Information recommendation method, electronic device, readable medium, and program product

Abstract

The present disclosure provides an information recommendation method, an electronic device, a readable medium and a program product. The method of the present disclosure comprises: obtaining a home region of a target object for which a transaction user has completed a transaction; obtaining a home region of a candidate object operated by the transaction user before the transaction target object; adding a similarity value of the home region of the target object and the home region of the candidate object to a similarity evaluation matrix; and in response to a recommendation request of a transaction user carrying a home region identifier, determining a recommended region based on the similarity evaluation matrix to recommend the recommended region to the transaction user.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and more particularly to information recommendation methods, electronic devices, readable media, and program products. Background Technology

[0002] With the development of computer technology, it has become increasingly convenient for users to obtain information. For example, users can search for desired products using computers or mobile phones.

[0003] In existing technologies, taking residential community search as an example, recommendation methods mainly rely on calculating the similarity of item attributes. These attributes are often static attributes selected by staff based on experience. However, these static attributes cannot accurately reflect the perceived similarity among users. In other words, the accuracy of product recommendations using static attributes varies. Although some technical solutions utilize user behavior data for product recommendations, the complexity and noise of this data lead to inaccurate recommendations. Therefore, a solution is needed to improve the accuracy of recommendation information. Summary of the Invention

[0004] This disclosure provides information recommendation methods, electronic devices, readable media, and program products.

[0005] According to a first aspect of this disclosure, an information recommendation method is provided. The method specifically includes: obtaining the region of the target object whose transaction has been completed by a user; obtaining the region of candidate objects that the user has previously processed before the target object; adding the similarity scores between the target object's region and the candidate objects' regions to a similarity evaluation matrix; and, in response to a recommendation request from a user carrying a region identifier, determining a recommendation region based on the similarity evaluation matrix to recommend the region to the user.

[0006] According to at least one embodiment of this disclosure, adding the similarity values ​​of the home regions of the target object and the candidate objects to a similarity evaluation matrix includes: obtaining the home region identifiers of the target object and the candidate objects; establishing a primary-secondary correspondence between the home region identifiers of the target object and the candidate objects; adding the home region identifiers of the target object and the candidate objects to different rows and columns in the similarity evaluation matrix; and increasing the similarity value by 1 at the intersection of the home region identifiers of the target object and the candidate objects in the similarity evaluation matrix.

[0007] According to at least one embodiment of this disclosure, a primary-secondary correspondence is established between the home region identifier of the target object and the home region identifier of the candidate object; the home region identifier of the target object and the home region identifier of the candidate object are respectively added to different rows and columns of the similarity evaluation matrix, including: adding the home region identifier of the target object to the column or row representing the primary identifier in the similarity evaluation matrix, and adding the home region identifier of the candidate object to the row or column representing the secondary identifier in the similarity evaluation matrix.

[0008] According to at least one embodiment of this disclosure, in response to a recommendation request from a user seeking a transaction carrying a home region identifier, determining a recommended region based on a similarity evaluation matrix includes: in response to a recommendation request from a user seeking a transaction carrying a home region identifier, searching for the existence of a home region identifier in the row or column of the secondary identifier in the similarity evaluation matrix; if it exists, searching for a recommended home region identifier in the column or row of the primary identifier that has a primary-secondary correspondence with the home region identifier in the similarity evaluation matrix; and recommending the recommended region corresponding to the recommended home region identifier to the user seeking a transaction.

[0009] According to at least one embodiment of this disclosure, the method further includes: filtering the target object's region identifiers corresponding to each intersection position based on the similarity value within each intersection position of the candidate object's region identifier, to obtain multiple target object region identifiers with similarity values ​​greater than a similarity value threshold; using the region identifiers of multiple target objects as a recommended region set similar to the candidate object's region; or, sorting the target object's region identifiers corresponding to each intersection position based on the similarity value within each intersection position of the candidate object's region identifier; selecting the region identifiers of the top N target objects from the sorting results as a recommended region set similar to the candidate object's region; where N is a positive integer.

[0010] According to at least one embodiment of this disclosure, in response to a recommendation request from a user seeking a transaction carrying a region identifier, determining a recommended region based on a similarity evaluation matrix includes: in response to a recommendation request from a user seeking a transaction carrying a region identifier, searching for the existence of a region identifier in the row or column of the sub-identifier in the similarity evaluation matrix; if it exists, finding at least one recommended region identifier from the set of recommended regions corresponding to the region identifier; and recommending the recommended region corresponding to the recommended region identifier to the user seeking a transaction.

[0011] According to at least one embodiment of this disclosure, the method further includes: obtaining a first and second home region identifier with a primary-secondary correspondence contained in the historical data of newly transacted users; determining whether the similarity evaluation matrix contains a primary-secondary correspondence between the first and second home region identifiers; if it does, increasing the existing similarity value at the intersection of the first and second home region identifiers in the similarity evaluation matrix by 1.

[0012] According to at least one embodiment of this disclosure, the method further includes: if the first attribution region identifier and the second attribution region identifier are not included, then the first attribution region identifier and the second attribution region identifier are added to the newly added rows and columns of the similarity evaluation matrix, respectively; if the first attribution region identifier is included but the second attribution region identifier is not included, then the second attribution region identifier is added to the newly added rows or columns of the similarity evaluation matrix; if the first attribution region identifier is not included but the second attribution region identifier is included, then the first attribution region identifier is added to the newly added columns or rows of the similarity evaluation matrix; and the similarity value at the intersection of the first attribution region identifier and the second attribution region identifier in the similarity evaluation matrix is ​​increased by 1.

[0013] According to at least one embodiment of this disclosure, obtaining the affiliation region of candidate objects that a transaction user has previously performed pending transactions on before the target object includes: obtaining transaction data and attention data from the transaction user's historical data; wherein, attention data includes: pending transaction data and non-pending transaction data; filtering out candidate objects that have previously performed pending transactions based on the attention data; wherein pending transactions include: showing or on-site inspection; and determining the affiliation region of the candidate objects to which the candidate objects belong.

[0014] According to a second aspect of this disclosure, an electronic device is provided, comprising: a memory storing execution instructions; and a processor executing the execution instructions stored in the memory, such that the processor performs the method described in the first aspect of any embodiment of this disclosure.

[0015] According to a third aspect of this disclosure, a readable storage medium is provided, wherein executable instructions are stored therein, which, when executed by a processor, are used to implement the method described in the first aspect of any embodiment of this disclosure.

