Address positioning method and device

Abstract

The present disclosure provides an address positioning method and device, and relates to the field of address information processing. The method comprises the following steps: performing address segmentation on a to-be-processed address text to obtain an address segmentation set; performing node matching on the address segmentation set and a node name set of a geographical hierarchical tree to obtain a target node set of successful matching; forming a candidate path set based on the target node set; selecting an optimal path from the candidate path set; and determining a standardized address of the to-be-processed address text based on the optimal path. The present disclosure takes into account the hidden geographical hierarchical information in the to-be-processed address text, maintains the hierarchical relationship between different geographical entities, and enhances the accuracy and robustness of the address positioning method.

Description

Technical Field

[0001] This disclosure relates to the field of address information processing, and in particular to an address location method and apparatus. Background Technology

[0002] With the digital development of the surveying and mapping industry and the arrival of the big data era, massive amounts of geographic data have been generated, including administrative divisions, points of interest, and roads and buildings. This has led to a wealth of upper-level applications, such as map service providers offering location search services based on urban geographic data.

[0003] Address location is a common type of geographic query, which involves inferring the geographic entity that the text refers to and outputting its standard address, given a descriptive text related to an address.

[0004] In related technologies, address location methods are based on text search engines, which directly search for geographical entity information that is similar to the text in existing data. Summary of the Invention

[0005] The inventors discovered that the address placement methods in related technologies are based solely on text search technology, using the similarity between text characters to place addresses. Therefore, the accuracy and robustness of the address placement methods in related technologies are poor.

[0006] One technical problem to be solved by the embodiments of this disclosure is to enhance the accuracy and robustness of the address placement method.

[0007] According to one aspect of this disclosure, an address location method is proposed, comprising: performing address segmentation on the address text to be processed to obtain an address segmentation set; performing node matching between the address segmentation set and a set of node names of a geographic hierarchy tree to obtain a set of successfully matched target nodes; forming a candidate path set based on the target node set; selecting the optimal path from the candidate path set; and determining the standardized address of the address text based on the optimal path.

[0008] In some embodiments, forming a candidate path set based on the target node set includes: traversing all nodes in the target node set one by one and calculating the path from the root node to each node in the target node set; and adding the path as a candidate path to the candidate path set if the path is not included in the candidate path set.

[0009] In some embodiments, selecting the optimal path from the candidate path set includes: selecting the candidate path with the longest length from the candidate path set as the optimal path.

[0010] In some embodiments, selecting the optimal path from the candidate path set further includes: if the candidate path set includes multiple longest candidate paths, selecting the candidate path that includes the most successfully matched nodes from the multiple longest candidate paths as the optimal path.

[0011] In some embodiments, selecting the optimal path from the candidate path set further includes: when the candidate path set includes multiple candidate paths with the longest length and the most successfully matched nodes, selecting the candidate path with the largest sum of the number of levels of successfully matched nodes from the multiple candidate paths with the longest length and the most successfully matched nodes as the optimal path.

[0012] In some embodiments, matching the address segmentation set with the node name set of the geographic hierarchy tree to obtain a successfully matched target node set includes: matching the address segmentation set with the node name set of the geographic hierarchy tree at the cell level and above to obtain a successfully matched target node set.

[0013] In some embodiments, determining the standardized address of the address text based on the optimal path includes: based on the subtree of the optimal path, continuing to perform node matching at the cell-level and below with the node name set of the geographic hierarchy tree; and combining the optimal path and the cell-level and below node matching results as the standardized address of the address text.

[0014] In some embodiments, performing address segmentation on the address text to be processed to obtain an address segmentation set includes: using the node name set of the geographic hierarchy tree to perform address segmentation on the address text to be processed to obtain an address segmentation set.

[0015] In some embodiments, matching the address segmentation set with the node name set of the geographic hierarchy tree includes: constructing a regular expression for a specific node name in the node name set that contains numbers, letters, or specific characters, and matching the address segmentation set with the regular expression of the specific node name.

[0016] In some embodiments, matching the address word set with the node name set of the geographic hierarchy tree includes: using the longest common substring similarity matching method to match the address word set with the node name set of the geographic hierarchy tree.

[0017] In some embodiments, matching the address word set with the node name set of the geographic hierarchy tree includes: using a similarity matching method based on pinyin to match the address word set with the node name set of the geographic hierarchy tree.

