event interval according to the co-occurrence of the two types of change. Furthermore, the anchor point pair set between questionnaire items and operational indicators is screened based on the characteristic that these linkage segments remain continuous within the event interval but no longer appear outside the event interval. This ensures that the obtained anchor point pairs can reflect the dominant correlation before and after the service event, while avoiding interference from accidental resonances in non-event contexts.

[0032] Within the event interval, generate linked segments based on the correspondence between questionnaire change types and indicator change types, including: In the park's operation cloud platform, service dimension tags are configured for questionnaire items, operation dimension tags are configured for operation indicators, and a dimension mapping table is established to record the pairing relationship between service dimension tags and operation dimension tags. For each questionnaire item in the time slice sequence, a questionnaire change table is generated based on the comparison of the number of responses to each option recorded in the questionnaire statistics table in adjacent time slices. In the questionnaire change table, records where the number of responses to high-scoring options decreases and the number of responses to medium- or low-scoring options increases are classified as questionnaire change types, and the corresponding service dimension tags are recorded. For each operational indicator in the time slice sequence, the operational status of each time slice is determined based on the position of the operational indicator value in the preset operational segment within the adjacent time slices. An indicator change table is generated, and records in the indicator change table that have changed from normal operation status to critical operation status or from critical operation status to alarm operation status are classified as indicator change types, and the corresponding operational dimension labels are registered. For each questionnaire question and each operational indicator combination within each event interval, the service dimension label and the operational dimension label are confirmed to be able to be paired in the dimension mapping table. Within the adjacent time slices covered by the event interval, when there are questionnaire change type records in the questionnaire change table and indicator change type records in the indicator change table, the time range corresponding to the time slice sequence formed by the consecutive arrangement of the corresponding adjacent time slices is determined as the linkage segment.

[0033] It should be noted that when generating linkage segments based on the correspondence between questionnaire change types and indicator change types within an event interval, the process begins by configuring service dimension labels for questionnaire items and operational dimension labels for operational indicators in the park's operation cloud platform, and establishing a dimension mapping table. This conceptually links questionnaire items such as "air conditioning temperature comfort" and "elevator waiting time" with operational indicators such as "overall air conditioning load" and "elevator passenger flow" under the same service or operational dimension. Then, for each questionnaire item in the time slice sequence, the questionnaire statistics table records of adjacent time slices are compared, and the situation of "high scores decreasing and medium and low scores increasing" is abstracted as a questionnaire change type. Next, for each operational indicator, the operational indicator status of adjacent time slices is compared, and the situation of "from normal to tense or alarm" is abstracted as an indicator change type. Finally, within each event interval, only for the pairable question and indicator combinations in the dimension mapping table, it is checked whether the questionnaire change type and indicator change type appear simultaneously in adjacent time slices covered by the event interval. The time ranges corresponding to these consecutively appearing adjacent time slices are marked as linkage segments. In this way, complex numerical fluctuations and response changes are converged into a small number of "questionnaire deterioration - indicator deterioration" linkage intervals with business meaning.

[0034] For example, the service dimension label for the service event type "air conditioning cooling complaint" can be configured as "indoor comfort," and the operational dimension label for the operational indicator "overall air conditioning load rate" can be configured as "cooling load." The dimension mapping table records that "indoor comfort" can be paired with "cooling load." In the time slots 08:00, 09:00, and 10:00, assuming that in the 09:00 time slot, the number of high-scoring responses to questions related to air conditioning experience in the questionnaire statistics table decreases from 20 to 8, while the number of medium- and low-scoring responses increases from 5 to 18, and the indicator change table records that the "overall air conditioning load rate" changes from a normal operating state to a tense operating state in the 09:00 time slot, then the adjacent time slot sequence containing the 09:00 time slot can be identified as a linkage segment. This linkage segment represents the synchronous occurrence of the deterioration of the user's subjective perception and the increase in equipment operating pressure in time within this event interval, providing a candidate link with causal clues for subsequent selection of anchor point pairs.