[0016] According to a fourth aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the method described in the first aspect of any embodiment of this disclosure. Attached Figure Description

[0017] The accompanying drawings illustrate exemplary embodiments of the present disclosure and, together with the description thereof, serve to explain the principles of the present disclosure. These drawings are included to provide a further understanding of the present disclosure and are incorporated in and constitute a part of this specification.

[0018] Figure 1 This is a flowchart illustrating the information recommendation method provided in the embodiments of this disclosure.

[0019] Figure 2 This is a flowchart illustrating the similarity evaluation matrix construction method provided in this embodiment of the disclosure.

[0020] Figure 3 This is a schematic diagram illustrating a similarity evaluation matrix as an example of an embodiment of this disclosure.

[0021] Figure 4 This is a schematic flowchart illustrating a method for adding identifiers to a matrix according to an embodiment of this disclosure.

[0022] Figure 5 The diagram illustrates two matrix construction methods exemplified by embodiments of this disclosure.

[0023] Figure 6 This is a flowchart illustrating a method for finding recommendations, as provided in an embodiment of this disclosure.

[0024] Figure 7 This is a schematic flowchart of the matrix processing method provided in an embodiment of the present disclosure.

[0025] Figure 8 This is a flowchart illustrating another method for finding recommendations provided in an embodiment of this disclosure.

[0026] Figure 9 This is a flowchart illustrating a matrix-based method for adding similarity values, as provided in an embodiment of this disclosure.

[0027] Figure 10 A flowchart illustrating the method for determining the home region of candidate objects provided in this embodiment of the disclosure.

[0028] Figure 11 This is a schematic block diagram of an information recommendation device according to one embodiment of the present disclosure.

[0029] Figure 12 This is a schematic block diagram of an electronic device according to one embodiment of the present disclosure. Detailed Implementation

[0030] The present disclosure will now be described in further detail with reference to the accompanying drawings and examples. It should be understood that the specific examples described herein are for illustrative purposes only and are not intended to limit the scope of the disclosure. Furthermore, it should be noted that, for ease of description, only the parts relevant to the present disclosure are shown in the accompanying drawings.

[0031] It should be noted that, where there is no conflict, the embodiments and features described in this disclosure can be combined with each other. The technical solutions of this disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0032] Figure 1 This is a flowchart illustrating the information recommendation method provided in an embodiment of this disclosure. Figure 1 The method shown includes steps S101 to S104. This method can be executed by the server.

[0033] Specifically, Figure 1 The method shown includes step S101: obtaining the region to which the target object of the completed transaction by the transacting user belongs.

[0034] It's important to clarify that "transactional user" refers to a user who has completed at least one transaction, such as a user who has completed a home purchase transaction in the real estate sector. "Target object" refers to the specific product or service the user has transacted with, such as a completed property in the real estate sector. The "target object's area" refers to the region or area where the target object is located, such as the neighborhood where the completed property (i.e., the target object) is situated. Of course, the target object can also refer to other products besides real estate. For example, when a user is dining out, restaurant recommendations can be made, and the user might choose a restaurant. Historical data could include restaurants the user has added to their favorites and the commercial area where those restaurants are located. This can also be applied to recommending parking spaces in a parking lot.

[0035] Generally, each successful user corresponds to one target object and one region belonging to that target object. Of course, one successful user can also correspond to multiple target objects and multiple regions belonging to those target objects. The similarity evaluation matrix in subsequent steps is constructed jointly by the regions belonging to the target objects of multiple successful users. Here, we can use one successful user as an example to illustrate the process of matrix construction, adding similarity values, and evaluating similarity; multiple successful users are implemented in the same way.

[0036] In practical applications, transaction records of multiple users are extracted from historical data in user behavior log databases. For example, in a real estate platform, the transaction information of a user is queried, and then mapped to the corresponding target object's region, such as the neighborhood to which the transaction property belongs, through a basic property information database. By collecting regional data representing the final decision results of multiple users, the final region chosen by the user group after comprehensive evaluation is revealed, providing real user behavior data for uncovering implicit similarities between regions.

[0037] Step S102: Obtain the region to which the candidate objects that the transaction user has previously operated on before the transaction target object belong.

[0038] It's important to clarify that "alternative targets" refer to objects that the user intends to purchase before the target (e.g., various goods), such as properties for viewing in the real estate sector. The "regional affiliation" of an alternative target refers to the area where it belongs, such as the neighborhood where the property for viewing (i.e., the alternative target) is located. "Intended purchase action" refers to the specific actions the user takes to indicate their intention to buy. For example, actions taken by the user after initial browsing can represent specific behaviors indicating their intention to purchase, such as property viewings or in-store inquiries in the real estate sector. These actions reflect a deeper interest in the item and potential purchase intent more effectively than simple clicks. Of course, alternative targets can also be goods other than real estate, such as restaurant recommendations or parking space recommendations when a user goes out to eat. The target and alternative targets are the same type of item or commodity.

[0039] In practical applications, based on the complete behavioral sequence of the same user who made a transaction, we trace the pre-transaction actions they performed (only pre-transaction actions have similarity calculation value, indicating that the user's willingness to transact is relatively high. Other non-pre-transaction actions can be ignored, such as clicking or adding to favorites. Therefore, the pre-transaction action can be understood as the last step before the transaction to learn about the product or item. The pre-transaction action varies in different scenarios. In the real estate scenario, it can be the "showing" action, which is to inspect the property in person).

[0040] For example, in the real estate sector, a user might initially view multiple properties as potential buyers. The regions (e.g., neighborhoods) of these potential properties are recorded in the user's historical data. This historical data includes the user's behavioral changes from the initial viewing to the final transaction. By utilizing the user's behavioral sequence from the initial viewing (the region of the potential properties) to the final transaction (the region of the target property), we can implicitly reflect the user group's subjective perception that these regions have a high degree of regional similarity. This is not limited to quantifiable similarities between two regions based on static attributes (such as location, average price, and building age), but can also include much unquantifiable information. By leveraging the underlying logic of a user's journey from initial viewing to final transaction, such as complex personal factors (e.g., commuting convenience, school district matching, family needs, appreciation trends, subway planning, etc.), we can objectively reflect the user group's viewpoints by statistically analyzing the number of times a large number of users view properties before making a purchase. This method more accurately captures deep-seated connections than traditional static attribute methods and avoids ignoring the user's subjective judgment.