[0018] In some embodiments, the method further includes: adding all geographic entities to the node set of the geographic hierarchy tree; traversing each geographic entity in the node set of the geographic hierarchy tree, connecting each geographic entity to its parent geographic entity to form an edge, and adding each edge to the edge set of the geographic hierarchy tree; and constructing the geographic hierarchy tree based on the node set and edge set of the geographic hierarchy tree.

[0019] In some embodiments, the method further includes providing an address search service using the standardized address of the address text.

[0020] According to another aspect of this disclosure, an address placement apparatus is proposed, comprising: a module that performs the address placement method of any embodiment.

[0021] According to another aspect of this disclosure, an address placement apparatus is proposed, comprising: a memory; and a processor coupled to the memory, the processor being configured to execute the address placement method of any embodiment based on instructions stored in the memory.

[0022] According to another aspect of this disclosure, a computer-readable storage medium is provided on which a computer program is stored, which, when executed by a processor, implements the address placement method of any of the foregoing embodiments.

[0023] The address location scheme of this disclosure introduces a geographic hierarchy tree, takes into account the geographic hierarchy information hidden in the text, maintains the hierarchical relationship between different geographic entities, and enhances the accuracy and robustness of the address location method. Attached Figure Description

[0024] The accompanying drawings, which form part of this specification, illustrate embodiments of this disclosure and, together with the specification, serve to explain the principles of this disclosure.

[0025] This disclosure will become clearer with reference to the accompanying drawings and the following detailed description, wherein:

[0026] Figure 1 A flowchart of an address placement method according to some embodiments of the present disclosure is shown.

[0027] Figure 2 A schematic diagram of a geographic hierarchy tree according to some embodiments of the present disclosure is shown.

[0028] Figure 3 A flowchart illustrating the construction of a geographic hierarchy tree according to some embodiments of this disclosure is shown.

[0029] Figure 4 A flowchart of a method for selecting the optimal path according to some embodiments of the present disclosure is shown.

[0030] Figure 5 A schematic diagram of optimal path selection rules according to some embodiments of the present disclosure is shown.

[0031] Figure 6 A schematic diagram of an address placement device according to some embodiments of the present disclosure is shown.

[0032] Figure 7 A schematic diagram of an address placement device according to some embodiments of the present invention is shown. Detailed Implementation

[0033] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the present disclosure.

[0034] At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn according to actual scale.

[0035] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit this disclosure or its application or use.

[0036] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0037] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0038] It should be noted that similar labels and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be discussed further in subsequent figures.

[0039] Figure 1 A flowchart of an address placement method according to some embodiments of the present disclosure is shown.

[0040] like Figure 1As shown, the address placement method includes steps S130-S170, and may further include at least one of steps S110 and S120 as needed. Step S110: Input the address text to be processed; Step S120: Construct a geographic hierarchy tree; Step S130: Perform address segmentation on the address text to be processed to obtain an address segmentation set; Step S140: Perform node matching between the address segmentation set and the node name set of the geographic hierarchy tree to obtain a set of successfully matched target nodes; Step S150: Form a candidate path set based on the target node set; Step S160: Select the optimal path from the candidate path set; and Step S170: Determine the standardized address of the address text based on the optimal path. For example, the address placement method is executed by an address placement device.

[0041] In the above embodiments, by matching the address segmentation set obtained from the address text to be processed with the node name set of the geographic hierarchy tree, and forming a candidate path set based on the target node set, and then selecting the optimal path from the candidate path set, the hidden geographic hierarchy information in the text is taken into account, the hierarchical relationship between different geographic entities is maintained, and the accuracy and robustness of the address location method are enhanced.

[0042] In step S110, input the address text to be processed.

[0043] In some embodiments, address text is typically a sentence of natural language describing the specific location of a geographic entity. For example, "Room XXX, Building 2, Xingyue International Community, Tongzhou District, Beijing" can be referred to as address text, denoted by the lowercase letter 'q'. Here, XXX refers to the specific room number.

[0044] In step S120, a geographic hierarchy tree is constructed. For example, as shown... Figure 2 As shown, Figure 2 A schematic diagram of a geographic hierarchy tree according to some embodiments of the present disclosure is shown.