[0035] Based on the scenario where linked segments are continuous within an event interval but interrupted outside the event interval, the set of anchor point pairs between questionnaire items and operational metrics is determined, including: For each questionnaire question and each operational indicator combination, summarize the corresponding linkage segments of the questionnaire question and the operational indicator in all event intervals, and establish a linkage position table. The linkage position table records the service event record identifier to which each linkage segment belongs and the position information of the time slice covered by the linkage segment in the event interval, whether it is in the early stage, middle stage or late stage of the event. In the linked location table, for each questionnaire question and each operational indicator combination, find linked segments with the same service event type and location information belonging to the same location category, group these linked segments into linked groups, and record the event interval identifiers involved in the linked groups. Remove all time slices covered by event intervals from the time slice sequence. Within the time slices after removal, check whether there are records of combinations of questionnaire change type and indicator change type corresponding to the linkage group based on the questionnaire change table and indicator change table. If there are no combination records within this range, register the corresponding questionnaire question and operation indicator combination as anchor point pairs, and form an anchor point pair set with all the registered anchor point pairs.

[0036] In this embodiment, after obtaining the linkage segments within multiple event intervals, the linkage segments of each questionnaire question and each operational indicator combination in all event intervals are summarized and a linkage position table is established. In one table, the number of times the question and indicator are linked under different service events, as well as whether each linkage occurs in the early, middle, or late stages of the event, can be seen simultaneously. Then, linkage segments with the same service event type and the same location information are found in the linkage position table, and they are grouped into linkage groups and the event interval identifiers involved are recorded. This can be regarded as a linkage pattern that recurs stably in the same type of service event and the same time position. Finally, all time slices covered by event intervals are removed from the time slice sequence, and only in the "non-event background" time range are the questionnaire change table and indicator change table checked to see if there are still combinations of change types with the same linkage group.

[0037] Only when it is confirmed that there is no corresponding combination within the range, the questionnaire item and the operational indicator combination are registered as anchor point pairs. The resulting set of anchor point pairs not only reflects the high correlation that occurs repeatedly in the context of service events, but also has passed the exclusion test of non-event intervals. This can effectively avoid misjudging background co-fluctuations as key nodes on the problem path, and improve the consistency of the alignment between questionnaire items and operational indicators and the traceability of the service problem location path from a technical perspective.

[0038] S30: Within the set of anchor points in each event interval, select the questionnaire items in the time slice before the event and the operational indicators in the time slice after the event, and connect the selected questionnaire items and operational indicators in chronological order to generate a service question path. It is understandable that selecting questionnaire items from the pre-event time slice and operational metrics from the post-event time slice within the anchor pair set in each event interval is to connect the changes in questionnaires related to "deteriorating user perception" and the changes in metrics related to "deteriorating operational status" in chronological order within the same service event time context. This constructs a service problem path describing the evolution of service issues. This service problem path uses the time slice sequence as the main line, taking the pre-event questionnaire items as the starting point and the post-event operational metrics as the ending point. This allows subsequent visualization to trace back the changes in user perception and equipment operation along the path. Compared to existing technologies that only provide the correlation values ​​between questions and metrics, this path-based representation is more conducive to supporting causal chain analysis and cross-departmental collaborative processing for service problem localization.

[0039] Within the set of anchor points in each event interval, questionnaire items from the pre-event time slice and operational metrics from the post-event time slice are selected. These are then connected chronologically to generate a service question path, including: For each set of anchor point pairs in each event interval, an anchor point table is established according to the questionnaire question identifier, the operation indicator identifier, and the time slice identifier. In the anchor point table, each anchor point record is marked with its time slice relative to the time of the service event, indicating whether it is in the pre-event or post-event stage. In the anchor table, anchor records with the location category of the pre-event segment are grouped into a pre-questionnaire set, and anchor records with the location category of the post-event segment are grouped into a post-indicator set. The anchor records in the pre-questionnaire set and the post-indicator set are arranged in time slice order to obtain the pre-anchor sequence and the post-anchor sequence. For each event interval, questionnaire item identifiers are extracted in the time slice order from the previous anchor sequence, and operational indicator identifiers are extracted in the time slice order from the subsequent anchor sequence. The questionnaire item identifier sequence and the operational indicator identifier sequence are connected in chronological order to form a path node sequence, and the cross-event path node sequence is used as the updated service issue path.