[0041] Step S103: Add the similarity values ​​between the target object's region and the candidate object's region to the similarity evaluation matrix.

[0042] It should be noted that the similarity evaluation matrix is ​​an asymmetric matrix constructed by statistically analyzing the home regions of candidate objects with primary and secondary correspondences and the home regions of the target object. It is used to quantify the implicit similarity association between regions.

[0043] In practical applications, an asymmetric matrix is ​​initialized by statistically analyzing the regions of candidate objects and the target object that have a primary-secondary correspondence in the historical data of multiple transacting users. For example, in this similarity evaluation matrix, the rows represent the region identifiers (secondary identifiers) of candidate objects, and the columns represent the region identifiers (primary identifiers) of the target object. For each pair (region of candidate object, region of target object), the frequency of occurrence of this primary-secondary correspondence is accumulated at the intersection position to form the similarity evaluation matrix.

[0044] It should be noted that the similarity score mentioned above is determined based on the number of users with the same primary and secondary correspondence who completed transactions. Specifically, for any pair of target object's home region and candidate object's home region, the total number of users who executed pending transactions in the candidate object's home region and ultimately traded the target object's home region, based on all historical data of all users who completed transactions, is the similarity score between the target object's home region and the candidate object's home region.

[0045] It should be noted that the row and column relationships in the similarity evaluation matrix can be interchanged. For example, if C users complete transactions in area B after completing transactions in area A, the similarity values ​​at the corresponding intersections in the matrix increase by C times, resulting in the form M[Alternative][Target] = C times. Of course, the matrix can also be designed in the form M[Target][Alternative] = C times.

[0046] Furthermore, this matrix is ​​not reconstructed for each recommendation; instead, it can be pre-built based on historical data (such as behavioral logs from the past 180 days) and incrementally updated as new behavioral data becomes available. For example, it checks whether the corresponding row and column already exist in the matrix. If they do, similarity values ​​are added directly at the intersection; otherwise, the row and column are added and similarity values ​​are added, thus ensuring the real-time performance and computational efficiency of the matrix.

[0047] It's important to note that when constructing the similarity assessment matrix, a region-level similarity assessment matrix is ​​used instead of an object-level matrix (such as specific properties). This is because regions (such as neighborhoods) are the core aggregation units for user decision-making, encompassing quantifiable and non-quantifiable similar characteristics of multiple objects (such as surrounding amenities, commuting routes, career advancement trends, and the quality of the resident population). A neighborhood may contain many properties; by aggregating at a finer granularity based on objects, a matrix that more accurately represents the similarity of neighborhoods can be obtained.

[0048] Step S104: In response to a recommendation request from a user with a home region identifier, determine a recommendation region based on a similarity evaluation matrix to recommend the recommendation region to the user.

[0049] It's important to note that a recommendation request is a user-triggered request carrying a regional identifier, used to recommend similar areas. For example, a request triggered by a user's clicks, views, or online viewing of a specific area; these requests might originate automatically when the user browses a region's details page on the platform. A regional identifier is an identifier used to uniquely identify a region, such as a neighborhood ID.

[0050] When a user triggers a recommendation request, such as clicking, viewing, or watching a details page for a specific region on the platform, the request carries a region identifier (such as the currently viewed community ID). The system uses this identifier as the region identifier for candidate objects, searches the corresponding row of the matrix, extracts the region identifier of the target object with the highest frequency of use, and returns it to the user. This step enables real-time recommendations, ensuring a fast response, while inheriting the aforementioned creative logic: the recommendation results are based on the collective behavioral patterns of users, prioritizing regions that, although their attributes differ greatly, are considered similar in behavior.

[0051] Based on the publicly available solutions described above, this approach directly analyzes user behavior from the perspective of pre-transaction actions to final transactions, focusing on fine-grained behavioral paths. It aggregates a large amount of object-level behavioral data (e.g., property listings) to a regional (e.g., neighborhood) dimension, constructing a similarity matrix between the regions of candidate properties and the target property (this value represents the frequency of primary and secondary correspondences). This bottom-up approach doesn't require delving into the complex reasons behind user behavior; it naturally reflects the genuine similarities between different regions perceived by the user group. Specifically, by first collecting behavioral data at the object level and then mapping it to the regional level for statistical analysis, it combines fine-grained behavioral data collection with coarse-grained recommendation, ensuring both the richness and accuracy of the behavioral data while meeting users' needs for precise regional recommendations. This effectively solves the problem of low recommendation accuracy caused by relying solely on quantifiable static attributes such as geographical location, average price, and building age to calculate neighborhood similarity in existing technologies. It also effectively addresses the inability to capture the deep-seated connections arising from unquantifiable factors such as appreciation trends, investment value, and resident quality during actual transactions, leading to inaccurate recommendation results.

[0052] In one or more embodiments of this disclosure, such as Figure 2 This is a flowchart illustrating the similarity evaluation matrix construction method provided in this embodiment of the disclosure. Figure 2 As shown, step S103, which involves adding the similarity values ​​of the target object's region and the candidate object's region to the similarity evaluation matrix, includes: Step S1031: Obtaining the region identifiers of the target object and the candidate objects. Step S1032: Establishing a primary-secondary correspondence between the region identifiers of the target object and the candidate objects. Step S1033: Adding the region identifiers of the target object and the candidate objects to different rows and columns of the similarity evaluation matrix. And, Step S1034: Increasing the similarity value by 1 at the intersection of the region identifiers of the target object and the candidate objects in the similarity evaluation matrix.

[0053] The primary-secondary correspondence mentioned here refers to a directional relationship established based on user behavior sequences. The region identifier of the candidate object (corresponding to the pending transaction behavior) serves as the secondary identifier, while the region identifier of the target object (corresponding to the successful transaction behavior) serves as the primary identifier. This reflects a unidirectional behavioral logic from the region of the candidate object to the region of the target object. The intersection position refers to the cell coordinates where a specific row and column in the similarity evaluation matrix intersect, used to store the similarity value indicating the frequency of this primary-secondary correspondence.

[0054] In practical applications, based on the obtained home regions of the target and candidate properties, their home region identifiers can be extracted by querying the database. For example, in a real estate platform, the ID of the home region of the target property to which the sold property belongs can be extracted from the database storing relevant property information as the home region identifier of the target property, and the ID of the home region of the candidate property to which the property being viewed belongs can be extracted as the home region identifier of the candidate property. This provides an efficient indexing foundation for matrix-based data processing.

