Address identification method, device, equipment and storage medium

Abstract

The embodiments of the present disclosure disclose an address recognition method, apparatus, device and storage medium. The method includes: identifying the longest mentioned address entity in the target text; determining at least one candidate address entity corresponding to the mentioned address entity from the knowledge graph, wherein the attribute information of each candidate address entity includes subordinate geographic information of the candidate address entity, and the subordinate geographic information is used to represent the order of subordinate relationships between the candidate address entity and each superior address entity of the candidate address entity; determining the target address entity in the target text according to the attribute information of each candidate address entity. The method can efficiently and accurately identify address entities in text.

Description

Technical Field

[0001] The embodiments of the present disclosure relate to the field of artificial intelligence technology, and in particular, to an address recognition method, apparatus, device, and storage medium. Background Art

[0002] When extracting information from text, the recognition of address information is very important. For example, in a news headline "An important event occurred in Chaoyang District", in this event detection task, it is necessary to extract the specific location of the event from the news headline. However, the recognition process of traditional address recognition technology is relatively complicated, especially when there are nested address entities in the text, its recognition efficiency is not high. Summary of the invention

[0003] The present disclosure provides an address recognition method, apparatus, device and storage medium, which can efficiently and accurately recognize address entities in text.

[0004] In a first aspect, an embodiment of the present disclosure provides an address identification method, including:

[0005] Identify the longest address entity mention in the target text;

[0006] Determine at least one candidate address entity corresponding to the mentioned address entity from the knowledge graph, wherein the attribute information of each candidate address entity includes subordinate geographic information of the candidate address entity, and the subordinate geographic information is used to indicate the subordinate relationship order between the candidate address entity and each superior address entity of the candidate address entity;

[0007] According to the attribute information of each candidate address entity, a target address entity in the target text is determined.

[0008] In a second aspect, an embodiment of the present disclosure provides an address recognition device, including:

[0009] A recognition module is used to identify the longest mention address entity in the target text;

[0010] A first determination module is used to determine at least one candidate address entity corresponding to the mentioned address entity from the knowledge graph, wherein the attribute information of each candidate address entity includes subordinate geographic information of the candidate address entity, and the subordinate geographic information is used to represent the subordinate relationship order between the candidate address entity and each superior address entity of the candidate address entity;

[0011] The second determination module is used to determine the target address entity in the target text according to the attribute information of each candidate address entity.

[0012] In a third aspect, an embodiment of the present disclosure provides an address identification device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the address identification method provided in the first aspect of the embodiment of the present disclosure are implemented.

[0013] In a fourth aspect, an embodiment of the present disclosure provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the address identification method provided in the first aspect of the embodiment of the present disclosure are implemented.

[0014] The address recognition method, apparatus, device and storage medium provided by the embodiments of the present disclosure include subordinate geographic information of each candidate address entity in the attribute information of each candidate address entity, and the subordinate geographic information is used to represent the order of subordinate relationships between the candidate address entity and each superior address entity of the candidate address entity, that is, the attribute information of each candidate address entity includes detailed and complete address information of the candidate address entity. Therefore, after recalling the candidate address entity corresponding to the mentioned address entity from the knowledge graph, the detailed and accurate target address entity can be further determined through the attribute information of each candidate address entity, thereby improving the accuracy of address recognition. At the same time, when there are nested address entities in the target text, it is not necessary to identify multiple mentioned address entities, and only the longest mentioned address entity in the target text needs to be identified, which simplifies the information processing in the address recognition process and improves the efficiency of address recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0016] Figure 1 A schematic diagram of a flow chart of an address identification method provided in an embodiment of the present disclosure;

[0017] Figure 2 Another flowchart of the address identification method provided by the embodiment of the present disclosure;

[0018] Figure 3 A schematic diagram of another flow chart of the address identification method provided in the embodiment of the present disclosure;

[0019] Figure 4 A schematic diagram of the structure of an address recognition device provided in an embodiment of the present disclosure;

[0020] Figure 5 A schematic diagram of the structure of an address recognition device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0021] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0022] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0023] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0024] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0025] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0026] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0027] In order to make the purpose, technical solution and advantages of the present disclosure more clear, the embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. It should be noted that the embodiments and features in the embodiments of the present disclosure can be combined with each other arbitrarily without conflict.