[0045] In some embodiments, the geographic hierarchy is based on administrative divisions. For example, from highest to lowest, the geographic hierarchy can be divided into the following nine categories: Province, City, District / County, Street / Township, Community / Village, Residential, Building, Unit, and House. Higher and lower levels have an inclusive relationship. The lowercase letter 'l' is used to represent the geographic hierarchy, and the subscript indicates the level; for example, the smaller the subscript, the higher the level. For instance, l1 represents a province, and l9 represents a house number, etc., with the order l1>l2>l3>l4>l5>l6>l7>l8>l9 defined. For example, Beijing is a city-level administrative region, which administers multiple district-level administrative regions, such as Dongcheng District, Xicheng District, and Chaoyang District.

[0046] In some embodiments, a geographic hierarchy tree is essentially a tree data structure. A geographic hierarchy based on administrative region affiliation is a multi-branch tree where nodes are geographic entities, each node can contain zero or more child nodes at the next lower level, and edges represent the hierarchical geographic entity affiliation relationships. Since the scope of the community level is difficult to determine, and the scope of the district level is too large, points of interest are attached to street-level nodes. The geographic hierarchy tree is represented by the uppercase letter T, and formally expressed as T = ...<Vt,Et> , where Vt={s1,s2,…,sn} is the set of nodes, Et={e1,e2,…,em} is the set of edges, n represents the number of nodes, and m represents the number of edges.

[0047] In some embodiments, geographic entities are typically geographic spatial ranges based on administrative divisions, represented by lowercase 's', while uppercase 'S' represents a set of geographic entities. A geographic entity contains three attributes: parent, name, and level, i.e., s = (parent, name, level). For example, the geographic entity of Dongcheng District can be represented as s = (sp, Dongcheng District, l3), where sp represents the parent geographic entity of Dongcheng District, i.e., the geographic entity "Beijing".

[0048] It is worth noting that, in addition to the geographic entities based on administrative divisions mentioned above, there are also geographic entities of the Point of Interest (POI) type. This is because in reality, many POIs often do not have a clear administrative hierarchy, such as business districts, schools, or parks. However, these locations are often hotspots for user searches, so it is necessary to include them in the geographic hierarchy tree in order to participate in subsequent address placement.

[0049] The method for constructing a geographic hierarchy tree is as follows: Figure 3 As shown, Figure 3A flowchart illustrating the construction of a geographic hierarchy tree according to some embodiments of this disclosure is shown.

[0050] like Figure 3 As shown, the method for constructing a geographic hierarchy tree in this embodiment includes steps S310-S330.

[0051] In some embodiments, the definition of a geographic hierarchy tree is as follows: Given a set of geographic entities S = {s1, s2, ..., sn}, construct a geographic hierarchy tree T =<Vt,Et> Where n is the number of address entities, Vt represents the set of nodes in the geographic hierarchy tree, and Et represents the set of edges in the geographic hierarchy tree. The geographic hierarchy tree is essentially a multi-branch tree, where nodes represent geographic entities and edges represent the parent-child relationship between two entities. Therefore, the data structure of the geographic hierarchy tree implicitly stores the containment and included relationships between pairs of geographic entities.

[0052] In step S310, all geographic entities are added to the node set of the geographic hierarchy tree, i.e., Vt←{s1,…,sn}.

[0053] In some embodiments, each geographic entity is a node in the geographic hierarchy tree, and all geographic entities form a set of nodes in the geographic hierarchy tree.

[0054] In step S320, each geographic entity in the node set of the geographic hierarchy tree is traversed, and each geographic entity is connected to its parent geographic entity to form an edge. The resulting edges are then added to the edge set of the geographic hierarchy tree.

[0055] In some embodiments, each geographic entity s is traversed to obtain its parent geographic entity s.parent, and the two are connected to form an edge, i.e., e = (s.parent, s), and added to the edge set of the geographic hierarchy tree, i.e., Eg←Eg∪{e}.

[0056] In step S330, the geographic hierarchy tree is constructed based on the set of nodes and the set of edges of the geographic hierarchy tree.

[0057] Thus, through the above steps, a geographical hierarchy tree was constructed by adding points and then connecting edges.

[0058] In step S130, the address text to be processed is segmented into words to obtain a set of address segments.

[0059] In some embodiments, address word segmentation refers to the recognition and extraction of geographical element names from address text, that is, splitting a natural language sentence related to an address into multiple consecutive words with geographical meanings. The formal definition of address word segmentation is as follows: Given a set of geographical element names W, and then given an address text q, split the address text q into a sequence of multiple consecutive existing geographical element names Wq, where Wq = <w1, w2,.., wh>, w represents each word after word segmentation, and h represents the number of words after segmentation. Here, W represents the set of names of known geographical elements. For example, W = {"Beijing City", "Tongzhou District", "Xingyue International Community",...}, where the geographical element names can be administrative division names or names of points of interest such as communities and business districts. For example, the address word segmentation result of the address text "Room XXX, Building 2, Xingyue International Community, Tongzhou District, Beijing City" is "Beijing City / Tongzhou District / Xingyue International Community / Building 2 / XXX Room", where the slashes indicate the splitting positions.