[0055] Based on the temporal sequence of user behavior, the regional identifier of the candidate object is defined as the secondary identifier (representing the region where the user had pending transactions in the past), and the regional identifier of the target object is defined as the primary identifier (representing the region where the user ultimately completed the transaction), establishing a directional correspondence from secondary to primary. This primary-secondary correspondence accurately represents the user's subjective similarity judgment: the user's behavior from pending transactions in one region to final transactions in another region implicitly indicates that the user considers the two similar or substitutable. Even if the static attributes differ greatly, this relationship reflects the collective user perspective through user behavior patterns. The logical relationship of unquantifiable neighborhood similarity is quantified through user behavior.

[0056] In one alternative approach, an empty matrix is ​​initialized, and the region identifiers of candidate objects are added to the rows (as secondary identifiers), while the region identifier of the target object is added to the columns (as primary identifiers). Alternatively, the region identifiers of candidate objects are added to the columns (as secondary identifiers), and the region identifier of the target object is added to the rows (as primary identifiers). It is necessary to ensure that the rows and columns represent different roles. This matrix is ​​asymmetric because user behavior is directional, such as the direction of behavior from candidate A to target B, thus reflecting the user's actual behavior path. It cannot be recorded in the same matrix in reverse order, because subsequent matrix-based queries and recommendations are also strictly directional; see subsequent embodiments for details.

[0057] For each primary-secondary correspondence, increment the corresponding intersection position in the matrix by 1 (representing the occurrence of a completely identical primary-secondary correspondence). If a similarity value already exists, increment it (it can be a uniform increment or a differentiated non-uniform increment as needed, such as incrementing by 0.5 each time after exceeding a certain similarity value). This quantifies the underlying logic of unquantifiable user behavior; a higher number of increments indicates a higher similarity.

[0058] To better illustrate the matrix construction process, an example will be provided: Figure 3 This is a schematic diagram illustrating a similarity evaluation matrix as exemplified by an embodiment of this disclosure. For example... Figure 3As shown, assume there is data from three users who have completed transactions. User a1's candidate objects for the transaction are located in "Community X" (identifier X) and "Community Y" (identifier Y), while the target object for the final transaction is located in "Community Z" (identifier Z). User a2's transaction is in "Community X", and the transaction is in "Community W" (identifier W). User a3's transaction is in "Community Y", and the transaction is in "Community Z". First, obtain the identifiers of the target object's community as Z and W; and the identifiers of the candidate objects' community as X and Y. Establish primary and secondary correspondences: X with Z, Y with Z, and X with W. Add the identifiers to the matrix: add X and Y (secondary identifiers) to the rows, and add Z and W (primary identifiers) to the columns. Then, increment at the intersection positions: +1 at the intersection position of M[X][Z], +1 at the intersection position of M[Y][Z], +1 at the intersection position of M[X][W], and, assuming there was already 1 at the intersection position of M[Y][Z], the final M[Y][Z] = 2. The resulting matrix is ​​asymmetric; for example, M[X][Z] = 1, M[Z][X] = 0.

[0059] Based on the publicly available solutions mentioned above, by refining the matrix construction process and establishing an asymmetric primary-secondary correspondence, user behavior can be accurately quantified and used to assess the similarity of different regions.

[0060] In one or more embodiments of this disclosure, such as Figure 4 This is a schematic flowchart illustrating a method for adding identifiers to a matrix according to an embodiment of this disclosure. Figure 4 As shown, step S1302: Establish a primary-secondary correspondence between the home region identifier of the target object and the home region identifiers of the candidate objects. Step S1303: Add the home region identifiers of the target object and the candidate objects to different rows and columns of the similarity evaluation matrix, including: step S401: Add the home region identifier of the target object to the column or row representing the primary identifier in the similarity evaluation matrix. And, step S402: Add the home region identifiers of the candidate objects to the row or column representing the secondary identifiers in the similarity evaluation matrix.

[0061] In practical applications, there are two ways to add the region identifier when constructing a matrix, such as... Figure 5 Schematic diagrams illustrating two matrix construction methods exemplified in embodiments of this disclosure, such as... Figure 5 As shown: One approach is to establish a primary-secondary correspondence after obtaining the region identifiers. In this approach, the region identifier of the target object is assigned as the primary identifier to a column (horizontal axis) of the matrix, and the region identifiers of the candidate objects are assigned as secondary identifiers to a row (vertical axis). The intersection M[row: region identifier of candidate object][column: region identifier of target object] stores the number of successful primary-secondary correspondence matches.

[0062] Another approach is to assign the target object's region identifier as the primary identifier to the row, and the candidate object's region identifier as the secondary identifier to the column. Thus, the intersection position M[row: target object's region identifier][column: candidate object's region identifier] stores the number of successful matches of the secondary correspondence.

[0063] This ensures the matrix reflects the unidirectional behavior from pending execution to successful transaction, facilitating the rapid extraction of the region to which similar target objects belong from the corresponding columns. However, in practical applications, the query logic needs to be adjusted to match different matrices.

[0064] Based on the publicly available schemes described above, by refining the row and column allocation of primary and secondary correspondences, two flexible methods for constructing asymmetric similarity evaluation matrices are provided. This ensures that all primary and secondary correspondences in a matrix are consistent with the row and column correspondences. That is, if all rows are primary identifiers, then all columns are secondary identifiers; conversely, if all rows are secondary identifiers, then all columns are primary identifiers.

[0065] In one or more embodiments of this disclosure, such as Figure 6 This is a flowchart illustrating a method for finding recommendations provided in an embodiment of this disclosure. Figure 6 As shown, step S104, responding to a recommendation request from a user seeking a transaction carrying a region identifier, and determining the recommended region based on a similarity evaluation matrix, includes: Step S601: In response to a recommendation request from a user seeking a transaction carrying a region identifier, searching for the existence of a region identifier in the row or column of the secondary identifier in the similarity evaluation matrix. Step S602: If it exists, searching for the recommended region identifier in the column or row of the primary identifier that has a primary-secondary correspondence with the region identifier in the similarity evaluation matrix. Step S603: Recommending the recommended region corresponding to the recommended region identifier to the user seeking a transaction.