[0040] Within each event interval, an anchor table is established based on the anchor pair set, containing questionnaire question identifiers, operational indicator identifiers, and time slice identifiers. Each anchor record is labeled with its time slice position category, indicating whether it is in the pre-event or post-event phase. This allows for a clear distinction in the data structure between "which questionnaire questions were identified as anchors before the event" and "which operational indicators were identified as anchors after the event." Subsequently, anchor records with the position category of the pre-event phase are grouped into a pre-questionnaire set, and anchor records with the position category of the post-event phase are grouped into a post-indicator set. These are then sorted according to the time slice order to obtain the pre-anchor sequence and the post-anchor sequence.

[0041] When constructing the service issue path, for each event interval, the questionnaire question identifier in the preceding anchor sequence and the operational indicator identifier in the following anchor sequence are extracted sequentially along the time slice. The two sequences are then connected in chronological order to form a path node sequence, so that the path gradually points from the user's subjective feedback to the objective operational anomaly, forming a time-directed service issue path. Compared with a simple unordered set of anchor points, this processing logic allows those skilled in the art to intuitively present the temporal order of problem triggering, impact diffusion, and result manifestation in subsequent visualizations.

[0042] The service question path, which connects selected questionnaire questions and operational metrics in chronological order, also includes: For each service issue path corresponding to each event interval, record the path number, the sequence of questionnaire questions in the path, and the sequence of operational indicators according to the event interval identifier, and generate a path record table; Based on the service event type, service issue paths with the same service event type are grouped into the same path group. Within the same path group, paths with the same starting questionnaire question identifier are searched, and the starting part of these paths is extracted as a prefix segment. Paths with the same ending operational indicator identifier are searched, and the ending part of these paths is extracted as a suffix segment. Prefix and suffix segments are connected according to the order of the corresponding event intervals in the time slice sequence and the connection relationship of path nodes in the same path group to form a cross-event path node sequence, and the cross-event path node sequence is used as the service problem path.

[0043] Understandably, after generating service issue paths for each event interval, a path record table is generated by recording path numbers according to the event interval identifier, as well as the sequence of questionnaire question identifiers and operational indicator identifiers within the path. This allows for the global archiving and comparison of paths for different service events. Based on this, paths are divided into similar path groups according to the service event type. Within each similar path group, paths with the same starting questionnaire question identifier are searched, and their starting parts are extracted as prefix segments. Simultaneously, paths with the same ending operational indicator identifier are searched, and their ending parts are extracted as suffix segments. Then, according to the chronological order of the event intervals in the time slice sequence and the connection relationship between path nodes, the prefix and suffix segments are concatenated into a cross-event path node sequence, which serves as the service issue path. This allows for the depiction of common evolutionary patterns of similar service issues across multiple event intervals on the timeline. Compared to analyzing issue links only within a single event, this cross-event path approach helps operations personnel discover key questionnaire questions and key operational indicator nodes that recur frequently over a long period, providing a visually supported basis for subsequent optimization of questionnaire design and adjustment of operational strategies.

[0044] S40: In the service issue path, for each path, obtain the questionnaire view and indicator view on the path, as well as the event interval corresponding to the path, and record the visual access trajectory according to the order of the questionnaire view and indicator view in the service issue path.