[0060] In some embodiments, a general natural language word segmentation method can be adopted. For example, an address word segmentation model based on the Hidden Markov Model (HMM) is used to perform address word segmentation on the address text to be processed, and an address word segmentation set is obtained.

[0061] In some embodiments, the set of node names of the geographical hierarchy tree can be utilized, and the set of node names of the geographical hierarchy tree is used as the set of known geographical element names W to perform address word segmentation on the address text to be processed, and an address word segmentation set is obtained. For example, if segmented by the general word segmentation method, "Xingyue International Community" is segmented into Xingyue / International / Community. If segmented using the set of node names of the geographical hierarchy tree, then "Xingyue International Community" is segmented into Xingyue International Community, making the effect of address word segmentation more accurate.

[0062] In step S140, the address word segmentation set is matched with the set of node names of the geographical hierarchy tree to obtain a set of target nodes with successful matches.

[0063] In some embodiments, the definition of node matching is as follows: Given a geographical hierarchy tree T = <Vt, Et>, and then given the sequence Wq = <w1, w2,.., wh> after the address text q is segmented, map Wq to a set of multiple geographical hierarchy tree nodes Vq = <s1, s2,.., sk> based on name similarity, where k is the number of successfully matched nodes.

[0064] Furthermore, because address levels below the community level have low recognizability, the address segmentation set is matched with the node name set of the geographic hierarchy tree at the community level and above to obtain a set of successfully matched target nodes. For example, each community has Building 1 and Building 2, each building has Unit 1 and Unit 2, and each unit has No. 101 and No. 102, etc. If nodes below the community level are added to the matching, the number of successfully matched nodes will surge, which will have a certain impact on the performance and accuracy of subsequent path selection steps. For example, if a user enters "Building 7, Baiduizi Community, Haidian District, Beijing", but there is no Building 7 in that community, and several other communities contain Building 7, according to the path selection rules below, the address may be identified as another community. However, this situation is mostly due to the user entering the building number incorrectly; their real purpose is still to find "Baiduizi Community, Haidian District, Beijing". In summary, the processing scheme of this embodiment is to match nodes at the community level and above. As for nodes below the community level, the building unit number information can be matched based on the subtree of the path after the path selection is completed.

[0065] In some embodiments, for example, in real-world situations, the addresses entered by users often contain aliases, and the actual addresses are not filled in strictly according to the standard address names. For example, "Beijing" is written as "Beijing" and "Chaoyang District" is written as "Chaoyang". These situations cannot be avoided, so appropriate processing is required during matching. For example, standardized suffixes such as province, city, district, street, and community can be removed.

[0066] In some embodiments, a regular expression is constructed for a specific node name in the node name set that contains numbers, letters, or specific characters. The address segmentation set is then matched against the regular expression for the specific node name. For example, in reality, the addresses entered by users often contain aliases, and the actual addresses are not strictly filled in according to the standard address names, such as "Building No. 10" being written as "Building 10". These situations cannot be avoided, so appropriate processing is required during matching, such as using the regular expression "([0-9]+|[a-zA-Z])(No. Building|Building|Building|Building)" to extract the building number, etc.

[0067] In some embodiments, the longest common substring similarity matching method is used to match the address word segmentation set with the node name set of the geographic hierarchy tree. For example, in reality, users often mistype the addresses they enter. Due to errors such as mistakes, users may write the address information incorrectly, such as writing "Anzhenli Yiqu" as "Anzhen Yiqu". In this case, the longest common substring can be used for approximate matching. For example, the length of the longest common substring between "Anzhenli Yiqu" and "Anzhen Yiqu" is 4, so the similarity between the two can be simply defined as 4 / 5 = 0.8.

[0068] In some embodiments, a similarity matching method based on pinyin is used to match the address word segmentation set with the node name set of the geographic hierarchy tree. For example, in reality, users often miswrite addresses. Due to errors such as mistakes, users may write the address information incorrectly, such as writing "Anzhen Street" as "Anzhen Street". In such cases, pinyin can be used for approximate matching. For example, both "Anzhen Street" and "Anzhen Street" are pronounced "anzhenjiedao" in pinyin, so their similarity is high.

