Geographic name space-time derivation relation network construction method and system based on semantic driving

Abstract

The invention discloses a geographical name space-time derivation relation network construction method and system based on semantic driving, and relates to the technical field of geographic information, and the method comprises the steps: obtaining to-be-processed geographical name data; any two place names and proper name information thereof in the place name data to be processed are extracted; calculating a semantic relation, a topological relation and a distance relation of the two extracted place names; for any two geographical names, based on a semantic relationship, a topological relationship and a distance relationship between the two geographical names, judging whether a geographical name space-time derivation relationship exists between the two geographical names or not by adopting a geographical name space-time derivation relationship judgment method; constructing a place name space-time derivation relationship network based on place names with a place name space-time derivation relationship in the place name data to be processed; in the place name space-time derivation relation network, the space-time derivation relation edge is constructed according to the directivity from the native place name to the derived place name. The efficiency and the accuracy of geographic information processing can be remarkably improved.

Description

Technical Field

[0001] The present application relates to the field of geographic information technology, and in particular to a semantically driven method and system for constructing a spatiotemporal derivation relationship network of place names. Background Art

[0002] Place names are a representative and special type of geographic data in geographic information systems. They provide an intuitive way to identify and access specific geographic locations, thereby enhancing the retrieval, analysis and visualization capabilities of geographic data. The way place names are named usually reflects many factors such as geography, history, culture, society and language. Among them, place names can be divided into native place names and related place names according to the relationship between place names. Native place names are named for new places, and related place names are named for new places using native place names. If the relationship between two place names is a derivative relationship, the newly named place name is a derived place name. The derivative relationship refers to the process of naming a new place name by adding, merging and tightening existing place names when naming newly discovered features to establish a relationship between the two place names. Among the derivative relationships, the spatiotemporal derivative relationship is particularly prominent, because it can not only reflect the similarity between the two place names in place names, but also be associated with the location of geographical entities, indicating the spatial proximity between the two place names, and plays an important role in semantic expression and geographic information retrieval.

[0003] However, existing research lacks the discussion of the temporal and spatial derivation relationship of place names, and fails to explore the use of this relationship to enhance the relevance between place names and improve the performance of spatial retrieval or geographic question-answering systems. With the continuous expansion of the application scenarios of place names, place names play a vital role in geographic information systems. They not only serve as the core reference for spatial data positioning, but also carry rich semantic information. As a basic attribute of place names, the temporal and spatial derivation relationship is an important part of the semantics of place names and an important part of people's understanding and use of place names. However, the temporal and spatial derivation relationship has not been fully studied, which restricts the further exploration and application of place names and their semantics. Summary of the invention

[0004] The purpose of this application is to provide a method and system for constructing a spatiotemporal derived relationship network of place names based on semantics, which can improve the accuracy of query operations.

[0005] To achieve the above objectives, this application provides the following solutions:

[0006] In the first aspect, the present application provides a method for constructing a spatiotemporal derivation relationship network of place names based on semantics, including:

[0007] Get the place name data to be processed.

[0008] Extract any two place names and their specific name information from the place name data to be processed.

[0009] The semantic relationship, topological relationship and distance relationship are calculated between the two extracted place names.

[0010] For any two place names, based on the semantic relationship, topological relationship and distance relationship between the two place names, the method of judging the spatiotemporal derivation relationship of place names is used to judge whether there is a spatiotemporal derivation relationship between the two place names.

[0011] Based on the place names with spatiotemporal derivation relationships in the place name data to be processed, a place name spatiotemporal derivation relationship network is constructed; in the place name spatiotemporal derivation relationship network, the spatiotemporal derivation relationship edges are constructed based on the directionality from the original place name to the derived place name.

[0012] In the second aspect, the present application provides a semantically driven spatial-temporal derivation relationship network construction system for place names, including:

[0013] The data acquisition module is used to acquire the place name data to be processed.

[0014] The feature extraction module is used to extract any two place names and their specific name information from the place name data to be processed.

[0015] The relationship calculation module is used to calculate the semantic relationship, topological relationship and distance relationship between the two extracted place names.

[0016] The judgment module is used to judge whether there is a spatiotemporal derivation relationship between any two place names based on the semantic relationship, topological relationship and distance relationship between the two place names and the judgment method of spatiotemporal derivation relationship between place names.

[0017] A construction module is used to construct a place name spatiotemporal derivation relationship network based on the place names with spatiotemporal derivation relationships in the place name data to be processed; in the place name spatiotemporal derivation relationship network, the spatiotemporal derivation relationship edges are constructed based on the directionality from the original place name to the derived place name.

