A public facility personalized service recommendation method and system based on big data analysis

Abstract

This application relates to the field of service recommendation technology, providing a method and system for personalized public facility service recommendation based on big data analysis. The method includes: determining multiple candidate activity areas for the user based on the user profile and activity tag information of the activity area; determining the expected environmental information and noise impact information of the candidate activity areas on users in adjacent activity areas based on the current environmental information of the candidate activity areas; determining a recommendation index for the candidate activity areas based on the normal environmental information range, the expected environmental information, and the noise impact information on users in adjacent activity areas; and selecting the service corresponding to the candidate activity area with the highest recommendation index from among the multiple candidate activity areas as the service recommended to the user. This solution can improve the accuracy of personalized public facility service recommendations.

Description

Technical Field

[0001] This application relates to the field of service recommendation technology, and in particular to a method and system for personalized service recommendation of public facilities based on big data analysis. Background Technology

[0002] To better serve the public, existing public facilities such as libraries and civic centers have widely adopted service recommendation systems. These systems typically provide personalized service suggestions by collecting users' personal information and behavioral data.

[0003] However, this recommendation method, centered on individual user preferences, often overlooks the unique nature of public facilities as shared spaces. When a system independently recommends services to a large number of users, these recommendations may lead to user congregation in specific areas, altering the local environment, such as causing noise or congestion, thus affecting the experience of other users and even conflicting with the facility's intended functions. This phenomenon causes a discrepancy between the initial intention and actual effect of personalized recommendations, failing to effectively maintain the overall service quality and order of the public environment. Summary of the Invention

[0004] This application provides a method and system for personalized service recommendation of public facilities based on big data analysis. It aims to address the technical problem that existing public facility service recommendation systems, when providing personalized services, consider user preferences in isolation while ignoring the characteristics of public spaces. This can lead to recommendations that may cause localized environmental conflicts, negatively impacting the user experience of other users and failing to effectively maintain the overall service quality and order of the public environment. To achieve the above objective, this application adopts the following technical solution:

[0005] Firstly, a method for recommending personalized public facility services based on big data analysis is provided, including: obtaining a user profile, current environmental information of each activity area in multiple activity areas of the public facility, activity tag information of each activity area, and normal environmental information range of each activity area; determining multiple candidate activity areas for the user based on the user profile and activity tag information of the activity areas; for each candidate activity area, determining the expected environmental information of the candidate activity area after the user's activity in the candidate activity area and the noise impact information on users in adjacent activity areas based on the current environmental information of the candidate activity area; the noise impact information is used to indicate whether there is no noise impact or no noise impact; for each candidate activity area, determining the recommendation index of the candidate activity area based on the normal environmental information range of the candidate activity area, the expected environmental information of the candidate activity area, and the noise impact information on users in adjacent activity areas; and selecting the service corresponding to the candidate activity area with the highest recommendation index among the multiple candidate activity areas as the service recommended to the user.

[0006] Furthermore, based on the user's user profile and the activity tag information of the activity area, multiple candidate activity areas for the user are determined from multiple activity areas, including: for each activity area in the multiple activity areas, determining the matching index of the activity area based on the user's user profile and the activity tag information of the activity area; and selecting activity areas with a matching index greater than a preset matching index threshold as candidate activity areas.

[0007] Based on this, the matching index of the activity area is determined according to the user's user profile and the activity tag information of the activity area. This includes: inputting the user's user profile and the activity tag information of the activity area into a preset deep learning model to obtain the matching degree output by the preset deep learning model, and using the matching degree as the initial matching index of the activity area; the preset deep learning model is used to determine the matching degree between the user's user profile and the activity tag information of the activity area; for each of the multiple activity areas, the number of times the user was active in the activity area in the historical time period and the first correspondence relationship are obtained; the first correspondence relationship includes a one-to-one correspondence between multiple frequency ranges and multiple first adjustment coefficients; the first adjustment coefficient corresponding to the frequency range of the activity in the first correspondence relationship is used as the target first adjustment coefficient of the activity area; the product of the initial matching index of the activity area and the target first adjustment coefficient is used as the matching index of the activity area.

[0008] In some preferred embodiments, the environmental information includes pedestrian traffic and sound decibel levels. For each of the multiple candidate activity areas, based on the current environmental information of the candidate activity area, the expected environmental information of the candidate activity area after a user's activity in the candidate activity area, as well as the noise impact information on users in adjacent activity areas, are determined. This includes: for each of the multiple candidate activity areas, determining whether the activity tag information of the candidate activity area includes a sports service tag or a communication service tag; if the activity tag information of the candidate activity area includes a sports service tag or a communication service tag, increasing the sound decibel level in the current environmental information of the candidate activity area by a first preset sound decibel level to obtain the sound decibel level in the expected environmental information of the candidate activity area; if the activity tag information of the candidate activity area does not include a sports service tag or a communication service tag... When the noise level in the current environmental information of the candidate activity area is greater than the second preset noise level, the noise level in the expected environmental information of the candidate activity area is increased by a second preset noise level. If the first preset noise level is greater than the second preset noise level, it is determined whether the activity tag information of the candidate activity area includes a group activity service tag. If the activity tag information of the candidate activity area includes a group activity service tag, the pedestrian flow in the current environmental information of the candidate activity area is increased by a preset pedestrian flow to obtain the pedestrian flow in the expected environmental information of the candidate activity area. If the activity tag information of the candidate activity area does not include a group activity service tag, the sum of the pedestrian flow in the current environmental information of the candidate activity area and 1 is used as the pedestrian flow in the expected environmental information of the candidate activity area. Based on the expected environmental information of the candidate activity area, the noise impact information on users in adjacent activity areas is determined.

[0009] Furthermore, noise impact information on users in adjacent activity areas is determined based on the expected environmental information of the candidate activity area, including: obtaining the noise sensitivity index of the adjacent activity area; determining the transmitted sound decibel value to the adjacent activity area based on the sound decibel value in the expected environmental information of the candidate activity area; and determining the noise impact information on users in the adjacent activity area based on the noise sensitivity index and the transmitted sound decibel value of the adjacent activity area.

[0010] As a technological improvement, noise impact information on users in adjacent activity areas is determined based on the noise sensitivity index and transmitted sound decibel value of adjacent activity areas. This includes: obtaining the current sound decibel value and a second correspondence between adjacent activity areas; the second correspondence includes a one-to-one correspondence between multiple noise sensitivity index ranges and multiple first adjustment coefficients; using the first adjustment coefficient corresponding to the noise sensitivity index range of the adjacent activity area in the second correspondence as the target first adjustment coefficient; using the product of the transmitted sound decibel value and the target first adjustment coefficient as the user-perceived transmitted sound decibel value; using the sum of the user-perceived transmitted sound decibel value and the current sound decibel value as the total sound decibel value of the adjacent activity areas; when the total sound decibel value of the adjacent activity areas is less than the preset sound decibel value of the adjacent activity areas, the noise impact information on users in the adjacent activity areas is determined to indicate no noise impact; when the total sound decibel value of the adjacent activity areas is greater than or equal to the preset sound decibel value of the adjacent activity areas, the noise impact information on users in the adjacent activity areas is determined to indicate noise impact.

