Social graph construction and abnormal behavior early warning method for the elderly

Abstract

The present invention relates to the field of information processing technology, and specifically discloses a method for constructing a social graph and warning of abnormal behavior for the elderly, wherein the method for constructing a social graph for the elderly comprises the following steps: S1, continuously collecting location data of the elderly; obtaining a pre-constructed place semantic knowledge graph; for each two elderly people, determining whether there is a location co-occurrence relationship between the two, and calculating the location co-occurrence strength when there is a location co-occurrence relationship; calculating the social interaction intensity score between the elderly based on the location co-occurrence strength and the social function weight of the place; constructing a social graph of the elderly community based on the social interaction intensity score and the place function information; the method for constructing a social graph of the elderly community can effectively mine social interaction information from location data, and then construct a social graph of the elderly community, so as to more accurately identify and quantify the real social interactions between the elderly, and provide data support for optimizing care services and activity arrangements for nursing homes.

Description

Technical Field

[0001] The present application relates to the field of information processing technology, and more specifically, to a method for constructing a social graph and providing early warning of abnormal behavior for the elderly. Background Art

[0002] Currently, indoor positioning systems based on low-power wearable tags are widely deployed in nursing homes to ensure the safety of elderly individuals with limited mobility. These systems primarily record the location of elderly individuals for rapid location tracking in emergency situations. However, with the increasing demand for more sophisticated elderly care services, institutions require a deeper understanding of elderly individuals' social interaction patterns to optimize care and activity arrangements, particularly to predict the risk of social isolation and prevent abnormal behavior. Direct observation or manual recording of social interactions is time-consuming and inefficient, and may infringe on the privacy of elderly individuals. Leveraging knowledge graph technology to mine social information from location data has become a potential solution.

[0003] Although knowledge graph technology has the ability to enhance semantics, inferring social interactions directly from basic positioning data still faces many challenges. Basic positioning data can only reflect the proximity of physical locations, which is not entirely equivalent to real social interaction. For example, overlapping locations in hallways or non-social places may not represent the occurrence of social behavior. In addition, nursing homes usually lack professional technical personnel and have extremely high requirements for data privacy protection. How to cost-effectively mine information that can accurately reflect the real social interactions of the elderly from existing low-power positioning data without increasing additional hardware investment and infringing on the privacy of the elderly, and construct a social graph of the community, has become a technical challenge that currently faces a common and urgent solution for nursing homes.

[0004] There is currently no effective technical solution to the above problems. Summary of the Invention

[0005] The purpose of this application is to provide a method for constructing a social graph and warning of abnormal behavior of the elderly, so as to more deeply explore the social interaction relationships between the elderly based on their location data, and thus construct a more accurate social graph.

[0006] In a first aspect, the present application provides a method for constructing a social graph of the elderly, for analyzing the social relationships of the elderly group, the method comprising the following steps:

[0007] S1, continuously collect location data of the elderly;

[0008] S2. Obtain a pre-built place semantic knowledge graph, where the graph nodes include place area codes, and each place area code is preset with a place social function weight, a place type adjustment coefficient, and place function information;

[0009] S3. For each pair of elderly people, analyze the co-occurrence frequency and duration of the same location area code corresponding to their location data over multiple consecutive time periods to determine whether there is a location co-occurrence relationship between the two. If a location co-occurrence relationship exists, calculate the location co-occurrence intensity based on the co-occurrence frequency, the duration, and the location type adjustment coefficient.

[0010] S4. Calculating a social interaction intensity score among the elderly based on the location co-occurrence intensity and the social function weight of the venue;

[0011] S5. Construct a social graph of the elderly community based on the social interaction intensity score and the venue function information.

[0012] The social graph construction method for the elderly in this application uses a place semantic knowledge graph to semantically enhance the location data of the elderly, overcomes the limitation that simple location proximity is not equivalent to social interaction, more accurately mines the real social interaction information between the elderly, and constructs a community social graph for analyzing the social relationships of the elderly group. Among them, this method uses the continuously collected location data of the elderly as the data basis, combined with a pre-constructed place semantic knowledge graph that can give location data location semantic information, to analyze the location co-occurrence relationship between different elderly people, and calculate the social interaction intensity score. It can effectively mine social interaction information from the location data of the elderly, and then combine the place function information to construct an elderly community social graph that can be used to analyze the social relationships of the elderly group, so as to more accurately identify and quantify the real social interactions between the elderly, and provide data support for the optimization of care services and activity arrangements for nursing homes.

[0013] The method for constructing a social graph of the elderly, wherein the step of analyzing, for each two elderly people, the co-occurrence frequency and duration of the same location area codes corresponding to their location data in multiple consecutive time periods to determine whether there is a location co-occurrence relationship between the two includes:

[0014] S31. For each pair of elderly people, count the co-occurrence frequency and duration of the location data of the two elderly people corresponding to the same location area code in multiple consecutive time periods, and preliminarily determine whether there is a location co-occurrence relationship between the two based on the co-occurrence frequency and the duration;

[0015] S32. When it is preliminarily determined that there is a location co-occurrence relationship between the two elderly people, the effective co-occurrence frequency and effective co-occurrence duration during the multiple consecutive time periods in which the distance between the two elderly people is less than the interaction distance threshold are counted based on the location data;

[0016] S33. Finally determine whether there is a location co-occurrence relationship between the two based on the effective co-occurrence frequency and the effective co-occurrence duration.

[0017] The above-mentioned processing method can exclude the situation where the place area codes are the same but the actual distance is far by introducing the interaction distance threshold, the effective co-occurrence frequency, and the effective co-occurrence duration, and more accurately determine whether there is a location co-occurrence relationship between the elderly, and can effectively avoid the problem of inaccurate judgment caused by judging based solely on the same place area codes. Because just the same place area codes do not mean that there is actual social interaction between the two elderly people. When judging whether there is a location co-occurrence relationship between the elderly, the above-mentioned processing method not only takes into account the co-occurrence frequency and duration of the place area codes, but also takes into account the actual distance between the elderly, so that the judgment of the location co-occurrence relationship is more accurate and more in line with the actual social interaction of the elderly. Among them, step S31 plays the role of preliminary screening, step S32 makes accurate judgments through the interaction distance threshold, and step S33 gives the final judgment result. These three steps work together to improve the accuracy of the judgment of location co-occurrence relationships, making the subsequently constructed social graph more accurate and reliable.