[0018] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0019] The present application provides a method and system for constructing a spatiotemporal derivation relationship network of place names based on semantics-driven, and the method includes: first, it is necessary to obtain the place name data to be processed as the basis for subsequent analysis. From the obtained place name data, extract any two place names and their corresponding special name information to provide data support for subsequent semantic analysis. For the two extracted place names, in-depth calculations of semantic relationships, topological relationships and distance relationships are performed. Based on the semantic relationships, topological relationships and distance relationships calculated above, a method for judging the spatiotemporal derivation relationship of place names is used to accurately judge whether there is a spatiotemporal derivation relationship between any two place names. According to the judgment result, place names with spatiotemporal derivation relationships of place names are screened out from the set of place name data to be processed, and a spatiotemporal derivation relationship network of place names is constructed accordingly. In this network, the spatiotemporal derivation relationship edge is constructed strictly in accordance with the principle of pointing from the original place name to the derived place name. The spatiotemporal derivation relationship network of place names constructed using the method proposed in the present application can effectively improve the accuracy of query operations and provide users with more accurate place name information retrieval services. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 A flowchart of a method for constructing a spatiotemporal derivation relationship network of place names based on semantics-driven is provided in one embodiment of the present application.

[0022] Figure 2 A spatiotemporal derived relationship network provided in one embodiment of the present application.

[0023] Figure 3 A schematic diagram of the structure of a system for constructing a spatiotemporal derivation relationship network of place names based on semantics driven according to an embodiment of the present application. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0025] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0026] Embodiment 1

[0027] like Figure 1 As shown, this embodiment provides a method for constructing a spatiotemporal derivation relationship network of place names based on semantics, including:

[0028] Step 101: Obtain place name data to be processed.

[0029] Step 102: extract any two place names and their specific name information from the place name data to be processed.

[0030] Step 103: Calculate the semantic relationship, topological relationship and distance relationship between the two extracted place names.

[0031] Step 104: For any two place names, based on the semantic relationship, topological relationship and distance relationship between the two place names, a method for determining the spatiotemporal derivation relationship of place names is used to determine whether there is a spatiotemporal derivation relationship between the two place names.

[0032] Step 105: construct a place name spatiotemporal derivation relationship network based on the place names with spatiotemporal derivation relationships in the place name data to be processed; in the place name spatiotemporal derivation relationship network, the spatiotemporal derivation relationship edges are constructed based on the directionality from the original place names to the derived place names.

[0033] Before executing step 101, the following steps may also be included:

[0034] Collect place names in the target area.

[0035] The place names are semantically expressed, and the semantically expressed place names are stored to obtain place name data to be processed.

[0036] The semantic expression of the place name specifically includes:

[0037] Perform attribute description on the place name to obtain an attribute description set of the place name; the attribute description set includes the place name ID 、 Categories of place names, proper place names and common place names.

[0038] The place name is spatially described to obtain a spatial description set of the place name; the spatial description set includes spatial coordinates, administrative divisions, areas, geographic element types and boundary data of geometric objects.

[0039] A time description is performed on the place name to obtain a time description set of the place name; the time description set includes the time point of the place name change, the new name after the place name change, and the new category after the category change.

[0040] Specifically, the semantic expression of place names can be as follows:

[0041] (1) Attribute description:

[0042] The attribute description is intended to describe the basic information and meaning of a place name. The attribute description is defined as a set that contains the following elements:

[0043] Attribute={ID,C,S,G(1).

[0044] In the formula, ID is the place name ID; C is the category of the place name, that is, the category to which the geographical entity referred to by the place name belongs in the category classification system; S is the proper name of the place name, which is the word used to distinguish the geographical entities in the place name; G is the common name of the place name, which is the word used to distinguish the categories of geographical entities in the place name.

[0045] (2) Space description:

[0046] Spatial description is the absolute spatial location information of the geographical entity represented by the place name. The spatial description set is defined as follows:

[0047] Location={L,AD,A,ET,B}(2).

[0048] In the formula, L is the spatial coordinate, which represents the center point of the geographical entity referred to by the place name in the form of a coordinate point; AD is the administrative division, which refers to the administrative division where the geographical entity is located; A is the area, which is the area occupied by the geographical entity; ET is the feature type, which classifies and organizes the geographical entity according to its geometric shape and characteristics; B is the boundary data of the geometric object.