[0069] In some embodiments, in reality, the addresses entered by users may contain contradictions. For example, the address "Anzhenli No. 1, Chaoyang District, Haidian District, Beijing" contains information from both "Haidian District" and "Chaoyang District." According to the construction principles of the geographic hierarchy tree, an address cannot belong to two districts simultaneously, thus creating a contradiction. However, this situation is relatively easy to handle. Although erroneous information cannot be eliminated in the node matching step, other hierarchical information in the address text can be used to assist in the judgment during the optimal path selection step. For example, "Anzhenli No. 1" actually belongs to "Anzhen Street, Chaoyang District." Therefore, candidate paths for "Anzhenli No. 1" that include Haidian District do not meet the subsequent optimal path selection rules and can be directly eliminated.

[0070] In some embodiments, in reality, because nodes at different levels may have the same or similar node names, an address segmentation may simultaneously match multiple nodes in the set of node names in the geographic hierarchy tree during the matching process. This can lead to multiple candidate paths for a single address text. To resolve this, the optimal path is selected from the set of candidate paths (i.e., step S140). See [link to relevant documentation]. Figure 4 .

[0071] In some embodiments, in reality, users often omit information when filling in addresses. Most users don't know the name of the community or even the street where the address is located, such as "Building 10, Anzhenli First District, Chaoyang District, Beijing," which lacks street and community information. This situation can be resolved in the step of forming the candidate path set (i.e., step S130). This is because the missing information corresponding to the address can be traced back during the step of forming the candidate path set.

[0072] In some of the above embodiments, the set of node names of the geographic hierarchy tree is used as the data basis for address word segmentation, i.e., W = {s1.name, s2.name, ..., sn.name, ...}. At the same time, word segmentation matching is considered from multiple dimensions such as aliases, pinyin, and similarity of address names. Therefore, theoretically, the more complete the geographic hierarchy tree is constructed, the more accurate the word segmentation matching is, which enhances the accuracy and robustness of the address location method.

[0073] In step S150, a candidate path set is formed based on the target node set, as detailed in steps S151-S152.

[0074] In step S151, all nodes in the target node set are traversed one by one, and the path from the root node to each node in the target node set is calculated.

[0075] In some embodiments, the root node is the highest-level node in the geographic hierarchy tree. Every non-root node has one and only one path from the root node to itself. Simultaneously, each node in the geographic hierarchy tree implicitly contains a candidate path from the highest-level node to itself. For example, the node "Xingfu Community" has a candidate path "Beijing / Beijing / Dongcheng District / Donghuamen Street / Duofu Lane Community / Xingfu Community". Thus, calculating the path from the root node to the matching node directly yields the implicit candidate path information for each matching node. It is worth noting that the full path for "Xingfu Community" contains two instances of "Beijing". This is because Beijing is a municipality directly under the central government and does not have the concept of a province. To adapt to the nine-level geographic hierarchy tree, for municipalities directly under the central government, the provincial and municipal nodes are set as the same geographic entity.

[0076] In some embodiments, the path of a node is defined as follows: For a geographic hierarchy tree T and two nodes si and sj, without loss of generality assuming that the level of si is not lower than that of sj, the path pathij from node si to node sj =<si,si+1,…,sj> In this case, the level of the preceding node is always higher than that of the following node; that is, for all k, i ≤ k < j, the level of sk is higher than sk+1. For example... Figure 1 As shown, the path from node "City 2" to "Point of Interest 1" is "City 2 / District 1 / Street 1 / Point of Interest 1", while there is no path that satisfies the condition for nodes "Building 1" and "Building 2", that is, there is no path that satisfies the condition that the level of the previous node must be higher than that of the next node.

[0077] In step S152, if the path is not included in the candidate path set, the path is added to the candidate path set as a candidate path.

[0078] In some embodiments, a candidate path set is initialized, and then, if the path is not included in the candidate path set, the path is added to the candidate path set as a candidate path.

[0079] In step S160, the optimal path is selected from the candidate path set. See details... Figure 4 Steps S410-S450 in the process. Figure 4 A flowchart of a method for selecting the optimal path according to some embodiments of the present disclosure is shown.

