CN111858787A - POI information acquisition method and device - Google Patents

POI information acquisition method and device Download PDF

Info

Publication number
CN111858787A
CN111858787A CN201910907458.8A CN201910907458A CN111858787A CN 111858787 A CN111858787 A CN 111858787A CN 201910907458 A CN201910907458 A CN 201910907458A CN 111858787 A CN111858787 A CN 111858787A
Authority
CN
China
Prior art keywords
target
named entity
poi
poi information
information
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201910907458.8A
Other languages
Chinese (zh)
Inventor
陶言祺
陈欢
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Didi Infinity Technology and Development Co Ltd
Original Assignee
Beijing Didi Infinity Technology and Development Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Didi Infinity Technology and Development Co Ltd filed Critical Beijing Didi Infinity Technology and Development Co Ltd
Priority to CN201910907458.8A priority Critical patent/CN111858787A/en
Publication of CN111858787A publication Critical patent/CN111858787A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/29—Geographical information databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/906—Clustering; Classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Remote Sensing (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The application provides a method and a device for obtaining POI information; the method comprises the following steps: acquiring retrieval information input by a user through a client, inputting the retrieval information into a pre-trained named entity recognition model, and recognizing a named entity in the retrieval information; if the identified named entities comprise target named entities used for representing geographic positions, determining a target retrieval area according to the target named entities; the target retrieval area takes POI information corresponding to the target named entity as a center; and retrieving and acquiring at least one piece of target POI information corresponding to the retrieval information from the target retrieval area according to other named entities except the target named entity in the named entities. The method can acquire the target POI information by identifying the category of the entity in the retrieval information and searching step by step through the identified target named entity and other named entities which can represent the geographic position, thereby improving the accuracy rate of POI information acquisition.

Description

POI information acquisition method and device
Technical Field
The application relates to the technical field of information retrieval, in particular to a method and a device for obtaining POI information.
Background
Information retrieval occupies an important position in many fields. In many application scenarios, a plurality of search results are obtained and displayed for user search according to search information input by a user. For example, in the field of online car booking, a user inputs retrieval information for describing a destination of the user through a human-computer interaction interface of a client, and the client of the client acquires the POI according to the retrieval information input by the user. In the database, the storage form of a point of interest (POI) is generally: the name of the building (or target site) is associated with the geographic location coordinates corresponding to the name. Generally, the retrieval result is obtained by matching the key words in the retrieval information input by the user with each POI in a certain area range stored in the database, and obtaining the POI successfully matched as the retrieval result. The keywords are generally named entities in the search information, and the named entities mainly include names of people, places, organizations, proper nouns, and the like. The named entity type of a POI is typically a place name, organization name, or proper noun.
In practice, however, the search information entered by the user at the time of searching will typically include two or more named entities; when a retrieval result corresponding to retrieval information is obtained in a current retrieval strategy, the potential meaning of the retrieval information often cannot be excavated, the obtained retrieval result often is not a result which a user wants to really obtain, and the POI obtaining accuracy is low.
Disclosure of Invention
In view of this, an object of the present application is to provide a method and an apparatus for acquiring POI information, which are capable of acquiring target POI information by identifying categories of entities in search information and performing step-by-step search through identified target named entities and other named entities that can represent geographic locations, so as to improve accuracy of POI information acquisition.
In a first aspect, a method for obtaining POI information is provided, where the method includes:
acquiring retrieval information input by a user through a client, inputting the retrieval information into a pre-trained named entity recognition model, and recognizing a named entity in the retrieval information;
if the identified named entities comprise target named entities used for representing geographic positions, determining a target retrieval area according to the target named entities; the target retrieval area takes POI information corresponding to the target named entity as a center;
and retrieving and acquiring at least one piece of target POI information corresponding to the retrieval information from the target retrieval area according to other named entities except the target named entity in the named entities.
In an alternative embodiment, the named entity recognition model is trained in the following way:
Obtaining a plurality of sample texts and a label labeling sequence corresponding to each sample text; the label labeling sequence comprises a plurality of entity type labeling labels which are in one-to-one correspondence with a plurality of character strings forming the sample text;
for each sample text, inputting the sample text into a conditional random field model, and determining a score corresponding to each label prediction sequence in a plurality of label prediction sequences corresponding to the sample text; the tag annotation sequence is one of a plurality of the tag prediction sequences;
and training the conditional random field model according to the score to obtain the named entity recognition model.
In an alternative embodiment, the determining a score corresponding to each of a plurality of predicted tag sequences corresponding to the sample text includes:
based on the conditional random field model, aiming at each label prediction sequence, determining the transition probability corresponding to each two adjacent prediction labels in the label prediction sequence according to a probability transition matrix;
and determining the score corresponding to the label prediction sequence according to the transition probability corresponding to the prediction labels adjacent to each two positions in the label prediction sequence.