[0028] It should be noted that the execution subject of the following method embodiment may be an address recognition device, which may be implemented as part or all of an electronic device through software, hardware, or a combination of software and hardware. Optionally, the electronic device may be a client, including but not limited to a smart phone, a tablet computer, an e-book reader, and a vehicle-mounted terminal. Of course, the electronic device may also be an independent server or a server cluster, and the embodiments of the present disclosure do not limit the specific form of the electronic device. The following method embodiment is described by taking the execution subject as an electronic device as an example.

[0029] Figure 1 A flowchart of an address recognition method provided by an embodiment of the present disclosure. This embodiment relates to a specific process of how an electronic device can efficiently and accurately recognize an address entity in a target text. Figure 1 As shown, the method may include:

[0030] S101, identifying the longest mentioned address entity in the target text.

[0031] The target text refers to the text that needs to be recognized. The address entity mentioned refers to the entity used to describe the address information in the target text, such as a park, a road, a city, a district, etc.

[0032] The method for identifying the address entity mentioned in the target text may be any address entity recognition method, for example, the address entity mentioned in the target text may be recognized by using a named entity recognition (NER) technology. When the target text includes multiple nested address entities, the electronic device only needs to recognize the address entity mentioned with the longest length.

[0033] For example, suppose the target text is a piece of news, and the news is "Chaoyang District, Beijing holds an employment service event". At this time, there is a nested address entity "Chaoyang District, Beijing" in the target text. The electronic device only needs to identify the longest mention address entity "Chaoyang District, Beijing" from the target text through certain address recognition technology, and does not need to identify multiple other non-longest mention address entities in the target text, such as "Beijing" and "Chaoyang District". In this way, the information processing in the address recognition process is simplified and the efficiency of address recognition is improved.

[0034] S102: Determine at least one candidate address entity corresponding to the mentioned address entity from the knowledge graph.

[0035] The attribute information of each candidate address entity includes subordinate geographic information of the candidate address entity, and the subordinate geographic information is used to indicate the order of subordinate relationships between the candidate address entity and each superior address entity of the candidate address entity.

[0036] The knowledge graph may be an address knowledge graph, which may be a graph including all locations in the country or even the world, that is, the entity in the knowledge graph described in this embodiment refers to the address entity. The knowledge graph may include multiple address entities, as well as the relationship between two address entities, which can be described by a directional edge. For example, the entity "Shandong Province" and the entity "Qingdao City" in the knowledge graph both belong to address entities, and the attribute information of each address entity may include entity identification, entity name, and entity alias, etc. At the same time, the knowledge graph also includes an edge formed by the entity "Shandong Province" and the entity "Qingdao City", which may indicate that the entity "Qingdao City" is subordinate to the entity "Shandong Province".

[0037] In order to be able to efficiently and accurately identify address entities in the target text, this embodiment adds a new item to the attribute information of each address entity in the knowledge graph, namely, the subordinate geographic information of the address entity. The subordinate geographic information completely describes the order of subordinate relationships between the address entity and its superior address entities, which can be a text description information.

[0038] Taking the address entity "Jiaozhou City" in the knowledge graph as an example, its parent address entities may be the address entity "Qingdao City", the address entity "Shandong Province" and the address entity "China". At this time, the subordinate relationship between the address entity "Jiaozhou City" and the address entity "Qingdao City", the address entity "Shandong Province" and the address entity "China" can be described in text form to form the subordinate geographic information "Jiaozhou City, Qingdao City, Shandong Province, China" of the address entity "Jiaozhou City". Moreover, the subordinate geographic information "Jiaozhou City, Qingdao City, Shandong Province, China" exists as an item in the attribute information of the address entity "Jiaozhou City".

[0039] The above-mentioned candidate address entities refer to address entities related to the mentioned address entities in the knowledge graph. After obtaining the longest mentioned address entity in the target text, the electronic device can query the knowledge graph and use the ElasticSearch search method (wherein ElasticSearch is a Lucene-based search server for distributed full-text retrieval) to recall multiple candidate address entities from the knowledge graph. For example, if the address entity "Chaoyang" is mentioned, after querying the knowledge graph, multiple candidate address entities "Chaoyang District" and "Chaoyang City" can be recalled. In addition, the attribute information of each candidate address entity includes the subordinate geographic information of the candidate address entity.

[0040] S103: Determine a target address entity in the target text according to the attribute information of each candidate address entity.

[0041] The target address entity refers to the address entity in the final output target text. After recalling at least one candidate address entity from the knowledge graph, since the attribute information of each candidate address entity includes detailed and complete address information of the candidate address entity, the mentioned address entity can be linked to the closest candidate address entity through the attribute information of each candidate address entity, thereby obtaining the target address entity of the target text.