[0080] In some embodiments, after the aforementioned steps S130 and S140, multiple nodes have been matched on the geographic hierarchy tree. If there is one and only one path that contains all the matched nodes, i.e., the candidate path set contains only one path, then there is no path selection, and the path is directly returned, and the algorithm ends. However, due to the aforementioned issues such as address aliases, misspellings, omissions, and contradictions, multiple paths often become available. In this case, it is necessary to define some rules to determine which path to use, such as... Figure 5 As shown, Figure 5 A schematic diagram of optimal path selection rules according to some embodiments of the present disclosure is shown. The combination of circles and solid lines represents a geographic hierarchy tree, gray circles represent nodes in the tree where the address text is successfully matched, and dashed lines with arrows represent the paths formed by the nodes.

[0081] like Figure 4 As shown, the optimal path selection method includes: Step S410, start; Step S420, determine whether the candidate path set includes multiple longest candidate paths; Step S421, if the candidate path set does not include multiple longest candidate paths, then select the longest candidate path from the candidate path set as the optimal path; Step S430, if the candidate path set includes multiple longest candidate paths, determine whether the candidate path includes multiple candidate paths with the most successfully matched nodes; Step S431, if the candidate path does not include multiple candidate paths with the most successfully matched nodes, then select the candidate path with the most successfully matched nodes from the multiple longest candidate paths as the optimal path; Step S440, if the candidate path set includes multiple longest candidate paths that also include the most successfully matched nodes, then select the candidate path with the largest sum of the number of levels of successfully matched nodes from the multiple longest candidate paths that also include the most successfully matched nodes as the optimal path; Step S450, end.

[0082] In steps S420 and S421, the longest candidate path is selected from the candidate path set as the optimal path.

[0083] In some embodiments, the optimal path selection rule can be to choose the longest path over the shortest, that is, to prioritize the longer path. For example... Figure 5 As shown in (1), path 1 spans 3 nodes and has a length of 2, while path 2 spans 4 nodes and has a length of 3. Since the length of path 2 is greater than that of path 1, path 2 is ultimately chosen.

[0084] In steps S430 and S431, if the candidate path set includes multiple longest candidate paths, the candidate path with the most successfully matched nodes is selected from the multiple longest candidate paths as the optimal path.

[0085] In some embodiments, the optimal path selection rule can be to select the path with more matching nodes, i.e., when path lengths are equal, the path with more successfully matched nodes is preferred. For example... Figure 5 As shown in (2), both path 3 and path 4 span 4 nodes and have a length of 3, but path 4 contains 3 successfully matched nodes, while path 3 only contains 2 successfully matched nodes. Therefore, path 2 is ultimately chosen.

[0086] In step S440, if the candidate path set includes multiple candidate paths with the longest length and the most successfully matched nodes, the candidate path with the largest sum of the number of levels of successfully matched nodes is selected as the optimal path from the multiple candidate paths with the longest length and the most successfully matched nodes.

[0087] In some embodiments, the optimal path selection rule can be to choose the deepest path over the shallowest, that is, when the path length and the number of successfully matched nodes in the path are equal, the path with the deeper level of successfully matched nodes is preferred. For example... Figure 5 As shown in (3), both path 5 and path 6 span 4 nodes with a length of 3, and both contain 3 successfully matched nodes. However, because the sum of the number of levels of the successfully matched nodes in path 6 is larger, path 5 is 1+2+4=7, where 1, 2, and 4 are the number of levels of the 3 successfully matched nodes in path 5, and path 6 is 1+3+4=8. It is figuratively said that path 6 is deeper than path 5, so path 6 is finally chosen.

[0088] In step S170, the standardized address of the address text is determined based on the optimal path.

[0089] In some embodiments, determining the standardized address of the address text can be based on the subtree of the optimal path, and further performing node matching at the cell level with the node name set of the geographic hierarchy tree; the optimal path and the cell-level node matching results are combined as the standardized address of the address text.

[0090] The address location scheme of this disclosure introduces a geographic hierarchy tree, takes into account the geographic hierarchy information hidden in the text, maintains the hierarchical relationship between different geographic entities, and enhances the accuracy and robustness of the address location method.

[0091] Figure 6 A schematic diagram of an address placement device according to some embodiments of the present disclosure is shown.

[0092] like Figure 6 As shown, the address placement device of this embodiment includes a memory 61 and a processor 62 coupled to the memory 61. The processor 62 is configured to execute the method in any embodiment of this disclosure based on instructions stored in the memory 61.

[0093] The memory 61 may include, for example, system memory, fixed non-volatile storage media, etc. The system memory stores, for example, the operating system, application programs, boot loader, database, and other programs.