In an alternative embodiment, said training said conditional random field model based on said score comprises:
determining model loss according to the score corresponding to each label prediction sequence and the score of the label prediction sequence corresponding to the label labeling sequence;
adjusting parameters of the conditional random field model according to the model loss to complete the current round of training of the conditional random field model;
and determining the conditional random field model after multiple rounds of training as the named entity recognition model.
In an optional implementation, the determining a target retrieval region according to the target named entity includes:
retrieving according to the target named entity and acquiring at least one piece of POI information corresponding to the target named entity; the POI information comprises a POI name and geographical position information;
aiming at each piece of POI information, determining an area range corresponding to the POI information by taking geographical position information in the POI information as a center;
and determining the area range corresponding to each POI information as the target retrieval area.
In an optional implementation manner, the retrieving and acquiring at least one piece of POI information corresponding to the target named entity according to the target named entity includes:
Acquiring a plurality of candidate POI information related to the target named entity from a POI database by taking the target named entity as a retrieval key word;
determining the relevance of each alternative POI information and the target named entity;
and selecting at least one piece of POI information corresponding to the target named entity from the candidate POI information according to the degree of correlation between each piece of candidate POI information and the target named entity.
In an optional implementation manner, the determining the relevance of each candidate POI information to the target named entity includes:
determining the relevance of each candidate POI information and the target named entity based on the current position information of the client and the geographic position information in each candidate POI information;
alternatively, the first and second electrodes may be,
and determining the relevance of each piece of candidate POI information and the target named entity based on the target named entity and the number of the same characters in the POI names in each piece of candidate POI.
In an optional implementation manner, the retrieving and acquiring at least one piece of POI information corresponding to the target named entity according to the target named entity includes:
acquiring a plurality of candidate POI information related to the target named entity from a POI database by taking the target named entity as a retrieval key word;
Clustering the alternative POI information according to the geographical position information in the alternative POI information to form a plurality of POI groups; the distance between any one piece of alternative POI information in each POI group and at least one piece of other POI information in the POI group is not more than a preset distance threshold value;
and aiming at each POI group, determining POI information corresponding to the target named entity according to each alternative information in the POI group.
In an optional implementation manner, the determining, for each piece of alternative POI information included in each alternative POI group, POI information corresponding to the target named entity includes:
for each alternative POI group, if the number of alternative POI information in the alternative POI group is 1, determining the alternative POI information in the alternative POI group as the POI information corresponding to the target named entity;
if the number of the alternative POI information in the alternative POI group is larger than 1, determining the central POI information corresponding to the alternative POI group according to the geographical location information in each alternative POI information in the alternative POI group, and determining the central POI information as the POI information corresponding to the target named entity.
In an optional implementation manner, before determining the target retrieval region according to the target named entity, the method further includes:
Judging whether other named entities except the target named entity are included in the identified named entities;
the determining a target retrieval region according to the target named entity includes:
when the named entities which are identified comprise other named entities except the target named entity, determining the target retrieval area according to the target named entity;
and when the identified named entities do not comprise other named entities except the target named entity, searching according to the target named entity to obtain at least one piece of target POI information corresponding to the search information.
In a second aspect, there is provided a point of interest POI information acquisition apparatus, including:
the system comprises an identification module, a search module and a search module, wherein the identification module is used for acquiring search information input by a user through a client, inputting the search information into a pre-trained named entity identification model and identifying a named entity in the search information;
the determining module is used for determining a target retrieval area according to a target named entity when the identified named entity comprises the target named entity for representing the geographic position; the target retrieval area takes POI information corresponding to the target named entity as a center;
And the retrieval module is used for retrieving and acquiring at least one piece of target POI information corresponding to the retrieval information from the target retrieval area according to other named entities except the target named entity in the named entities.
In an alternative embodiment, the method further comprises: the model training module is used for training the named entity recognition model in the following way:
obtaining a plurality of sample texts and a label labeling sequence corresponding to each sample text; the label labeling sequence comprises a plurality of entity type labeling labels which are in one-to-one correspondence with a plurality of character strings forming the sample text;
for each sample text, inputting the sample text into a conditional random field model, and determining a score corresponding to each label prediction sequence in a plurality of label prediction sequences corresponding to the sample text; the tag annotation sequence is one of a plurality of the tag prediction sequences;
and training the conditional random field model according to the score to obtain the named entity recognition model.