[0066] After a user triggers a recommendation request, the user's currently followed region identifier is used as the query key. A search is then performed in the row or column of the similarity evaluation matrix containing the secondary identifier (corresponding to the region identifier of the candidate object, i.e., the stage before the transaction). For example, if a row in the matrix represents a secondary identifier, the matrix is ​​traversed to check if it contains that region identifier. It's important to note that the search must strictly follow the region identifier of the candidate object to find the target object's region identifier (primary identifier). That is, the search must start from the secondary identifier and find the primary identifier, not vice versa. This search method aligns with the actual user behavior from following to completing a transaction, exhibiting strict unidirectionality and directionality.

[0067] Because the actual user behavior path always starts from the pending transaction (the region to which the candidate object belongs) and ends with the transaction being completed (the region to which the target object belongs), reverse lookup (such as looking at the transaction backtracking) is the opposite of the actual behavior of the user group, and cannot accurately reproduce or express the user's unquantifiable behavioral logic.

[0068] If the attribution region identifier was successfully found in the previous step, the intersection of the rows or columns of that identifier is traversed. The attribution region identifiers of the target objects in the columns or rows where the main identifier has a non-zero or high similarity value are extracted as recommended attribution region identifiers. For example, in a real estate platform, if matrix M[Alternate A][Target B] = 5 and has the highest similarity value, then B is selected as the recommended attribution region identifier. The extracted recommended attribution region identifiers are mapped back to the actual region information and returned to the front-end application for display through the recommendation service interface. For example, in the real estate field, a list of Top N similar communities is returned for users to view.

[0069] If the region identifier is not found in the matrix (or the region identifier only appears in the primary identifier row or column, but not in the secondary identifier column or row), an empty string can be returned, and the user will be notified that the search failed. Alternatively, a different recommendation algorithm can be used to recommend regions to the user.

[0070] Based on the publicly available solutions mentioned above, it is clear that by refining the recommendation process and strictly searching along the path from secondary to primary recommendations, the user behavior path can be reflected more accurately and realistically, thereby improving the accuracy of recommendations.

[0071] In one or more embodiments of this disclosure, such as Figure 7 This is a flowchart illustrating the matrix processing method provided in an embodiment of this disclosure. Figure 7 As shown, the method further includes: Step S701: Based on the similarity values ​​within each intersection position of the candidate object's region identifier, filter the target object's region identifiers corresponding to each intersection position to obtain multiple target object region identifiers with similarity values ​​greater than a similarity value threshold. Step S702: Use the region identifiers of multiple target objects to form a recommended region set similar to the candidate object's region. Alternatively, Step S703: Based on the similarity values ​​within each intersection position of the candidate object's region identifier, sort the target object's region identifiers corresponding to each intersection position. Step S704: Select the top N target object region identifiers from the sorted results to form a recommended region set similar to the candidate object's region; where N is a positive integer.

[0072] In practical applications, after constructing the similarity evaluation matrix, for each candidate object's region identifier row or column (depending on the row and column allocation), iterate through all intersection positions and count the similarity values ​​(the number of times the primary and secondary correspondences occur). Retain the region identifiers of the corresponding target objects whose similarity values ​​exceed a similarity threshold. For example, if the threshold is 1, invalid associations with zero occurrences are excluded. This similarity threshold filters noise, ensuring the recommended region set contains only highly correlated region identifiers for more accurate recommendations. The filtered target object region identifiers are then mapped back to the actual target object's region, forming a set. Subsequent recommendations can be directly retrieved from this set without needing to consult the matrix, improving recommendation efficiency.

[0073] In addition to filtering noise by setting a similarity threshold, the similarity values ​​of the candidate object's region identifiers can be sorted in descending order based on the intersection similarity values, for example, those with higher similarity values ​​are ranked first. The region identifiers of the top N target objects in the sorting results are extracted and mapped to the target object's region, forming a Top N set. For example, when N=5, the regions of the top 5 high-frequency target objects are selected.

[0074] For example, suppose in the similarity evaluation matrix, the rows of candidate objects whose region identifier "A" have overlapping positions: M[A][B]=3, M[A][C]=1, M[A][D]=5, M[A][E]=0. Threshold filtering (threshold=2): obtain B(3) and D(5) which are greater than 2, forming the set {B,D}. Or sort: D(5)>B(3)>C(1)>E(0), select the first two items D and B of N=2, forming the Top 2 set {D,B}.

[0075] Based on the above-mentioned publicly available solutions, a set of recommended regions can be generated by filtering or sorting. This eliminates the need to traverse the matrix every time and allows for real-time recommendations using the set, effectively improving recommendation efficiency.

[0076] In one or more embodiments of this disclosure, such as Figure 8 This is a flowchart illustrating another method for finding recommendations provided in an embodiment of this disclosure. Figure 8 As shown, step S104, responding to a recommendation request from a user with a home region identifier, determines the recommended region based on a similarity evaluation matrix, including: Step S801: In response to a recommendation request from a user with a home region identifier, searching for the existence of a home region identifier in the row or column of the similarity evaluation matrix where the identifier is located. Step S802: If it exists, finding at least one recommended home region identifier from the set of recommended regions corresponding to the home region identifier. Step S803: Recommending the recommended region corresponding to the recommended home region identifier to the user.

[0077] In practical applications, after a user triggers a recommendation request, the user's currently followed region identifier is used as the query key. A search is then performed in the row or column of the similarity evaluation matrix containing the secondary identifier (corresponding to the region identifier of the candidate object, i.e., the stage before the transaction). For example, if a matrix row represents a secondary identifier, the matrix rows are traversed to check if the target object's region identifier is included. It's important to note that the search must strictly follow the candidate object's region (secondary identifier) ​​to find the target object's region (primary identifier). That is, the search must start from the secondary identifier and find the primary identifier, not vice versa. This search method aligns with the user's actual behavior from following to completing a transaction, exhibiting strict unidirectionality and directionality.

[0078] Secondly, if it exists, at least one recommended home region identifier is found from the recommended region set corresponding to the home region identifier. Specifically, if the home region identifier was successfully found in the previous step, one or more recommended home region identifiers are directly extracted from the recommended region set.

[0079] If the region identifier is not found in the matrix (or the region identifier appears only in the primary identifier row or column but not in the secondary identifier column or row), an empty value can be returned, and the user should be notified that the search failed. Alternatively, a different recommendation algorithm can be used to make recommendations to the user.