[0049] (3) Time description:

[0050] Place names are relatively stable, but they are not immutable. They will continue to evolve due to historical evolution, changes in the times, and development needs. The temporal evolution of place names is reflected in the renaming of place names. Place names will be affected by many factors, resulting in changes in their naming or the type of geographical entity. The change of place name information considers two situations: ① The category of the place name has not changed, only the name has changed. For example, "Home Inn" was renamed "Hanting Hotel", and both were in the "hotel" category. ② Both the category and name of the place name have changed, for example: "Home Inn" became "711 Convenience Store", and the category changed from "hotel" to "shopping". The definition of the time description set is:

[0051] Time = {NT, NN, NC} (3).

[0052] In the formula, NT is the time point of the new name change, that is, the time point when the place name information changes; NN is the new place name, that is, the new name after the place name changes; NC is the new category, that is, the new category after the category changes.

[0053] The storage of semantically expressed place names specifically includes:

[0054] Store place names as first-level nodes.

[0055] The attribute description set, the space description set and the time description set are connected to the primary node as the secondary nodes.

[0056] For any description set among the attribute description set, the spatial description set and the temporal description set, the attribute value in the description set is associated with its corresponding second-level node as a third-level node.

[0057] Specifically, the acquired place name data is first preprocessed, and according to the standard expression of place names, erroneous or incomplete data is corrected, and redundant and useless data is deleted to ensure the standardization and consistency of place name data. This embodiment uses the Neo4j graph database to store and express place name information and its temporal and spatial derivative relationships. Compared with traditional relational databases, the advantages of graph databases are reflected in intuitiveness, flexibility, high performance, etc. These characteristics make graph databases particularly suitable for processing complex network structures and rich semantic relationships.

[0058] The storage structure of place name information is divided into two main parts: the storage of place name semantic information and the storage of the place name spatiotemporal derivation relationship network. In terms of the expression of place name semantics, place names are stored as first-level nodes; their attribute description sets, spatial description sets, and temporal description sets are organized as second-level nodes and connected to the first-level nodes; the specific attribute values ​​in each description set are used as third-level nodes and associated with the corresponding second-level nodes. Through this hierarchical storage method, the attribute relationship of place names can be clearly expressed, from the first-level node to the second-level node, and so on. For the relevant information of the spatiotemporal derivation relationship of place names, the first-level nodes in the semantic expression can be directly used. In this network, the relationship between the original place name and the derived place name is represented by a directed edge, from the original place name to the derived place name, which intuitively reflects the derivation relationship between place names.

[0059] In some embodiments, when executing step 103, the specific steps may be as follows:

[0060] By using the judgment method of the spatiotemporal derivation relationship of place names, the corresponding attribute values ​​are extracted from the information of the semantic expression of the place names for calculation, and the results are stored as the relationship between the two place names.

[0061] Specifically, firstly, the place names and their proper noun information are extracted to calculate the semantic similarity, and it is determined whether there is a semantic inclusion relationship between the two place names. Secondly, the category information of the two place names is extracted, and the category relationship between the two categories in the category ontology is used for judgment. If the place names are semantically similar and the categories are different, it means that the two place names meet the semantic requirements of the spatiotemporal derivation relationship of place names, and a semantic relationship is generated between the two place names; then the topological relationship between the two place names is calculated and expressed, where the topological relationship includes inclusion, adjacency and separation; finally, the spatial distance between the place names is calculated using the spatial coordinate information of the two objects using a spatial measurement method, and the distance between the two places is expressed as a distance relationship. The distance relationship is obtained by calculating the spatial distance between the place names based on the spatial coordinates of the place names using a spatial measurement method.

[0062] In some embodiments, when executing steps 104-105, the specific steps may be as follows:

[0063] For any two place names, based on the semantic relationship, topological relationship and distance relationship between the two place names, a method for determining the spatiotemporal derivation relationship of place names is used to determine whether there is a spatiotemporal derivation relationship between the two place names, specifically including:

[0064] If there is no semantic relationship between two place names, they are directly judged as common place names.

[0065] If there is a semantic relationship between two place names and their topological relationship is inclusion or adjacency, it is determined that there is a spatiotemporal derivation relationship between the two place names.

[0066] If there is a semantic relationship between two place names and the topological relationship values ​​are separated, it is determined whether there is a spatiotemporal derivation relationship between the two place names based on the category of the place names and the distance relationship between the two place names.