In an optional implementation manner, the model training module is configured to determine a score corresponding to each of a plurality of predicted tag sequences corresponding to the sample text by:
Based on the conditional random field model, aiming at each label prediction sequence, determining the transition probability corresponding to each two adjacent prediction labels in the label prediction sequence according to a probability transition matrix;
and determining the score corresponding to the label prediction sequence according to the transition probability corresponding to the prediction labels adjacent to each two positions in the label prediction sequence.
In an alternative embodiment, the model training module is configured to train the conditional random field model based on the score by:
determining model loss according to the score corresponding to each label prediction sequence and the score of the label prediction sequence corresponding to the label labeling sequence;
adjusting parameters of the conditional random field model according to the model loss to complete the current round of training of the conditional random field model;
and determining the conditional random field model after multiple rounds of training as the named entity recognition model.
In an optional implementation manner, the determining module is configured to determine a target retrieval area according to the target named entity by using the following manner:
retrieving according to the target named entity and acquiring at least one piece of POI information corresponding to the target named entity; the POI information comprises a POI name and geographical position information;
Aiming at each piece of POI information, determining an area range corresponding to the POI information by taking geographical position information in the POI information as a center;
and determining the area range corresponding to each POI information as the target retrieval area.
In an optional implementation manner, the determining module is configured to retrieve and obtain at least one piece of POI information corresponding to the target named entity according to the target named entity by using the following manners:
acquiring a plurality of candidate POI information related to the target named entity from a POI database by taking the target named entity as a retrieval key word;
determining the relevance of each alternative POI information and the target named entity;
and selecting at least one piece of POI information corresponding to the target named entity from the candidate POI information according to the degree of correlation between each piece of candidate POI information and the target named entity.
In an optional embodiment, the determining module is configured to determine the relevance of each candidate POI information to the target named entity by using the following method:
determining the relevance of each candidate POI information and the target named entity based on the current position information of the client and the geographic position information in each candidate POI information;
Alternatively, the first and second electrodes may be,
and determining the relevance of each piece of candidate POI information and the target named entity based on the target named entity and the number of the same characters in the POI names in each piece of candidate POI.
In an optional implementation manner, the determining module is configured to retrieve and obtain at least one piece of POI information corresponding to the target named entity according to the target named entity by using the following manners:
acquiring a plurality of candidate POI information related to the target named entity from a POI database by taking the target named entity as a retrieval key word;
clustering the alternative POI information according to the geographical position information in the alternative POI information to form a plurality of POI groups; the distance between any one piece of alternative POI information in each POI group and at least one piece of other POI information in the POI group is not more than a preset distance threshold value;
and aiming at each POI group, determining POI information corresponding to the target named entity according to each alternative information in the POI group.
In an optional embodiment, the determining module is configured to determine, for each alternative POI information included in each alternative POI group, POI information corresponding to the target named entity by using the following method:
For each alternative POI group, if the number of alternative POI information in the alternative POI group is 1, determining the alternative POI information in the alternative POI group as the POI information corresponding to the target named entity;
if the number of the alternative POI information in the alternative POI group is larger than 1, determining the central POI information corresponding to the alternative POI group according to the geographical location information in each alternative POI information in the alternative POI group, and determining the central POI information as the POI information corresponding to the target named entity.
In an alternative embodiment, the method further comprises: a detection module, configured to determine whether the named entities identified include other named entities except the target named entity before determining a target retrieval region according to the target named entity;
the determining module is specifically configured to:
when the named entities which are identified comprise other named entities except the target named entity, determining the target retrieval area according to the target named entity;
and when the identified named entities do not comprise other named entities except the target named entity, searching according to the target named entity to obtain at least one piece of target POI information corresponding to the search information.
In a third aspect, an embodiment of the present application further provides a computer device, including: a processor, a memory and a bus, the memory storing machine-readable instructions executable by the processor, the processor and the memory communicating via the bus when the computer device is running, the machine-readable instructions when executed by the processor performing the steps of the first aspect described above, or any possible implementation of the first aspect.
In a fourth aspect, this embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and the computer program is executed by a processor to perform the steps in the first aspect described above or any possible implementation manner of the first aspect.
According to the embodiment of the application, through the identification of the named entity in the retrieval information, a target retrieval area is determined according to a target named entity representing the geographic position, and the target retrieval area is an area taking POI (point of interest) corresponding to the target named entity as a center; and then, searching and acquiring at least one piece of target POI information from the target searching region by other named entities except the target named entity in the identified named entities, so that the problems that the acquisition accuracy of the searching result is low, the acquisition of the searching result is less and even the searching result cannot be acquired when the searching information comprises two or more named entities can be solved, and the effects of improving the acquisition accuracy of the searching result and increasing the acquisition of the searching result are achieved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.