[0011] To improve the solution, for each of the multiple candidate activity areas, a recommendation index is determined based on the normal environmental information range of the candidate activity area, the expected environmental information of the candidate activity area, and the noise impact information on users in adjacent activity areas. This includes: determining whether the expected environmental information of the candidate activity area exceeds the normal environmental information range of the candidate activity area; when the expected environmental information of the candidate activity area exceeds the normal environmental information range, determining the environmental recommendation index of the candidate activity area as 0; when the expected environmental information of the candidate activity area does not exceed the normal environmental information range of the candidate activity area, taking the difference between the normal environmental information range of the candidate activity area and the expected environmental information as the environmental information difference of the candidate activity area; obtaining a third correspondence relationship; the three correspondence relationships include a one-to-one correspondence between multiple environmental information difference ranges and multiple environmental recommendation indices; taking the environmental recommendation index corresponding to the environmental information difference range of the candidate activity area in the third correspondence relationship as the environmental recommendation index of the candidate activity area; and determining the recommendation index of the candidate activity area based on the environmental recommendation index of the candidate activity area and the noise impact information on users in adjacent activity areas.

[0012] As a further improvement, the recommendation index of the candidate activity area is determined based on the environmental recommendation index of the candidate activity area and the noise impact information on users in adjacent activity areas. This includes: for each adjacent activity area, determining whether the noise impact information of the adjacent activity area indicates noise impact; when the noise impact information of the adjacent activity area indicates no noise impact, determining the sound recommendation index corresponding to the adjacent activity area as the preset maximum sound recommendation index; when the noise impact information of the adjacent activity area indicates noise impact, obtaining the pedestrian flow and a fourth correspondence relationship in the adjacent activity area; the fourth correspondence relationship includes a one-to-one correspondence between multiple pedestrian flow ranges and multiple sound recommendation indices; using the sound recommendation index corresponding to the pedestrian flow range in the adjacent activity area in the fourth correspondence relationship as the sound recommendation index of the adjacent activity area; determining the sound recommendation index of the candidate activity area based on the sound recommendation indices of multiple adjacent activity areas; and using the sum of the sound recommendation index and the environmental recommendation index of the candidate activity area as the recommendation index of the candidate activity area.

[0013] To optimize the structure, the sound recommendation index of the candidate activity area is determined based on the sound recommendation index of multiple adjacent activity areas. This includes: obtaining the number of noise complaints against each adjacent activity area in the historical time period; normalizing the noise complaint volume of multiple adjacent activity areas and using the normalization result as the weight of each adjacent activity area; and using the weight of multiple adjacent activity areas as the weighted sum of the sound recommendation indices of the candidate activity area.

[0014] Secondly, this application also discloses a personalized service recommendation system for public facilities based on big data analysis, comprising: an acquisition device and a processing device; the acquisition device is used to acquire a user profile, current environmental information of each activity area in multiple activity areas of the public facility, activity tag information of each activity area, and normal environmental information range of each activity area; the processing device is used to determine multiple candidate activity areas for the user in multiple activity areas based on the user profile; the processing device is used to determine, for each candidate activity area in multiple candidate activity areas, the expected environmental information of the candidate activity area after the user's activity in the candidate activity area and the noise impact information on users in adjacent activity areas, based on the current environmental information of the candidate activity area; the noise impact information is used to indicate whether there is no noise impact or no noise impact; the processing device is used to determine, for each candidate activity area in multiple candidate activity areas, the recommendation index of the candidate activity area based on the normal environmental information range of the candidate activity area, the expected environmental information of the candidate activity area, and the noise impact information on users in adjacent activity areas; the processing device is used to select the service corresponding to the candidate activity area with the highest recommendation index in multiple candidate activity areas as the service recommended to the user.

[0015] Beneficial effects

[0016] This application provides a method for personalized service recommendation for public facilities based on big data analysis. The method acquires user profiles, environmental information of the public facility activity area, activity tag information, and the normal environmental information range. Based on this, it determines the user's potential activity areas according to the user profile and activity tag information. For each potential activity area, the method further predicts the expected environmental information of the area after the user's activity and the noise impact on adjacent areas. Subsequently, considering the normal environmental information range, expected environmental information, and noise impact information of the potential activity area, a recommendation index for the potential activity area is calculated. Finally, the service corresponding to the potential activity area with the highest recommendation index is recommended to the user.

[0017] Through the above technical solution, this application effectively addresses the technical problem that existing public facility service recommendation systems, when providing personalized services, neglect the characteristics of public spaces by considering user preferences in isolation. This leads to potential local environmental conflicts arising from the recommendation behavior, affecting the user experience of other users and failing to effectively maintain the overall service quality and order of the public environment. Specifically, this application introduces the assessment of expected environmental information and noise impact information, enabling the recommendation system to anticipate and avoid potential environmental conflicts, such as excessive noise or crowd congestion, while providing personalized services to users. This comprehensive consideration not only improves the service experience of individual users but, more importantly, maintains the overall service quality and order of public facilities as shared spaces, achieving a win-win situation where personalized recommendations and the public environment coexist harmoniously. Therefore, this application has significantly superior technical effects compared to existing technologies. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a method for recommending personalized public facility services based on big data analysis, provided for this application;

[0019] Figure 2 A flowchart illustrating yet another method for recommending personalized public facility services based on big data analysis provided in this application;

[0020] Figure 3 This is a schematic diagram of the structure of a personalized service recommendation system for public facilities based on big data analysis, provided for this application. Detailed Implementation

[0021] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0022] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0023] Traditional public facility service recommendation systems typically focus on individual user preferences, providing personalized service suggestions by collecting users' personal information and behavioral data. However, this approach often overlooks the unique nature of public facilities as shared spaces. When the system independently recommends services to a large number of users, it can lead to user congestion in specific areas, altering the local environment, such as causing noise or congestion, thus affecting the experience of other users and even conflicting with the facility's intended functions. This phenomenon causes a discrepancy between the initial intention and actual effect of personalized recommendations, failing to effectively maintain the overall service quality and order of the public environment.

[0024] In this regard, such as Figure 1 As shown, this application proposes a method for recommending personalized public facility services based on big data analysis, including:

[0025] S101. Obtain the user profile, the current environmental information of each activity area in multiple activity areas of public facilities, the activity tag information of each activity area, and the normal environmental information range of each activity area.

[0026] S102. Based on the user profile and activity tag information of the activity area, determine multiple candidate activity areas for the user in multiple activity areas.

[0027] S103. For each of the multiple candidate activity areas, based on the current environmental information of the candidate activity area, determine the expected environmental information of the candidate activity area after the user's activity in the candidate activity area, as well as the noise impact information on users in adjacent activity areas; the noise impact information is used to indicate whether there is no noise impact or no noise impact.

[0028] S104. For each of the multiple candidate activity areas, determine the recommendation index of the candidate activity area based on the normal environmental information range of the candidate activity area, the expected environmental information of the candidate activity area, and the noise impact information on users in adjacent activity areas.

[0029] S105. The service corresponding to the candidate activity area with the highest recommendation index among multiple candidate activity areas shall be the service recommended to the user.

[0030] This application, by comprehensively considering user preferences, real-time environmental information of the activity area, expected environmental changes, and potential noise impact on adjacent areas, can provide users with more reasonable personalized service recommendations that also take into account the overall benefits of the public environment. This method effectively avoids the problems of local environmental degradation and decreased user experience that may occur with traditional recommendation systems, thereby improving the service quality and user satisfaction of public facilities.

[0031] To better understand the technical solution proposed in this application, some key terms involved will be explained first.

[0032] A "user profile" refers to a set of user characteristics constructed by collecting and analyzing user behavior data, preferences, historical activity records, and other information, used to depict a user's interests, needs, and habits. For example, a user profile may include a user's age, occupation, hobbies (such as reading, sports, and art), historically visited activity areas, and preferred environments (such as quiet or lively).