[0080] Based on the aforementioned publicly available solutions, refining the recommendation process and strictly following the search from secondary to primary categories can more accurately and realistically reflect user behavior paths, thereby improving recommendation accuracy. Furthermore, grouping the regions to which recommendable target objects belong into a single set can effectively improve recommendation efficiency.

[0081] In one or more embodiments of this disclosure, such as Figure 9 This is a flowchart illustrating a matrix-based method for adding similarity values, as provided in an embodiment of this disclosure. Figure 9 As shown, the specific steps include: Step S901: Obtain the first and second home region identifiers with a primary-secondary correspondence contained in the historical data of newly transacted users. Step S902: Determine whether the similarity evaluation matrix contains a primary-secondary correspondence between the first and second home region identifiers. Step S903: If it does, increase the existing similarity value by 1 at the intersection of the first and second home region identifiers in the similarity evaluation matrix.

[0082] In practical applications, if a new user completes a transaction, their latest behavioral data can be extracted from the user behavior log database. For example, in a real estate platform, the relationship between their pending transactions (the region of the candidate object) and subsequent transactions (the region of the target object) can be filtered and mapped to the corresponding second-home region identifier (secondary identifier) ​​and first-home region identifier (primary identifier). It is important to note that since the matrix already contains some data, the newly added data must maintain consistency with the data types of the existing matrix rows and columns. In other words, the new primary and secondary identifiers must match the primary identifier row or column and the secondary identifier column or row in the original matrix.

[0083] Check whether the second attribution region identifier already exists in the row or column where the secondary identifier is located in the existing similarity evaluation matrix, and whether the first attribution region identifier also exists in the column or row where the primary identifier is located.

[0084] If the determination result indicates that the matrix already contains the first and second belonging region identifiers, then the intersection position (i.e., the cell in the matrix) corresponding to the primary and secondary identifiers is located, and the existing similarity value is incremented by 1. For example, the existing similarity value of the intersection position M[P][Q] (originally 2) is increased by 1 to become 3.

[0085] Based on the publicly available solutions described above, it is clear that by acquiring historical data of newly transacted users and updating the similarity values ​​after confirming their existence, no additional identifiers are needed. Furthermore, during the update process, it is crucial to ensure that the primary and secondary correspondences remain consistent with the original matrix.

[0086] In one or more embodiments of this disclosure, the method further includes: Step S904: If neither the first nor the second attributable region identifier is included, then the first and second attributable region identifiers are added to the newly added rows and columns of the similarity evaluation matrix, respectively. Step S905: If the first attributable region identifier is included but the second attributable region identifier is not included, then the second attributable region identifier is added to the newly added rows or columns of the similarity evaluation matrix. Step S906: If the first attributable region identifier is not included but the second attributable region identifier is included, then the first attributable region identifier is added to the newly added columns or rows of the similarity evaluation matrix. And, Step S907: The similarity value at the intersection of the first and second attributable region identifiers in the similarity evaluation matrix is ​​increased by 1.

[0087] If the primary and secondary correspondence does not exist, and the matrix contains neither a primary identifier nor a secondary identifier, then a new row (or column) is added to add the secondary identifier, and a new column (or row) is added to add the primary identifier.

[0088] If the result of the judgment is that the matrix already contains the first home region identifier (primary identifier) ​​but lacks the second home region identifier (secondary identifier), then only the row or column corresponding to the secondary identifier is added, and the second home region identifier is inserted.

[0089] If the result of the judgment is that the matrix already contains the second home region identifier (secondary identifier) ​​but lacks the first home region identifier (primary identifier), then only the column or row corresponding to the primary identifier is added, and the first home region identifier is inserted.

[0090] After completing any of the above expansion scenarios, locate the intersection of the newly added or existing primary and secondary correspondences, and increment the similarity value by 1 (from 0 to 1 if it is a new position), which represents adding a primary and secondary correspondence.

[0091] Based on the above-mentioned publicly available solutions, it can be seen that by establishing a primary-secondary correspondence in the matrix and maintaining the addition of subsequent correspondences, the efficiency and accuracy of maintaining the primary-secondary correspondence can be improved.

[0092] In one or more embodiments of this disclosure, such as Figure 10 This is a flowchart illustrating the method for determining the home region of candidate objects provided in embodiments of this disclosure. Figure 10 As shown, step S102, obtaining the region of the candidate object that the user has previously operated on before the target object, includes: Step S1021: Obtaining transaction data and attention data from the user's historical data; wherein, attention data includes: data of pending transactions and data of non-pending transactions. Step S1022: Filtering out candidate objects that have previously operated on based on the attention data; wherein pending transactions include: showing or on-site inspection. Step S1023: Determining the region of the candidate object to which it belongs.

[0093] It should be noted that non-transactional operation data refers to shallow records in the data of interest that are coarse-grained and do not necessarily represent the intention to make a transaction, such as simple clicks or browsing data; showing refers to the on-site viewing behavior that users actually arrange or participate in, such as scheduling a viewing in the real estate field; on-site inspection refers to the investigation behavior that users conduct in person at the site, such as visiting the property in the real estate field. Similar to showing, both indicate deep interest and show that users have a high willingness to make a transaction about the property.

[0094] Extract complete historical data of users who have completed transactions from user behavior log databases. For example, in real estate platforms, collect transaction data (such as transaction property IDs) and attention data (such as click records and viewing records) from the past 180 days. It's important to note that user attention data is divided into pending transaction data (viewings, etc.) and non-pending transaction data (clicks, etc.). Non-pending transaction data (such as clicks) is often considered noise, representing more random user behavior and failing to reflect the user's true intention to complete the transaction. Pending transaction data (such as viewings) represents the stage where users are actively engaged, indicating a higher level of intent to complete the transaction. Therefore, pending transaction data can be used as a basis for calculating similarity.

[0095] Select pending transaction data from the data of interest, such as filtering viewing or on-site inspection events by behavior type tags, and excluding non-pending transaction data, so as to obtain the candidate objects that the user has performed pending transaction before the transaction (such as viewing properties).

[0096] Further, by querying the database, the region to which the selected candidates belong can be found. For example, in the real estate field, the ID of the property being viewed can be mapped to the ID of the candidate community to form the region to which the candidate belongs.

[0097] Based on the publicly available solutions described above, by extracting the user's previous pending transaction data from a large amount of complex attention data, and using this as the basis for subsequent calculations of the similarity between the user and the target object's region, the similarity between different regions can be represented more accurately from the user's perspective. The similarity assessment results obtained in this way are also more accurate.