[0018] In the method for constructing a social graph for the elderly, the process of determining the interaction distance threshold includes:

[0019] A1. According to the venue function information corresponding to the venue area code, obtain the interaction distance threshold from the venue function information and distance threshold mapping relationship table.

[0020] In this way, the interaction distance threshold can be adaptively adjusted in different locations, allowing for more accurate determination of location co-occurrence relationships among seniors. This improves the accuracy of location co-occurrence determination and effectively enhances the accuracy of constructing social graphs for senior communities.

[0021] In the method for constructing a social graph for the elderly, each location area code is also preset with location area information; and the process of determining the interaction distance threshold further includes:

[0022] A2. Adjust the interaction distance threshold according to the venue area information corresponding to the venue area code.

[0023] In the method for constructing a social graph for the elderly, the step of calculating the location co-occurrence intensity based on the co-occurrence frequency, the duration, and the venue type adjustment coefficient includes:

[0024] S34, splitting the duration according to the duration to obtain the sub-co-occurrence durations of each two elderly people in multiple consecutive time periods;

[0025] S35. Adjust the coefficient according to the venue type and set a time decay factor sequence for each time period regarding time decay;

[0026] S36. Based on the time decay factor sequence, weight the sub-co-occurrence durations in each time period to obtain a weighted co-occurrence duration;

[0027] S37. Multiply the weighted co-occurrence duration by the co-occurrence frequency to obtain the position co-occurrence strength.

[0028] The method for constructing a social graph for the elderly, wherein step S4 includes:

[0029] S41. Obtain the age score and physical ability score of the elderly;

[0030] S42. Adjust the weight of the social function of the venue according to the age score and physical ability score based on the location co-occurrence relationship;

[0031] S43. Calculate the social interaction intensity score between the elderly based on the adjusted venue social function weight and the location co-occurrence intensity.

[0032] The method for constructing a social graph for the elderly, wherein step S5 comprises:

[0033] S51. Constructing an initial social graph of the elderly community based on the social interaction intensity score, where the graph nodes represent elderly people and the edge weights are initialized to the social interaction intensity score;

[0034] S52. Based on the guidance of the venue area coding, the social interaction type corresponding to the venue function information is edge-matched on the initial elderly community social graph to form the elderly community social graph.

[0035] In a second aspect, the present application also provides a method for early warning of abnormal behavior of the elderly, the method comprising the following steps:

[0036] B1. Analyze the elderly community social graph obtained based on the elderly social graph construction method provided in the first aspect based on a community discovery algorithm to obtain community structure information;

[0037] B2. Obtaining community parameter information based on the community structure information;

[0038] B3. Dynamically assessing the social isolation risk level and its changing trend of each elderly person based on their position in the community structure information and the community parameter information;

[0039] B4. Predict whether the social behavior of each elderly person is abnormal based on the social isolation risk level and its changing trend.

[0040] The abnormal behavior early warning method for the elderly of the present application is based on the analysis of the elderly community social graph obtained by the social graph construction method for the elderly provided in the first aspect to identify the community structure information in the elderly group and extract the community parameter information, so as to dynamically evaluate the social isolation risk level and its changing trend of each elderly person for each node in the community structure and in combination with the community parameter information to determine whether there is a social isolation risk, thereby predicting whether the elderly’s future social behavior is abnormal, realizing early warning of abnormal behavior of the elderly, providing technical support for elderly care institutions, and helping to improve the quality of elderly care services.

[0041] The method for early warning of abnormal behavior of the elderly, wherein step B1 comprises:

[0042] B11. Use the modularity-optimized Louvain algorithm to perform community discovery on the elderly community social graph, wherein the resolution parameter of the Louvain algorithm is adaptively optimized in the following manner:

[0043] B111. Construct a candidate parameter set for the resolution parameter, use a grid search method to select different resolution parameter values ​​from the candidate parameter set, and run the Louvain algorithm to divide the elderly community social graph into different groups;

[0044] B112. Calculate the silhouette coefficient of each community division result, and select the resolution parameter corresponding to the maximum silhouette coefficient as the optimal resolution parameter;

[0045] B113. Extract community structure information based on the community segmentation results under the optimal resolution parameters.

[0046] Through the above processing method, the resolution parameters of the Louvain algorithm can be adaptively adjusted to the optimal state, thereby improving the accuracy and robustness of community discovery.

[0047] The method for warning of abnormal behavior of the elderly is described, wherein the community parameter information includes community size, community connection strength and node centrality index, wherein the community size is represented by the number of nodes in the community, the community connection strength is represented by the average weight of the connecting edges between nodes in the community, and the node centrality index is represented by the degree centrality index.

[0048] From the above, it can be seen that the present application provides a method for constructing a social graph and warning of abnormal behavior for the elderly, wherein the method for constructing a social graph for the elderly uses a place semantic knowledge graph to semantically enhance the location data of the elderly, overcomes the limitation that simple location proximity is not equivalent to social interaction, more accurately mines the real social interaction information between the elderly, and constructs a community social graph for analyzing the social relationships of the elderly group. Among them, the method uses the continuously collected location data of the elderly as the data basis, combined with a pre-constructed place semantic knowledge graph that can give location data location semantic information, to analyze the location co-occurrence relationship between different elderly people, and calculate the social interaction intensity score. It can effectively mine social interaction information from the location data of the elderly, and then combine the place function information to construct an elderly community social graph that can be used to analyze the social relationships of the elderly group, so as to more accurately identify and quantify the real social interactions between the elderly, and provide data support for the optimization of care services and activity arrangements for nursing homes. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A flowchart of a method for constructing a social graph for the elderly provided in an embodiment of the present application.

[0050] Figure 2 A flowchart of a method for early warning of abnormal behavior of the elderly provided in an embodiment of the present application. DETAILED DESCRIPTION

[0051] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work fall within the scope of protection of the present application.