[0033] "Activity areas" refer to physical spaces within public facilities that have specific functions or uses, such as reading areas, study rooms, discussion areas, and coffee areas in libraries, or gyms, dance studios, and meeting rooms in civic centers. Each activity area may accommodate different types of activities and user groups.

[0034] "Current environmental information" refers to real-time environmental status data of the activity area at a specific moment, such as pedestrian traffic, sound level (decibels), temperature, and humidity. This information can be collected in real time using devices such as sensors, cameras, and Wi-Fi probes.

[0035] "Activity tag information" refers to a set of tags that describe the main functions of an activity area and the activities suitable for it, such as "quiet reading," "group discussion," "fitness exercise," and "art appreciation." These tags help the system understand the attributes of the activity area.

[0036] "Normal environmental information range" refers to the reasonable range that environmental information (such as pedestrian traffic and sound decibel levels) should maintain in each activity area under normal operating conditions. Exceeding this range may indicate environmental abnormalities or a decline in service quality.

[0037] "Expected environmental information" refers to the prediction of potential changes in the environmental information of a selected activity area after a user enters or engages in activities there. For example, if a user plans to hold a group activity in a certain area, the system will predict how the pedestrian traffic and sound decibel levels in that area will increase.

[0038] "Noise impact information" refers to the assessment of the potential noise impact on adjacent activity areas after a user's activity in the selected activity area, indicating whether this impact is "no noise impact" or "noise impact".

[0039] The method proposed in this application aims to achieve personalized service recommendations for public facilities through refined data analysis and multi-dimensional considerations.

[0040] First, the system needs to acquire a series of basic data, including user profiles, current environmental information of various activity areas within the public facility, activity tag information, and the normal environmental information range. Acquiring this data is fundamental for subsequent analysis and recommendations. For example, user profiles can be obtained through information filled in during user registration, historical behavioral data (such as borrowing records and activity participation records), and questionnaires. Current environmental information of activity areas can be collected in real time by sensors deployed in various areas (such as pedestrian flow sensors and sound sensors). Activity tag information can be pre-set by the facility management or automatically generated through analysis of area functions and historical activity data. The normal environmental information range is set by the facility management based on the functional positioning and management requirements of each area.

[0041] After obtaining the aforementioned basic data, the system will initially filter out several candidate activity areas based on the user's profile and the activity tags of the activity areas. The purpose of this step is to narrow down the recommendation scope, considering only those areas that initially match the user's preferences and needs. For example, if the user's profile indicates a preference for quiet reading, the system will prioritize activity areas tagged with "quiet reading" or "study." If the user's profile indicates a preference for group discussions, the system will prioritize activity areas tagged with "group discussion" or "meeting."

[0042] Next, for each candidate activity area, the system predicts the expected environmental information that might be generated after a user's activity in that area, based on its current environmental information, and assesses the potential noise impact on adjacent activity areas. This step is one of the core innovations of this application, as it incorporates the dynamic impact of user activity on the environment. For example, if a user plans to hold a group discussion in a discussion area with currently low foot traffic, the system predicts that the foot traffic and sound decibel level in that area will increase, forming the expected environmental information. Simultaneously, the system also assesses the potential noise impact of this increase on the adjacent reading area. The noise impact information clearly indicates whether it is "no noise impact" or "noise impact".

[0043] Subsequently, for each candidate activity area, the system comprehensively considers the normal environmental information range, expected environmental information, and noise impact information on adjacent activity areas to determine the recommendation index for that candidate activity area. The recommendation index is a quantitative indicator used to measure the suitability of the activity area for the current user and its friendliness to the public environment. For example, if the expected environmental information of a candidate activity area exceeds its normal environmental information range (such as excessive pedestrian traffic or noise), or if it has a "noise impact" on adjacent areas, its recommendation index may be significantly reduced. Conversely, if the expected environmental information is within the normal range and has no noise impact on adjacent areas, the recommendation index will be higher.

[0044] Finally, the system will select the service corresponding to the activity area with the highest recommendation index from among the multiple candidate activity areas as the final service recommended to the user. In this way, the system not only considers the user's own preferences, but also takes into account the overall environmental capacity of public facilities and the potential impact on other users, thus achieving a more intelligent and responsible personalized service recommendation.

[0045] The overall working principle of this application lies in its ability to overcome the limitations of traditional recommendation systems that only focus on individual user preferences, by incorporating public facilities as dynamic environmental factors of shared spaces into the recommendation decision-making process. Specifically, when a user requests service recommendations, the system first analyzes the user's profile and activity tag information of the activity area to initially filter out candidate activity areas related to the user's needs. This initial screening ensures the initial relevance of the recommendations.

[0046] Subsequently, the system conducts in-depth environmental impact predictions for each candidate activity area. It not only considers how the area's own environment (such as pedestrian traffic and sound levels) will change after a user enters, but more importantly, it predicts the potential noise impact of these changes on adjacent activity areas. For example, if a user plans to engage in an activity that may generate significant noise, the system simulates the propagation of that noise within public facilities and assesses its impact on surrounding quiet areas. This proactive environmental impact assessment allows the system to anticipate potential environmental conflicts before making recommendations.

[0047] After obtaining expected environmental information and noise impact information, the system integrates these dynamic environmental factors with the normal environmental information range of the activity area to calculate a recommendation index for each candidate activity area. This recommendation index is a multi-dimensional comprehensive score that not only reflects the area's suitability for the current user, but more importantly, it quantifies the degree to which the recommendation behavior maintains the overall order of the public environment and the user experience. For example, even if an area highly matches the user's preferences, but is expected to cause environmental overload or severe noise interference, its recommendation index will be significantly reduced.

[0048] Ultimately, the system selects the service corresponding to the candidate activity area with the highest recommendation index. Through this mechanism, this application effectively solves the problems of localized environmental degradation and decreased user experience caused by isolated recommendations in existing technologies. It ensures that the recommended services not only meet the user's personalized needs but also maintain the overall service quality and harmonious environment of public facilities, achieving a balance between personalized services and public interests.

[0049] Specifically, such as Figure 2As shown, in the above-mentioned method for recommending personalized public facilities services based on big data analysis, the step of determining multiple candidate activity areas for users in multiple activity areas can be further refined as follows.

[0050] S201. For each of the multiple activity areas, determine the matching index of the activity area based on the user's user profile and the activity tag information of the activity area.

[0051] S202. Select the active regions whose matching index is greater than the preset matching index threshold as candidate active regions.

[0052] The matching index of an activity area refers to the degree of fit or relevance between a user's profile and the activity tag information of that activity area. This index can reflect a user's preference for or potential interest in that activity area. For example, a user's profile may include information such as their age, gender, occupation, and hobbies, while the activity tag information of an activity area may include the type of services offered in that area (such as fitness, reading, socializing, entertainment, etc.), themes, or characteristics. By comprehensively analyzing this information, the degree of matching between a user and a specific activity area can be calculated. The preset matching index threshold is a pre-set value used to filter out activity areas that highly match the user's interests. When the matching index of an activity area exceeds this threshold, it indicates that the activity area has sufficient relevance to the user's interests or needs, and is therefore considered a potential activity area for the user.

[0053] The proposed solution calculates a matching index for each activity area and sets a filtering threshold, effectively identifying potential activity areas that highly align with user interests and needs from a large pool of activity areas. This matching index-based filtering mechanism makes the recommendation process more targeted, avoiding the recommendation of activity areas that users are not interested in or that are unsuitable for them, thereby improving the accuracy of recommendations and user satisfaction.