[0045] In this embodiment, for each service issue path, the questionnaire view and indicator view on the path, as well as the event interval corresponding to the path, are obtained. The visualized access trajectory is recorded according to the order of the questionnaire view and indicator view in the service issue path. This is to further map the abstract service issue path constructed based on the questionnaire question identifier and the operation indicator identifier to the visualized view object actually used on the front end of the park operation cloud platform. This ensures that each path node corresponds to a specific questionnaire report page or indicator monitoring page. At the same time, the event interval information to which the path node belongs and its position category within the event interval are retained. By arranging these view identifiers according to the path order and generating a visualized access trajectory, an analysis path that can be directly accessed sequentially in the interface can be provided to the operation personnel. Compared with the technical solution that only provides the question and indicator codes, the visualized access trajectory in this embodiment can directly drive the front end view jump and linkage, which greatly reduces the cost of manual retrieval between multiple reports.

[0046] Within the service issue path, for each path, obtain the questionnaire view and indicator view along the path, as well as the corresponding event interval. Record the visualized access trajectory according to the order of the questionnaire view and indicator view within the service issue path, including: For each service issue path, the preset view configuration table is queried based on the questionnaire question identifier and the operation indicator identifier. The questionnaire question identifier is mapped to the questionnaire view identifier, and the operation indicator identifier is mapped to the indicator view identifier, generating a path view table indexed by the order of the path nodes. In the path view table, each view identifier is associated with a service question path, an event interval identifier, and its position within the event interval (before, during, or after the event). The questionnaire view identifiers and indicator view identifiers are then arranged consecutively according to the path node order to form a view segment set. For each service issue path, the view identifiers are read sequentially from the path nodes in the view segment set, and the sequence of view identifiers is associated with the corresponding event interval identifiers and recorded as a visual access trajectory.

[0047] It should be noted that when querying the preset view configuration table based on the questionnaire question identifier and the operation indicator identifier, the questionnaire view identifier corresponding to each questionnaire question can be maintained on the front end, and the indicator view identifier corresponding to each operation indicator can be maintained on the front end. Through this mapping relationship, the abstract identifier in the path is converted into a specific callable view identifier. On this basis, a path view table is generated with the path node order as the index. In the path view table, the event interval identifier associated with the service question path and the position information of the view identifier in the event interval (before, middle or after the event) are added to each view identifier. Then, the questionnaire view identifier and indicator view identifier are arranged continuously according to the path node order to form a view segment set.

[0048] In this example, three nodes in a service issue path can be mapped sequentially to "Air Conditioner Satisfaction Questionnaire View Icon", "Air Conditioner Maintenance Work Order Statistics View Icon", and "Air Conditioner Energy Consumption Index View Icon". Event interval identifier E1 and corresponding front, middle and back position information are uniformly attached. By reading these view icons in order in the view segment set and combining them with the event interval identifier to record as a visual access trajectory, it can be directly used to drive the front end to display the questionnaire and index information related to the same service event in the order of the path, so that the path is not only connected at the data level, but also has executability at the visualization level.

[0049] The visualization of the access path is recorded according to the order of the questionnaire view and indicator view in the service issue path, and also includes: Extract path identifiers, view identifiers, and interaction types from the front-end access records of the park's operation cloud platform to generate an operation log table indexed by access order. Interaction types include view opening, view switching, and view returning. Associate the path identifier and view identifier in the operation log table with the path node order in the path view table, filter out operation records whose access order includes the path node order, and generate a view operation table indexed by the path node order. For each service issue path, the view identifier and its interaction type are read sequentially from the path node in the view operation table. The view identifier sequence and interaction type sequence are combined with the event interval identifier and recorded as a visual access trajectory.

[0050] Furthermore, after generating the theoretically visualized access trajectory, an operation log table is generated by extracting path identifiers, view identifiers, and interaction types from the front-end access records of the park operation cloud platform. The path identifiers and view identifiers in the operation log table are associated with the path node order in the path view table. This allows the filtering of operation records that actually contain the path node order in the access order, forming a view operation table indexed by the path node order.