[0052] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.

[0053] First, please refer to Figure 1Some embodiments of the present application provide a method for constructing a social graph of the elderly, which is used to analyze the social relationships of the elderly group. The method includes the following steps:

[0054] S1, continuously collect location data of the elderly;

[0055] S2. Obtain a pre-built place semantic knowledge graph, where the graph nodes include place area codes, and each place area code is preset with a place social function weight, a place type adjustment coefficient, and place function information;

[0056] S3. For each pair of elderly people, analyze the co-occurrence frequency and duration of the location data corresponding to the same location area code in multiple consecutive time periods to determine whether there is a location co-occurrence relationship between the two. If a location co-occurrence relationship exists, calculate the location co-occurrence intensity based on the co-occurrence frequency, duration, and location type adjustment coefficient;

[0057] S4. Calculate the social interaction intensity score among the elderly based on the location co-occurrence intensity and the weight of the social function of the venue;

[0058] S5. Construct a social graph of the elderly community based on the social interaction intensity score and venue function information.

[0059] Specifically, in step S1, the location data can be continuously collected through low-power wearable tags and a positioning system, or collected through a wearable device with a positioning function carried by the elderly.

[0060] More specifically, in step S2, a place semantic knowledge graph is pre-constructed, and the place area codes in the graph can be, for example, room numbers or area identifiers of different areas in a nursing home.

[0061] More specifically, in step S2, a place semantic knowledge graph is pre-constructed. The place semantic knowledge graph includes a plurality of graph nodes and edges connecting the graph nodes. It can be constructed by collecting relevant place data through an existing knowledge graph generation algorithm; the place area codes in the graph can be, for example, room numbers or area identifiers of different areas in a nursing home. The weight of the social function of the place can be pre-set based on the social attributes of the place. For example, the social function weight of the restaurant is higher than that of the corridor. The place type adjustment coefficient can be set according to the type of place. For example, for non-social places, the place type adjustment coefficient can be set to a lower value to reduce the impact of location co-occurrence in these places on the strength of social relationships. The place function information can include a description of the use of the place, such as "restaurant", "activity room", etc.

[0062] More specifically, step S3 is used to determine location co-occurrence relationships and calculate location co-occurrence strength. Specifically, for each pair of senior citizens, the co-occurrence frequency and duration of locations with the same location area code corresponding to their location data are counted over multiple consecutive time periods, such as 10-minute, half-hour, hour, or day time periods. The co-occurrence frequency refers to the number of time periods with the same location area code for the two senior citizens within these multiple consecutive time periods, while the duration can refer to the total length of time with the same location area code for the two senior citizens within the statistical period. Location co-occurrence relationships are determined based on thresholds set for the co-occurrence frequency and duration. A location co-occurrence relationship is determined to exist when both meet preset conditions. For example, a location co-occurrence relationship is determined to exist when the co-occurrence frequency exceeds a certain number of times and the duration exceeds a certain duration. The calculation of location co-occurrence strength takes into account the co-occurrence frequency, duration, and location type adjustment coefficient. The location type adjustment coefficient can be set based on the actual social function of the location. For example, a lower value can be set for locations with weaker social functions, while a higher value can be set for locations with weaker social functions. The calculation formula for the location co-occurrence intensity can be simply written as: location co-occurrence intensity = co-occurrence frequency * duration * venue type adjustment coefficient.

[0063] More specifically, step S4 is used to calculate the social interaction strength score. The calculation process can be simply expressed as: social interaction strength score = location co-occurrence strength * venue social function weight. The venue social function weight can have different preset values ​​for different venues to reflect the differences in social attributes of different venues.

[0064] More specifically, step S5 is used to construct a social graph for the elderly community, where graph nodes represent elderly people, and edges can represent social relationships or types of social interactions between elderly people. This can be determined by venue function information, or the edge attributes can be directly configured as venue function information. The edge weight can be a social interaction intensity score. Thus, a social graph for the elderly community is constructed and used to analyze the social relationships of the elderly group.

[0065] The social graph construction method for the elderly in the embodiment of the present application uses a place semantic knowledge graph to semantically enhance the location data of the elderly, overcomes the limitation that simple location proximity is not equivalent to social interaction, more accurately mines the real social interaction information between the elderly, and constructs a community social graph for analyzing the social relationships of the elderly group. Among them, the method uses the continuously collected location data of the elderly as the data basis, combined with a pre-constructed place semantic knowledge graph that can give location data location semantic information, to analyze the location co-occurrence relationship between different elderly people, and calculate the social interaction intensity score. It can effectively mine social interaction information from the location data of the elderly, and then combine the place function information to construct an elderly community social graph that can be used to analyze the social relationships of the elderly group, so as to more accurately identify and quantify the real social interactions between the elderly, and provide data support for the optimization of care services and activity arrangements for nursing homes.

[0066] In some preferred embodiments, for each two elderly people, analyzing the co-occurrence frequency and duration of the same location area codes corresponding to their location data in multiple consecutive time periods to determine whether there is a location co-occurrence relationship between the two includes:

[0067] S31. For each pair of elderly people, count the co-occurrence frequency and duration of the location data of the two elderly people corresponding to the same location area code in multiple consecutive time periods, and preliminarily determine whether there is a location co-occurrence relationship between the two based on the co-occurrence frequency and duration;

[0068] S32. When it is preliminarily determined that there is a location co-occurrence relationship between the two elderly people, the effective co-occurrence frequency and effective co-occurrence duration during the multiple consecutive time periods in which the distance between the two elderly people is less than the interaction distance threshold are counted based on the location data;

[0069] S33. Finally determine whether there is a position co-occurrence relationship between the two based on the effective co-occurrence frequency and the effective co-occurrence duration.

[0070] Specifically, step S31 is used to analyze the characteristics of the location data of each two elderly people over multiple consecutive time periods, mainly to count the co-occurrence frequency and duration of locations with the same area code. A preliminary determination of whether there is a location co-occurrence relationship between the two can be made based on a preset frequency threshold and duration threshold. For example, if the co-occurrence frequency exceeds the frequency threshold and the duration exceeds the duration threshold, then a preliminary determination is made that there is a location co-occurrence relationship between the two.