[0054] This application further proposes steps for determining the matching index of an activity area based on the user's user profile and the activity tag information of the activity area, including:

[0055] The user profile and activity tag information of the activity area are input into a preset deep learning model to obtain the matching degree output by the preset deep learning model, and the matching degree is used as the initial matching index of the activity area. The preset deep learning model is used to determine the matching degree between the user profile and the activity tag information of the activity area. For each of the multiple activity areas, the number of times the user was active in the activity area in the historical time period and the first correspondence relationship are obtained. The first correspondence relationship includes a one-to-one correspondence between multiple frequency ranges and multiple first adjustment coefficients. The first adjustment coefficient corresponding to the frequency range of the activity in the first correspondence relationship is used as the target first adjustment coefficient of the activity area. The product of the initial matching index of the activity area and the target first adjustment coefficient is used as the matching index of the activity area.

[0056] Specifically, in the above methods, a user profile can be understood as a data set describing a user's interests, preferences, behavioral habits, and other characteristics, such as age, gender, occupation, hobbies, and historical activity records. Activity area tag information refers to a set of tags describing the types of services offered, environmental characteristics, and target audience of the activity area, such as "fitness," "reading," "social," "quiet," and "child-friendly."

[0057] The pre-defined deep learning model can be a neural network model, such as a convolutional neural network, a recurrent neural network, a Transformer model, or a variant thereof, which is trained to learn the complex matching relationship between user profiles and activity tag information. By learning from a large amount of historical user behavior data, this model can identify the potential correlation between user profiles and activity tag information, thereby outputting a quantified degree of matching. This degree of matching reflects the user's potential interest or suitability for a specific activity area and is used as an initial matching index for that activity area.

[0058] Furthermore, to more comprehensively assess the user's fit with the activity area, this application also considers the number of times the user has been active within the activity area during a historical time period. This historical activity count reflects the user's actual engagement and loyalty to the activity area. A first correspondence is pre-defined, comprising a one-to-one correspondence between multiple frequency ranges and multiple first adjustment coefficients. For example, when a user's historical activity count in a certain activity area falls within the range of "0-5 times," the corresponding first adjustment coefficient might be 0.8; when it falls within the range of "6-20 times," the corresponding first adjustment coefficient might be 1.0; and when it falls within the range of "more than 21 times," the corresponding first adjustment coefficient might be 1.2. These adjustment coefficients are intended to correct the initial matching index based on the user's historical engagement.

[0059] Therefore, the system locates the corresponding first adjustment coefficient in the first correspondence relationship based on the range of times a user has been active within a specific activity area, and uses this as the target first adjustment coefficient for the activity area. Finally, the initial matching index of the activity area is multiplied by this target first adjustment coefficient to obtain the final matching index for that activity area. This multiplication operation directly weights the historical activity counts on the matching index, allowing the matching index to more accurately reflect the user's actual preferences and engagement.

[0060] This application's solution utilizes a pre-defined deep learning model to calculate the initial matching index, enabling a more intelligent and comprehensive capture of the complex relationship between user profiles and activity tag information, overcoming the limitations of traditional matching based on simple rules or keywords. Furthermore, it incorporates the frequency of user activity within an activity area over a historical time period and adjusts the initial matching index using a first adjustment coefficient in the first correspondence relationship. This ensures that the final matching index not only considers the user's potential interests but also incorporates their actual behavioral data. It is precisely this combination of intelligent matching through deep learning and personalized correction based on historical behavior that allows the matching index of an activity area to more accurately reflect the user's true preferences and suitability for that area.

[0061] This application further proposes that the aforementioned environmental information includes pedestrian traffic and sound decibel levels. For each of the multiple candidate activity areas, based on the current environmental information of the candidate activity area, it determines the expected environmental information of the candidate activity area after a user's activity there, as well as the noise impact information on users in adjacent activity areas, including:

[0062] For each of the multiple candidate activity areas, determine whether the activity tag information of the candidate activity area includes a sports service tag or a communication service tag; if the activity tag information of the candidate activity area includes a sports service tag or a communication service tag, increase the sound decibel value in the current environmental information of the candidate activity area by a first preset sound decibel value to obtain the sound decibel value in the expected environmental information of the candidate activity area; if the activity tag information of the candidate activity area does not include a sports service tag or a communication service tag, increase the sound decibel value in the current environmental information of the candidate activity area by a second preset sound decibel value to obtain the sound decibel value in the expected environmental information of the candidate activity area; the first The preset sound decibel value is greater than the second preset sound decibel value; it is determined whether the activity label information of the candidate activity area includes a group activity service label; if the activity label information of the candidate activity area includes a group activity service label, the current environmental information of the candidate activity area is increased by a preset population flow to obtain the expected population flow in the environmental information of the candidate activity area; if the activity label information of the candidate activity area does not include a group activity service label, the sum of the current environmental information of the candidate activity area and 1 is used as the expected population flow in the environmental information of the candidate activity area; based on the expected environmental information of the candidate activity area, the noise impact information on users in adjacent activity areas is determined.

[0063] Specifically, environmental information can be understood as various indicators describing the current state of an activity area, such as pedestrian traffic and sound decibel levels. Pedestrian traffic refers to the number of people in a specific activity area, and sound decibel levels refer to the environmental noise level in that area. Among these, sports service tags refer to tags related to physical exercise and fitness activities, such as basketball, badminton, and running; communication service tags refer to tags related to interpersonal interaction, discussion, and meetings, such as salons, lectures, and group discussions. These activities typically generate high to moderate levels of noise.

[0064] The first and second preset sound decibel levels are preset increments used to quantify the impact of different types of activities on sound decibel levels. The first preset sound decibel level is set higher than the second preset sound decibel level to distinguish the differences in sound generation between sports or communication services and non-sports or non-communication services. For example, sports or communication services typically generate more noise, thus requiring a larger preset sound decibel level. Group activity service labels refer to labels associated with activities involving multiple people, such as team building, group games, and large gatherings. These activities typically significantly increase population density in an area.

[0065] Preset pedestrian flow is a preset increment used to quantify the impact of group activities on pedestrian traffic. Its purpose is to accurately predict changes in pedestrian traffic within the activity area after a group activity occurs. Through the above steps, based on the activity tag information of the candidate activity area, the sound decibel value and pedestrian flow in the current environmental information can be adjusted in a targeted manner, thereby obtaining more accurate expected environmental information. Subsequently, based on this expected environmental information, the potential noise impact of the activity on users in adjacent activity areas can be further assessed. Noise impact information can indicate no noise impact or noise impact.

[0066] This application's solution addresses the issue of insufficient accuracy in activity impact assessment in the basic solution by refining the process for determining expected environmental information. Specifically, different types of activities have significantly different environmental impacts; for example, sports and social activities are often accompanied by higher sound levels, while group activities significantly increase pedestrian traffic. This application, by determining whether the activity label information of the candidate activity area includes sports service labels, social service labels, or group activity service labels, can specifically adjust the sound level and pedestrian traffic in the current environmental information.

[0067] For example, when an activity tag includes a sports service tag or a communication service tag, a larger initial preset sound level in decibels is used to predict a higher noise level; when an activity tag includes a group activity service tag, a preset population density is used to predict a denser population distribution. This activity-type-based differentiated adjustment mechanism makes the predicted environmental information more closely resemble the actual situation, thus providing more accurate input for subsequent noise impact assessments. It is precisely because the accuracy of the predicted environmental information is improved that the judgment of noise impact information generated by users in adjacent activity areas becomes more reliable, thereby enabling the recommended services to better balance user needs with the harmony of public facilities and the environment.