Fig. 1 is a schematic diagram illustrating an architecture of a service system provided in an embodiment of the present application;
fig. 2 shows a flowchart of a POI information acquisition method provided in an embodiment of the present application;
fig. 3 is a flowchart illustrating a specific method for training a named entity recognition model in the POI information acquisition method according to the embodiment of the present application;
fig. 4 is a flowchart illustrating a specific method for determining a score corresponding to each predicted sequence of tags in the POI information acquisition method according to the embodiment of the present application;
fig. 5 is a flowchart illustrating a specific method for determining a target search area according to a target named entity in the POI information acquisition method according to the embodiment of the present application;
fig. 6 is a schematic structural diagram illustrating a POI information acquiring apparatus according to an embodiment of the present application;
Fig. 7 shows a schematic structural diagram of a computer device provided in an embodiment of the present application.
Detailed Description
In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it should be understood that the drawings in the present application are for illustrative and descriptive purposes only and are not used to limit the scope of protection of the present application. Additionally, it should be understood that the schematic drawings are not necessarily drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of the present application. It should be understood that the operations of the flow diagrams may be performed out of order, and steps without logical context may be performed in reverse order or simultaneously. One skilled in the art, under the guidance of this application, may add one or more other operations to, or remove one or more operations from, the flowchart.
In addition, the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present application without making any creative effort, shall fall within the protection scope of the present application.
To enable those skilled in the art to utilize the present disclosure, the following embodiments are presented in conjunction with a specific application scenario, "network appointment". It will be apparent to those skilled in the art that the general principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope of the application. Although the present application is described primarily in the context of retrieving information for an input networked car booking platform to obtain POIs corresponding to the retrieved information, it should be appreciated that this is merely one exemplary embodiment. The embodiment of the application can also be used in other fields, such as POI position inquiry based on map software.
It should be noted that in the embodiments of the present application, the term "comprising" is used to indicate the presence of the features stated hereinafter, but does not exclude the addition of further features.
One aspect of the application relates to a POI information acquisition system, through the recognition of named entities in retrieval information, firstly, a target retrieval area is determined according to a target named entity representing a geographic position, and the target retrieval area is an area taking POI corresponding to the target named entity as a center; and then, searching and acquiring at least one piece of target POI information from the target searching region by other named entities except the target named entity in the identified named entities, so that the problems that the acquisition accuracy of the searching result is low, the acquisition of the searching result is less and even the searching result cannot be acquired when the searching information comprises two or more named entities can be solved, and the effects of improving the acquisition accuracy of the searching result and increasing the acquisition of the searching result are achieved.
It is worth noting that before the application is filed, when the retrieval information input by the user comprises two or more named entities, the retrieval result obtained based on the current retrieval strategy has the problems of low obtaining accuracy rate and less obtained retrieval results. For example, the search information input by the user is "wiegano kendyi", wherein two named entities of "wiegano" and "kendyi" are included; the search result obtained by the current keyword search method is the kendyn of the whole area or the wiya building, and the real intention of the user is the kendyn around the 'wiya building'. For another example, the retrieval information input by the user is 'new rural circular express delivery', wherein two keywords of 'new rural' and 'circular express delivery' are included, the retrieval result obtained by the current keyword retrieval mode is a whole-area new rural area or a whole-area circular express delivery, and the real intention of the user is a circular express delivery near the new rural area.
Fig. 1 is a schematic architecture diagram of a service system 100 for acquiring POI information according to an embodiment of the present disclosure. For example, the service system 100 may be an online transportation service platform for transportation services such as taxi cab, designated drive service, express, carpool, bus service, driver rental, or shift service, or any combination thereof, and may also be a map navigation platform. The service system 100 may include one or more of a server 110, a network 120, a client 130, and a database 140.
In some embodiments, the server 110 may include a processor. The processor may process information and/or data related to the service request to perform one or more of the functions described herein. For example, the processor may perform POI retrieval based on retrieval information obtained from the client 130, and acquire a target POI corresponding to the retrieval information. In some embodiments, a processor may include one or more processing cores (e.g., a single-core processor (S) or a multi-core processor (S)). Merely by way of example, a Processor may include a Central Processing Unit (CPU), an Application Specific Integrated Circuit (ASIC), an Application specific Instruction Set Processor (ASIP), a Graphics Processing Unit (GPU), a Physical Processing Unit (PPU), a Digital Signal Processor (DSP), a Field Programmable Gate Array (FPGA), a Programmable Logic Device (PLD), a controller, a microcontroller Unit, a Reduced Instruction Set computer (Reduced Instruction Set Computing, RISC), a microprocessor, or the like, or any combination thereof.
