User identification method, device and equipment based on multi-source bus data and storage medium

Abstract

The invention relates to a user identification method and device based on multi-source bus data, equipment and a storage medium. The method comprises the following steps: performing semantic context analysis on trajectory pairs based on travel time consistency and finite time space accessibility, and extracting a to-be-matched trajectory pair set with set co-occurrence times and reasonable semantics; performing individual access popularity, group access popularity and path connectivity analysis on a co-occurrence point corresponding to each track pair in the to-be-matched track pair set, and respectively defining contribution factors of the individual access popularity, the group access popularity and the path connectivity for performing uniqueness identification on the same user according to analysis results; and constructing a total matching degree calculation formula according to the contribution factor, and calculating a matching track of the target track by using the total matching degree calculation formula. According to the embodiment of the invention, the same user can be accurately identified in a track asynchronous scene, and the problem that the mass users are difficult to distinguish in a public traffic scene is effectively solved.

Description

Technical Field

[0001] This application belongs to the field of spatiotemporal data processing and public transportation technology, and specifically relates to a user identification method, device, equipment and storage medium based on multi-source public transportation data. Background Technology

[0002] Resident travel information is fundamental to urban population movement modeling and a prerequisite for resident travel service recommendations, urban traffic management, and epidemic simulation and control. Urban public transportation, as a crucial component of resident travel, is limited by the accessibility of a single transportation network. A complete trip often requires the use of multiple modes of transport, such as combining subways and buses, or subways and shared bicycles, to solve the "last mile" problem. Linking and connecting cross-trajectory data sources belonging to different modes of transport is essential to obtaining an individual's complete travel information. However, in most cities, different modes of transport belong to different operating units, and their travel data is collected and stored independently. The same user may have different identifiers in different datasets.

[0003] The problem of identifying the same user across trajectory data sources can be reduced to aligning or calculating the similarity of trajectories at different sampling rates and spatiotemporal resolutions. Given a target trajectory in one dataset, the most similar trajectory is retrieved from another dataset, and these two trajectories are considered to belong to the same user. Currently, there is a wealth of research on trajectory similarity calculation within the same value space, including methods such as longest common subsequence, edit distance, and dynamic time warping. Examples include two trajectories collected by a user while driving via a navigation app and a mobile base station; or two trajectories collected by a user while riding public transportation via a ticket reader and a vehicle-installed Wi-Fi device. However, existing methods primarily focus on measuring the similarity between synchronous trajectories with the same value space, neglecting the scenario of calculating the similarity of asynchronous trajectories with different value spaces. Summary of the Invention

[0004] This application provides a user identification method, apparatus, device, and storage medium based on multi-source public transport data, aiming to at least partially solve one of the aforementioned technical problems in the prior art.

[0005] To address the above problems, this application provides the following technical solution:

[0006] A user identification method based on multi-source public transport data includes:

[0007] Obtain multi-source trajectory datasets of at least two traffic subsystems, and generate trajectory pairs based on the multi-source trajectory datasets, wherein the trajectory pairs include a target trajectory and a trajectory to be matched.

[0008] Based on travel time consistency and limited time and space accessibility, semantic context analysis is performed on the trajectory pairs to extract a set of trajectory pairs to be matched that have a set of co-occurrence times and reasonable semantics.

[0009] For each trajectory pair in the set of trajectory pairs to be matched, perform individual access popularity, group access popularity and path connectivity analysis on the co-occurrence points corresponding to them, and define the contribution factors of individual access popularity, group access popularity and path connectivity for unique identification of the same user based on the analysis results.

[0010] Based on the contribution factor, a total matching degree calculation formula is constructed. The total matching degree calculation formula is used to calculate the similarity between the target trajectory and each trajectory to be matched in the set of trajectories to be matched. The trajectory to be matched with the highest similarity is taken as the matching trajectory of the target trajectory, and a travel trajectory pair of the same user is generated.

[0011] The technical solution adopted in this application embodiment further includes: after obtaining multi-source trajectory datasets of at least two traffic subsystems and generating trajectory pairs based on the multi-source trajectory datasets, it further includes:

[0012] Based on the spatiotemporal co-occurrence relationship, trajectory pairs with co-occurrence points are extracted, and trajectory pairs with a number of co-occurrence points greater than a set co-occurrence point threshold are selected to generate a candidate trajectory pair set. From the candidate trajectory pair set, potential matching trajectories that may match the target trajectory are extracted to generate a candidate trajectory set.

[0013] The technical solution adopted in this application embodiment also includes: the trajectory T is the user Tu's use of the traffic subsystem G over a period of time. x The resulting time-ordered travel sequence TL = {tl1, tl2, ..., tl |L|}, where tl i .m=m x The step involves extracting trajectory pairs with co-occurrence points based on spatiotemporal co-occurrence relationships, filtering out trajectory pairs with a number of co-occurrence points greater than a set threshold, generating a candidate trajectory pair set, and extracting potential matching trajectories from the candidate trajectory pair set to generate a candidate trajectory set. Specifically:

[0014] For a travel pair {tl,tl′} (tl.m ≠ tl′.m) consisting of two travel sequences from different transportation subsystems, if the starting point of the previous travel tl and the ending point of the subsequent travel tl′ simultaneously satisfy a set spatiotemporal condition in terms of their spatiotemporal relationship, then the travel pair tl,tl′ is called a travel pair with a co-occurrence point. The spatiotemporal condition is as follows:

[0015] (1) Spatially, the endpoint tl.s of the previous trip tl. dand the starting point of the next trip, tl′.s o The spherical distance between them is Dis(tl.s) d ,tl′.s o )≤λ d , where λ d This is the spatial proximity threshold;

[0016] (2) In terms of time, the start time of the later trip is tl′.t o The end time of the previous trip (tl) k .t d The time difference between them is Δt≤λ t , where λ t The time proximity threshold;