[0068] The above-mentioned information on the noise impact on users in adjacent activity areas, determined based on the expected environmental information of the candidate activity area, includes:

[0069] Obtain the noise sensitivity index of adjacent activity areas; determine the transmitted sound decibel value to adjacent activity areas based on the sound decibel value in the expected environmental information of the candidate activity areas; determine the noise impact information on users in adjacent activity areas based on the noise sensitivity index and transmitted sound decibel value of adjacent activity areas.

[0070] Specifically, when determining the noise impact on users in adjacent activity areas, it is first necessary to obtain the noise sensitivity index of the adjacent activity areas. The noise sensitivity index can be understood as the tolerance or sensitivity of adjacent activity areas to noise. For example, a library or rest area may have a higher noise sensitivity index, while a gym or open communication area may have a lower noise sensitivity index. This index can be preset or dynamically adjusted based on historical data, user feedback, or the functional attributes of the area.

[0071] Furthermore, based on the expected environmental information of the candidate activity area (in decibels), it is necessary to determine the transmitted sound decibels to adjacent activity areas. The sound decibels in the expected environmental information reflect the noise intensity that may be generated when a user is active in the candidate activity area. The transmitted sound decibels refer to the actual sound decibels reaching the adjacent activity areas after the noise intensity has propagated and attenuated (e.g., through walls, distance, etc.). This can be calculated or estimated using physical models, empirical formulas, or actual measurement data. For example, factors such as the distance between the candidate activity area and adjacent activity areas, sound insulation materials, and spatial layout can be considered to calculate sound attenuation.

[0072] Finally, based on the noise sensitivity index and transmitted sound decibel value of adjacent activity areas, the noise impact information on users in adjacent activity areas is determined. This noise impact information indicates whether there is no noise impact or a noise impact. For example, when the transmitted sound decibel value exceeds the threshold set by the noise sensitivity index of the adjacent activity area, it can be determined that there is a noise impact; otherwise, it is determined that there is no noise impact.

[0073] This application's solution introduces a noise sensitivity index for adjacent activity areas and calculates the decibel value of transmitted sound. This allows the assessment of noise impact to go beyond solely relying on the noise intensity generated by the candidate activity area itself. Instead, it comprehensively considers the noise propagation characteristics and the actual noise tolerance capacity of adjacent areas. Therefore, it can more comprehensively and accurately simulate and predict the potential noise interference that users may cause to the surrounding environment when active in a specific activity area, thus providing a more reliable basis for subsequent recommendation index calculations.

[0074] Specifically, according to the above method, noise impact information on users in adjacent activity areas is determined based on the noise sensitivity index and transmitted sound decibel value of adjacent activity areas, including:

[0075] The system obtains the current sound decibel value and a second correspondence between adjacent activity areas. The second correspondence includes a one-to-one correspondence between multiple noise sensitivity index ranges and multiple first adjustment coefficients. The first adjustment coefficient corresponding to the noise sensitivity index range of the adjacent activity area in the second correspondence is used as the target first adjustment coefficient. The product of the transmitted sound decibel value and the target first adjustment coefficient is used as the user-perceived transmitted sound decibel value. The sum of the user-perceived transmitted sound decibel value and the current sound decibel value is used as the total sound decibel value of the adjacent activity areas. When the total sound decibel value of the adjacent activity areas is less than the preset sound decibel value of the adjacent activity areas, the system determines that there is no noise impact on the user in the adjacent activity areas. When the total sound decibel value of the adjacent activity areas is greater than or equal to the preset sound decibel value of the adjacent activity areas, the system determines that there is a noise impact on the user in the adjacent activity areas.

[0076] The first step, obtaining the current decibel level of adjacent activity areas, involves using sound sensors or environmental monitoring equipment deployed within these areas to collect ambient sound decibel levels in real time or periodically, reflecting the current noise background in those areas. The second correspondence can be understood as a pre-defined lookup table or a function model, designed to dynamically adjust the impact of transmitted sound decibel levels on user perception based on the noise sensitivity index of adjacent activity areas. For example, for areas with a high noise sensitivity index, the corresponding first adjustment coefficient may be larger, meaning that even a small transmitted sound decibel level may be strongly perceived by the user; conversely, for areas with a low noise sensitivity index, the corresponding first adjustment coefficient may be smaller.

[0077] The target first adjustment coefficient is a specific adjustment factor found in the second correspondence based on the specific noise sensitivity index of adjacent activity areas. The user-perceived transmitted sound decibel value is obtained by multiplying the transmitted sound decibel value generated by the candidate activity area by the target first adjustment coefficient, reflecting the actual perceived increase in noise intensity after considering the noise sensitivity of adjacent activity areas. The total sound decibel value of adjacent activity areas is obtained by superimposing the user-perceived transmitted sound decibel value with the current sound decibel value of adjacent activity areas, thus obtaining the overall sound level that the area may reach after user activity. The preset sound decibel value is an acceptable noise upper limit threshold set for each adjacent activity area. Once the total sound decibel value exceeds this threshold, it is considered to have a negative noise impact on users in that area.

[0078] This application's solution, by introducing the current decibel levels of adjacent activity areas and an adjustment mechanism based on a noise sensitivity index, makes the determination of noise impact information more precise and realistic. Specifically, firstly, the current decibel levels of adjacent activity areas are obtained, providing a realistic background noise benchmark for assessing the impact of new noise. Secondly, through a second correspondence, a target first adjustment coefficient is determined based on the noise sensitivity index of adjacent activity areas. This coefficient is used to quantify the differences in noise perception among users with different levels of sensitivity. Subsequently, the transmitted sound decibel level is multiplied by the target first adjustment coefficient to obtain the user-perceived transmitted sound decibel level. This makes the noise assessment no longer a simple physical addition but incorporates the user's subjective perception factor.

[0079] Finally, the user-perceived sound decibel value is added to the current sound decibel value to obtain the total sound decibel value of adjacent activity areas. This total decibel value is then compared with a preset sound decibel value to accurately determine whether noise is affecting the area. This method effectively avoids the problem of small amounts of newly added noise being misjudged as having an impact when the background noise is already high, or small amounts of newly added noise being misjudged as having no impact when the background noise is low but the user is highly sensitive.

[0080] This application further proposes a step for determining a recommendation index for each of multiple candidate activity areas, based on the normal environmental information range of the candidate activity area, the expected environmental information of the candidate activity area, and the noise impact information on users in adjacent activity areas. The steps include:

[0081] Determine whether the expected environmental information of the candidate activity area exceeds the normal environmental information range of the candidate activity area; if the expected environmental information of the candidate activity area exceeds the normal environmental information range of the candidate activity area, determine the environmental recommendation index of the candidate activity area as 0; if the expected environmental information of the candidate activity area does not exceed the normal environmental information range of the candidate activity area, take the difference between the normal environmental information range of the candidate activity area and the expected environmental information range of the candidate activity area as the environmental information difference of the candidate activity area; obtain the third correspondence relationship; the three correspondence relationship includes a one-to-one correspondence between multiple environmental information difference ranges and multiple environmental recommendation indices; take the environmental recommendation index corresponding to the environmental information difference range of the candidate activity area in the third correspondence relationship as the environmental recommendation index of the candidate activity area; determine the recommendation index of the candidate activity area based on the environmental recommendation index of the candidate activity area and the noise impact information generated on users in adjacent activity areas.