In some embodiments, the device type corresponding to the client 130 may be a mobile device, such as a smart home device, a wearable device, a smart mobile device, a virtual reality device, an augmented reality device, or the like, or may be a tablet computer, a laptop computer, or a built-in device in a motor vehicle, or the like.
In some embodiments, a database 140 may be connected to network 120 to communicate with one or more components (e.g., server 110, client 130, etc.) in service system 100. One or more components in service system 100 may access data or instructions stored in database 140 via network 120. In some embodiments, the database 140 may be directly connected to one or more components in the service system 100, or the database 140 may be part of the server 110.
The following describes in detail a POI information acquisition method provided in an embodiment of the present application with reference to the content described in the service system 100 shown in fig. 1. It should be noted that the method can be performed by both the client 130 and the server 110.
Referring to fig. 2, a schematic flow chart of a POI information acquisition method provided in the embodiment of the present application is shown, where the method includes:
S201: the method comprises the steps of obtaining retrieval information input by a user through a client, inputting the retrieval information into a pre-trained named entity recognition model, and recognizing a named entity in the retrieval information.
S202: if the identified named entities comprise target named entities used for representing geographic positions, determining a target retrieval area according to the target named entities; and the target retrieval area takes POI information corresponding to the target named entity as a center.
S203: and retrieving and acquiring at least one piece of target POI information corresponding to the retrieval information from the target retrieval area according to other named entities except the target named entity in the named entities.
The following describes the above-described steps S201 to S203.
I: in the above S201, in a specific implementation, the retrieval information is generally input by the user through a human-computer interaction interface of the client; in some cases, the search information may be obtained through other means. Taking the example that a user inputs retrieval information through a human-computer interaction interface, if the execution main body of the POI information acquisition method is a server, the client sends the retrieval information to the server based on the connection between the client and the server after receiving the retrieval information input by the user through the human-computer interaction interface; after receiving the retrieval information, the server inputs the retrieval information into a named entity identification model embedded in the server, identifies the named entity in the retrieval information, and then retrieves the POI based on the identified named entity. If the execution main body of the POI information acquisition method is the client side, the client side inputs the retrieval information into the named entity identification model embedded in the client side after receiving the retrieval information input by the user, identifies the named entity in the retrieval information, and then sends the identified named entity to the server to realize the retrieval of the POI information.
Specifically, referring to fig. 3, an embodiment of the present application provides a specific method for training a named entity recognition model, including:
s301: obtaining a plurality of sample texts and a label labeling sequence corresponding to each sample text; the label labeling sequence comprises a plurality of entity type labeling labels which are in one-to-one correspondence with a plurality of character strings forming the sample text.
In a specific implementation, the sample text is generally history retrieval information; each historical retrieval information comprises at least one named entity. The sample text may also include names of buildings (or target places) for different POIs retrieved from the POI database.
Sample text is generally composed of at least one string; each character string forming the sample text can be formed by only one character, namely each character independently forms one character string; or may be composed of at least one character, that is, a character string is composed according to the characters constituting the entity.
The label labeling sequence corresponding to each sample text is a sequence formed by entity types labeled for each character in each sample text. In the present application, entity types generally include: geographical location, other entities except geographical location, non-named entities; the other instance types except the geographic position can be divided in more detail according to actual needs. Under different conditions, corresponding entity types may be different for the same entity; for example, the sample text is "wiya building kendyi", when the entity type of "wiya building" is geographic location; if the sample text is "wimedia building in the middle guan village," the entity type of "wimedia building" is other than the geographical location.
The label labeling sequence corresponding to each sample text, and a plurality of entity type labeling labels which are in one-to-one correspondence with a plurality of character strings forming the sample text; for example, if only one character is included in each of the character strings constituting the sample text, for example, if the entity type "geographical location" is denoted by a, the entity type "other than geographical location" is denoted by b; the entity type "non-named entity" is denoted by c, and for the sample text "viia mansion kendyki", the corresponding tag label sequence is: a. a, b. When each character string constituting the sample text is composed of at least one character, the sample text "viia mansion kendyi" in which "viia", "mansion" and "kendyi" constitute one character string, respectively, and the corresponding label mark sequences are a, and b.
S302: for each sample text, inputting the sample text into a conditional random field model, and determining a score corresponding to each label prediction sequence in a plurality of label prediction sequences corresponding to the sample text; the tag annotation sequence is one of a plurality of the tag prediction sequences.
S303: and training the conditional random field model according to the score to obtain the named entity recognition model.