[0071] More specifically, step S31 is equivalent to quickly screening out elderly people who may have a location co-occurrence relationship based on the consistency of the venue area codes, and is therefore a preliminary judgment step. After screening out these elderly people who may have a location co-occurrence relationship, steps S32 and S33 conduct a detailed analysis. This is done by counting the effective co-occurrence frequency and effective co-occurrence duration of two elderly people within multiple consecutive time periods when their distance is less than the interaction distance threshold, to accurately determine whether a location co-occurrence relationship exists based on the distance relationship between the elderly people. The interaction distance threshold represents the maximum distance between the elderly people at which social interaction can occur; the effective co-occurrence frequency refers to the number of time periods within multiple consecutive time periods when the distance between the two elderly people is less than the interaction distance threshold and the venue area codes are the same; and the effective co-occurrence duration refers to the total duration of time when the distance between the two elderly people is less than the interaction distance threshold and the venue area codes are the same. The interaction distance threshold can be determined based on venue function information. The final judgment in step S33 can be made based on preset effective frequency thresholds and effective duration thresholds. For example, if the effective co-occurrence frequency exceeds the effective frequency threshold and the effective co-occurrence duration exceeds the effective duration threshold, a location co-occurrence relationship is ultimately determined. The effective frequency threshold and the effective duration threshold may be the same as or different from the frequency threshold and the duration threshold determined initially.

[0072] More specifically, the above-mentioned processing method can exclude the situation where the place area codes are the same but the actual distance is far by introducing the interaction distance threshold, the effective co-occurrence frequency, and the effective co-occurrence duration, and more accurately determine whether there is a location co-occurrence relationship between the elderly, and can effectively avoid the problem of inaccurate judgment caused by judgment based solely on the same place area codes. Because just the same place area codes do not mean that there is actual social interaction between the two elderly people. When judging whether there is a location co-occurrence relationship between the elderly, the above-mentioned processing method not only takes into account the co-occurrence frequency and duration of the place area codes, but also takes into account the actual distance between the elderly, so that the judgment of the location co-occurrence relationship is more accurate and more in line with the actual social interaction situation of the elderly. Among them, step S31 plays the role of preliminary screening, step S32 makes accurate judgments through the interaction distance threshold, and step S33 gives the final judgment result. These three steps work together to improve the accuracy of the judgment of location co-occurrence relationships, making the subsequently constructed social graph more accurate and reliable.

[0073] In some preferred embodiments, the process of determining the interaction distance threshold includes:

[0074] A1. According to the venue function information corresponding to the venue area code, obtain the interaction distance threshold from the venue function information and distance threshold mapping relationship table.

[0075] Specifically, the table mapping place function information to distance thresholds is a pre-established mapping table that records the interaction distance thresholds corresponding to different place function information. Place function information can be extracted based on the place region encoding in the place semantic knowledge graph. The interaction distance threshold is the maximum distance used to determine whether there is a valid co-occurrence relationship between elderly people.

[0076] More specifically, in step S32, the rationality of the interaction distance threshold affects the accuracy of location co-occurrence relationship determination. A mapping table between place function information and distance thresholds stores the correspondence between place function information and interaction distance thresholds. When determining the interaction distance threshold, the corresponding place function information is first retrieved from the place semantic knowledge graph based on the place area code. The place function information is then used as an index to search the corresponding interaction distance threshold in the mapping table between place function information and distance thresholds. For example, the mapping table between place function information and distance thresholds can be configured such that: when the place function information is "public activity area," the interaction distance threshold is set to 5 meters; when the place function information is "rest area," the interaction distance threshold is set to 3 meters; and when the place function information is "corridor," the interaction distance threshold is set to 2 meters. In this way, the interaction distance threshold can be adaptively adjusted in different places, thereby more accurately determining whether location co-occurrence relationships exist among elderly people. This improves the accuracy of location co-occurrence relationship determination and effectively enhances the accuracy of constructing social graphs for elderly communities.

[0077] In some preferred embodiments, each venue area code is also preset with venue area information; and the process of determining the interaction distance threshold further includes:

[0078] A2. Adjust the interaction distance threshold according to the venue area information corresponding to the venue area code.

[0079] Specifically, venue area information is preset in the venue region code of the venue semantic knowledge graph. It can be the actual floor area of ​​the venue or the area scale of the corresponding functional venue. After obtaining the initial interaction distance threshold from the venue function information and distance threshold mapping relationship table based on the venue function information, step A2 further adjusts the initial interaction distance threshold based on the venue area information. Specifically, if the venue area is large, the interaction distance threshold can be appropriately increased; conversely, if the venue area is small, the interaction distance threshold can be appropriately reduced. The adjustment can be achieved by presetting an area adjustment coefficient. For example, a corresponding relationship table between venue area ranges and area adjustment coefficients can be established, or a functional relationship can be used to calculate the area adjustment coefficient based on the venue area size. The initial interaction distance threshold is then multiplied by the area adjustment coefficient to obtain the final interaction distance threshold. Step A2 adjusts the interaction distance threshold based on the venue area information so that the setting of the interaction distance threshold can be more consistent with the actual venue conditions, thereby more accurately determining whether there is a real social interaction relationship between the elderly.

[0080] In some preferred embodiments, the step of calculating the location co-occurrence intensity based on the co-occurrence frequency, duration, and venue type adjustment coefficient includes:

[0081] S34, splitting the co-occurrence duration of each two elderly people in multiple consecutive time periods according to the duration;

[0082] S35. Adjust the coefficient according to the venue type and set a time decay factor sequence for each time period regarding time decay;

[0083] S36. Based on the time decay factor sequence, weight the sub-co-occurrence durations in each time period to obtain a weighted co-occurrence duration;

[0084] S37. Multiply the weighted co-occurrence duration by the co-occurrence frequency to obtain the position co-occurrence strength.

[0085] Specifically, step S34 is used to further analyze the location data of every two elderly people in multiple consecutive time periods to obtain the duration of time in which the location area codes of the two are the same in each time period, that is, the sub-co-occurrence duration.