[0082] Specifically, when determining whether the expected environmental information of a candidate activity area exceeds the normal environmental information range, various indicators in the expected environmental information (such as pedestrian traffic and sound decibel levels) can be compared with the corresponding upper and lower limits of the normal environmental information range. For example, if the expected pedestrian traffic is higher than the upper limit of normal pedestrian traffic, or the expected sound decibel level is higher than the upper limit of normal sound decibel level, then the expected environmental information is considered to exceed the normal environmental information range. When the expected environmental information exceeds the normal range, the environmental recommendation index for the activity area is directly set to 0, meaning that the area is not suitable for recommendation under the current or expected environment.

[0083] Furthermore, when the expected environmental information does not exceed the normal environmental information range, it is necessary to calculate the environmental information difference of the candidate activity area. The environmental information difference can be understood as the "distance" or "degree of deviation" between the expected environmental information and the normal environmental information range. For example, the difference between the expected pedestrian flow and the upper limit of normal pedestrian flow, or the difference between the expected sound decibel value and the upper limit of normal sound decibel value, can be calculated, and the maximum value or weighted average of these differences can be taken as the comprehensive environmental information difference. The smaller the environmental information difference, the closer the expected environmental information is to the ideal state within the normal range.

[0084] The third correspondence is a pre-established mapping table used to quantify environmental information differences into environmental recommendation indices. This correspondence can be configured based on actual operational experience, user feedback, or expert knowledge. Its purpose is to discretize continuous environmental information differences and assign corresponding recommendation weights. For example, when the environmental information difference is within a small range, the corresponding environmental recommendation index is higher; when the environmental information difference is large but still within a normal range, the corresponding environmental recommendation index is relatively lower.

[0085] Therefore, by querying the third correspondence, the environmental recommendation index of the candidate activity area can be accurately obtained based on the calculated environmental information difference. This environmental recommendation index reflects the environmental suitability score of the candidate activity area.

[0086] Finally, the recommendation index for the candidate activity area is determined by comprehensively considering the area's environmental recommendation index and information on its noise impact on users in adjacent activity areas. This means that the final recommendation index not only considers the suitability of the activity area itself but also takes into account its potential impact on the surrounding environment and users, thus providing a more comprehensive and responsible service recommendation.

[0087] This application's solution addresses the potential coarseness in the basic solution's recommendation index calculation by introducing a refined judgment and quantification of the range between expected and normal environmental information. Specifically, firstly, by determining whether the expected environmental information exceeds the normal range, unsuitable activity areas can be quickly filtered out, preventing the recommendation of services that may lead to environmental degradation or a decline in user experience. Secondly, when the environmental information is within the normal range, by calculating the environmental information difference and combining it with a pre-defined third correspondence, the suitability of the environment can be quantified into a specific environmental recommendation index.

[0088] This quantification method makes the calculation of the recommendation index more precise and objective, distinguishing different levels of environmental suitability. For example, the closer the expected environmental information is to the ideal state within the normal range, the higher its environmental recommendation index. Finally, combining this environmental recommendation index with information on the noise impact on adjacent activity areas ensures that the final recommendation index not only reflects the attractiveness of the activity area itself but also fully considers its potential negative impact on the surrounding environment, thereby achieving more comprehensive and responsible personalized service recommendations.

[0089] According to the above method, the recommendation index of the candidate activity area is determined based on the environmental recommendation index of the candidate activity area and the noise impact information on users in adjacent activity areas, including:

[0090] For each adjacent activity area, determine whether the noise impact information of the adjacent activity area indicates noise impact; when the noise impact information of the adjacent activity area indicates no noise impact, determine the sound recommendation index corresponding to the adjacent activity area as the preset maximum sound recommendation index; when the noise impact information of the adjacent activity area indicates noise impact, obtain the pedestrian flow and the fourth correspondence relationship in the adjacent activity area; the fourth correspondence relationship includes a one-to-one correspondence between multiple pedestrian flow ranges and multiple sound recommendation indices; use the sound recommendation index corresponding to the pedestrian flow range in the adjacent activity area in the fourth correspondence relationship as the sound recommendation index of the adjacent activity area; determine the sound recommendation index of the candidate activity area based on the sound recommendation indices of multiple adjacent activity areas of the candidate activity area; use the sum of the sound recommendation index and the environmental recommendation index of the candidate activity area as the recommendation index of the candidate activity area.

[0091] Specifically, when determining the recommendation index for a candidate activity area, an independent noise impact assessment must first be conducted for each adjacent activity area. This assessment is performed by determining whether the noise impact information of adjacent activity areas indicates the presence of noise impact. Noise impact information can be understood as a binary or multi-valued indicator used to show whether activities in the candidate activity area will cause perceptible noise interference to users in adjacent activity areas.

[0092] Specifically, when the noise impact information for adjacent activity areas indicates no noise impact, it means that the activities in the candidate activity area will not cause negative noise interference to users in the adjacent activity areas. In this case, to reflect this positive or neutral impact, the sound recommendation index corresponding to the adjacent activity area is determined to be a preset maximum sound recommendation index. This preset maximum sound recommendation index can be a fixed value, such as 100 points, indicating no deductions or bonuses in this aspect.

[0093] Furthermore, when noise impact information from adjacent activity areas indicates a noise impact, a more refined assessment of the degree of this noise impact is required. Specifically, the current pedestrian traffic within that adjacent activity area will be obtained. Pedestrian traffic can be understood as the number of active users within that adjacent activity area at a given moment or time period.

[0094] Simultaneously, the system acquires a fourth correspondence, which predefines a one-to-one correspondence between multiple pedestrian flow ranges and multiple sound recommendation indices. For example, pedestrian flow ranges can be divided into "0-10 people," "11-30 people," "31-50 people," etc., with each range corresponding to a specific sound recommendation index. By matching the pedestrian flow of the current adjacent activity area with the fourth correspondence, the range of that pedestrian flow can be determined, and the corresponding sound recommendation index can be obtained as the sound recommendation index for that adjacent activity area. The purpose is that when noise impact exists, a larger pedestrian flow may mean more affected users, or that the area may have a higher tolerance for noise. Therefore, it is necessary to adjust the sound recommendation index according to pedestrian flow to more accurately reflect the actual impact of noise.

[0095] Therefore, after obtaining the sound recommendation index of each adjacent activity area of ​​the candidate activity area, it is necessary to determine the overall sound recommendation index of the candidate activity area based on the sound recommendation indices of multiple adjacent activity areas. This usually involves some form of aggregation of the sound recommendation indices of multiple adjacent activity areas, such as averaging, weighted averaging, or taking the minimum value, to comprehensively reflect the overall noise impact of the candidate activity area on the surrounding environment.

[0096] Finally, the sound recommendation index and the environmental recommendation index of the selected activity areas are summed to obtain the final recommendation index for the activity areas. The environmental recommendation index mainly reflects the environmental suitability of the selected activity area itself, while the sound recommendation index reflects its noise impact on the surrounding environment. The combination of the two provides a comprehensive and balanced recommendation assessment.

[0097] This application's solution effectively addresses the potential for coarseness in noise impact assessment in basic solutions by introducing an independent and detailed evaluation mechanism for the noise impact of each adjacent activity area. Specifically, when it is determined that there is no noise impact from adjacent activity areas, a preset maximum sound recommendation index is directly assigned. This ensures positive feedback for noise-free conditions and avoids recommendation bias caused by insufficient noise assessment. When noise impact exists, the solution further considers the pedestrian flow in adjacent activity areas and combines it with a fourth correspondence to determine the sound recommendation index.