In specific implementation, if all character strings forming the sample text comprise one character, directly descending and inputting the sample text into the conditional random field model; if the character string forming the sample text comprises a plurality of characters, before each sample text is input into the conditional random field model, word segmentation processing is carried out on the sample text, each entity in the sample text is predicted by using the conditional random field model with each word segmentation as a unit, and the conditional random field model is trained on the basis of a prediction result and a label labeling sequence.
Specifically, the plurality of tag prediction sequences corresponding to the sample text refers to all tag sequences that may be obtained from the current sample text.
For example, if the types of named entities include: the geographic location, other categories than geographic location, and non-target named entities are denoted A, B and C, respectively. For the sample text, after the sample text is participled, the resulting character string includes: s1, S2, and S3. The predicted sequence of labels corresponding to the sample text is: the total of 8 types of (A, A, A), (A, A, B), (A, B, A), (A, B, B), (B, A, A), (B, A, B), (B, B, A), (B, B, B). Similarly, if the sample text is segmented to obtain 3 character strings, and there are 5 categories of named entities, the possible tag prediction sequences include: 3 5And (4) seed preparation.
Referring to fig. 4, an embodiment of the present application further provides a specific method for determining a score corresponding to each of a plurality of predicted tag sequences corresponding to the sample text, where the specific method includes:
s401: and determining the transition probability corresponding to each two adjacent predicted labels in the label prediction sequence according to the probability transition matrix aiming at each label prediction sequence based on the conditional random field model.
S402: and determining the score corresponding to the label prediction sequence according to the transition probability corresponding to the prediction labels adjacent to each two positions in the label prediction sequence.
In one implementation, a probability transition matrix A exists in the conditional random field model,
Figure BDA0002213697680000111
each element in the probability transition matrix, representing a label ygTransfer label yhWherein g is 1, 2, … …, nclass；h＝1，2，……，nclass，nclassThe number of named entity types.
That is, Agh＝p(yt＝yh|yt-1＝yg)。
For example, if there are 5 classes, the resulting probability transition matrix is:
Figure BDA0002213697680000121
here, the probability transition matrix may be obtained in advance according to the crawled corpus, or may be obtained through initialization, and in the process of training the conditional random field model, the probability transition matrix is also used as a target for parameter adjustment.
For the input ith sample text Xi＝(xi1，xi2，……，xim) Wherein m is the number of character strings in the sample text; corresponding tag prediction sequence Yi＝(yi1,yi2,……，yim) The score of (a) is:
Figure BDA0002213697680000122
after the score corresponding to each label prediction sequence is determined, the model loss can be determined according to the score corresponding to each label prediction sequence and the score of the label prediction sequence corresponding to the label marking sequence; adjusting parameters of the conditional random field model according to the model loss to complete the current round of training of the conditional random field model; and determining the conditional random field model after multiple rounds of training as the named entity recognition model.
In addition, other models can be used to identify named entities in the search information.
II: in S202, after identifying each named entity in the search information based on S201, it is first determined whether the identified named entities include a target named entity for characterizing a geographic location. If it is determined that the identified named entity includes the target named entity, the target search area may be determined based on the target named entity.
Specifically, referring to fig. 5, an embodiment of the present application provides a specific method for determining a target search area according to a target named entity, including:
S501: retrieving according to the target named entity and acquiring at least one piece of POI information corresponding to the target named entity; the POI information comprises a POI name and geographical position information.
In a specific implementation, the at least one piece of POI information corresponding to the target named entity may be determined in any one of the following two manners:
(1): acquiring a plurality of candidate POI information related to the target named entity from a POI database by taking the target named entity as a retrieval key word; determining the relevance of each alternative POI information and the target named entity; and selecting at least one piece of POI information corresponding to the target named entity from the candidate POI information according to the degree of correlation between each piece of candidate POI information and the target named entity.
Here, there are various ways to determine the relevance between the candidate POI information and the target named entity, for example, determining the distance between the POI corresponding to each candidate POI information and the user side according to the geographic location information in the candidate POI information and the current location information of the client, and then determining the relevance between each candidate POI information and the target named entity based on the distance. The smaller the distance between the POI corresponding to the alternative POI information and the user side is, the larger the correlation degree is correspondingly; the larger the distance, the smaller the correlation accordingly.
In addition, the relevance between each piece of candidate POI information and the target naming entity may also be determined based on the number of characters in the target naming entity that are the same as the number of characters in the POI name in each piece of candidate POI. The greater the number of identical characters, the greater the degree of correlation.
After the degree of correlation between each piece of alternative POI information and the target named entity is determined, a preset number of alternative POI information can be selected from the alternative POI information as POI information corresponding to the target named entity according to the degree of correlation between each piece of POI information and the target named entity; or selecting POI information with the relevance larger than a preset threshold value from the alternative POI information as POI information corresponding to the target named entity; or a certain amount of POI information may be selected from the candidate POI information as the POI information corresponding to the target named entity according to the percentage of the amount.