[0086] It should be noted that in the implementation method of using steps S31 to S33 to determine the location co-occurrence relationship, since it has been confirmed that there is a location co-occurrence relationship between the two elderly people, step S34 can use a complete and better reflection of the duration of the co-occurrence of the two elderly people in the place coded in the same place area to split and obtain sub-co-occurrence durations for deeper analysis.

[0087] More specifically, in step S35, the time decay factor sequence is linked to the venue type adjustment coefficient. Social interactions in different venues may have different temporal decay characteristics, so the time decay factor sequence needs to be considered in the venue type adjustment coefficient. The time decay factor can be a linear or exponential decay factor, representing a higher weight for more recent sub-co-occurrence durations. The introduction of the time decay factor sequence reduces the influence of earlier location co-occurrences, allowing the scoring to prioritize more recent social interactions.

[0088] More specifically, in step S36, the temporal decay factor sequence is used to weight the sub-co-occurrence durations within each time period to obtain a weighted co-occurrence duration. This weighting process reflects that sub-co-occurrence durations in recent time periods have a greater impact on the location co-occurrence strength, while sub-co-occurrence durations in more distant time periods have a smaller impact.

[0089] More specifically, in step S37, the weighted co-occurrence duration replaces the original duration, so that the calculated position co-occurrence intensity can more accurately reflect the social interaction intensity with time-series attenuation characteristics, thereby reflecting the timeliness of social interaction and obtaining a position co-occurrence intensity that is more in line with the actual social interaction situation.

[0090] More specifically, the location co-occurrence intensity finally obtained by the above processing method comprehensively considers the co-occurrence frequency, the time length considering time decay and venue type adjustment coefficient, and the overall duration, thereby more comprehensively and finely quantifying the strength of the location co-occurrence relationship between the elderly, so that it can more accurately and meticulously reflect the intensity of social interaction between the elderly based on location data, making the subsequent construction of social graphs and abnormal behavior warning analysis more reliable and effective.

[0091] In some preferred embodiments, step S4 includes:

[0092] S41. Obtain the age score and physical ability score of the elderly;

[0093] S42. Adjust the weight of the social function of the venue according to the age score and physical ability score based on the location co-occurrence relationship;

[0094] S43. Calculate the social interaction intensity score among the elderly based on the adjusted venue social function weights and location co-occurrence intensity.

[0095] Specifically, in step S41, the elderly person's age score and physical ability score quantify their individual characteristics. The age score and physical ability score can be obtained in any manner. For example, they can be obtained by retrieval of pre-recorded score data from the elderly person's personal profile or through real-time assessment by a health assessment system.

[0096] More specifically, step S42 is used to adjust the weight of the social function of the place based on the location co-occurrence relationship between the elderly and their age scores and physical ability scores. Specifically, when two elderly people are determined to have a location co-occurrence relationship, the social graph construction method for the elderly in the embodiment of the present application will further consider the age and physical ability differences between the two elderly people. The adjustment of the weight of the social function of the place can be achieved in a variety of ways. For example, an adjustment model can be preset, which takes the age difference coefficient and the physical ability difference coefficient as input and outputs a comprehensive adjustment coefficient. The social function weight of the place is adjusted according to this comprehensive adjustment coefficient.

[0097] More specifically, in step S43, since the weight of the social function of the venue has been adjusted according to the individual characteristics of the elderly, the final social interaction intensity score can more accurately reflect the actual social interaction level among the elderly.

[0098] More specifically, in order to more accurately assess the intensity of social interactions among the elderly, the social graph construction method for the elderly in the embodiment of the present application further introduces age scores and physical ability scores to adjust the weights of social functions of places, so that the impact of individual differences among the elderly on the intensity of social interactions is taken into account, thereby more accurately assessing the intensity of social interactions among the elderly;

[0099] In some preferred embodiments, step S42 includes:

[0100] S421. For every two elderly people who have a location co-occurrence relationship, calculate the age difference coefficient and the physical ability difference coefficient based on the age score and the physical ability score;

[0101] S422. Based on the age difference coefficient and the physical ability difference coefficient, a comprehensive adjustment coefficient is calculated using the weighted average method;

[0102] S423. Adjust the weight of the social function of the venue based on the comprehensive adjustment coefficient.

[0103] Specifically, in step S421, as an implementation method, the age difference coefficient is obtained by taking the absolute value of the age difference between the two elderly people and then normalizing it. For example, the age difference coefficient can be calculated as the absolute value of the age difference divided by the maximum possible age difference. The physical ability difference coefficient can be obtained by quantifying the normalized difference in the physical ability scores of the two elderly people based on a preset physical ability scoring standard. For example, the physical ability difference coefficient can be calculated as the absolute value of the difference in physical ability scores divided by the maximum possible difference in physical ability scores. The age difference coefficient and the physical ability difference coefficient calculated in step S421 can quantify the degree of difference between the elderly in age and physical ability, and provide a quantitative basis for the subsequent adjustment of the weight of the social function of the venue.

[0104] More specifically, in step S422, the comprehensive adjustment coefficient is a weighted average of the age difference coefficient and the physical ability difference coefficient. The weighting coefficient used in the weighted processing can be flexibly adjusted according to the actual application scenario. For example, in a scenario where physical ability differences are more important, the weight of the physical ability difference coefficient can be set higher. This processing method can flexibly adjust the influence of age and physical ability differences on the comprehensive adjustment coefficient according to actual conditions, making the adjustment process more refined.

[0105] More specifically, step S423 uses a comprehensive adjustment coefficient to adjust the weight of the social function of the venue, which can be a multiplication adjustment or an addition adjustment, preferably a multiplication adjustment. This step uses a comprehensive adjustment coefficient to adjust the weight of the social function of the venue so that the weight of the social function of the venue can more accurately reflect the impact of the age and physical ability differences of the elderly on social interaction.