[0017] Traverse the trip pairs that are adjacent to each other and belong to different transportation subsystems, and extract the number of co-occurrence points of the trip pairs. And set the co-occurrence threshold λ o Filter out the number of all co-occurrence points Greater than the co-occurrence threshold λ o For each travel pair, generate a set of candidate trajectory pairs. and from the candidate trajectory set Extract the possible trajectory T that is related to the target trajectory. x Matching trajectory T y Generate a candidate trajectory set T y ={T y |C x,y ≥λ c}

[0018] The technical solution adopted in this application embodiment further includes: performing semantic context analysis on the trajectory pairs based on travel time consistency and finite time-space accessibility to extract a set of trajectory pairs to be matched with a set number of co-occurrences and reasonable semantics, specifically:

[0019] Regarding the consistency of travel time, this is achieved by statistically analyzing the target trajectory T. x and the trajectory to be matched T y ∈T y Number of times the travel time is inconsistent The proportion of travel time inconsistency is calculated using the following formula:

[0020]

[0021] The greater the inconsistency in travel time, the more effective the trajectory is for T. x ,T y The lower the probability that they belong to the same user, the lower the λ. c R is the noise tolerance collision threshold. t ≥λc As trajectory pair T x ,T y The first criterion for determining semantic rationality;

[0022] For the aforementioned finite time-space reachability, from the trajectory pair T x ,T y The number of semantically unreasonable adjacent travel pairs was statistically analyzed. And define the conflict ratio as:

[0023]

[0024] The greater the conflict ratio, the better the trajectory affects T. x ,T y The lower the probability that they belong to the same user, the lower the λ. s As the conflict threshold, R s ≥λ s As trajectory pair T x ,T y The second criterion for determining semantic rationality;

[0025] Based on the first and second judgment conditions of semantic rationality, for a given target trajectory T x Trajectories with unreasonable semantics are filtered out from the candidate trajectory pair set to obtain a set of trajectory pairs to be matched that have a certain number of co-occurrences and are semantically reasonable. Then, the filtered set of trajectory pairs to be matched, T, is extracted from the set of trajectory pairs to be matched. y ',in

[0026] The technical solution adopted in this application embodiment further includes: performing individual access popularity, group access popularity, and path connectivity analysis on the co-occurrence points corresponding to each trajectory pair in the set of trajectory pairs to be matched, and defining contribution factors for unique identification of the same user based on the analysis results, namely, individual access popularity, group access popularity, and path connectivity, further including:

[0027] The target trajectory T x and the set of trajectories to be matched T y The trajectory T to be matched in ′ y All co-occurring points are grouped according to their location, and the target trajectory T is obtained. x and the trajectory to be matched T y The set of all co-occurrence point locations contained in the data is S = (s1, s2, ..., s...). K ), where K is the number of co-occurrence points, C = (c1, c2, ..., c K ) represents T x and T y The number of co-occurrences at each co-occurrence point in S, Cx =(c x,1 ,c x,2 ,…,c x,K ) and C y =(c y,1 ,c y,2 ,…,c y,K ) represent the target trajectory T respectively x and the trajectory to be matched T y The number of co-occurrences at each location S, c x,k (c y,k ) represents the trajectory T x (T y The origin or destination and location s in the trip) k Satisfying the spatial proximity threshold λ d Number of trips;

[0028] For the target trajectory T x and the trajectory to be matched T y Each co-occurrence point s k ∈S and co-occurrence count c x,1 From the perspectives of travel patterns and statistical models, the co-occurrence point s is defined. k c x,1 The contribution of each co-occurrence to uniquely identify the same user; wherein, the contribution from the perspective of travel mode is calculated by using the Sigmoid function, using The calculation occurs at site s i The contribution of the qth co-occurrence to uniquely identify the same user, wherein the contribution decreases as q increases, and the parameter γ1 is used to control the number of decays; at all co-occurrence locations s k The total contribution of ∈S is:

[0029]

[0030] From a statistical perspective, based on the two trajectories at each co-occurrence point location s k co-occurrence count c x,k and number of trips c y,k The contribution of each co-occurrence point is calculated using the following formula:

[0031]

[0032] in This indicates the contribution that each co-occurrence point can provide.

[0033] The technical solution adopted in this application embodiment further includes: performing individual access popularity, group access popularity, and path connectivity analysis on the co-occurrence points corresponding to each trajectory pair in the set of trajectory pairs to be matched, and defining contribution factors for unique identification of the same user based on the analysis results, specifically:

[0034] The contributions from the travel mode perspective and the statistical mode perspective are fused, and the weights of the two contributions are controlled by θ1∈

[01] to obtain the co-occurrence point s. k The contribution factor M of individual visit popularity in uniquely identifying the same user u (s k ):

[0035] M u (s k )=(1-θ1)M c (s k )+θ1M h (s k )

[0036] Using co-occurrence point s k Total number of travelers This serves as the group access popularity of this site, and defines a contribution factor M based on the group access popularity view to uniquely identify the same user. d (s k ):

[0037]

[0038] Wherein, γ2 is a parameterized exponential distribution used to control the rate at which the contribution factor decreases as the popularity of group visits increases;

[0039] For each trip pair (tl, tl′) with co-occurring points, calculate the number C of co-occurring points where the two trips at all co-occurring point connections conform to the transfer path. f The contribution factor M of path connectivity is calculated using the number of co-occurring points. f :

[0040]

[0041] The technical solution adopted in this application embodiment further includes: constructing a total matching degree calculation formula based on the contribution factor, calculating the similarity between the target trajectory and each trajectory to be matched in the set of trajectories to be matched using the total matching degree calculation formula, and taking the trajectory to be matched with the highest similarity as the matching trajectory of the target trajectory to generate a travel trajectory pair for the same user, specifically:

[0042] The contribution factors of individual access popularity, group access popularity, and path connectivity to user identification are divided into two categories: point level and line level. The contribution factors of the two categories are combined using an addition operation, and the weights of the two categories are controlled by the parameter θ2. The final formula for calculating the total matching degree is as follows:

[0043]

[0044] Another technical solution adopted in this application embodiment is: a user identification device based on multi-source public transport data, comprising:

[0045] Data acquisition module: used to acquire multi-source trajectory datasets of at least two traffic subsystems, and generate trajectory pairs based on the multi-source trajectory datasets, wherein the trajectory pairs include a target trajectory and a trajectory to be matched;

[0046] Semantic analysis module: used to perform semantic context analysis on the trajectory pairs based on travel time consistency and limited time and space accessibility, and extract a set of trajectory pairs to be matched with a set number of co-occurrences and reasonable semantics;

[0047] Contribution factor calculation module: used to perform individual access popularity, group access popularity and path connectivity analysis on the co-occurrence points corresponding to each trajectory pair in the set of trajectory pairs to be matched, and define contribution factors for unique identification of the same user based on the analysis results of individual access popularity, group access popularity and path connectivity.