[0098] This dynamic adjustment mechanism based on pedestrian traffic makes noise impact assessments more realistic. For example, in quiet areas with low pedestrian traffic, even slight noise may result in a low sound recommendation index; while in areas with high pedestrian traffic, the same noise level may lead to a relatively high sound recommendation index because the area may have a higher tolerance for noise or the noise's impact may be diluted. In this way, the noise impact of each adjacent activity area is quantified into a specific sound recommendation index, and these indices are then combined to form the overall sound recommendation index for the candidate activity area. Finally, this sound recommendation index is added to the candidate activity area's own environmental recommendation index, ensuring that the final recommendation index not only considers the candidate activity area's own environmental conditions but also fully and meticulously considers its noise impact on the surrounding environment, making the recommendation results more comprehensive and accurate.

[0099] This application further proposes a method for determining the sound recommendation index of a candidate activity area based on the sound recommendation indexes of multiple adjacent activity areas, including:

[0100] For each adjacent activity area, obtain the number of noise complaints against the adjacent activity area within a historical time period; normalize the noise complaint volume of multiple adjacent activity areas, and use the normalization result as the weight of each adjacent activity area; based on the weight of multiple adjacent activity areas, use the weighted sum of the sound recommendation index of multiple adjacent activity areas of the candidate activity area as the sound recommendation index of the candidate activity area.

[0101] Specifically, acquiring the number of noise complaints against adjacent activity areas over a historical period refers to the system continuously collecting and storing user noise complaint records received by various activity areas within a public facility over a past period (e.g., the past week, month, or year). These complaint volumes can serve as an important indicator of an area's noise sensitivity or tolerance. This data can be obtained, for example, through user feedback systems, customer service records, or dedicated complaint channels. Normalizing the noise complaint volumes of multiple adjacent activity areas and using the normalized result as the weight for each adjacent activity area can be understood as a process to make the noise complaint volumes of different activity areas comparable and to convert them into weights. Normalization maps the original complaint volumes to a fixed range (e.g., 0 to 1), eliminating the influence of dimensions. For example, a max-min normalization method can be used, dividing the noise complaint volume of each adjacent activity area by the maximum noise complaint volume among all adjacent activity areas to obtain a value between 0 and 1. This normalized value can be used as the weight of the adjacent activity area when calculating the sound recommendation index of the candidate activity area. The higher the number of complaints, the greater the weight, indicating that noise should be given more importance.

[0102] In practical applications, the sound recommendation index of a candidate activity area is obtained by weighting the sound recommendation indices of multiple adjacent activity areas. Specifically, the sound recommendation index of each adjacent activity area is multiplied by its corresponding weight, and then the products of all adjacent activity areas are summed to obtain the final sound recommendation index of the candidate activity area. This weighted summation method can more comprehensively reflect the overall noise impact of the candidate activity area on the surrounding environment, especially after considering the subjective feelings of users and regional sensitivity reflected in historical complaint data.

[0103] This application's solution addresses the issue that relying solely on current environmental information may not fully reflect the noise sensitivity of adjacent activity areas by incorporating historical noise complaint volumes as weights. Specifically, when a candidate activity area is used by a user, the noise it generates not only affects the current environment of adjacent activity areas but also interacts with their historical noise sensitivity. This historical sensitivity can be quantified by obtaining historical noise complaint volumes. Normalizing these complaint volumes and using them as weights ensures that, when calculating the overall sound recommendation index for candidate activity areas, adjacent activity areas with a higher historical noise complaint rate have a greater impact on the final result. This means that if an activity in a candidate activity area might affect a neighboring area with a historically prominent noise problem, the overall sound recommendation index for that candidate activity area will be correspondingly lowered, thus avoiding recommending services that might generate more complaints to users.

[0104] This application also discloses a personalized service recommendation system for public facilities based on big data analysis, comprising: an acquisition device and a processing device; the acquisition device is used to acquire a user profile, current environmental information of each activity area in multiple activity areas of the public facility, activity tag information of each activity area, and normal environmental information range of each activity area; the processing device is used to determine multiple candidate activity areas for the user in multiple activity areas based on the user profile; the processing device is used to determine, for each candidate activity area in multiple candidate activity areas, the expected environmental information of the candidate activity area after the user's activity in the candidate activity area and the noise impact information on users in adjacent activity areas, based on the current environmental information of the candidate activity area; the noise impact information is used to indicate whether there is no noise impact or no noise impact; the processing device is used to determine, for each candidate activity area in multiple candidate activity areas, the recommendation index of the candidate activity area based on the normal environmental information range of the candidate activity area, the expected environmental information of the candidate activity area, and the noise impact information on users in adjacent activity areas; the processing device is used to select the service corresponding to the candidate activity area with the highest recommendation index in multiple candidate activity areas as the service recommended to the user.

[0105] Specifically, the acquisition device can be understood as a module responsible for data collection and input. In some implementations, the acquisition device may include, but is not limited to, various sensors (such as people flow sensors, sound sensors, temperature sensors, etc.), network interface modules, data receiving modules, and interfaces for interacting with user terminal devices.

[0106] The processing unit can be understood as the core computing module responsible for data analysis, logical judgment, and recommendation decisions. In some implementations, the processing unit may consist of one or more central processing units, graphics processing units, memory, storage devices, and software programs and algorithm models running on these hardware components. The processing unit receives data provided by the acquisition device and executes various processing tasks according to preset logic and algorithms.

[0107] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A public facility personalized service recommendation method based on big data analysis, characterized by, include: Obtain the user profile, the current environmental information of each activity area in the multiple activity areas of the public facility, the activity tag information of each activity area, and the normal environmental information range of each activity area; Based on the user profile and activity tag information of the activity area, multiple candidate activity areas for the user in multiple activity areas are determined; For each of the multiple candidate activity areas, based on the current environmental information of the candidate activity area, determine the expected environmental information of the candidate activity area after a user's activity in the candidate activity area, as well as the noise impact information on users in adjacent activity areas; Noise impact information is used to indicate whether there is no noise impact or a noise impact. For each of the multiple candidate activity areas, the recommendation index of the candidate activity area is determined based on the normal environmental information range of the candidate activity area, the expected environmental information of the candidate activity area, and the noise impact information on users in adjacent activity areas. The service corresponding to the candidate activity area with the highest recommendation index among multiple candidate activity areas will be the service recommended to the user.

2. The method for recommending personalized public facility services based on big data analysis according to claim 1, characterized in that, Based on the user's user profile and activity tag information of the activity area, multiple candidate activity areas for the user in multiple activity areas are determined, including: For each of the multiple activity areas, the matching index of the activity area is determined based on the user's user profile and the activity tag information of the activity area; Active regions with a matching index greater than a preset matching index threshold are selected as candidate active regions.

3. The method for recommending personalized public facility services based on big data analysis according to claim 2, characterized in that, The matching index of the activity area is determined based on the user's user profile and the activity tag information of the activity area, including: The user profile and activity tag information of the activity area are input into the preset deep learning model to obtain the matching degree output by the preset deep learning model, and the matching degree is used as the initial matching index of the activity area; the preset deep learning model is used to determine the matching degree between the user profile and the activity tag information of the activity area. For each of the multiple activity regions, obtain the number of times a user was active within the activity region in a historical time period and a first correspondence relationship; the first correspondence relationship includes a one-to-one correspondence relationship between multiple frequency ranges and multiple first adjustment coefficients. The first adjustment coefficient corresponding to the range of the number of activities in the first correspondence is used as the target first adjustment coefficient of the activity area. The product of the initial matching index of the active region and the target first adjustment coefficient is used as the matching index of the active region.