[0051] In this example, the view operation table can be used to record the access order and interaction type of an operations and maintenance personnel when handling event interval E1: first opening the "Air Conditioner Satisfaction Questionnaire View," then switching to the "Air Conditioner Maintenance Work Order Statistics View," and finally returning to the "Air Conditioner Energy Consumption Index View." Ultimately, the view identifier sequence and interaction type sequence are combined with the event interval identifier and recorded as a visual access trajectory. This can verify whether the service problem path generated by the system is actually adopted and followed, and it can also precipitate real access behavior as path usage data, providing a quantitative basis for optimizing path recommendations, evaluating the latency of visual linkage response, and improving the efficiency of service problem location. Compared with the existing technology that only builds static paths offline, this embodiment achieves closed-loop tracking of service problem paths at the visualization level by combining front-end operation logs.

[0052] Example 2 Please see Figure 2 As shown, based on the same inventive concept, this embodiment discloses a system for multi-dimensional item relevance modeling and visualization analysis in questionnaire data. For details not covered in this embodiment, please refer to the relevant sections of Embodiment 1. The system includes: Sequence building module: It is used to receive questionnaire data, operational indicator data and service event data, combine questionnaire data and operational indicator data into time slice sequence according to a unified time step, and set the event interval around the time of each service event to cover the time slices before and after. Anchor point generation module: used to classify questionnaire change types according to changes in responses in adjacent time slices, classify indicator change types according to changes in values ​​in adjacent time slices, generate linkage segments in the event interval according to the correspondence between questionnaire change types and indicator change types, and determine the set of anchor point pairs between questionnaire items and operational indicators based on the case that linkage segments are continuous within the event interval and interrupted outside the event interval. Path generation module: Used to select questionnaire questions in the time slice before the event and operational indicators in the time slice after the event within the set of anchor point pairs in each event interval, and generate service question paths by connecting the selected questionnaire questions and operational indicators in chronological order. Visual Recording Module: Used to obtain the questionnaire view and indicator view on each path in the service issue path, as well as the event interval corresponding to the path, and record the visual access trajectory according to the order of the questionnaire view and indicator view in the service issue path.

[0053] The accompanying drawings of the embodiments of this invention only involve the structures involved in the embodiments of this invention. Other structures can refer to the general design. In the absence of conflict, the features of the same embodiment and different embodiments of this invention can be combined with each other. The above are only specific implementations of this invention, but the protection scope of this invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this invention should be included within the protection scope of this invention. Therefore, the protection scope of this invention should be determined by the protection scope of the claims.

Claims

1. A method for multi-dimensional item relevance modeling and visualization analysis in questionnaires, characterized in that, include: Receive questionnaire data, operational indicator data, and service event data. Combine the questionnaire data and operational indicator data into a time slice sequence with a unified time step. Set the event interval in the time slice sequence around the time of each service event, covering the time slices before and after it. The questionnaire change type is classified according to the change in responses in adjacent time slices, and the indicator change type is classified according to the change in values ​​in adjacent time slices. Linkage segments are generated in the event interval according to the correspondence between the questionnaire change type and the indicator change type. The anchor point pair set between the questionnaire questions and the operational indicators is determined by the case that the linkage segments are continuous in the event interval and interrupted outside the event interval. Within the set of anchor points in each event interval, select questionnaire items in the time slice before the event and operational indicators in the time slice after the event, and connect the selected questionnaire items and operational indicators in chronological order to generate a service question path. In the service issue path, for each path, obtain the questionnaire view and indicator view on the path, as well as the event interval corresponding to the path, and record the visual access trajectory according to the order of the questionnaire view and indicator view in the service issue path.

2. The method for multi-dimensional item relevance modeling and visualization analysis in questionnaires according to claim 1, characterized in that, In the time-slice sequence, set the event interval around the occurrence time of each service event, covering the time slices before and after it, including: From the questionnaire data, the number of responses to questionnaire questions corresponding to service event types within each time segment in the park dimension is counted, a questionnaire statistics table is generated, and the time segments in the questionnaire statistics table that contain responses to questionnaire questions are grouped into a questionnaire segment set. From the operational indicator data, the numerical status of operational indicators corresponding to service event types within each time slice is statistically analyzed at the park level, an indicator status table is generated, and the time slices in which the numerical status of operational indicators in the indicator status table changes are combined into an indicator slice set. For each service event record in the service event data, with the time slice where the service event occurred as the center, a search is performed simultaneously forward and backward along the questionnaire slice set and the indicator slice set in the time slice sequence. Time slices that continuously belong to the questionnaire slice set or the indicator slice set are formed into time slice segments, and the time range corresponding to the time slice segment is determined as the event interval of the service event record in the time slice sequence.