[0067] Specifically, in this embodiment, based on the semantic relationship, topological relationship and distance relationship generated in the first step, it is determined whether there is a spatiotemporal derivation relationship between the two place names. If there is no semantic relationship between the two place names, the two place names are directly determined to be ordinary place names; if there is a semantic relationship between the two place names and their topological relationship values ​​are contained or adjacent, it is determined that there is a spatiotemporal derivation relationship between the two place names; if there is a semantic relationship between the two place names and the topological relationship values ​​are separated, it is necessary to introduce a decision tree model to further determine whether there is a spatiotemporal derivation relationship between the place names based on the category of the original place names and the distance value between the two.

[0068] Then, the place names with spatiotemporal derivation relationships are used to generate spatiotemporal derivation relationship edges, with the original place names pointing to the derived place names, thus completing the construction of the spatiotemporal derivation relationship network of place names.

[0069] In addition, based on the spatiotemporal derivation relationship network of place names constructed in this application, in some embodiments, spatial relationship reasoning based on the spatiotemporal derivation relationship network of place names can also be performed, which can be specifically as follows:

[0070] 1) Reasoning rules for spatial relations:

[0071] In order to improve the accuracy of geographic spatial queries, this embodiment proposes an innovative method that combines qualitative description and quantitative distance. The essence of this method is to identify and effectively utilize the derivative relationship between place names, and on this basis deduce the spatial proximity between land features, thereby enhancing the expressiveness and adaptability of traditional quantitative query methods. Specifically, the consideration of the spatiotemporal derivative relationship of place names is not only limited to the similarity at the semantic level, but also deeply explores the spatial proximity between different geographic entities. By accurately identifying the derivative relationship between place names, other place name entities that have a derivative relationship with the place name entity to be queried can be screened out, thereby realizing accurate reasoning of spatial proximity, and finally returning more practical query results.

[0072] The spatial proximity relationship reasoning process based on the spatiotemporal derived relationship network of place names can be regarded as screening and confirming derived place names that meet specific conditions in the derived relationship network, so as to achieve accurate positioning of the query results.

[0073] R(t p ,t d )∩C i ∈Result(4).

[0074] In the formula, R(t p ,t d ) represents the native place name that satisfies the temporal and spatial derivation relationship t p and derived place names d , C i Indicates the constraint category that the derived place name needs to satisfy. Therefore, the reasoning rules for spatial proximity are as follows: ① Identify all spatiotemporal derived relations of the question entity; ② Determine whether the category of the identified derived place name satisfies the constraint category in the question. If so, the derived place name is the result of reasoning.

[0075] 2) Spatial proximity reasoning:

[0076] Through semantic analysis of the question, when it is determined that the query intention is a qualitative spatial query, the geographic entities and constraint information obtained in the feature extraction process are inferred according to the inference rules.

[0077] First, the spatiotemporal derivation relationship of place names is identified for geographic entities, and all place names with a derivation relationship with the place name are identified. A place name spatiotemporal derivation relationship subnetwork with the place name as the core is constructed. Place names with a derivation relationship in the subnetwork will be used as spatially adjacent entities that meet the qualitative query. Then, based on the entity type of the question query, the place names in the derivation relationship subnetwork are classified and filtered to determine the entities that match the query intent, and all results that meet the category filtering are returned. Finally, the place names that meet the category conditions meet the spatial constraints, so it is inferred that the geographic entities referred to by the two place names have a spatial proximity relationship, and the geographic entity referred to by the place name is returned as the inference result.

[0078] Specifically, take the question "What schools are near Fish Creek?" as an example. Through intent recognition and feature extraction, we get "Fish Creek, nearby, school". First, Fish Creek is used as a place name to identify the spatiotemporal derivation relationship of place names, forming a subnetwork of spatiotemporal derivation relationship of place names about Fish Creek. Then, in this network, we search for derived place names with the category "school" for the category label. Finally, because place names with spatiotemporal derivation relationships of place names meet the spatial constraints of derived recognition, the place names that meet the query conditions in the network are returned as spatial proximity results.

[0079] 3) Spatial location reasoning based on the spatiotemporal derivation relationship network of place names:

[0080] Among them, spatial fuzzy position reasoning is: using the semantic relationship and spatial relationship contained in the constructed spatiotemporal derivation relationship network of place names to reason about the location of missing geographical entities in the place name database, and studying and discussing the influencing factors that need to be considered in the reasoning process. Position is a basic and important feature among the many feature dimensions of space. Cognitive spatial position can answer the question of "where" among the six major questions in geography and provide spatial reference for the answers to other questions in geography.