4. The method for recommending personalized public facility services based on big data analysis according to claim 1, characterized in that, Environmental information includes pedestrian traffic and sound decibel levels. For each of the multiple candidate activity areas, based on the current environmental information of the candidate activity area, the expected environmental information of the candidate activity area after a user's activity there, as well as the noise impact information on users in adjacent activity areas, are determined, including: For each of the multiple candidate activity areas, determine whether the activity tag information of the candidate activity area includes a sports service tag or a communication service tag; When the activity tag information of the candidate activity area includes a sports service tag or a communication service tag, the sound decibel value in the current environmental information of the candidate activity area is increased by a first preset sound decibel value to obtain the sound decibel value in the expected environmental information of the candidate activity area. When the activity tag information of the candidate activity area does not include a sports service tag or a communication service tag, the sound decibel value in the current environmental information of the candidate activity area is increased by a second preset sound decibel value to obtain the expected sound decibel value in the environmental information of the candidate activity area; the first preset sound decibel value is greater than the second preset sound decibel value. Determine whether the activity tag information for the candidate activity area includes group activity service tags; When the activity tag information of the candidate activity area includes the group activity service tag, the preset number of people in the current environmental information of the candidate activity area will be added to obtain the expected number of people in the environmental information of the candidate activity area. When the activity tag information of the candidate activity area does not include the group activity service tag, the sum of the current environmental information of the candidate activity area and 1 will be used as the expected environmental information of the candidate activity area. Based on the expected environmental information of the candidate activity area, determine the noise impact information on users in adjacent activity areas.

5. The method for recommending personalized public facility services based on big data analysis according to claim 4, characterized in that, Based on the expected environmental information of the candidate activity area, determine the noise impact information on users in adjacent activity areas, including: Obtain the noise sensitivity index of adjacent activity areas; The transmitted sound decibel value to the adjacent activity area is determined based on the sound decibel value in the expected environmental information of the candidate activity area. The noise impact information on users in adjacent activity areas is determined based on the noise sensitivity index and transmitted sound decibel value of adjacent activity areas.

6. The method for recommending personalized public facility services based on big data analysis according to claim 5, characterized in that, Based on the noise sensitivity index and transmitted sound decibel value of adjacent activity areas, information on the noise impact on users in adjacent activity areas is determined, including: Obtain the current sound decibel value and second correspondence between adjacent active areas; the second correspondence includes a one-to-one correspondence between multiple noise sensitivity index ranges and multiple first adjustment coefficients. The first adjustment coefficient corresponding to the noise sensitivity index range of adjacent active areas in the second correspondence is taken as the target first adjustment coefficient. The product of the transmitted sound decibel value and the target first adjustment factor is used as the user-perceived transmitted sound decibel value. The sum of the user-perceived sound decibel value and the current sound decibel value is used as the total sound decibel value of the adjacent active area. When the total sound decibel value of adjacent activity areas is less than the preset sound decibel value of adjacent activity areas, the noise impact information for users in adjacent activity areas is determined to indicate no noise impact. When the total sound decibel value of adjacent activity areas is greater than or equal to the preset sound decibel value of adjacent activity areas, the noise impact information for users in adjacent activity areas is determined to indicate that there is a noise impact.

7. The method for recommending personalized public facility services based on big data analysis according to claim 1, characterized in that, For each of the multiple candidate activity areas, a recommendation index is determined based on the normal environmental information range of the candidate activity area, the expected environmental information of the candidate activity area, and the noise impact information on users in adjacent activity areas. This includes: Determine whether the expected environmental information of the candidate activity area exceeds the normal environmental information range of the candidate activity area; When the expected environmental information of the candidate activity area exceeds the normal environmental information range of the candidate activity area, the environmental recommendation index of the candidate activity area is determined to be 0. When the expected environmental information of the candidate activity area does not exceed the normal environmental information range of the candidate activity area, the difference between the normal environmental information range of the candidate activity area and the expected environmental information range of the candidate activity area is taken as the environmental information difference of the candidate activity area. Obtain the third correspondence; the third correspondence includes a one-to-one correspondence between multiple environmental information difference ranges and multiple environmental recommendation indices; The environmental recommendation index corresponding to the range of environmental information difference values ​​of the candidate activity areas in the third correspondence relationship is used as the environmental recommendation index of the candidate activity areas. The recommendation index of the candidate activity area is determined based on the environmental recommendation index of the candidate activity area and the noise impact information on users in adjacent activity areas.

8. The method for recommending personalized public facility services based on big data analysis according to claim 7, characterized in that, The recommendation index for the candidate activity area is determined based on the environmental recommendation index of the candidate activity area and information on the noise impact on users in adjacent activity areas, including: For each adjacent activity area, determine whether the noise impact information of the adjacent activity area indicates a noise impact; When the noise impact information of adjacent activity areas indicates no noise impact, the sound recommendation index corresponding to the adjacent activity area is determined to be the preset maximum sound recommendation index; When noise impact information in adjacent activity areas indicates noise impact, obtain the pedestrian flow in adjacent activity areas and the fourth correspondence relationship; the fourth correspondence relationship includes a one-to-one correspondence relationship between multiple pedestrian flow ranges and multiple sound recommendation indices. The sound recommendation index corresponding to the pedestrian flow range of adjacent activity areas in the fourth correspondence is used as the sound recommendation index of adjacent activity areas. The sound recommendation index of the candidate activity area is determined based on the sound recommendation index of multiple adjacent activity areas of the candidate activity area; The sum of the sound recommendation index and the environmental recommendation index of the candidate activity area is used as the recommendation index of the candidate activity area.

9. The method for recommending personalized public facility services based on big data analysis according to claim 8, characterized in that, The sound recommendation index of the candidate activity area is determined based on the sound recommendation index of multiple adjacent activity areas, including: For each adjacent activity area, obtain the number of noise complaints against the adjacent activity areas within a historical time period; The noise complaint volume of multiple adjacent activity areas is normalized, and the normalization result is used as the weight of each adjacent activity area. Based on the weights of multiple adjacent activity areas, the weighted sum of the voice recommendation indices of multiple adjacent activity areas of the candidate activity area is used as the voice recommendation index of the candidate activity area.

10. A personalized service recommendation system for public facilities based on big data analysis, characterized in that, include: Acquisition device and processing device; The acquisition device is used to acquire the user profile of the user, the current environmental information of each activity area in the multiple activity areas of the public facility, the activity tag information of each activity area, and the normal environmental information range of each activity area; The processing device is used to determine multiple candidate activity areas for a user in multiple activity areas based on the user profile of the user and the activity tag information of the activity area; The processing device is used to determine, for each of a plurality of candidate activity areas, the expected environmental information of the candidate activity area after a user's activity in the candidate activity area and the noise impact information on users in adjacent activity areas, based on the current environmental information of the candidate activity area. Noise impact information is used to indicate whether there is no noise impact or a noise impact. The processing device is used to determine a recommendation index for each of a plurality of candidate activity areas based on the normal environmental information range of the candidate activity area, the expected environmental information of the candidate activity area, and the noise impact information on users in adjacent activity areas. The processing device is used to select the service corresponding to the candidate activity area with the highest recommendation index among multiple candidate activity areas as the service recommended to the user.

Images