[0094] Figure 7 A schematic diagram of an address placement device according to some embodiments of the present disclosure is shown.

[0095] like Figure 7 As shown, the address placement device of this embodiment includes a memory 710 and a processor 720 coupled to the memory 710. The processor 720 is configured to execute the method in any of the foregoing embodiments based on instructions stored in the memory 710.

[0096] The memory 710 may include, for example, system memory, fixed non-volatile storage media, etc. The system memory may store, for example, the operating system, application programs, boot loader, and other programs.

[0097] The address placement device may also include an input / output interface 730, a network interface 740, and a storage interface 750. These interfaces 730, 740, and 750, as well as the memory 710 and processor 720, can be connected via, for example, a bus 760. The input / output interface 730 provides a connection interface for input / output devices such as a monitor, mouse, keyboard, touchscreen, microphone, and speakers. The network interface 740 provides a connection interface for various networked devices. The storage interface 750 provides a connection interface for external storage devices such as SD cards and USB flash drives.

[0098] (1) An address placement method, comprising:

[0099] The address text to be processed is segmented into words to obtain a set of address segments;

[0100] The address word segmentation set is matched with the node name set of the geographic hierarchy tree to obtain the set of target nodes that are successfully matched;

[0101] A candidate path set is formed based on the target node set;

[0102] Select the optimal path from the set of candidate paths; and

[0103] The standardized address of the address text is determined based on the optimal path.

[0104] (2) According to (1), forming a candidate path set based on the target node set includes:

[0105] Iterate through all nodes in the target node set one by one, and calculate the path from the root node to each node in the target node set; and

[0106] If the path is not included in the candidate path set, the path is added to the candidate path set as a candidate path.

[0107] (3) According to (1-2), selecting the optimal path from the candidate path set includes:

[0108] The longest candidate path is selected from the set of candidate paths as the optimal path.

[0109] (4) According to (1-3), selecting the optimal path from the candidate path set further includes:

[0110] If the candidate path set includes multiple longest candidate paths, the candidate path containing the most successfully matched nodes is selected from the multiple longest candidate paths as the optimal path.

[0111] (5) According to (1-4), selecting the optimal path from the candidate path set further includes:

[0112] If the candidate path set includes multiple candidate paths with the longest length and the most successfully matched nodes, the candidate path with the largest sum of the number of levels of successfully matched nodes is selected as the optimal path from among the multiple candidate paths with the longest length and the most successfully matched nodes.

[0113] (6) According to (1-5), the set of address word segments is matched with the set of node names of the geographic hierarchy tree to obtain a set of successfully matched target nodes, including:

[0114] The address segmentation set is matched with the node name set of the geographic hierarchy tree at the cell level and above to obtain the target node set that is successfully matched.

[0115] (7) According to (1-6), determining the standardized address of the address text based on the optimal path includes:

[0116] Based on the subtree of the optimal path, the address word segmentation set is further matched with the node name set of the geographic hierarchy tree at the sub-cell level.

[0117] The optimal path and the node matching results at the cell level and below are combined to form the standardized address of the address text.

[0118] (8) According to (1-7), the address text to be processed is segmented into words, and the resulting set of address segments includes:

[0119] Using the set of node names in the geographic hierarchy tree, address text to be processed is segmented into words to obtain a set of address segments.

[0120] (9) According to (1-8), the node matching between the address segmentation set and the node name set of the geographic hierarchy tree includes:

[0121] For a specific node name in the set of node names that contains numbers, letters, or specific characters, a regular expression is constructed, and the set of address tokens is matched with the regular expression of the specific node name.

[0122] (10) According to (1-9), the node matching between the address segmentation set and the node name set of the geographic hierarchy tree includes:

[0123] The address word segmentation set is matched with the node name set of the geographic hierarchy tree using the longest common substring similarity matching method; or,

[0124] Using a similarity matching method based on pinyin, the address word segmentation set is matched with the node name set of the geographic hierarchy tree.

[0125] (11) According to (1-10), it also includes:

[0126] Add all geographic entities to the node set of the geographic hierarchy tree;

[0127] Traverse each geographic entity in the node set of the geographic hierarchy tree, connect each geographic entity to its parent geographic entity to form an edge, and add each edge to the edge set of the geographic hierarchy tree.

[0128] The geographic hierarchy tree is constructed based on the set of nodes and the set of edges of the geographic hierarchy tree.