[0048] Similarity calculation module: used to construct a total matching degree calculation formula based on the contribution factor, use the total matching degree calculation formula to calculate the similarity between the target trajectory and each trajectory to be matched in the set of trajectories to be matched, and take the trajectory to be matched with the highest similarity as the matching trajectory of the target trajectory to generate a travel trajectory pair of the same user.

[0049] Another technical solution adopted in this application embodiment is: a device, the device including a processor and a memory coupled to the processor, wherein,

[0050] The memory stores program instructions for implementing the user identification method based on multi-source public transport data;

[0051] The processor is used to execute the program instructions stored in the memory to control the user identification method based on multi-source public transport data.

[0052] Another technical solution adopted in this application embodiment is: a storage medium storing program instructions that can be run by a processor, the program instructions being used to execute the user identification method based on multi-source public transport data.

[0053] Compared to existing technologies, the beneficial effects of the embodiments of this application are as follows: The user identification method, device, equipment, and storage medium based on multi-source public transport data in this application define a method for identifying the same user based on multi-source trajectory data by combining global semantic rationality and local co-occurrence point differences. At the level of global semantic rationality, it obtains matching trajectory pairs with consistent semantic context by combining individual travel time consistency and limited time and space accessibility. At the level of local co-occurrence point differences, it comprehensively considers features such as individual access popularity, group access popularity, co-occurrence frequency, and path connectivity of co-occurrence points to perform adaptive calculation of contribution factors, and constructs a total matching degree calculation formula based on the contribution factors. The trajectory pair with the highest similarity is then calculated using the total matching degree calculation formula. This application can accurately identify the same user in asynchronous trajectory scenarios, effectively solving the problem of difficulty in distinguishing users in public transport scenarios with massive numbers of users due to similar travel patterns, limited space of occurrence location values, and sparse and low-quality trajectories. Attached Figure Description

[0054] Figure 1 This is a flowchart of a user identification method based on multi-source public transport data according to an embodiment of this application;

[0055] Figure 2 This is a schematic diagram comparing the number of co-occurrence points of users who are correctly matched and incorrectly matched, according to an embodiment of this application.

[0056] Figure 3 This is a schematic diagram of the user identification device based on multi-source public transport data according to an embodiment of this application;

[0057] Figure 4 This is a schematic diagram of the device structure according to an embodiment of this application;

[0058] Figure 5 This is a schematic diagram of the structure of the storage medium according to an embodiment of this application. Detailed Implementation

[0059] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0060] The terms "first," "second," and "third" in this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of this application are only used to explain the relative positional relationships and movements between components in a specific orientation (as shown in the figures). If the specific orientation changes, the directional indications also change accordingly. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0061] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0062] Specifically, please refer to Figure 1 This is a flowchart of a user identification method based on multi-source public transport data according to an embodiment of this application. The user identification method based on multi-source public transport data according to an embodiment of this application includes the following steps:

[0063] S100: Obtain multi-source trajectory datasets from various different traffic subsystems, and generate trajectory pairs consisting of the target trajectory and the trajectory to be matched based on the multi-source trajectory datasets;

[0064] In this step, let's assume we use graph G. x = (V, E) represents the mode of travel as m x The traffic subsystem, where V = {v1, v2, ..., v n} represents the set of stations such as bus stops, subway stations, or shared bicycle stops, expressed in latitude and longitude; E = {(v i ,v j )|v i ,v j ∈V},(v i ,v j ) indicates at site v i and vj The two are reachable. The trajectory is represented by T, indicating the time period during which user Tu uses a certain transportation subsystem G. x The resulting time-ordered travel sequence TL = {tl1, tl2, ..., tl |L|}, where tl i .m=m x Each trip in the trip sequence includes the user's trip identifier tl and the starting position tl.o∈G. x V, End position tl.d∈G x .V、Start time tl.t o End time tl.t d (tl.t o <tl.t d ) and travel mode tl.m. Given any two transportation subsystems G within the same time period. x and G y The trajectory dataset D to which it belongs x ={T x},D y ={T y The two trajectories T in} x ∈D x and T y ∈D y If these two trajectories belong to the same user, then T y It is T x The purpose of this application embodiment is to construct the function F(T) for matching trajectories. x D y ), combining factors such as co-occurrence points in the two trajectories, from the trajectory dataset D x ={T x},D y ={T y} Filter out those with the target trajectory T x The most similar matching trajectory T y It generates matching trajectory pairs of the same user using different modes of travel, thereby uniquely identifying the same user using different modes of travel.

[0065] S110: Extract trajectory pairs with co-occurrence points based on spatiotemporal co-occurrence relationships, filter out trajectory pairs with a number of co-occurrence points greater than a set co-occurrence point threshold, generate a candidate trajectory pair set, and extract potential matching trajectories that may match the target trajectory from the candidate trajectory pair set to generate a candidate trajectory set.