[0081] In the urban space, the main forms of location information include: coordinates, postal codes, telephone numbers, IP addresses, place names, and addresses. Except for the coordinate type, other location data need to be spatially located before they can be converted to the urban space. Among these types of data that need to be converted, place name data is a relatively standardized data in the form of natural language with a wide range of usage scenarios. It is widely included in the data of citizens' lives, government management, and corporate institutions.

[0082] The basis of spatial location reasoning is to refine the spatial relationship between geographic entities and different reference objects. This refinement mainly utilizes the semantic and spatial relationships in the spatiotemporal derived relationship network.

[0083] The process of the method includes the following three steps:

[0084] The first step is to obtain place name data that have a spatiotemporal derivative relationship with the features to be inferred from the place name database; the second step is to use semantic similarity to obtain the spatial similarity set of the features to be inferred from the spatiotemporal derivative relationship network; the third step is to use the spatiotemporal derivative relationship to predict the location.

[0085] For example, the place name "Mount Royal University Lot 5" does not exist in the place name database, so it is impossible to obtain the geographical location of the place name. Using the location reasoning method of the present application, first search for place name data in the place name database that has a place name spatiotemporal derivation relationship with the place name; then obtain the spatial similarity set of the place name: Mount Royal University, Mount Royal University Lot 1, Mount Royal University Lot 2, MountRoyal University Lot 3, Mount Royal University Lot 4; according to the definition of the derivation relationship, it can be clearly determined that Mount Royal University Lot 5 is located within the constraint range of Mount Royal University, and then according to the distribution mode of Lot 1, Lot 2, Lot 3, and Lot 4 in a straight line, it is determined that Mount Royal University Lot 5 is located next to Mount Royal University Lot 4.

[0086] This embodiment is described by way of example as follows.

[0087] Taking the place name data of area A as an example, the place name data of area A is publicly obtained from OpenStreetMap, including OpenStreetMapID, place name, longitude and latitude, category code, etc. The downloaded data is pre-processed, and after removing the data containing special characters, there are a total of 52,400 valid place name data.

[0088] First, the attributes, space, and time of the place names are expressed in the form of triples (entity, attribute, attribute value) to express the semantics of the place names. For example, the "Slave Lake Station" node represents a place name, and the semantic expression of its place name is decomposed into three basic dimensions: attribute description, spatial description, and time description. The three description sets are represented by secondary nodes, including attribute description set, spatial description set, and time description set. Specific third-level nodes are mounted under each secondary node to store specific attribute values. For example, "bus_stop", "Station", "Slave Lake", etc. are expressed as the values ​​of the attributes such as category, common name, and proper name in the attribute description set.

[0089] The relationship between place name attributes is represented by directed edges, and each node and edge is assigned a corresponding label. For example, category labels are used to clearly identify the categories of all place names, which helps to quickly identify and retrieve specific types of place name information. This multi-level, labeled expression method can not only fully express the semantic information of place names, but also improve the efficiency and accuracy of place name information retrieval.

[0090] Then, the Neo4j graph database is used to store the place name information of area A, and the place names of area A are identified through the judgment method of the spatiotemporal derivation relationship of place names, and a complex spatiotemporal derivation relationship network of place names in area A is constructed, such as Figure 2 shown.

[0091] The network divides place names into two categories: native place names and derived place names. The spatiotemporal derivation relationship of place names is divided into primary derivation, secondary derivation, etc. according to the number of derivations. In particular, when there is a secondary derivation relationship, the derived place name that serves as the primary derivation relationship will be used as the native place name in the secondary derivation relationship. In the network, nodes are used to represent place names, and the edges between nodes represent the spatiotemporal derivation relationship of place names. The direction of the edge is from the native place name to the derived place name. Place names with continuous derivation relationships will be expressed in the form of place name entity chains. In this way, place names with continuous derivation relationships can be connected instead of existing in the form of independent triples. This not only facilitates tracing back to their most original native place names, but also formalizes the complex connections between place names.

[0092] Through analysis, it can be seen that the construction of the association network of the spatiotemporal derivation relationship of place names is of great significance for revealing the deep connection between native place names and derived place names. On the one hand, as a newly defined relationship connecting two place names, it increases the correlation between place names and enriches the semantic expression of place names. On the other hand, the spatiotemporal derivation relationship of place names indicates the semantic similarity and spatial proximity of two place names with this relationship, so the recognition and expression of the spatiotemporal derivation relationship of place names can effectively improve the application of place names in information retrieval and other aspects.