[0106] In some preferred embodiments, step S5 includes:

[0107] S51. Based on the social interaction intensity scores, construct an initial social graph of the elderly community, where the graph nodes represent the elderly and the edge weights are initialized to the social interaction intensity scores;

[0108] S52. Based on the guidance of the venue area coding, the social interaction types corresponding to the venue function information are edge-matched on the initial elderly community social graph to form the elderly community social graph.

[0109] Specifically, in step S51, the graph nodes of the initial elderly community social graph are set to represent individual elderly people, and the weights of the edges connecting two elderly nodes are initialized to the social interaction intensity scores calculated in step S4. Thus, the initial graph can preliminarily quantify the intensity of social interactions among elderly people.

[0110] More specifically, step S52 uses the location region codes as a guide to process the edges of the initial senior citizen community social graph constructed in step S51. Specifically, for each edge in the graph, the location region codes corresponding to the two locations where the elderly individuals connected by that edge interacted are retrieved. Subsequently, based on the location region codes, the pre-constructed location semantic knowledge graph is searched to obtain the location function information corresponding to the location region codes. Location function information reflects the type of social interaction at that location. For example, a park might correspond to the social function of "leisure and entertainment," while a restaurant might correspond to the social function of "dining and socializing." The social interaction type corresponding to the location function information is then matched to the edge of the initial senior citizen community social graph as the attribute information for that edge. Through the processing in step S52, the resulting senior citizen community social graph is enriched with edge information. It not only includes weight information for the social interaction intensity scores, but also includes information on the social interaction type corresponding to the location function information. This provides a more comprehensive and detailed reflection of the social interaction patterns among senior citizens, providing a richer data foundation for subsequent community analysis and abnormal behavior warnings.

[0111] It should be noted that all calculation processes involved in the method for constructing a social graph for the elderly in the embodiment of the present application are dimensionless calculations.

[0112] Second, please refer to Figure 2 Some embodiments of the present application also provide a method for early warning of abnormal behavior of the elderly, the method comprising the following steps:

[0113] B1. Analyze the elderly community social graph obtained based on the elderly social graph construction method provided in the first aspect based on a community discovery algorithm to obtain community structure information;

[0114] B2. Obtain community parameter information based on community structure information;

[0115] B3. Based on the location of the elderly in the community structure information, combined with community parameter information, dynamically assess the social isolation risk level and its changing trend of each elderly person;

[0116] B4. Predict whether there are any abnormalities in the social behaviors of each elderly person based on the social isolation risk level and its changing trend.

[0117] Specifically, in step B1, a community discovery algorithm is used to analyze the social graph of the elderly community to identify the community structure in the graph. The community structure information reflects the community division within the elderly group and the relationship between communities. The community discovery algorithm can adopt existing algorithms such as the Leiden algorithm and the modularity optimization algorithm. The community structure information can be reflected in the community division results, for example, the identifier of the community to which each elderly person belongs. As an implementation method, the Louvain algorithm is adopted as the community discovery algorithm. The Louvain algorithm discovers the community structure in the graph by iteratively optimizing the modularity, and the modularity is used to evaluate the quality of the community division.

[0118] More specifically, in step B2, the community parameter information can quantitatively describe the characteristics of the community.

[0119] More specifically, step B3 dynamically assesses each elderly person's social isolation risk level and its changing trend based on their position in the community structure and community parameter information. The social isolation risk level reflects the degree of connection between the elderly person and their community, while the changing trend reflects how this connection changes over time.

[0120] More specifically, step B4 predicts whether each elderly person's social behavior is abnormal based on their social isolation risk level and its changing trends. For example, if an elderly person's social isolation risk level continues to rise, the prediction indicates a high probability of abnormal social behavior in the future. Through in-depth analysis of the social graph of elderly communities, the transition from social relationship graphs to abnormal behavior early warning is achieved. This overcomes the problem that simply constructing a social graph cannot directly be used for abnormal behavior early warning, providing a proactive and intelligent security measure for elderly care institutions.

[0121] The abnormal behavior warning method for the elderly in the embodiment of the present application is based on the analysis of the elderly community social graph obtained by the social graph construction method for the elderly provided in the first aspect to identify the community structure information in the elderly group and extract community parameter information, so as to dynamically evaluate the social isolation risk level and its changing trend of each elderly person for each node in the community structure and in combination with the community parameter information to determine whether there is a social isolation risk, thereby predicting whether the elderly’s future social behavior is abnormal, realizing early warning of abnormal behavior of the elderly, providing technical support for elderly care institutions, and helping to improve the quality of elderly care services.

[0122] In some preferred embodiments, step B1 comprises:

[0123] B11. The Louvain algorithm with modularity optimization is used to discover communities in the social graph of elderly communities. The resolution parameters of the Louvain algorithm are adaptively optimized in the following way:

[0124] B111. Construct a candidate parameter set for the resolution parameter, use a grid search method to select different resolution parameter values ​​from the candidate parameter set, and run the Louvain algorithm to divide the social graph of the elderly community into communities;

[0125] B112. Calculate the silhouette coefficient of each community division result, and select the resolution parameter corresponding to the maximum silhouette coefficient as the optimal resolution parameter;

[0126] B113. Extract community structure information based on the community segmentation results under the optimal resolution parameters.

[0127] Specifically, step B11 proposes a method for adaptively optimizing the resolution parameters of the Louvain algorithm to perform community discovery on the social graph of the elderly community. Step B111 first creates a set of candidate resolution parameter parameters, and then uses a grid search method to select different resolution parameter values ​​from the candidate parameter set so that the Louvain algorithm runs under each selected resolution parameter value, thereby obtaining multiple community division results. Step B112 calculates the silhouette coefficient of the community division for each community division result. The silhouette coefficient is used as an indicator to evaluate the quality of community division, and its numerical value reflects the quality of the community division effect. The selection of the optimal resolution parameter is achieved by comparing the silhouette coefficients of the community division under different resolution parameters. The resolution parameter corresponding to the maximum silhouette coefficient is determined as the optimal resolution parameter. After the optimal resolution parameter is determined, step B113 extracts the community division result under this parameter as community structure information for subsequent analysis.

[0128] More specifically, through the above processing method, the resolution parameter of the Louvain algorithm can be adaptively adjusted to the optimal state, thereby improving the accuracy and robustness of community discovery.