[0066] In this step, because the accessibility of public transportation is limited by its network, transfers between multiple modes of transportation are necessary. Trajectories of the same user collected from different modes of transportation may co-occur at connection points or transfer points. Therefore, co-occurrence points can be used as a basic condition for two trajectories belonging to the same user. Specifically, for a travel pair {tl,tl′} (tl.m ≠ tl′.m) consisting of two travel sequences from different transportation subsystems, if the starting point of the previous travel tl and the ending point of the subsequent travel tl′ simultaneously satisfy the set spatiotemporal conditions, then the travel pair tl,tl′ is called a travel pair with co-occurrence points, and the ending point of the previous travel tl is defined as... d As asynchronous neighboring co-occurrence points of the travel pair {tl,tl′}, they are simply referred to as co-occurrence points. The specific spatiotemporal conditions set are as follows:

[0067] (1) Spatially, the endpoint tl.s of the previous trip tl. d and the starting point of the next trip, tl′.s o The spherical distance between them is Dis(tl.s) d ,tl′.s o )≤λ d , where λ d The spatial proximity threshold can be set according to the specific conditions of the urban transportation network facilities;

[0068] (2) In terms of time, the start time of the later trip is tl′.t o The end time of the previous trip (tl) k .t d The time difference between them is Δt≤λ t , where λ t The time proximity threshold can be set according to the specific conditions of the city's transportation network facilities.

[0069] Furthermore, after extracting trip pairs with co-occurring points, if the number of co-occurring points for a trip pair is too small, it is difficult to determine whether two trajectories belong to the same user. Therefore, this embodiment of the application extracts the number of co-occurring points for trip pairs by traversing adjacent trip pairs belonging to different traffic subsystems. And set the co-occurrence threshold λ o Filter out the number of all co-occurrence points Greater than the co-occurrence threshold λ o For each travel pair, generate a set of candidate trajectory pairs. For the set of candidate trajectory pairs D xy Each target trajectory T in x The corresponding trajectory T to be matched y There may be more than one; for convenience, this application embodiment uses a set of candidate trajectory pairs. Extract possible trajectory T related to the target trajectory x Matching trajectory T y Generate a candidate trajectory set T y ={T y |C x,y ≥λ c}

[0070] S120: Based on travel time consistency and limited time and space accessibility, perform semantic context analysis on the candidate trajectory pair set, extract the set of trajectory pairs to be matched with a certain number of co-occurrences and reasonable semantics, and extract the filtered set of trajectory pairs to be matched from the set of trajectory pairs to be matched;

[0071] In this step, regarding the consistency of travel time, the location of an individual at a given moment is unique, meaning that for any two consecutive trips tl of the same individual... k and tl k+1 In terms of time, there are two types of overlapping relationships: overlapping and nesting. Overlapping indicates the previous trip (TL). k The end time should not be later than the next trip. k+1 End time tl k+1 , i.e. tl k .t d -tl k+1 .t d ≥0. Nested representation indicates the last trip's time limit. k+1 The start time should not be later than the previous trip. k The end time, i.e., tl k .t d -tl k+1 .t o ≥0. Meanwhile, due to noise in the trajectory dataset, the travel times of the same user may overlap to some extent. To improve the user recall rate of the algorithm, a small number of time conflicts can be allowed when the co-occurrence frequency is sufficiently high. Therefore, this application embodiment uses statistical analysis of the target trajectory T... x and the trajectory to be matched T y ∈T y Number of times the travel time is inconsistent And the proportion of travel time inconsistency is calculated using formula (1):

[0072]

[0073] The greater the inconsistency in travel time, the greater the impact of trajectory on T. x ,T y The lower the probability that they belong to the same user, the less likely they are to be involved. A noise tolerance conflict threshold λ is also set. c , will R t ≥λ c As trajectory pair Tx ,T y The first criterion for determining semantic rationality.

[0074] For finite time and space accessibility, in any adjacent travel pair (tl) k ,tl k+1 In the previous trip, tl k The endpoint tl k .s d After arrival, one trip tl k+1 The starting point tl k+1 .s o It needs to meet the requirement of meeting the maximum speed limit for urban vehicles within the time interval tl k+1 .t o -tl k+1 .t o Within the city limits, the maximum speed limit is set to 35 m / s, based on the highest speed a car can travel within the city. Trips exceeding this speed are considered anomalies. Similarly, due to data noise, the same user may experience some spatial accessibility conflicts in both traffic network trajectories. Therefore, a small number of spatial accessibility conflicts are allowed provided there are a sufficient number of co-occurrences. Based on this, embodiments of this application consider trajectory pair T... x ,T y The number of semantically unreasonable adjacent travel pairs was statistically analyzed. And define the conflict ratio as:

[0075]

[0076] The greater the conflict ratio, the better the trajectory affects T. x ,T y The lower the probability of conflicts originating from the same user, the less likely they are to occur. A conflict threshold λ is also set. s , will R s ≥λ s As trajectory pair T x ,T y The second criterion for determining semantic rationality.

[0077] Based on the above two semantic rationality judgment conditions, for a given target trajectory T x The process involves filtering out semantically unreasonable trajectories from the candidate trajectory pair set, ultimately obtaining a set of trajectory pairs to be matched that have a certain number of co-occurrences and are semantically reasonable. Finally, the filtered set of trajectory pairs to be matched, T, is extracted from this set. y ',in T y ′ could be null or include multiple trajectories.

[0078] S130: Perform individual access popularity, group access popularity, and path connectivity analysis on the co-occurrence points corresponding to each trajectory pair in the set of trajectory pairs to be matched, and define the contribution factors of individual access popularity, group access popularity, and path connectivity for unique identification of the same user based on the analysis results.