[0042] The address recognition method provided by the embodiment of the present disclosure includes subordinate geographic information of each candidate address entity in the attribute information of each candidate address entity, and the subordinate geographic information is used to represent the order of subordinate relationships between the candidate address entity and each superior address entity of the candidate address entity, that is, the attribute information of each candidate address entity includes detailed and complete address information of the candidate address entity. Therefore, after recalling the candidate address entity corresponding to the mentioned address entity from the knowledge graph, the detailed and accurate target address entity can be further determined through the attribute information of each candidate address entity, thereby improving the accuracy of address recognition. At the same time, when there are nested address entities in the target text, it is not necessary to identify multiple mentioned address entities, and it is only necessary to identify the longest mentioned address entity in the target text, which simplifies the information processing in the address recognition process and improves the efficiency of address recognition.

[0043] In one embodiment, a specific process for adding subordinate geographic information to the attribute information of each address entity in the knowledge graph is also provided. Based on the above embodiment, optionally, Figure 2 As shown, before the above S101, the method may further include:

[0044] S201. For each reference address entity in the knowledge graph, obtain all parent address entities of the reference address entity from the knowledge graph.

[0045] The reference address entity refers to the address entity in the knowledge graph. The reference address entities in the knowledge graph can be distinguished according to the administrative division level. The reference address entities with relatively low administrative division levels are subordinate to the reference address entities with relatively high administrative division levels, and the knowledge graph includes such subordinate relationships between reference address entities. Therefore, for each reference address entity in the knowledge graph, the parent address entity of the reference address entity can be obtained by querying the knowledge graph, and the parent address entity of the parent address entity can be queried, and so on, so as to obtain all the parent address entities of the reference address entity.

[0046] S202: Combine all upper-level address entities and the reference address entity to obtain subordinate geographic information of the reference address entity, and add the subordinate geographic information to the attribute information of the reference address entity.

[0047] Among them, after obtaining all the superior address entities of the reference address entity, all the superior address entities and the reference address entity are combined according to the subordinate relationship between the address entities, so as to obtain the subordinate geographic information of the reference address entity. At the same time, the subordinate geographic information is added as an item to the attribute information of the reference address entity, so as to convert the subordinate relationship between the reference address entity and each superior address entity of the reference address entity into a text description. Optionally, the electronic device can combine all the superior address entities with the reference address entity in the order of administrative division level from large to small to obtain the subordinate geographic information of the reference address entity. Of course, the electronic device can also combine all the superior address entities with the reference address entity in the order of administrative division level from small to large, and this embodiment does not limit this.

[0048] Exemplarily, taking the reference address entity "Jiaozhou City" in the knowledge graph as an example, the knowledge graph is queried to obtain that the parent address entity of the reference address entity "Jiaozhou City" is "Qingdao City", the parent address entity of "Qingdao City" is "Shandong Province", and the parent address entity of "Shandong Province" is "China". In this way, all parent address entities of the reference address entity "Jiaozhou City" are obtained, including "Qingdao City", "Shandong Province" and "China". Then, in order from large to small administrative division levels, "Jiaozhou City", "Qingdao City", "Shandong Province" and "China" are combined to form the subordinate geographic information of the reference address entity "Jiaozhou City" "Jiaozhou City, Qingdao City, Shandong Province, China". Since the subordinate geographic information contains the description information of all the parent address entities of the reference address entity, the subordinate geographic information describes the address information of the reference address entity "Jiaozhou City" in a relatively detailed and complete manner. Further, the subordinate geographic information "Jiaozhou City, Qingdao City, Shandong Province, China" is added to the attribute information of the reference address entity "Jiaozhou City".

[0049] In this embodiment, by querying the knowledge graph, all the parent address entities of the reference address entity in the knowledge graph are obtained, all the parent address entities are combined with the reference address entity to form the subordinate geographic information of the reference address entity, and the subordinate geographic information is added as an item to the attribute information of the reference address entity. In this way, the attribute information of each reference address entity in the knowledge graph contains detailed and complete address information of the reference address entity. When performing address recognition on the target text, the attribute information of each reference address entity in the knowledge graph can be used to determine a detailed and accurate target address entity from the target text, thereby improving the accuracy of address recognition.

[0050] In practical applications, there may be multiple nested address entities in the target text. In this case, in order to improve the address recognition efficiency of the target text, based on the above embodiment, optionally, the above S101 can be: identifying the longest mentioned address entity in the target text through a preset entity recognition network.