3. The method for multi-dimensional item relevance modeling and visualization analysis in questionnaires according to claim 2, characterized in that, Setting an event interval in the time-slice sequence that covers the time slices before and after each service event occurrence time also includes: At the park level, the event intervals corresponding to each service event record are summarized, and a service attribution table is established. The service attribution table is indexed by time slices and records a list of service event record identifiers containing that time slice. In the service attribution table, for time slices that have multiple service event records, based on the service event type corresponding to the time slice, time slices that are adjacent and have all the same service event type are grouped into the same type of segment, and time slices that are adjacent and do not all have the same service event type are grouped into different types of segment. For each service event record, within its corresponding event interval, the time slice in the same type of segment is defined as the primary impact time slice of the service event record, and the time slice in the different type of segment is defined as the secondary impact time slice of the service event record. The time range covered by the primary impact time slice and the secondary impact time slice is used as the updated event interval of the service event record in the time slice sequence.

4. The method for multi-dimensional item relevance modeling and visualization analysis in questionnaires according to claim 1, characterized in that, Within the event interval, generate linked segments based on the correspondence between questionnaire change types and indicator change types, including: In the park's operation cloud platform, service dimension tags are configured for questionnaire items, operation dimension tags are configured for operation indicators, and a dimension mapping table is established to record the pairing relationship between service dimension tags and operation dimension tags. For each questionnaire item in the time slice sequence, a questionnaire change table is generated based on the comparison of the number of responses to each option recorded in the questionnaire statistics table in adjacent time slices. In the questionnaire change table, records where the number of responses to high-scoring options decreases and the number of responses to medium- or low-scoring options increases are classified as questionnaire change types, and the corresponding service dimension tags are recorded. For each operational indicator in the time slice sequence, the operational status of each time slice is determined based on the position of the operational indicator value in the preset operational segment within the adjacent time slices. An indicator change table is generated, and records in the indicator change table that have changed from normal operation status to critical operation status or from critical operation status to alarm operation status are classified as indicator change types, and the corresponding operational dimension labels are registered. For each questionnaire question and each operational indicator combination within each event interval, the service dimension label and the operational dimension label are confirmed to be able to be paired in the dimension mapping table. Within the adjacent time slices covered by the event interval, when there are questionnaire change type records in the questionnaire change table and indicator change type records in the indicator change table, the time range corresponding to the time slice sequence formed by the consecutive arrangement of the corresponding adjacent time slices is determined as the linkage segment.

5. The method for multi-dimensional item relevance modeling and visualization analysis in questionnaires according to claim 4, characterized in that, Based on the scenario where linked segments are continuous within an event interval but interrupted outside the event interval, the set of anchor point pairs between questionnaire items and operational metrics is determined, including: For each questionnaire question and each operational indicator combination, summarize the corresponding linkage segments of the questionnaire question and the operational indicator in all event intervals, and establish a linkage position table. The linkage position table records the service event record identifier to which each linkage segment belongs and the position information of the time slice covered by the linkage segment in the event interval, whether it is in the early stage, middle stage or late stage of the event. In the linked location table, for each questionnaire question and each operational indicator combination, find linked segments with the same service event type and location information belonging to the same location category, group these linked segments into linked groups, and record the event interval identifiers involved in the linked groups. Remove all time slices covered by event intervals from the time slice sequence. Within the time slices after removal, check whether there are records of combinations of questionnaire change type and indicator change type corresponding to the linkage group based on the questionnaire change table and indicator change table. If there are no combination records within this range, register the corresponding questionnaire question and operation indicator combination as anchor point pairs, and form an anchor point pair set with all the registered anchor point pairs.