[0129] In some preferred embodiments, the community parameter information includes community size, community connection strength and node centrality index, wherein the community size is represented by the number of nodes in the community, the community connection strength is represented by the average weight of the connecting edges between nodes in the community, and the node centrality index is represented by the degree centrality index.

[0130] Specifically, in step B2, community parameter information is obtained based on the community structure information obtained in step B1. Community parameter information is a quantitative description of the community's structural characteristics, including community size, community connection strength, and node centrality indicators. Community size is determined by counting the number of elderly nodes within the community; community connection strength is quantified by calculating the weighted average of all connected edges within the community, with the edge weight representing the intensity of social interaction between elderly people; and node centrality indicators use degree centrality, which refers to the number of other elderly nodes directly connected to a specific elderly node in the community's social graph. Thus, community size can intuitively reflect the size of the community, community connection strength can reflect the closeness of connections between community members, and node centrality can quantify the social activity of a specific elderly person within the community. Using these three community parameter information, the social graph of the elderly community can be quantitatively analyzed from different dimensions of the community. Elderly care institutions can quantify the structural characteristics of the community and the position of individuals within the community, providing data support for subsequent social isolation risk assessment and abnormal behavior warning.

[0131] In some preferred embodiments, step B3 includes:

[0132] B31. Analyze and obtain the node type of each elderly person in the elderly community social graph based on their position in the community structure information. Node types are divided into: isolated nodes, edge nodes, and core nodes.

[0133] B32. Based on the node type, query the relationship between the node type and the community parameters to obtain the community size threshold, community connection strength threshold, and node centrality index threshold corresponding to each elderly person;

[0134] B33. Determine whether at least one of the following three conditions holds true: the size of each elderly person's community is smaller than the community size threshold, the community connection strength is smaller than the community connection strength threshold, and the node centrality index is smaller than the node centrality index threshold;

[0135] B34. If the result of step B33 is yes, it is determined that the elderly person is at risk of social isolation.

[0136] B35. For elderly people at risk of social isolation, based on community structure information, extract the elderly person's connection strength, number of connections, and type of connection partners in the community. Utilize a pre-trained risk assessment model to integrate connection strength, number of connections, and type of connection partners to calculate the elderly person's social isolation risk level.

[0137] B36. Use time series analysis to analyze the social isolation risk level of the elderly to obtain the changing trend of social isolation risk level.

[0138] Specifically, in step B31, based on the position of the elderly in the community structure information, the node types of the elderly in the community social graph are divided into isolated nodes, edge nodes, and core nodes. This division can reflect the different levels of social participation of the elderly in the community. The node division process can use a variety of graph theory indicators to determine the node type. For example, the division can be based on degree centrality. Nodes with extremely low degree centrality can be classified as isolated nodes, nodes between isolated nodes and core nodes can be classified as edge nodes, and nodes with high degree centrality can be regarded as core nodes.

[0139] More specifically, the thresholds set in step B32 serve as a benchmark for assessing the social isolation risk of the elderly. The thresholds can be set based on the statistical distribution of community structure parameters. For example, the community size, community connection strength, and node centrality indicators of all communities can be statistically analyzed, and then corresponding thresholds can be set based on the statistical distribution of different node types. As a preferred embodiment, these thresholds can be manually set based on experience or expert knowledge, combined with actual community social graph data.

[0140] More specifically, steps B33 and B34 compare the community size, community connection strength and node centrality indicators of the elderly community with their respective corresponding thresholds to determine whether at least one indicator is lower than the threshold. If so, it is preliminarily determined that the elderly person is at risk of social isolation.

[0141] More specifically, step B35 performs a detailed assessment of elderly individuals preliminarily identified as at risk of social isolation. This involves inputting community structure information, such as the strength of the elderly individual's connections within the community, the number of connections (the number of other elderly individuals connected within the community structure information), and the type of connection partners (the node types of other elderly individuals connected within the community structure information), into a pre-trained risk assessment model. The model then integrates this information and outputs the elderly individual's social isolation risk level. The assessment model can utilize a machine learning classification model, such as a support vector machine or random forest. The model's training data can utilize historical community social data and social isolation risk assessment results for the elderly individual. The model's input features include community structure information, such as the strength of the elderly individual's connections within the community, the number of connections, and the type of connection partners.

[0142] More specifically, step B36 uses a time series analysis method to perform a time series analysis on the elderly's social isolation risk level, thereby obtaining a trend in risk level changes. The time series analysis method can use methods such as an autoregressive integrated moving average model to model and predict the time series data of the elderly's social isolation risk level, thereby obtaining a trend in risk level changes.

[0143] More specifically, this approach can effectively assess the social isolation risk level and its changing trends among seniors, providing early warning information to nursing homes so they can implement timely intervention measures to improve their social well-being. This approach can conduct a multi-level, dynamic risk assessment for seniors with different social profiles, combining community parameters and individual connection characteristics, providing data support for early warning of abnormal behavior.

[0144] In some preferred embodiments, step B4 includes:

[0145] B41. Based on the pre-built abnormal behavior prediction model, the social isolation risk level and the trend of change in the social isolation risk level of the elderly who are at risk of social isolation are input to predict the probability of the target elderly participating in social activities in the future time window. If the probability is lower than the preset threshold, it is determined that the social activities of the target elderly have abnormal behavior.

[0146] Specifically, the abnormal behavior prediction model can adopt a machine learning algorithm. For example, the model can be trained using historical data, which includes the social isolation risk level of the elderly and its changing trend, as well as the corresponding social activity participation.

[0147] More specifically, the social isolation risk level and the trend of changes in the social isolation risk level are used as input to an abnormal behavior prediction model. This abnormal behavior prediction model predicts the probability of the target elderly person participating in social activities within a future time window. The future time window can be set according to actual needs, for example, it can be set to the next week or the next month. A preset threshold is set as the standard for determining whether social activities are abnormal. The threshold setting can be based on empirical data or expert knowledge. For example, if the average probability of elderly people participating in normal social activities is 50%, the preset threshold can be set to 30%. When the probability predicted by the model is lower than the preset threshold, the system determines that the elderly person's social activities have exhibited abnormal behavior, indicating that the elderly person's willingness or ability to participate in social activities in the future time window has significantly decreased.