[0051] The entity recognition network is trained by the sample text and the longest sample address entity mentioned in the sample text. The entity recognition network can be implemented by a machine learning network, such as using a conditional random field (CRF) model as the basic model of the entity recognition network.

[0052] Since subordinate geographic information of the reference address entity is generated for each reference address entity in the knowledge graph, and the subordinate geographic information is added as an item to the attribute information of the reference address entity, that is, the attribute information of each reference address entity in the knowledge graph contains detailed and complete address information of the reference address entity, therefore, when training the entity recognition network, only the longest sample mention address entity in each sample text needs to be annotated, and there is no need to annotate the sample text multiple times. Taking the sample text "Employment service activities held in Chaoyang District, Beijing" as an example, there are nested address entities in the sample text. When annotating the address entity of the sample text, only the longest sample mention address entity "Chaoyang District, Beijing" needs to be annotated, and the address entity "Beijing" and the address entity "Chaoyang District" no longer need to be annotated, which greatly reduces the number of address entities that need to be annotated, reduces the error rate of address entity annotation, and thus improves the training efficiency and recognition effect of the entity recognition network.

[0053] After all sample texts are annotated, the sample texts are used as the input of the entity recognition network, and the longest sample address entity mentioned in the sample text is used as the expected output of the entity recognition network. The entity recognition network is trained using a preset loss function to obtain a converged entity recognition network. In this way, when address recognition is performed on the target text, the target text is input into the pre-trained entity recognition network to extract the longest address entity mentioned in the target text.

[0054] In this embodiment, the longest address entity mentioned in the target text is identified by an entity recognition network, and the entity recognition network is trained by the sample text and the longest sample address entity mentioned in the sample text. During the training process, only the longest sample address entity mentioned in each sample text needs to be annotated, and the sample text does not need to be annotated multiple times, which greatly reduces the number of address entities that need to be annotated and reduces the error rate of address entity annotation, thereby improving the training efficiency and recognition effect of the entity recognition network.

[0055] In one embodiment, an optional implementation method of determining a target address entity in a target text based on the attribute information of each candidate address entity is also provided. Figure 3 , the above S103 may include:

[0056] S301. Calculate the similarity between the mentioned address entity and the at least one candidate address entity according to the attribute information of each candidate address entity.

[0057] The attribute information of the candidate address entity may include multiple items, and the subordinate geographic information is one of the multiple items. For example, the multiple items may be the name, category, alias, and subordinate geographic information of the candidate address entity. The above category refers to the level of address entity to which the candidate address entity belongs, such as district-level address entity, prefecture-level address entity, provincial-level address entity, and national-level address entity. For example, the category of "Chaoyang District" is a district-level address entity.

[0058] Similarity can be understood as the degree of similarity between the mentioned address entity and the candidate address entity. The higher the similarity between the two, the closer the candidate address entity is to the real address entity. Therefore, the electronic device can calculate the similarity between the mentioned address entity and the candidate address entity based on the name, category, alias and subordinate geographic information of each candidate address entity.

[0059] Optionally, the electronic device may calculate the similarity between the mentioned address entity and at least one candidate address entity based on the following process. Specifically, the above S301 may include:

[0060] S3011. Generate first text information corresponding to the candidate address entity according to the multiple items.

[0061] Among them, multiple items included in the attribute information of the candidate address entity are spliced ​​to generate the first text information corresponding to the candidate address entity. For example, assuming that the multiple items included in the attribute information of the candidate address entity are name, category, and subordinate geographic information, in this way, multiple items in the attribute information of candidate address entity A are spliced, and the generated first text information A is "Chaoyang District, district-level geographic entity, Chaoyang District, Beijing, China", and multiple items in the attribute information of candidate address entity B are spliced, and the generated first text information B is "Chaoyang City, city-level geographic entity, Chaoyang City, Jilin Province, China".

[0062] S3012: Obtain second text information corresponding to the mentioned address entity.

[0063] As an optional implementation, the above-mentioned method of obtaining the second text information may be: determining the target text as the second text information corresponding to the mentioned address entity.

[0064] In this manner, the target text containing the mentioned address entity is directly determined as the second text information corresponding to the mentioned address entity.

[0065] As another optional implementation, the above method of obtaining the second text information can be: obtaining associated text that is associated with the target text; combining the target text and the associated text to obtain a combined text; and determining the combined text as the second text information corresponding to the mentioned address entity.