[0098] Based on any of the above embodiments, this disclosure also provides an information recommendation device. Figure 11 This is a schematic block diagram of an information recommendation device according to one embodiment of the present disclosure. Figure 11 As shown, the information recommendation device includes: an acquisition module 1101, used to acquire the home region of the target object to which at least one completed transaction by a user belongs. The acquisition module 1101 is also used to acquire the home regions of candidate objects that the user previously operated on before the target object. A matrix construction module 1102 is used to add the similarity values ​​between the home regions of the target object and the home regions of the candidate objects to a similarity evaluation matrix. A recommendation module 1103 is used to determine the recommendation region based on the similarity evaluation matrix in response to a recommendation request from a user carrying a home region identifier.

[0099] The matrix construction module 1102 is used to obtain the home region identifier of the target object and the home region identifier of the candidate object; establish the primary and secondary correspondence between the home region identifier of the target object and the home region identifier of the candidate object; add the home region identifier of the target object and the home region identifier of the candidate object to different rows and columns in the similarity evaluation matrix; and increase the similarity value by 1 at the intersection position of the home region identifier of the target object and the home region identifier of the candidate object in the similarity evaluation matrix.

[0100] The matrix construction module 1102 is used to add the home region identifier of the target object to the column or row representing the primary identifier in the similarity evaluation matrix, and to add the home region identifier of the candidate object to the row or column representing the secondary identifier in the similarity evaluation matrix.

[0101] The recommendation module 1103 is used to respond to a recommendation request from a user who is about to complete a transaction and carries a home region identifier. It searches for the existence of a home region identifier in the row or column of the secondary identifier in the similarity evaluation matrix. If it exists, it searches for the recommended home region identifier in the column or row of the primary identifier that has a primary-secondary correspondence with the home region identifier in the similarity evaluation matrix. It then recommends the recommended region corresponding to the recommended home region identifier to the user who is about to complete a transaction.

[0102] The matrix construction module 1102 is used to filter the target object's region identifiers corresponding to each intersection position based on the similarity value within each intersection position of the candidate object's region identifier, obtaining multiple target object region identifiers with similarity values ​​greater than a similarity value threshold; the region identifiers of the target objects corresponding to the multiple target object region identifiers are used as a recommended region set similar to the candidate object's region identifier; or, based on the size of the similarity value within each intersection position of the candidate object's region identifier, the target object's region identifiers corresponding to each intersection position are sorted; the region identifiers of the N target objects with the highest sorting are selected as a recommended region set similar to the candidate object's region identifier; where N is a positive integer.

[0103] The recommendation module 1103 is used to respond to a recommendation request from a user who is about to complete a transaction and carries a home region identifier, and to search for whether a home region identifier exists in the row or column of the sub-identifier in the similarity evaluation matrix; if it exists, it searches for at least one recommended home region identifier from the set of recommended regions corresponding to the home region identifier; and recommends the recommended region corresponding to the recommended home region identifier to the user who is about to complete a transaction.

[0104] The matrix construction module 1102 is used to obtain the first and second home region identifiers with a primary-secondary correspondence contained in the historical data of new transaction users; determine whether the similarity evaluation matrix contains a primary-secondary correspondence between the first and second home region identifiers; if it does, the existing similarity value at the intersection of the first and second home region identifiers in the similarity evaluation matrix is ​​increased by 1.

[0105] The matrix construction module 1102 is used to add the first and second attribution region identifiers to the newly added rows and columns of the similarity evaluation matrix if the first attribution region identifier and the second attribution region identifier are not included; add the second attribution region identifier to the newly added rows or columns of the similarity evaluation matrix if the first attribution region identifier is included but the second attribution region identifier is not included; add the first attribution region identifier to the newly added columns or rows of the similarity evaluation matrix if the first attribution region identifier is not included but the second attribution region identifier is included; and increase the similarity value by 1 at the intersection of the first attribution region identifier and the second attribution region identifier in the similarity evaluation matrix.

[0106] The acquisition module 1101 is used to acquire transaction data and attention data from the historical data of transacting users; where attention data includes: pending transaction data and non-pending transaction data; based on the attention data, candidate objects that have undergone pending transaction operations are filtered out; where pending transaction operations include: showing or on-site inspection; and the region to which the candidate objects belong is determined.

[0107] The specific implementation process of the functions and roles of each module in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0108] The execution subject of the information recommendation method in the specific embodiments of this disclosure can be an electronic device such as a server (including a local server or a cloud computing platform).

[0109] Therefore, based on any of the above embodiments, this disclosure also provides an electronic device that can execute the information recommendation method of any of the embodiments described above.

[0110] Figure 12 This is a schematic block diagram of an electronic device according to one embodiment of the present disclosure.

[0111] The hardware architecture of the electronic device 1000 can be implemented using a bus architecture. The bus architecture can include any number of interconnect buses and bridges, depending on the specific application of the hardware and overall design constraints. Bus 1100 connects various circuits, including one or more processors 1200, memory 1300, and / or hardware modules. Bus 1100 can also connect various other circuits 1400, such as peripheral devices, voltage regulators, power management circuits, external antennas, etc.

[0112] Bus 1100 can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, this diagram uses only one connection line, but this does not imply that there is only one bus or one type of bus.

[0113] This disclosure also provides a readable storage medium storing a computer program that, when executed by a processor, is used to implement the methods described above. A "readable storage medium" can be any means capable of containing, storing, communicating, propagating, or transmitting a program for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples of a readable storage medium include: an electrical connection with one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable read-only memory (CDROM), etc.

[0114] This disclosure also provides a computer program product, the methods of which can be implemented wholly or partially through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented wholly or partially as a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instructions are loaded and executed, all or part of the processes or functions of this disclosure are performed.

[0115] Computer programs or instructions can be stored in a readable storage medium or transferred from one readable storage medium to another. For example, the computer program or instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The readable storage medium can be any available medium capable of access, or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; an optical medium, such as a digital video optical disc; or a semiconductor medium, such as a solid-state drive. The computer-readable storage medium can be a volatile or non-volatile storage medium, or it can include both volatile and non-volatile types of storage media.