[0148] In this document, relational terms such as first and second, etc. are used merely to distinguish one entity or operation from another entity or operation, but do not necessarily require or imply any actual relationship or order between these entities or operations.

[0149] The above description is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, various modifications and variations of the present application are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for constructing a social graph of the elderly, used to analyze the social relationships of the elderly group, characterized in that: The method comprises the following steps: S1, continuously collect location data of the elderly; S2. Obtain a pre-built place semantic knowledge graph, where the graph nodes include place area codes, and each place area code is preset with a place social function weight, a place type adjustment coefficient, and place function information; S3. For each pair of elderly people, analyze the co-occurrence frequency and duration of the same location area code corresponding to their location data over multiple consecutive time periods to determine whether there is a location co-occurrence relationship between the two. If a location co-occurrence relationship exists, calculate the location co-occurrence intensity based on the co-occurrence frequency, the duration, and the location type adjustment coefficient. S4. Calculating a social interaction intensity score among the elderly based on the location co-occurrence intensity and the social function weight of the venue; S5. Constructing a social graph of the elderly community based on the social interaction intensity score and the venue function information; Step S4 includes: S41. Obtain the age score and physical ability score of the elderly; S42. Adjust the weight of the social function of the venue according to the age score and physical ability score based on the location co-occurrence relationship; S43, calculating a social interaction intensity score between the elderly based on the adjusted location social function weight and the location co-occurrence intensity; Step S42 includes: S421. For every two elderly people who have a location co-occurrence relationship, calculate the age difference coefficient and the physical ability difference coefficient based on the age score and the physical ability score; S422. Based on the age difference coefficient and the physical ability difference coefficient, a comprehensive adjustment coefficient is calculated using the weighted average method; S423. Adjust the weight of the social function of the venue based on the comprehensive adjustment coefficient.

2. A method for constructing a social graph for the elderly according to claim 1, characterized in that: The step of analyzing, for each two elderly people, the co-occurrence frequency and duration of the same location area codes corresponding to their location data in multiple consecutive time periods to determine whether there is a location co-occurrence relationship between the two elderly people includes: S31. For each pair of elderly people, count the co-occurrence frequency and duration of the location data of the two elderly people corresponding to the same location area code in multiple consecutive time periods, and preliminarily determine whether there is a location co-occurrence relationship between the two based on the co-occurrence frequency and the duration; S32. When it is preliminarily determined that there is a location co-occurrence relationship between the two elderly people, the effective co-occurrence frequency and effective co-occurrence duration during the multiple consecutive time periods in which the distance between the two elderly people is less than the interaction distance threshold are counted based on the location data; S33. Finally determine whether there is a location co-occurrence relationship between the two based on the effective co-occurrence frequency and the effective co-occurrence duration.

3. A method for constructing a social graph for the elderly according to claim 2, characterized in that: The process of determining the interaction distance threshold includes: A1. According to the venue function information corresponding to the venue area code, obtain the interaction distance threshold from the venue function information and distance threshold mapping relationship table.

4. A method for constructing a social graph for the elderly according to claim 3, characterized in that: Each location area code is also preset with location area information; the process of determining the interaction distance threshold further includes: A2. Adjust the interaction distance threshold according to the venue area information corresponding to the venue area code.

5. The method for constructing a social graph for the elderly according to claim 1, characterized in that: The step of calculating the location co-occurrence intensity according to the co-occurrence frequency, the duration, and the venue type adjustment coefficient includes: S34, splitting the duration according to the duration to obtain the sub-co-occurrence durations of each two elderly people in multiple consecutive time periods; S35. Adjust the coefficient based on the venue type and set a time decay factor sequence for each time period regarding time decay; S36. Based on the time decay factor sequence, weight the sub-co-occurrence durations in each time period to obtain a weighted co-occurrence duration; S37. Multiply the weighted co-occurrence duration by the co-occurrence frequency to obtain the position co-occurrence strength.

6. The method for constructing a social graph for the elderly according to claim 1, characterized in that: Step S5 includes: S51. Constructing an initial social graph of the elderly community based on the social interaction intensity score, where the graph nodes represent elderly people and the edge weights are initialized to the social interaction intensity score; S52. Based on the guidance of the venue area coding, the social interaction type corresponding to the venue function information is edge-matched on the initial elderly community social graph to form the elderly community social graph.

7. A method for early warning of abnormal behavior of the elderly, characterized in that: The method comprises the following steps: B1. Analyzing the elderly community social graph obtained by the elderly social graph construction method according to any one of claims 1 to 6 based on a community discovery algorithm to obtain community structure information; B2. Obtaining community parameter information based on the community structure information; B3. Dynamically assessing the social isolation risk level and its changing trend of each elderly person based on their position in the community structure information and combining the community parameter information; B4. Predict whether the social behavior of each elderly person is abnormal based on the social isolation risk level and its changing trend.

8. The method for early warning of abnormal behavior of the elderly according to claim 7, characterized in that: Step B1 includes: B11. Use the modularity-optimized Louvain algorithm to perform community discovery on the elderly community social graph, wherein the resolution parameter of the Louvain algorithm is adaptively optimized in the following manner: B111. Construct a candidate parameter set for the resolution parameter, use a grid search method to select different resolution parameter values ​​from the candidate parameter set, and run the Louvain algorithm to divide the elderly community social graph into different groups; B112. Calculate the silhouette coefficient of each community division result, and select the resolution parameter corresponding to the maximum silhouette coefficient as the optimal resolution parameter; B113. Extract community structure information based on the community segmentation results under the optimal resolution parameters.

9. The method for early warning of abnormal behavior of the elderly according to claim 7, characterized in that: The community parameter information includes community size, community connection strength and node centrality index, wherein the community size is represented by the number of nodes in the community, the community connection strength is represented by the average weight of the connecting edges between nodes in the community, and the node centrality index is represented by the degree centrality index.

Images