(2): acquiring a plurality of candidate POI information related to the target named entity from a POI database by taking the target named entity as a retrieval key word; clustering the alternative POI information according to the geographical position information in the alternative POI information to form a plurality of POI groups; the distance between any one piece of alternative POI information in each POI group and at least one piece of other POI information in the POI group is not more than a preset distance threshold value; and aiming at each POI group, determining POI information corresponding to the target named entity according to each alternative information in the POI group.
Here, when candidate POIs are obtained according to a target named entity, there is also a great correlation between candidate POI information obtained in many cases, for example, the target named entity includes: "beijing south station", the corresponding POI information includes: "Beijing south station subway station", "Beijing south station south square", "Beijing south station north entrance", "Beijing south station east entrance", "Beijing south station west entrance", and the like ". The relevance between the candidate POI information and the target named entity is large, and POIs corresponding to the candidate POI information are distributed in a small area range.
In order to avoid overlapping of different target retrieval areas determined according to different POIs corresponding to a target named entity and further cause repeated retrieval in the target retrieval area, in the embodiment of the application, after a plurality of pieces of alternative POI information related to the target named entity are acquired from a POI database by taking the target named entity as a key word, the alternative POI information is clustered according to geographical position information carried in the alternative POI information to form a plurality of POI groups.
Then, for each alternative POI group, if the number of alternative POI information in the alternative POI group is 1, determining the alternative POI information in the alternative POI group as the POI information corresponding to the target named entity;
If the number of the alternative POI information in the alternative POI group is larger than 1, determining the central POI information corresponding to the alternative POI group according to the geographical location information in each alternative POI information in the alternative POI group, and determining the central POI information as the POI information corresponding to the target named entity.
Here, a central geographical location information may be determined first according to the geographical location information in each candidate POI information in the candidate POI group, and then a POI information closest to the central geographical location information may be determined as the central POI information according to the central geographical location information.
S502: and aiming at each piece of POI information, determining an area range corresponding to the POI information by taking the geographical position information in the POI information as a center.
S503: and determining the area range corresponding to each POI information as the target retrieval area.
Here, after the POI information corresponding to the target named entity is determined, an area range is determined as a target area range corresponding to the POI information centering on the geographical location information in the POI information.
Here, the shape of the target area range may be set according to actual needs, and is determined to be, for example, a circle, a rectangle, or the like.
III: in the above S203, after the target area is determined, at least one piece of target POI information can be obtained by retrieving and searching from the POI database according to other named entities as keywords except the target named entity in the named entities.
In addition, it should be noted that, in practice, when searching is performed based on the target named entity, there may be a case where the number of obtained candidate POI information is 0, and at this time, since the target search area cannot be determined according to the target named entity, at this time, the named entity identified from the search information may be directly used as a keyword to perform searching, so as to obtain the target POI information corresponding to the search information.
In addition, in another embodiment, when the named entities identified do not include other named entities except the target named entity, the target named entity is searched to obtain at least one piece of target POI information corresponding to the search information.
According to the embodiment of the application, through the identification of the named entity in the retrieval information, a target retrieval area is determined according to a target named entity representing the geographic position, and the target retrieval area is an area taking POI (point of interest) corresponding to the target named entity as a center; and then, according to other named entities except the target named entity in the identified named entities, at least one piece of target POI information is retrieved and obtained from the target retrieval area, so that the problems that when the retrieval information comprises two or more named entities, the retrieval result obtaining accuracy is low, the retrieval result obtaining is less, and even the retrieval result cannot be obtained can be solved, the retrieval result obtaining accuracy is improved, and the retrieval result obtaining effect is increased.
Based on the same inventive concept, a POI information device corresponding to POI information acquisition is also provided in the embodiments of the present application, and since the principle of solving the problem of the device in the embodiments of the present application is similar to the POI information acquisition method in the embodiments of the present application, the implementation of the device may refer to the implementation of the method, and repeated details are not repeated.
Referring to fig. 6, which is a schematic diagram of a POI information acquiring apparatus provided in an embodiment of the present application, the apparatus includes: an identification module 61, a determination module 62, and a retrieval module 63; wherein:
the identification module 61 is used for acquiring retrieval information input by a user through a client, inputting the retrieval information into a pre-trained named entity identification model, and identifying a named entity in the retrieval information;
a determining module 62, configured to determine a target retrieval area according to a target named entity when the identified named entity includes the target named entity for representing the geographic location; the target retrieval area takes POI information corresponding to the target named entity as a center;
and a retrieval module 63, configured to retrieve and obtain at least one piece of target POI information corresponding to the retrieval information from the target retrieval area according to a named entity other than the target named entity.