[0129] (12) According to (1-11), it also includes:

[0130] Address search service is provided using the standardized address of the address text.

[0131] 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 non-transitory storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0132] This concludes the detailed description of the present disclosure. To avoid obscuring the concept of this disclosure, some details known in the art have not been described. Those skilled in the art will fully understand how to implement the technical solutions disclosed herein based on the above description.

[0133] The methods and systems of this disclosure may be implemented in many ways. For example, they may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above-described order of steps for the methods is for illustrative purposes only, and the steps of the methods of this disclosure are not limited to the specific order described above unless otherwise specifically stated. Furthermore, in some embodiments, this disclosure may also be implemented as a program recorded on a recording medium, the program including machine-readable instructions for implementing the methods according to this disclosure. Thus, this disclosure also covers recording media storing programs for performing the methods according to this disclosure.

[0134] While specific embodiments of this disclosure have been described in detail by way of example, those skilled in the art should understand that the examples are for illustrative purposes only and not intended to limit the scope of this disclosure. Those skilled in the art should understand that modifications can be made to the above embodiments without departing from the scope and spirit of this disclosure. The scope of this disclosure is defined by the appended claims.

Claims

1. An address placement method, comprising: The address text to be processed is segmented into words to obtain a set of address words, including: using the set of node names of the geographic hierarchy tree to segment the address text to be processed into words to obtain a set of address words; The process of matching the address word set with the node name set of the geographic hierarchy tree to obtain a set of successfully matched target nodes includes: matching the address word set with the node name set of the geographic hierarchy tree at the cell level and above to obtain a set of successfully matched target nodes. A candidate path set is formed based on the target node set; Select the optimal path from the set of candidate paths; and Determining the standardized address of the address text based on the optimal path includes: based on the subtree of the optimal path, continuing to perform node matching at the cell-level and below with the node name set of the geographic hierarchy tree; combining the optimal path and the cell-level and below node matching results as the standardized address of the address text.

2. The address location method according to claim 1, wherein forming a candidate path set based on the target node set includes: Iterate through all nodes in the target node set one by one, and calculate the path from the root node to each node in the target node set; as well as If the path is not included in the candidate path set, the path is added to the candidate path set as a candidate path.

3. The address placement method according to claim 1, wherein selecting the optimal path from the candidate path set includes: The longest candidate path is selected from the set of candidate paths as the optimal path.

4. The address placement method according to claim 3, wherein selecting the optimal path from the candidate path set further includes: If the candidate path set includes multiple longest candidate paths, the candidate path containing the most successfully matched nodes is selected from the multiple longest candidate paths as the optimal path.

5. The address placement method according to claim 4, wherein selecting the optimal path from the candidate path set further includes: If the candidate path set includes multiple candidate paths with the longest length and the most successfully matched nodes, the candidate path with the largest sum of the number of levels of successfully matched nodes is selected as the optimal path from among the multiple candidate paths with the longest length and the most successfully matched nodes.

6. The address location method according to any one of claims 1-5, wherein matching the address segmentation set with the node name set of the geographic hierarchy tree includes: For a specific node name in the set of node names that contains numbers, letters, or specific characters, a regular expression is constructed, and the set of address tokens is matched with the regular expression of the specific node name.

7. The address location method according to any one of claims 1-5, wherein matching the address segmentation set with the node name set of the geographic hierarchy tree includes: The longest common substring similarity matching method is used to perform node matching between the address word segmentation set and the node name set of the geographic hierarchy tree; or, Using a similarity matching method based on pinyin, the address word segmentation set is matched with the node name set of the geographic hierarchy tree.

8. The address placement method according to any one of claims 1-5, further comprising: Add all geographic entities to the node set of the geographic hierarchy tree; Traverse each geographic entity in the node set of the geographic hierarchy tree, connect each geographic entity to its parent geographic entity to form an edge, and add each edge to the edge set of the geographic hierarchy tree. The geographic hierarchy tree is constructed based on the set of nodes and the set of edges of the geographic hierarchy tree.

9. The address placement method according to any one of claims 1-5, further comprising: Address search service is provided using the standardized address of the address text.

10. An address placement device, comprising: A module that performs the address placement method according to any one of claims 1-9.

11. An address placement device, comprising: Memory; as well as A processor coupled to the memory, the processor being configured to execute the address placement method of any one of claims 1-9 based on instructions stored in the memory.

12. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the address placement method according to any one of claims 1-9.

Images