6. The method for multi-dimensional item relevance modeling and visualization analysis in questionnaires according to claim 1, characterized in that, Within the set of anchor points in each event interval, questionnaire items from the pre-event time slice and operational metrics from the post-event time slice are selected. These are then connected chronologically to generate a service question path, including: For each set of anchor point pairs in each event interval, an anchor point table is established according to the questionnaire question identifier, the operation indicator identifier, and the time slice identifier. In the anchor point table, each anchor point record is marked with its time slice relative to the time of the service event, indicating whether it is in the pre-event or post-event stage. In the anchor table, anchor records with the location category of the pre-event segment are grouped into a pre-questionnaire set, and anchor records with the location category of the post-event segment are grouped into a post-indicator set. The anchor records in the pre-questionnaire set and the post-indicator set are arranged in time slice order to obtain the pre-anchor sequence and the post-anchor sequence. For each event interval, questionnaire item identifiers are extracted in the time slice order from the previous anchor sequence, and operational indicator identifiers are extracted in the time slice order from the subsequent anchor sequence. The questionnaire item identifier sequence and the operational indicator identifier sequence are connected in chronological order to form a path node sequence, and the cross-event path node sequence is used as the updated service issue path.

7. The method for multi-dimensional item relevance modeling and visualization analysis in questionnaires according to claim 6, characterized in that, The service question path, which connects selected questionnaire questions and operational metrics in chronological order, also includes: For each service issue path corresponding to each event interval, record the path number, the sequence of questionnaire questions in the path, and the sequence of operational indicators according to the event interval identifier, and generate a path record table; Based on the service event type, service issue paths with the same service event type are grouped into the same path group. Within the same path group, paths with the same starting questionnaire question identifier are searched, and the starting part of these paths is extracted as a prefix segment. Paths with the same ending operational indicator identifier are searched, and the ending part of these paths is extracted as a suffix segment. Prefix and suffix segments are connected according to the order of the corresponding event intervals in the time slice sequence and the connection relationship of path nodes in the same path group to form a cross-event path node sequence, and the cross-event path node sequence is used as the service problem path.

8. The method for multi-dimensional item relevance modeling and visualization analysis in questionnaires according to claim 1, characterized in that, Within the service issue path, for each path, obtain the questionnaire view and indicator view along the path, as well as the corresponding event interval. Record the visualized access trajectory according to the order of the questionnaire view and indicator view within the service issue path, including: For each service issue path, the preset view configuration table is queried based on the questionnaire question identifier and the operation indicator identifier. The questionnaire question identifier is mapped to the questionnaire view identifier, and the operation indicator identifier is mapped to the indicator view identifier, generating a path view table indexed by the order of the path nodes. In the path view table, each view identifier is associated with a service question path, an event interval identifier, and its position within the event interval (before, during, or after the event). The questionnaire view identifiers and indicator view identifiers are then arranged consecutively according to the path node order to form a view segment set. For each service issue path, the view identifiers are read sequentially from the path nodes in the view segment set, and the sequence of view identifiers is associated with the corresponding event interval identifiers and recorded as a visual access trajectory.

9. A method for multi-dimensional item relevance modeling and visualization analysis in questionnaires according to claim 8, characterized in that, The visualization of the access path is recorded according to the order of the questionnaire view and indicator view in the service issue path, and also includes: Extract path identifiers, view identifiers, and interaction types from the front-end access records of the park's operation cloud platform to generate an operation log table indexed by access order. Interaction types include view opening, view switching, and view returning. Associate the path identifier and view identifier in the operation log table with the path node order in the path view table, filter out operation records whose access order includes the path node order, and generate a view operation table indexed by the path node order. For each service issue path, the view identifier and its interaction type are read sequentially from the path node in the view operation table. The view identifier sequence and interaction type sequence are combined with the event interval identifier and recorded as a visual access trajectory.

Images