[0066] In this method, the associated text refers to the text that has a certain association relationship with the target text. Taking the target text as a piece of news as an example, the associated text can be the unit that publishes the news, such as Beijing Daily. In other words, the associated text of the target text can be obtained from multiple dimensions to obtain various text information related to the mentioned address entity, and the associated text is spliced ​​with the content of the target text to form a combined text, and the combined text is determined as the second text information corresponding to the mentioned address entity. Taking the target text as a piece of news, and the news is "Chaoyang District, Beijing holds an employment service event" as an example, the associated text obtained is "Beijing Daily releases", then the target text and the associated text are spliced ​​to obtain the combined text "Beijing Daily releases, Chaoyang District, Beijing holds an employment service event".

[0067] S3013: Calculate the text similarity between the first text information and the second text information, and determine the text similarity as the similarity between the mentioned address entity and the candidate address entity.

[0068] Among them, text similarity refers to the degree of similarity between the contents of two texts. Text similarity can indirectly reflect the similarity between address entities, that is, the higher the text similarity, the higher the similarity between the address entities corresponding to the two texts. Therefore, after obtaining the first text information corresponding to the candidate address entity and the second text information corresponding to the mentioned address entity, the similarity between the mentioned address entity and the candidate address entity can be determined by calculating the text similarity between the first text information and the second text information.

[0069] In a specific implementation, the first text information and the second text information can be input into a pre-trained neural network, and the text similarity between the first text information and the second text information can be predicted by the pre-trained neural network. The pre-trained neural network can be a BERT (Bidirectional Encoder Representation from Transformers) model. The pre-trained neural network is trained by the first sample text, the second sample text and the sample similarity, and the sample similarity is the similarity between the first sample text and the second sample text.

[0070] S302: Determine a target address entity in the target text according to the obtained similarities.

[0071] After obtaining the similarity between the mentioned address entity and at least one candidate address entity, the similarities can be sorted, and the target address entity in the target text can be determined based on the sorting result. Optionally, the subordinate geographic information corresponding to the candidate address entity with the highest similarity can be determined as the target address entity of the target text.

[0072] Exemplarily, assuming that the target text is "Chaoyang District holds employment service activities", the longest mentioned address entity identified from the target text is "Chaoyang District", the candidate address entity A recalled from the knowledge graph through the mentioned address entity is "Chaoyang District, Beijing", and the candidate address entity B is "Chaoyang City, Jilin Province". Based on multiple items in the attribute information of the candidate address entity A, the first text information A corresponding to the candidate address entity A is generated as "Chaoyang District, district-level geographic entity, Chaoyang District, Beijing, China". Based on multiple items in the attribute information of the candidate address entity B, the first text information B corresponding to the candidate address entity B is generated as "Chaoyang City, city-level geographic entity, Chaoyang City, Jilin Province, China". The text similarity A between the target text and the first text information A, and the text similarity B between the target text and the first text information B are calculated respectively, and the text similarity A is determined as the similarity C between the mentioned address entity and the candidate address entity A, and the text similarity B is determined as the similarity D between the mentioned address entity and the candidate address entity B. After calculation, it is found that the similarity C is higher than the similarity D. Therefore, the candidate address A can be determined as the target address entity in the target text, and the subordinate geographic information corresponding to the candidate address A can also be determined as the target address entity in the target text.

[0073] In this embodiment, the similarity between the mentioned address entity and at least one candidate address entity is calculated through the attribute information of each candidate address entity, and the target address entity in the target text is determined by comparing the similarity between the mentioned address entity and the candidate address entity, so that the determined target address entity is more accurate. At the same time, the similarity calculation between the mentioned address entity and the candidate address entity is converted into the similarity calculation between the corresponding text information. Since the similarity calculation technology between text information is relatively mature, the similarity calculation between address entities is simplified, and the efficiency and accuracy of address recognition are further improved.

[0074] Figure 4 A schematic diagram of the structure of an address recognition device provided in an embodiment of the present disclosure. Figure 4 As shown, the device may include: an identification module 401 , a first determination module 402 , and a second determination module 403 .

[0075] Specifically, the identification module 401 is used to identify the longest address mention entity in the target text;

[0076] The first determination module 402 is used to determine at least one candidate address entity corresponding to the mentioned address entity from the knowledge graph, wherein the attribute information of each candidate address entity includes subordinate geographic information of the candidate address entity, and the subordinate geographic information is used to represent the subordinate relationship order between the candidate address entity and each superior address entity of the candidate address entity;

[0077] The second determination module 403 is used to determine the target address entity in the target text according to the attribute information of each of the candidate address entities.