[0079] In this step, due to the large number of people, limited location space, limited individual activity locations, and strong travel patterns in urban public transportation, there may be a large number of co-occurrence points among different users. Simply relying on the number of co-occurrences based on trajectories is insufficient to accurately identify the uniqueness of the same user. Therefore, this application proposes an adaptive co-occurrence point contribution factor calculation method by comprehensively considering features such as individual access popularity, group access popularity, and path connectivity at the location of co-occurrence points. The specific calculation method includes the following steps:

[0080] S131: Set the target trajectory T x and the set of trajectories to be matched T y The trajectory T to be matched in ′ y All co-occurring points are grouped according to their location (station), and the target trajectory T is obtained. x and the trajectory to be matched T y The set of all co-occurring points contained in the data; specifically, user travel patterns are regular, and they may take or transfer between different modes of transportation multiple times at the same location. Therefore, there may be multiple co-occurring points at the same location, denoted by S = (s1, s2, ..., s...). K ) represents T x and T y The set of all co-occurrence locations contained in the data, where K is the number of co-occurrence locations, and C = (c1, c2, ..., c...). K ) represents T x and T y The number of co-occurrences at each co-occurrence point in S, C x =(c x,1 ,c x,2 ,…,c x,K ) and C y =(c y,1 ,c y,2 ,…,c y,K ) represent the target trajectory T respectively x and the trajectory to be matched T y The number of co-occurrences at each location S, c x,k (c y,k ) represents the trajectory T x (T y The origin or destination and location s in the trip) k Satisfying the spatial proximity threshold λ d Number of trips.

[0081] S132: For the target trajectory T x and the trajectory to be matched T y Each co-occurrence point s k ∈S and co-occurrence count c x,1 From the perspectives of travel patterns and statistical models, the co-occurrence point s is defined. k c x,1 The contribution of each co-occurrence to the unique identification of the same user;

[0082] From the perspective of travel patterns, statistics revealed that mismatched trajectory pairs share a common characteristic: their co-occurrence points are concentrated in a limited number of stations. Specifically, for example... Figure 2 The diagram shows a comparison of co-occurrence points between correctly matched and incorrectly matched user trajectories. It illustrates the co-occurrence point distribution of trajectory pairs that collided with themselves more than 10 times within a month, as well as incorrectly matched and correctly matched trajectory pairs. Figure 2 It can be observed that although mismatched trajectories co-occur more frequently than correctly matched trajectories, 98% of these co-occurrence points are concentrated at a single station. In contrast, nearly 90% of correctly matched trajectory pairs show at least two user co-occurrence points. This is because, assuming user A exits subway station S around 8:00 AM every day, while user B takes a bus from a nearby bus stop at the same time, then users A and B share the travel pattern of "appearing at station S around 8:00 AM every day." This type of travel is relatively regular and predictable; that is, if two users frequently meet at the same location within a certain period, the probability of them meeting again is relatively high. For highly predictable co-occurrence points, their contribution to identifying the same user can be appropriately reduced, which also increases the contribution of some random travel from another perspective.

[0083] The more times two trajectories co-occur at a certain location, the less effective adding another co-occurrence at that location will be in determining whether the two trajectories belong to the same user. This application's embodiment uses the Sigmoid function. The calculation occurs at site s i The contribution of the qth co-occurrence to uniquely identify the same user decreases as q increases. A parameter γ1 is set to control the number of decays, which can be configured according to the specific application scenario. Specifically, at all co-occurrence locations s... k The total contribution of ∈S is calculated using the following formula:

[0084]

[0085] From a statistical perspective, given a known number of co-occurrences, the fewer times two trajectories travel to the same location, the greater the likelihood that they belong to the same user. Therefore, this embodiment of the application comprehensively considers the two trajectories at each co-occurrence point s.k co-occurrence count c x,k and number of trips c y,k The contribution of each co-occurrence point is calculated using the following formula:

[0086]

[0087] in This represents the contribution that each co-occurrence point can provide, and its relationship with the two users at the co-occurrence point location s. k The total number of trips is inversely proportional to the number of co-occurrences.

[0088] S133: The contributions from the perspectives of travel patterns and statistical patterns are fused, and the weights of the two contributions are controlled by θ1∈

[01] to obtain the co-occurrence point s. k The contribution factor M of individual visit popularity in uniquely identifying the same user u (s k ):

[0089] M u (s k )=(1-θ1)M c (s k )+θ1M h (s k (5)

[0090] S134: Using co-occurrence points s k Total number of travelers This serves as the group access popularity of this site, and defines a contribution factor M based on the group access popularity view to uniquely identify the same user. d (s k );

[0091] Here, group visit popularity refers to the total number of visits by all users to this location, for each co-occurrence point s. k The greater the popularity of a user group at location S, the greater the likelihood of different users co-occurring at that location, thus reducing the possibility of uniquely identifying the same user at that location. Scholars Cao et al. proposed a rule to filter public areas in their cross-data source user identification model to avoid interference from "fellow travelers." The proposed public area refers to the area where the number of visiting users exceeds a threshold. For asynchronous trajectory scenarios, the number of co-occurrence points is inherently small. Directly filtering all high-popularity sites would reduce the number of co-occurrence points between trajectories. Therefore, this application embodiment uses co-occurrence point S. k Total number of travelers As a measure of the group access popularity of this site, the user identification contribution factor M is defined from the group access popularity view. d (s k ):

[0092]

[0093] in, The range of values ​​for M is relatively large. d (s k )along with The change should not be too rapid. γ2 is a parameterized exponential distribution used to control the rate at which the contribution factor decreases as the popularity of group visits increases. The specific configuration can be determined according to the actual application scenario.

[0094] S135: For each travel pair (tl, tl′) with co-occurring points, calculate the number C of co-occurring points where the two travels at all co-occurring point connections conform to the transfer path. f The contribution factor M of path connectivity is calculated using the number of co-occurring points. f ;

[0095] Specifically, under normal circumstances, if a user switches from one mode of public transportation to another at a certain location within a short period of time, it is highly likely to be a transfer. Therefore, if two trips tl and tl′ with co-occurrence points belong to the same user, the path connecting trips tl and tl′ should be close to or not deviate from tl.s. o and tl′.s d There are too many shortest paths, including tl.s o tl′.s represents the starting point of the first segment of a continuous trip with co-occurring points. d This represents the destination of the last segment of a continuous transfer trip with co-located points. Therefore, for each trip pair (tl, tl′) with co-occurring points, the endpoints of the two points tl and tl′ are calculated using the open-source algorithm OpenTripPlanner, from a connectivity or line perspective. o and tl′.s d Average travel time when using a combination of two transportation systems And use the following formula to calculate the abnormal status of travel time:

[0096]

[0097] Where A(tl,tl′)=0 indicates abnormal travel time, A(tl,tl′)=1 indicates travel time is within the normal range, σ t The criteria for controlling anomalies are set to 1.5 in this paper, based on the effective path selection method. That is, if the path exceeds the average path by 1.5 times, it is called an abnormal path.