[0093] In order to verify the effectiveness of the method proposed in this application, this application compares the method proposed in this application with the temporal and spatial derivation relationship of place names manually identified, and comprehensively evaluates the method proposed in the text. 10,000 data were randomly selected from the pre-processed place name data in area A for verification analysis. Among them, there are 216 pairs of place names with temporal and spatial derivation relationships of place names manually identified, and there are 202 pairs of place names displayed on the network using the method of this application. The specific analysis results of the network are shown in Table 3:

[0094] Table 3 Evaluation indicators and results of the spatiotemporal derivation relationship network of place names

[0095] Evaluation indicators result TP 199 FP 3 FN 14 Accuracy 98.5％ Recall 93.4％

[0096] By constructing a spatiotemporal derivation relationship network of place names, place names with spatiotemporal derivation relationships are identified and expressed. Specifically, the method of this application correctly identified 199 pairs of derived place names and misidentified 17 pairs of derived place names. Based on this, the accuracy of the method proposed in this application can reach 98.5% and the recall rate is 93.4%.

[0097] Further analysis of the method of this application found that although the method performed well in most cases, there are still some cases where the spatiotemporal derived relationship was not correctly identified. These cases mainly include the following aspects: ① Place name abbreviation problem: In some cases, derived place names are named in the abbreviation form of the original place name, which makes it difficult to identify in the process of semantic similarity calculation. For example: the derived place name "USask Bookstore" in which "USask" is the abbreviation of "University of Saskachewan". Due to the existence of abbreviations, the semantic relationship between the two cannot be accurately identified when making semantic constraint judgments, and the two place names are mistakenly judged to be ordinary place name relationships; ② The derived place names are not written in a standardized manner: For the same native place name, if the derived place name is not named according to the standard writing method of the native place name, it will also lead to incorrect recognition of the semantic constraint relationship. For example: the native place name is "7-Eleven (711 convenience store)", but its derived place names are named and written in "7-11CarWash (711 convenience store car wash store)" and "7Eleven stop (711 convenience store station)". Because of the problem of non-standard writing of derived place names, the ability of this embodiment to identify their semantic relationship is hindered; ③ Decision tree model: Although the decision tree model constructed in this application improves the efficiency of identifying the derived relationship of place names, it also has certain limitations. It cannot completely and accurately identify the spatiotemporal derived relationship between place names based on the spatial distance between place names, so there are some recognition errors when using the decision tree for empty constraint judgment.

[0098] Therefore, the method of this embodiment can achieve satisfactory results in most cases, but the existence of the above-mentioned problems suggests that this embodiment needs to further optimize the method of this application in future work to improve the accuracy and robustness of recognition.

[0099] Therefore, reasoning is performed based on the spatiotemporal derivation relationship network of place names, as follows:

[0100] 1) Spatial proximity reasoning:

[0101] In the field of geographic information retrieval, improving the accuracy of qualitative query results is an ongoing challenge. This application proposes a spatial relationship reasoning method based on the spatiotemporal derived relationship network of place names, aiming to improve the accuracy of qualitative query results, and uses Google Maps and Baidu Maps as references to verify the effectiveness of the reasoning method of this application.

[0102] Google Maps' "Nearby" search function allows users to search based on two quantitative conversion methods: one is based on walking distance, simulating the range a person can walk within a specified time; the other is based on car driving distance, determining the range that can be reached by car. Baidu Maps' "Nearby" search directly sets a quantitative distance around the center point to form a buffer zone for search.

[0103] Take the query "What schools are near McKenzie Towne?" as an example. First, use Google Maps to locate "McKenzie Towne", then use its "Nearby" search function to search for schools near it. The quantitative distance is set to schools that a person can reach within 15 minutes on foot. Baidu Maps follows the same steps to search, and the results returned are schools within 800 meters around "McKenzie Towne" as the center point. The query results are shown in Table 4.

[0104] Table 4. Search results of Google and Baidu maps

[0105]

[0106] The method proposed in this application is used for reasoning. According to the question "What schools are near McKenzie Towne?", the key information "McKenzie Towne, nearby, school" is extracted, and the qualitative spatial query is clarified based on the characteristic word "nearby". Therefore, for place names that have a spatiotemporal derivative relationship with McKenzie Towne, it can be inferred that they are spatially adjacent to McKenzie Towne. The process is as follows: First, in the constructed spatiotemporal derivative relationship network of place names in area A, place names that have a spatiotemporal derivative relationship with "McKenzie Towne" are searched, and there are a total of 8 pairs of derivative relationship pairs; then, among the place names with a derivative relationship, the place names with the place name category of "school" are filtered according to the category label; finally, the derived place name "McKenzie Towne School|Calgary Board of Education" that meets the category filter is returned.