[0078] The address recognition device provided by the embodiment of the present disclosure includes subordinate geographic information of each candidate address entity in the attribute information of each candidate address entity. The subordinate geographic information is used to represent the order of subordinate relationships between the candidate address entity and each superior address entity of the candidate address entity, that is, the attribute information of each candidate address entity includes detailed and complete address information of the candidate address entity. Therefore, after recalling the candidate address entity corresponding to the mentioned address entity from the knowledge graph, the detailed and accurate target address entity can be further determined through the attribute information of each candidate address entity, thereby improving the accuracy of address recognition. At the same time, when there are nested address entities in the target text, it is not necessary to identify multiple mentioned address entities. It is only necessary to identify the longest mentioned address entity in the target text, which simplifies the information processing in the address recognition process and improves the efficiency of address recognition.

[0079] On the basis of the above embodiment, optionally, the device may further include: an acquisition module, a combination module and an addition module.

[0080] Specifically, the acquisition module is used to acquire all parent address entities of each reference address entity in the knowledge graph before identifying the longest mentioned address entity in the target text;

[0081] The combination module is used to combine all the upper-level address entities and the reference address entity to obtain the subordinate geographic information of the reference address entity;

[0082] The adding module is used to add the subordinate geographic information to the attribute information of the reference address entity.

[0083] On the basis of the above embodiment, optionally, the combination module is specifically used to combine all the superior address entities with the reference address entity in order of administrative division levels from large to small to obtain the subordinate geographic information of the reference address entity.

[0084] Based on the above embodiment, optionally, the recognition module 401 is specifically used to recognize the longest mention address entity in the target text through a preset entity recognition network, wherein the entity recognition network is trained through sample text and the longest sample mention address entity in the sample text.

[0085] On the basis of the above embodiment, optionally, the second determination module 403 includes: a similarity calculation unit and a determination unit.

[0086] Specifically, the similarity calculation unit is used to calculate the similarity between the mentioned address entity and the at least one candidate address entity according to the attribute information of each candidate address entity;

[0087] The determination unit is used to determine the target address entity in the target text according to the obtained similarities.

[0088] Based on the above embodiment, optionally, the attribute information of the candidate entity includes multiple items, and the subordinate geographic information is one of the multiple items;

[0089] The similarity calculation unit includes: a generation subunit, an acquisition subunit and a determination subunit.

[0090] Specifically, the generating subunit is used to generate first text information corresponding to the candidate address entity according to the multiple items;

[0091] The acquisition subunit is used to acquire the second text information corresponding to the mentioned address entity;

[0092] The determination subunit is used to calculate the text similarity between the first text information and the second text information, and determine the text similarity as the similarity between the mentioned address entity and the candidate address entity.

[0093] Based on the above embodiment, optionally, the acquisition subunit is specifically used to determine the target text as the second text information corresponding to the mentioned address entity; or,

[0094] The acquisition subunit is specifically used to acquire associated text that is associated with the target text; combine the target text and the associated text to obtain a combined text; and determine the combined text as the second text information corresponding to the mentioned address entity.

[0095] Reference below Figure 5 , which shows an electronic device (eg, Figure 1 The electronic devices in the embodiments of the present disclosure may include but are not limited to mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0096] like Figure 5 As shown, the electronic device 500 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage device 506 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 are also stored. The processing device 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0097] Typically, the following devices may be connected to the I / O interface 505: an input device 506 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 509 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 506 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 509. The communication device 509 may allow the electronic device 500 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 5The electronic device 500 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0098] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 509, or installed from the storage device 506, or installed from the ROM 502. When the computer program is executed by the processing device 501, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.

[0099] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0100] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0101] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0102] The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: obtains at least two Internet Protocol addresses; sends a node evaluation request including the at least two Internet Protocol addresses to a node evaluation device, wherein the node evaluation device selects an Internet Protocol address from the at least two Internet Protocol addresses and returns it; receives the Internet Protocol address returned by the node evaluation device; wherein the obtained Internet Protocol address indicates an edge node in a content distribution network.

[0103] Alternatively, the computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device: receives a node evaluation request including at least two Internet Protocol addresses; selects an Internet Protocol address from the at least two Internet Protocol addresses; and returns the selected Internet Protocol address; wherein the received Internet Protocol address indicates an edge node in a content distribution network.