[0116] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, systems, or computer program products. Therefore, this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this disclosure can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0117] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0118] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0119] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0120] It is understood that before using the technical solutions disclosed in the embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained. Simultaneously, sensitive information should be de-identified (e.g., de-identification, anonymization, etc.). For example, in response to receiving a user's active request, a prompt message can be sent to the user to clearly inform them that the requested operation will require the acquisition and use of their personal information. This allows the user to choose whether to provide personal information to the software or hardware such as electronic devices, applications, servers, or storage media that perform the operations of this disclosure's technical solutions, based on the prompt message. As an optional but non-limiting implementation, the method of sending a prompt message to the user in response to receiving a user's active request can be, for example, a pop-up window, where the prompt message can be presented in text form. Furthermore, the pop-up window can also include a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0121] In the description of this specification, the references to terms such as "one embodiment / mode," "some embodiments / modes," "example," "specific example," or "some examples," etc., refer to specific features, structures, or characteristics described in connection with that embodiment / mode or example, which are included in at least one embodiment / mode or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment / mode or example. Moreover, the specific features, structures, or characteristics described may be combined in any suitable manner in one or more embodiments / modes or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments / modes or examples described in this specification, as well as the features of different embodiments / modes or examples.

[0122] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this disclosure, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0123] Those skilled in the art should understand that the above embodiments are merely for illustrating the present disclosure and are not intended to limit the scope of the disclosure. Those skilled in the art can make other changes or modifications based on the above disclosure, and these changes or modifications still fall within the scope of the present disclosure.

Claims

1. An information recommendation method, characterized in that, The method includes: Obtain the geographic region of the target object whose transaction has been completed by the transacting user; Obtain the region to which the user who completed the transaction previously operated on the candidate objects to be traded before trading the target object; Add the similarity scores between the target object's region and the candidate object's region to the similarity evaluation matrix; In response to a recommendation request from a user with a region identifier, a recommendation region is determined based on the similarity evaluation matrix to recommend the recommendation region to the user.

2. The information recommendation method according to claim 1, characterized in that, Adding the similarity scores between the target object's region and the candidate object's region to the similarity evaluation matrix includes: Obtain the region identifier of the target object and the region identifier of the candidate objects; Establish a primary-secondary correspondence between the home region identifiers of the target object and the home region identifiers of the candidate objects; The region identifiers of the target object and the candidate objects are added to different rows and columns of the similarity evaluation matrix, respectively; and... The similarity value at the intersection of the target object's home region identifier and the candidate object's home region identifier in the similarity evaluation matrix is ​​increased by 1.

3. The information recommendation method according to claim 2, characterized in that, The step involves establishing a primary-secondary correspondence between the home region identifiers of the target object and the home region identifiers of the candidate objects. The home region identifiers of the target object and the candidate objects are added to different rows and columns of the similarity evaluation matrix, including: The region identifier of the target object is added to the column or row representing the primary identifier in the similarity evaluation matrix, and the region identifier of the candidate object is added to the row or column representing the secondary identifier in the similarity evaluation matrix.

4. The information recommendation method according to claim 3, characterized in that, The step of responding to a recommendation request from a user with a region identifier and determining a recommendation region based on the similarity evaluation matrix to recommend the recommendation region to the user includes: In response to a recommendation request from a user with a region identifier, the system searches the row or column of the similarity evaluation matrix to see if the region identifier exists. If it exists, find the recommended home region identifier in the column or row of the main identifier that has the primary-secondary correspondence with the home region identifier from the similarity evaluation matrix; The recommended region corresponding to the recommended region identifier is recommended to the users who are about to complete a transaction.

5. The information recommendation method according to claim 3, characterized in that, Also includes: Based on the similarity values ​​at each intersection position in the row or column where the candidate object's home region identifier is located, the home region identifiers of the target object corresponding to each intersection position are filtered to obtain multiple target object home region identifiers whose similarity values ​​are greater than the similarity value threshold. The home regions of the target objects corresponding to the home region identifiers of multiple target objects are used as a recommended region set similar to the home regions of the candidate objects; or... Based on the similarity values ​​at each intersection position in the row or column where the candidate object's home region identifier is located, the home region identifiers of the target object corresponding to each intersection position are sorted. From the sorting results, select the home regions of the target objects corresponding to the top N home region identifiers of the target objects as a set of recommended regions similar to the home regions of the candidate objects; where N is a positive integer.

6. The information recommendation method according to claim 5, characterized in that, The step of responding to a recommendation request from a user with a region identifier and determining a recommendation region based on the similarity evaluation matrix to recommend the recommendation region to the user includes: In response to a recommendation request from a user with a region identifier, the system searches the row or column of the similarity evaluation matrix to see if the region identifier exists. If it exists, at least one recommended home region identifier is found from the recommended region set corresponding to the home region identifier; The recommended region corresponding to the recommended region identifier is recommended to the users who are about to complete a transaction.

7. The information recommendation method according to claim 2, characterized in that, Also includes: Obtain the primary and secondary region identifiers with a primary-secondary correspondence from the historical data of newly transacting users. Determine whether the similarity evaluation matrix contains a primary-secondary correspondence between the first attribution region identifier and the second attribution region identifier; If included, the existing similarity value at the intersection of the first and second attribution region identifiers in the similarity evaluation matrix is ​​increased by 1.

8. The information recommendation method according to claim 7, characterized in that, Also includes: If the first attribution region identifier and the second attribution region identifier are not included, then the first attribution region identifier and the second attribution region identifier are added to the new rows and columns of the similarity evaluation matrix, respectively. If the first attribution region identifier is included but the second attribution region identifier is not included, then the second attribution region identifier is added to a new row or column in the similarity evaluation matrix; If the first attribution region identifier is not included, but the second attribution region identifier is included, then the first attribution region identifier is added to a new column or row in the similarity evaluation matrix. as well as, The similarity value at the intersection of the first and second attribution region identifiers in the similarity evaluation matrix is ​​increased by 1.

9. The information recommendation method according to claim 1, characterized in that, The step of obtaining the region to which the user who completed the transaction had previously operated on the candidate objects before trading the target object includes: Obtain transaction data and attention data from the historical data of the transacting users; wherein, the attention data includes: pending transaction data and non-pending transaction data; Based on the aforementioned data of interest, the candidate objects that have undergone pending transactions are selected; wherein the pending transactions include: showing or on-site inspection. Determine the region to which the candidate object belongs.

10. An electronic device, characterized in that, include: The memory stores execution instructions; as well as A processor that executes execution instructions stored in the memory, causing the processor to perform the method of any one of claims 1 to 9.

Images