[0107] When the method of the present application is also used for reasoning, this embodiment searches for place names that have a spatiotemporal derivation relationship with "North Battleford" in the spatiotemporal derivation relationship network of place names in area A, and filters out place names with the category of "clinic" according to the category label, and obtains "North Battleford Medical Clinic", which is a clinic located in the vicinity of North Battleford.

[0108] In actual searches, this embodiment can observe that the results of the above two inferences are close to the outer edge of the quantitative search range of Google Maps and Baidu Maps, but cannot be returned as results. If the place name spatiotemporal derived relationship network is used in combination for reasoning, the place names with spatial proximity can be obtained, which can improve the fit between the search results and the actual situation. Therefore, the method proposed in this application is expanded on the basis of traditional quantitative queries, by introducing the spatiotemporal derived relationship of place names to infer the spatial proximity between two place names. This method improves the accuracy and practicality of the retrieval results.

[0109] In addition, this embodiment also notes that Google Maps and Baidu Maps often use a unified quantitative distance standard when processing "nearby" searches for different categories of geographic entities, without fully considering the differences in the existence and scope of influence of different types of land features. This "one-size-fits-all" approach may ignore the specificity and diversity of geographic entities. In response to this problem, the spatiotemporal derived relationship network of place names proposed in this application takes into account the characteristics of different types of land features during the construction process. By setting corresponding spatial proximity distance thresholds for different categories of land features, the spatial proximity between place names can be inferred more accurately, so that the results of geographic information retrieval are closer to the actual situation and can better meet the user's retrieval needs.

[0110] 2) Spatial fuzzy position reasoning:

[0111] This study uses Google Maps to verify the feasibility of location reasoning. To perform location reasoning on the place name "Little Red Deerfishing access", first search for place name data in the place name database that has a temporal and spatial derivation relationship with the place name; then obtain the spatial similarity set of the feature; finally, according to the definition of the derivation relationship, it can be clearly stated that "Little RedDeer fishing access" is within the constraint range of "Red Deer". By searching the location of "Little RedDeer fishing access" on Google Maps, it can be found that it is about 1,760 meters away from "Red Deer".

[0112] Embodiment 2

[0113] like Figure 3 As shown, this embodiment provides a semantically driven spatiotemporal derivation relationship network construction system for place names, including:

[0114] The data acquisition module 301 is used to acquire place name data to be processed.

[0115] The feature extraction module 302 is used to extract any two place names and their specific name information from the place name data to be processed.

[0116] The relationship calculation module 303 is used to calculate the semantic relationship, topological relationship and distance relationship between the two extracted place names.

[0117] The judgment module 304 is used to judge whether there is a spatiotemporal derivation relationship between any two place names based on the semantic relationship, topological relationship and distance relationship between the two place names and using a judgment method for spatiotemporal derivation relationship between place names.

[0118] Construction module 305 is used to construct a place name spatiotemporal derivation relationship network based on the place names with spatiotemporal derivation relationships in the place name data to be processed; in the place name spatiotemporal derivation relationship network, the spatiotemporal derivation relationship edges are constructed based on the directionality from the original place name to the derived place name.

[0119] The system also includes:

[0120] The place name collection module is used to collect place names in the target area.

[0121] The place name data storage module is used to express the place name semantically and store the place name after the semantic expression to obtain the place name data to be processed.

[0122] In summary, this application has the following technical effects:

[0123] This application precisely defines the spatiotemporal derived relationship of place names, clarifies its definition standards and identification technology, and then constructs a network model of spatiotemporal derived relationship of place names, and uses this network as a basis for reasoning research on spatial proximity relationships. Experiments and results show that using the theory and method proposed in this application, the accuracy rate of identifying the spatiotemporal derived relationship of place names in area A is as high as 98.5%, the recall rate is 93.4%, and the reasoning results significantly improve the accuracy of the query. This study not only enriches the theoretical system of spatiotemporal derived relationships of place names, but also solves the current problems of vague definition of spatiotemporal derived relationships and difficulties in automatic identification, and opens up new horizons and tools for application practice in the field of geographic information science.

[0124] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0125] The present application uses specific examples to illustrate the principles and implementation methods of the present application. The above examples are only used to help understand the method and core ideas of the present application. At the same time, for those skilled in the art, according to the ideas of the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present application.