[0104] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0105] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0106] The units involved in the embodiments described in the present disclosure may be implemented by software or hardware. The name of a unit does not limit the unit itself in some cases. For example, the first acquisition unit may also be described as a "unit for acquiring at least two Internet Protocol addresses".

[0107] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0108] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0109] In one embodiment, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0110] Identify the longest address entity mention in the target text;

[0111] Determine at least one candidate address entity corresponding to the mentioned address entity from the knowledge graph, wherein the attribute information of each candidate address entity includes subordinate geographic information of the candidate address entity, and the subordinate geographic information is used to indicate the subordinate relationship order between the candidate address entity and each superior address entity of the candidate address entity;

[0112] According to the attribute information of each candidate address entity, a target address entity in the target text is determined.

[0113] In one embodiment, when the processor executes the computer program, the following steps are also implemented: for each reference address entity in the knowledge graph, all parent address entities of the reference address entity are obtained from the knowledge graph; all parent address entities and the reference address entity are combined to obtain subordinate geographic information of the reference address entity, and the subordinate geographic information is added to the attribute information of the reference address entity.

[0114] In one embodiment, when the processor executes the computer program, the following steps are also implemented: in order of administrative division levels from large to small, all upper-level address entities are combined with the reference address entity to obtain subordinate geographic information of the reference address entity.

[0115] In one embodiment, when the processor executes the computer program, the following steps are also implemented: identifying the longest address mention entity in the target text through a preset entity recognition network, wherein the entity recognition network is trained through sample text and the longest sample address mention entity in the sample text.

[0116] In one embodiment, when the processor executes the computer program, the following steps are also implemented: calculating the similarity between the mentioned address entity and the at least one candidate address entity based on the attribute information of each candidate address entity; and determining the target address entity in the target text based on the obtained similarities.

[0117] In one embodiment, optionally, the attribute information of the candidate entity includes multiple items, and the subordinate geographic information is one of the multiple items; when the processor executes the computer program, the following steps are also implemented: based on the multiple items, generating first text information corresponding to the candidate address entity; obtaining second text information corresponding to the mentioned address entity; calculating the text similarity between the first text information and the second text information, and determining the text similarity as the similarity between the mentioned address entity and the candidate address entity.

[0118] In one embodiment, when the processor executes the computer program, the following steps are also implemented: determining the target text as the second text information corresponding to the mentioned address entity; or obtaining an associated text that has an associated relationship with the target text; combining the target text and the associated text to obtain a combined text; and determining the combined text as the second text information corresponding to the mentioned address entity.

[0119] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0120] Identify the longest address entity mention in the target text;

[0121] Determine at least one candidate address entity corresponding to the mentioned address entity from the knowledge graph, wherein the attribute information of each candidate address entity includes subordinate geographic information of the candidate address entity, and the subordinate geographic information is used to indicate the subordinate relationship order between the candidate address entity and each superior address entity of the candidate address entity;

[0122] According to the attribute information of each candidate address entity, a target address entity in the target text is determined.

[0123] The address recognition device, equipment and storage medium provided in the above embodiments can execute the address recognition method provided in any embodiment of the present disclosure, and have the corresponding functional modules and beneficial effects of executing the method. For technical details not described in detail in the above embodiments, please refer to the address recognition method provided in any embodiment of the present disclosure.

[0124] According to one or more embodiments of the present disclosure, there is provided an address identification method, comprising:

[0125] Identify the longest address entity mention in the target text;

[0126] Determine at least one candidate address entity corresponding to the mentioned address entity from the knowledge graph, wherein the attribute information of each candidate address entity includes subordinate geographic information of the candidate address entity, and the subordinate geographic information is used to indicate the subordinate relationship order between the candidate address entity and each superior address entity of the candidate address entity;

[0127] According to the attribute information of each candidate address entity, a target address entity in the target text is determined.

[0128] According to one or more embodiments of the present disclosure, an address identification method as above is provided, further comprising: for each reference address entity in the knowledge graph, obtaining all parent address entities of the reference address entity from the knowledge graph; combining all parent address entities and the reference address entity to obtain subordinate geographic information of the reference address entity, and adding the subordinate geographic information to the attribute information of the reference address entity.

[0129] According to one or more embodiments of the present disclosure, the above address identification method is provided, which also includes: combining all upper-level address entities with the reference address entity in order of administrative division level from large to small to obtain subordinate geographic information of the reference address entity.