[0098] By calculating the number C of co-occurring points at all points where two trips conform to the transfer path, we can determine the co-occurring points. fThe more times a user's connection to a transfer path is confirmed, the greater the contribution to unique identification of that user. This is calculated as the proportion of co-occurring points confirming transfer paths out of the total number of occurrences. The number of co-occurring points matching the transfer path (C) f The product of these factors yields the contribution factor M of path connectivity. f for:

[0099]

[0100] S140: Construct a formula for calculating the total matching degree based on the contribution factors, and use the formula to calculate the target trajectory T. x With the set of trajectories to be matched T y Each trajectory T to be matched in ′ y The similarity is used to determine the trajectory T to be matched with the highest similarity. y As the target trajectory T x The matching trajectory generates travel trajectory pairs for the same user;

[0101] In this step, given the differences in individual and group visit popularity and path connectivity at the locations of co-occurrence points between two trajectories, the weight for unique identification of the same user will also differ. This embodiment of the application dynamically calculates the trajectory pair (T) by comprehensively considering these influencing factors. x ,T y The contribution value of each co-occurring point in the data to uniquely identify the same user is accumulated to obtain the similarity between the two trajectories, which will be compared with T. x The most similar T y As its matching trajectory.

[0102] Specifically, the formula for calculating the overall matching degree is constructed as follows: First, the contribution factors of individual access popularity, group access popularity, and path connectivity to user identification are divided into point levels (including M). u M h ) and line level (path connectivity M) f The system considers two categories, combines their contribution factors using an addition operation, and controls the weights of the two categories using the parameter θ2. The final formula for calculating the overall matching degree is:

[0103]

[0104] The target trajectory T is calculated using formula (9). x With the set of trajectories to be matched T y Each T in ′ y The similarity score will be used to determine the T with the highest similarity score. y As T x The matching trajectory.

[0105] To verify the feasibility and effectiveness of the embodiments of this application, experiments were conducted using real data from buses and subways in Shenzhen, China. The experimental results show that the accuracy rate of using the embodiments of this application for unique identification of the same user is greater than 85%.

[0106] Based on the above, the user identification method based on multi-source public transport data in this application defines a method for identifying the same user from multi-source trajectory data by combining global semantic rationality and local co-occurrence point differences. At the level of global semantic rationality, it obtains matching trajectory pairs with consistent semantic context by combining individual travel time consistency and limited time-space reachability. At the level of local co-occurrence point differences, it comprehensively considers features such as individual access popularity, group access popularity, co-occurrence frequency, and path connectivity of co-occurrence points to adaptively calculate contribution factors, and constructs a total matching degree calculation formula based on the contribution factors. The total matching degree calculation formula is then used to calculate the trajectory pair with the highest similarity. This application can accurately identify the same user in asynchronous trajectory scenarios, effectively solving the problem of difficulty in distinguishing users in public transport scenarios with massive numbers of users due to similar travel patterns, limited space of occurrence locations, and sparse and low-quality trajectories.

[0107] Please see Figure 3 This is a schematic diagram of the user identification device based on multi-source public transport data according to an embodiment of this application. The user identification device 40 based on multi-source public transport data according to an embodiment of this application includes:

[0108] Data acquisition module 41: used to acquire multi-source trajectory datasets of at least two traffic subsystems, and generate trajectory pairs based on the multi-source trajectory datasets, wherein the trajectory pairs include a target trajectory and a trajectory to be matched;

[0109] Semantic analysis module 42: used to perform semantic context analysis on the trajectory pairs based on travel time consistency and limited time and space accessibility, and extract a set of trajectory pairs to be matched with a set number of co-occurrences and reasonable semantics;

[0110] Contribution factor calculation module 43: used to perform individual access popularity, group access popularity and path connectivity analysis on the co-occurrence points corresponding to each trajectory pair in the set of trajectory pairs to be matched, and define contribution factors for unique identification of the same user based on the analysis results of individual access popularity, group access popularity and path connectivity.

[0111] Similarity calculation module 44: is used to construct a total matching degree calculation formula based on the contribution factor, calculate the similarity between the target trajectory and each trajectory to be matched in the set of trajectory to be matched using the total matching degree calculation formula, and take the trajectory to be matched with the highest similarity as the matching trajectory of the target trajectory to generate a travel trajectory pair of the same user.

[0112] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0113] The apparatus provided in this application can be applied to the foregoing method embodiments. For details, please refer to the description of the above method embodiments, which will not be repeated here.

[0114] Please see Figure 4 This is a schematic diagram of the device structure according to an embodiment of this application. The device 50 includes:

[0115] Memory 51 storing executable program instructions;

[0116] Processor 52 connected to memory 51;

[0117] The processor 52 is used to call executable program instructions stored in the memory 51 and perform the following steps: acquire multi-source trajectory datasets of at least two traffic subsystems; generate trajectory pairs based on the multi-source trajectory datasets, wherein the trajectory pairs include a target trajectory and a trajectory to be matched; perform semantic context analysis on the trajectory pairs based on travel time consistency and limited time-space reachability, and extract a set of trajectory pairs to be matched that have a set number of co-occurrences and are semantically reasonable; perform individual access popularity, group access popularity, and path connectivity analysis on the co-occurrence points corresponding to each trajectory pair in the set of trajectory pairs to be matched, and define contribution factors for unique identification of the same user based on the analysis results; construct a total matching degree calculation formula based on the contribution factors; use the total matching degree calculation formula to calculate the similarity between the target trajectory and each trajectory to be matched in the set of trajectory pairs to be matched, and take the trajectory to be matched with the highest similarity as the matching trajectory of the target trajectory to generate a travel trajectory pair for the same user.