Claims

1. A semantically driven method for constructing a spatiotemporal derivation relationship network of place names, characterized in that: The method for constructing a spatiotemporal derivation relationship network of place names based on semantics drive includes: Get the place name data to be processed; Extract any two place names and their specific name information from the place name data to be processed; Calculate the semantic relationship, topological relationship and distance relationship between the two extracted place names; For any two place names, based on the semantic relationship, topological relationship and distance relationship between the two place names, a method for determining the spatiotemporal derivation relationship of place names is used to determine whether there is a spatiotemporal derivation relationship between the two place names. Based on the place names with spatiotemporal derivation relationships in the place name data to be processed, a place name spatiotemporal derivation relationship network is constructed; in the place name spatiotemporal derivation relationship network, the spatiotemporal derivation relationship edges are constructed based on the directionality from the original place name to the derived place name.

2. According to the method for constructing a semantically driven spatiotemporal derivation relationship network of place names according to claim 1, it is characterized in that: Before obtaining the place name data to be processed, it also includes: Collect place names in the target area; The place names are semantically expressed, and the semantically expressed place names are stored to obtain place name data to be processed.

3. According to the method for constructing a spatiotemporal derivation relationship network of place names driven by semantics in claim 2, it is characterized in that: The place name is semantically expressed, specifically including: Perform attribute description on the place name to obtain an attribute description set of the place name; the attribute description set includes the place name , categories of place names, proper place names and common place names; Performing spatial description on the place name to obtain a spatial description set of the place name; the spatial description set includes spatial coordinates, administrative divisions, areas, geographic element types and boundary data of geometric objects; A time description is performed on the place name to obtain a time description set of the place name; the time description set includes the time point of the place name change, the new name after the place name change, and the new category after the category change.

4. The method for constructing a spatiotemporal derivation relationship network of place names based on semantics drive according to claim 3 is characterized in that: The storage of semantically expressed place names specifically includes: Store place names as first-level nodes; Connect the attribute description set, the space description set and the time description set as the second-level nodes to the first-level nodes; For any description set among the attribute description set, the spatial description set and the temporal description set, the attribute value in the description set is associated with its corresponding second-level node as a third-level node.

5. The method for constructing a spatiotemporal derivation relationship network of place names based on semantics drive according to claim 4 is characterized in that: The semantic relationship includes a containment relationship and a category relationship.

6. The method for constructing a spatiotemporal derivation relationship network of place names based on semantics drive according to claim 5 is characterized in that: The topological relationships include inclusion, proximity and separation.

7. The method for constructing a spatiotemporal derivation relationship network of place names based on semantics drive according to claim 6 is characterized in that: The distance relationship is obtained by calculating the spatial distance between the place names based on the spatial coordinates of the place names using a spatial measurement method.

8. The method for constructing a spatiotemporal derivation relationship network of place names based on semantics drive according to claim 7 is characterized in that: For any two place names, based on the semantic relationship, topological relationship and distance relationship between the two place names, a method for determining the spatiotemporal derivation relationship of place names is used to determine whether there is a spatiotemporal derivation relationship between the two place names, specifically including: If there is no semantic relationship between two place names, they are directly judged as common place names; If there is a semantic relationship between two place names, and their topological relationship is inclusion or adjacency, then it is determined that there is a temporal and spatial derivation relationship between the two place names; If there is a semantic relationship between two place names and the topological relationship values ​​are separated, it is determined whether there is a spatiotemporal derivation relationship between the two place names based on the category of the place names and the distance relationship between the two place names.

9. A semantically driven spatial-temporal derivation relationship network construction system for place names, characterized by: include: A data acquisition module, used to acquire place name data to be processed; A feature extraction module is used to extract any two place names and their specific name information from the place name data to be processed; A relationship calculation module is used to calculate the semantic relationship, topological relationship and distance relationship between the two extracted place names; A judgment module is used to judge whether there is a spatiotemporal derivation relationship between any two place names based on the semantic relationship, topological relationship and distance relationship between the two place names and the judgment method of spatiotemporal derivation relationship between place names; A construction module, used for constructing a place name spatiotemporal derivation relationship network based on place names with spatiotemporal derivation relationship in the place name data to be processed; In the place name spatiotemporal derivation relationship network, the spatiotemporal derivation relationship edge is constructed based on the directionality from the original place name to the derived place name.

10. The semantically driven spatiotemporal derivation relationship network construction system of place names according to claim 9 is characterized in that: Also includes: A place name collection module is used to collect place names of the target area; The place name data storage module is used to express the place name semantically and store the place name after the semantic expression to obtain the place name data to be processed.

Images