[0130] According to one or more embodiments of the present disclosure, an address recognition method as above is provided, further comprising: identifying the longest mentioned address entity in the target text through a preset entity recognition network, wherein the entity recognition network is trained through sample text and the longest sample mentioned address entity in the sample text.

[0131] According to one or more embodiments of the present disclosure, an address recognition method as above is provided, further comprising: calculating the similarity between the mentioned address entity and the at least one candidate address entity based on the attribute information of each candidate address entity; and determining the target address entity in the target text based on the obtained similarities.

[0132] Optionally, the attribute information of the candidate entity includes multiple items, and the subordinate geographic information is one of the multiple items; according to one or more embodiments of the present disclosure, an address recognition method as above is provided, further comprising: generating first text information corresponding to the candidate address entity based on the multiple items; obtaining second text information corresponding to the mentioned address entity; calculating the text similarity between the first text information and the second text information, and determining the text similarity as the similarity between the mentioned address entity and the candidate address entity.

[0133] According to one or more embodiments of the present disclosure, an address recognition method as above is provided, further comprising: determining the target text as the second text information corresponding to the mentioned address entity; or, obtaining an associated text that has an associated relationship with the target text; combining the target text and the associated text to obtain a combined text; and determining the combined text as the second text information corresponding to the mentioned address entity.

[0134] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.

[0135] In addition, although each operation is described in a specific order, this should not be understood as requiring these operations to be performed in the specific order shown or in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0136] Although the subject matter has been described in language specific to structural features and / or methodological logical actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims.

Claims

1. An address recognition method, characterized in that: include: Identify the longest address entity mention in the target text; Determine at least one candidate address entity corresponding to the mentioned address entity from the knowledge graph, wherein the attribute information of each candidate address entity includes subordinate geographic information of the candidate address entity, the subordinate geographic information is used to indicate the subordinate relationship order between the candidate address entity and each superior address entity of the candidate address entity, and the attribute information of each candidate address entity includes multiple items, and the subordinate geographic information is one of the multiple items; Generating first text information corresponding to the candidate address entity according to the multiple items; Based on the target text, obtaining second text information corresponding to the mentioned address entity; Calculating text similarity between the first text information and the second text information, and determining the text similarity as the similarity between the mentioned address entity and the candidate address entity; According to the obtained similarities, a target address entity in the target text is determined.

2. The method according to claim 1, characterized in that Before identifying the longest mentioned address entity in the target text, the method further includes: For each reference address entity in the knowledge graph, obtain all parent address entities of the reference address entity from the knowledge graph; All upper-level address entities and the reference address entity are combined to obtain subordinate geographic information of the reference address entity, and the subordinate geographic information is added to the attribute information of the reference address entity.

3. The method according to claim 2, characterized in that The combining of all the upper-level address entities and the reference address entity to obtain the subordinate geographic information of the reference address entity includes: In descending order of administrative division levels, all superior address entities are combined with the reference address entity to obtain subordinate geographic information of the reference address entity.

4. The method according to claim 1, characterized in that The identifying the longest mentioned address entity in the target text includes: The longest address mention entity in the target text is identified by a preset entity recognition network, wherein the entity recognition network is trained by a sample text and the longest sample address mention entity in the sample text.

5. The method according to claim 1, characterized in that The acquiring the second text information corresponding to the mentioned address entity includes: Determine the target text as the second text information corresponding to the mentioned address entity; or, Acquire associated text that is associated with the target text; combine the target text and the associated text to obtain a combined text; and determine the combined text as the second text information corresponding to the mentioned address entity.

6. An address recognition device, characterized in that: include: A recognition module is used to identify the longest address entity mention in the target text; A first determination module is used to determine at least one candidate address entity corresponding to the mentioned address entity from the knowledge graph, wherein the attribute information of each candidate address entity includes subordinate geographic information of the candidate address entity, and the subordinate geographic information is used to represent the subordinate relationship order between the candidate address entity and each superior address entity of the candidate address entity, and the attribute information of each candidate address entity includes multiple items, and the subordinate geographic information is one of the multiple items; The second determination module is used to generate the first text information corresponding to the candidate address entity based on the multiple items; obtain the second text information corresponding to the mentioned address entity based on the target text; calculate the text similarity between the first text information and the second text information, and determine the text similarity as the similarity between the mentioned address entity and the candidate address entity; determine the target address entity in the target text based on the obtained similarities.

7. An address recognition device, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images