[0118] The processor 52 can also be referred to as a CPU (Central Processing Unit). The processor 52 may be an integrated circuit chip with signal processing capabilities. The processor 52 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor can be a microprocessor or any conventional processor.

[0119] Please see Figure 5This is a schematic diagram of the storage medium in an embodiment of this application. The storage medium in this embodiment stores program instructions 61 capable of implementing the following steps: acquiring multi-source trajectory datasets of at least two traffic subsystems; generating trajectory pairs based on the multi-source trajectory datasets, wherein each trajectory pair includes a target trajectory and a trajectory to be matched; performing semantic context analysis on the trajectory pairs based on travel time consistency and finite time-space reachability, and extracting a set of trajectory pairs to be matched that have a set number of co-occurrences and are semantically reasonable; performing individual access popularity, group access popularity, and path connectivity analysis on the co-occurrence points corresponding to each trajectory pair in the set of trajectory pairs to be matched, and defining contribution factors for unique identification of the same user based on the analysis results; constructing a total matching degree calculation formula based on the contribution factors; calculating the similarity between the target trajectory and each trajectory to be matched in the set of trajectory pairs using the total matching degree calculation formula; taking the trajectory to be matched with the highest similarity as the matching trajectory of the target trajectory, and generating a travel trajectory pair for the same user. The program instructions 61 can be stored in the aforementioned storage medium in the form of a software product. These instructions include several instructions to cause a device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program instructions, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks, or terminal devices such as computers, servers, mobile phones, and tablets. The server can be an independent server or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0120] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, apparatuses, or units, and may be electrical, mechanical, or other forms.

[0121] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A user identification method based on multi-source public transport data, characterized in that, include: Obtain multi-source trajectory datasets of at least two traffic subsystems, and generate trajectory pairs based on the multi-source trajectory datasets, wherein the trajectory pairs include a target trajectory and a trajectory to be matched. Based on travel time consistency and limited time and space accessibility, semantic context analysis is performed on the trajectory pairs to extract a set of trajectory pairs to be matched that have a set of co-occurrence times and reasonable semantics. For each trajectory pair in the set of trajectory pairs to be matched, perform individual access popularity, group access popularity and path connectivity analysis on the co-occurrence points corresponding to them, and define the contribution factors of individual access popularity, group access popularity and path connectivity for unique identification of the same user based on the analysis results. Based on the contribution factor, a total matching degree calculation formula is constructed. The total matching degree calculation formula is used to calculate the similarity between the target trajectory and each trajectory to be matched in the set of trajectories to be matched. The trajectory to be matched with the highest similarity is taken as the matching trajectory of the target trajectory, and a travel trajectory pair of the same user is generated.

2. The user identification method based on multi-source public transport data according to claim 1, characterized in that, After acquiring multi-source trajectory datasets from at least two traffic subsystems and generating trajectory pairs based on the multi-source trajectory datasets, the method further includes: Based on the spatiotemporal co-occurrence relationship, trajectory pairs with co-occurrence points are extracted, and trajectory pairs with a number of co-occurrence points greater than a set co-occurrence point threshold are selected to generate a candidate trajectory pair set. From the candidate trajectory pair set, potential matching trajectories that may match the target trajectory are extracted to generate a candidate trajectory set.

3. The user identification method based on multi-source public transport data according to claim 2, characterized in that, The trajectory T represents the user Tu's use of the traffic subsystem G over a period of time. x The resulting time-ordered travel sequence TL = {tl1, tl2, ..., tl |L| }, where tl i .m=m x The step involves extracting trajectory pairs with co-occurrence points based on spatiotemporal co-occurrence relationships, filtering out trajectory pairs with a number of co-occurrence points greater than a set threshold, generating a candidate trajectory pair set, and extracting potential matching trajectories from the candidate trajectory pair set to generate a candidate trajectory set. Specifically: For a travel pair {tl,tl′} (tl.m ≠ tl′.m) consisting of two travel sequences from different transportation subsystems, if the starting point of the previous travel tl and the ending point of the subsequent travel tl′ simultaneously satisfy a set spatiotemporal condition in terms of their spatiotemporal relationship, then the travel pair tl,tl′ is called a travel pair with a co-occurrence point. The spatiotemporal condition is as follows: (1) Spatially, the endpoint tl.s of the previous trip tl. d and the starting point of the next trip, tl′.s o The spherical distance between them is Dis(tl.s) d ,tl′.s o )≤λ d , where λ d This is the spatial proximity threshold; (2) In terms of time, the start time of the later trip is tl′.t o The end time of the previous trip (tl) k .t d The time difference between them is Δt≤λ t , where λ t The time proximity threshold; Traverse the trip pairs that are adjacent to each other and belong to different transportation subsystems, and extract the number of co-occurrence points of the trip pairs. And set the co-occurrence threshold λ o Filter out the number of all co-occurrence points Greater than the co-occurrence threshold λ o For each travel pair, generate a set of candidate trajectory pairs. and from the candidate trajectory set Extract the possible trajectory T that is related to the target trajectory. x Matching trajectory T y Generate a candidate trajectory set T y ={T y |C x,y ≥λ c } 4. The user identification method based on multi-source public transport data according to claim 3, characterized in that, The semantic context analysis of the trajectory pairs based on travel time consistency and finite time-space reachability extracts a set of trajectory pairs to be matched that have a set of co-occurrence frequencies and are semantically reasonable. Specifically: Regarding the consistency of travel time, this is achieved by statistically analyzing the target trajectory T. x and the trajectory to be matched T y ∈T y Number of times the travel time is inconsistent The proportion of travel time inconsistency is calculated using the following formula: The greater the inconsistency in travel time, the more effective the trajectory is for T. x ,T y The lower the probability that they belong to the same user, the lower the λ. c R is the noise tolerance collision threshold. t ≥λ c As trajectory pair T x ,T y The first criterion for determining semantic rationality; For the aforementioned finite time-space reachability, from the trajectory pair T x ,T y The number of semantically unreasonable adjacent travel pairs was statistically analyzed. And define the conflict ratio as: The greater the conflict ratio, the better the trajectory affects T. x ,T y The lower the probability that they belong to the same user, the lower the λ. s As the conflict threshold, R s ≥λ s As trajectory pair T x ,T y The second criterion for determining semantic rationality; Based on the first and second judgment conditions of semantic rationality, for a given target trajectory T x Trajectories with unreasonable semantics are filtered out from the candidate trajectory pair set to obtain a set of trajectory pairs to be matched that have a certain number of co-occurrences and are semantically reasonable. Then, the filtered set of trajectory pairs to be matched, T, is extracted from the set of trajectory pairs to be matched. y ',in 5. The user identification method based on multi-source public transport data according to any one of claims 1 to 4, characterized in that, The step of performing individual access popularity, group access popularity, and path connectivity analysis on the co-occurrence points corresponding to each trajectory pair in the set of trajectory pairs to be matched, and defining contribution factors for unique identification of the same user based on the analysis results, further includes: The target trajectory T x and the set of trajectories to be matched T y The trajectory T to be matched in ′ y All co-occurring points are grouped according to their location, and the target trajectory T is obtained. x and the trajectory to be matched T y The set of all co-occurrence point locations contained in the data is S = (s1, s2, ..., s...). K ), where K is the number of co-occurrence points, C = (c1, c2, ..., c K ) represents T x and T y The number of co-occurrences at each co-occurrence point in S, C x =(c x,1 ,c x,2 ,…,c x,K ) and C y =(c y,1 ,c y,2 ,…,c y,K ) represent the target trajectory T respectively x and the trajectory to be matched T y The number of co-occurrences at each location S, c x,k (c y,k ) represents the trajectory T x (T y The origin or destination and location s in the trip) k Satisfying the spatial proximity threshold λ d Number of trips; For the target trajectory T x and the trajectory to be matched T y Each co-occurrence point s k ∈S and co-occurrence count c x,1 From the perspectives of travel patterns and statistical models, the co-occurrence point s is defined. k c x,1 The contribution of each co-occurrence to uniquely identify the same user; wherein, the contribution from the perspective of travel mode is calculated by using the Sigmoid function, using The calculation occurs at site s i The contribution of the qth co-occurrence to uniquely identify the same user, wherein the contribution decreases as q increases, and the parameter γ1 is used to control the number of decays; at all co-occurrence locations s k The total contribution of ∈S is: From a statistical perspective, based on the two trajectories at each co-occurrence point location s k co-occurrence count c x,k and number of trips c y,k The contribution of each co-occurrence point is calculated using the following formula: in This indicates the contribution that each co-occurrence point can provide.

6. The user identification method based on multi-source public transport data according to claim 5, characterized in that, The method involves performing individual access popularity, group access popularity, and path connectivity analysis on the co-occurrence points corresponding to each trajectory pair in the set of trajectory pairs to be matched, and defining contribution factors for unique identification of the same user based on the analysis results, specifically: The contributions from the travel mode perspective and the statistical mode perspective are fused, and the weights of the two contributions are controlled by θ1∈[01] to obtain the co-occurrence point s. k The contribution factor M of individual visit popularity in uniquely identifying the same user u (s k ): M u (s k )=(1-θ1)M c (s k )+θ1M h (s k ) Using co-occurrence point s k Total number of travelers This serves as the group access popularity of this site, and defines a contribution factor M based on the group access popularity view to uniquely identify the same user. d (s k ): Wherein, γ2 is a parameterized exponential distribution used to control the rate at which the contribution factor decreases as the popularity of group visits increases; For each trip pair (tl, tl′) with co-occurring points, calculate the number C of co-occurring points where the two trips at all co-occurring point connections conform to the transfer path. f The contribution factor M of path connectivity is calculated using the number of co-occurring points. f :

7. The user identification method based on multi-source public transport data according to claim 6, characterized in that, The step involves constructing a total matching degree calculation formula based on the contribution factor, calculating the similarity between the target trajectory and each trajectory in the set of trajectories to be matched using the total matching degree calculation formula, and taking the trajectory with the highest similarity as the matching trajectory of the target trajectory to generate a travel trajectory pair for the same user. Specifically: The contribution factors of individual access popularity, group access popularity, and path connectivity to user identification are divided into two categories: point level and line level. The contribution factors of the two categories are combined using an addition operation, and the weights of the two categories are controlled by the parameter θ2. The final formula for calculating the total matching degree is as follows:

8. A user identification device based on multi-source public transport data, characterized in that, include: Data acquisition module: used to acquire multi-source trajectory datasets of at least two traffic subsystems, and generate trajectory pairs based on the multi-source trajectory datasets, wherein the trajectory pairs include a target trajectory and a trajectory to be matched; Semantic analysis module: used to perform semantic context analysis on the trajectory pairs based on travel time consistency and limited time and space accessibility, and extract a set of trajectory pairs to be matched with a set number of co-occurrences and reasonable semantics; Contribution factor calculation module: used to perform individual access popularity, group access popularity and path connectivity analysis on the co-occurrence points corresponding to each trajectory pair in the set of trajectory pairs to be matched, and define contribution factors for unique identification of the same user based on the analysis results of individual access popularity, group access popularity and path connectivity. Similarity calculation module: used to construct a total matching degree calculation formula based on the contribution factor, use the total matching degree calculation formula to calculate the similarity between the target trajectory and each trajectory to be matched in the set of trajectories to be matched, and take the trajectory to be matched with the highest similarity as the matching trajectory of the target trajectory to generate a travel trajectory pair of the same user.

9. A device, characterized in that, The device includes a processor and a memory coupled to the processor, wherein, The memory stores program instructions for implementing the user identification method based on multi-source public transport data as described in any one of claims 1-7; The processor is used to execute the program instructions stored in the memory to control the user identification method based on multi-source public transport data.

10. A storage medium, characterized in that, The device stores processor-executable program instructions for performing the user identification method based on multi-source public transport data as described in any one of claims 1 to 7.

Images