According to the embodiment of the application, through the identification of the named entity in the retrieval information, a target retrieval area is determined according to a target named entity representing the geographic position, and the target retrieval area is an area taking POI (point of interest) corresponding to the target named entity as a center; and then, according to other named entities except the target named entity in the identified named entities, at least one piece of target POI information is retrieved and obtained from the target retrieval area, so that the problems that when the retrieval information comprises two or more named entities, the retrieval result obtaining accuracy is low, the retrieval result obtaining is less, and even the retrieval result cannot be obtained can be solved, the retrieval result obtaining accuracy is improved, and the retrieval result obtaining effect is increased.
In a possible embodiment, the method further comprises: a model training module 64 for training the named entity recognition model in the following manner:
obtaining a plurality of sample texts and a label labeling sequence corresponding to each sample text; the label labeling sequence comprises a plurality of entity type labeling labels which are in one-to-one correspondence with a plurality of character strings forming the sample text;
for each sample text, inputting the sample text into a conditional random field model, and determining a score corresponding to each label prediction sequence in a plurality of label prediction sequences corresponding to the sample text; the tag annotation sequence is one of a plurality of the tag prediction sequences;
And training the conditional random field model according to the score to obtain the named entity recognition model.
In one possible embodiment, the model training module 64 is configured to determine a score corresponding to each predicted sequence of tags in the plurality of predicted sequences of tags corresponding to the sample text by:
based on the conditional random field model, aiming at each label prediction sequence, determining the transition probability corresponding to each two adjacent prediction labels in the label prediction sequence according to a probability transition matrix;
and determining the score corresponding to the label prediction sequence according to the transition probability corresponding to the prediction labels adjacent to each two positions in the label prediction sequence.
In one possible embodiment, the model training module 64 is configured to train the conditional random field model based on the score by:
determining model loss according to the score corresponding to each label prediction sequence and the score of the label prediction sequence corresponding to the label labeling sequence;
adjusting parameters of the conditional random field model according to the model loss to complete the current round of training of the conditional random field model;
And determining the conditional random field model after multiple rounds of training as the named entity recognition model.
In a possible implementation, the determining module 62 is configured to determine the target retrieval area according to the target named entity by:
retrieving according to the target named entity and acquiring at least one piece of POI information corresponding to the target named entity; the POI information comprises a POI name and geographical position information;
aiming at each piece of POI information, determining an area range corresponding to the POI information by taking geographical position information in the POI information as a center;
and determining the area range corresponding to each POI information as the target retrieval area.
In a possible implementation manner, the determining module 62 is configured to retrieve and obtain at least one piece of POI information corresponding to the target named entity according to the target named entity by:
acquiring a plurality of candidate POI information related to the target named entity from a POI database by taking the target named entity as a retrieval key word;
determining the relevance of each alternative POI information and the target named entity;
and selecting at least one piece of POI information corresponding to the target named entity from the candidate POI information according to the degree of correlation between each piece of candidate POI information and the target named entity.
In a possible embodiment, the determining module 62 is configured to determine the relevance of each candidate POI information to the target named entity by:
determining the relevance of each candidate POI information and the target named entity based on the current position information of the client and the geographic position information in each candidate POI information;
alternatively, the first and second electrodes may be,
and determining the relevance of each piece of candidate POI information and the target named entity based on the target named entity and the number of the same characters in the POI names in each piece of candidate POI.
In a possible implementation manner, the determining module 62 is configured to retrieve and obtain at least one piece of POI information corresponding to the target named entity according to the target named entity by:
acquiring a plurality of candidate POI information related to the target named entity from a POI database by taking the target named entity as a retrieval key word;
clustering the alternative POI information according to the geographical position information in the alternative POI information to form a plurality of POI groups; the distance between any one piece of alternative POI information in each POI group and at least one piece of other POI information in the POI group is not more than a preset distance threshold value;
And aiming at each POI group, determining POI information corresponding to the target named entity according to each alternative information in the POI group.
In a possible embodiment, the determining module 62 is configured to determine, for each alternative POI information included in each alternative POI group, POI information corresponding to the target named entity by using the following method:
for each alternative POI group, if the number of alternative POI information in the alternative POI group is 1, determining the alternative POI information in the alternative POI group as the POI information corresponding to the target named entity;
if the number of the alternative POI information in the alternative POI group is larger than 1, determining the central POI information corresponding to the alternative POI group according to the geographical location information in each alternative POI information in the alternative POI group, and determining the central POI information as the POI information corresponding to the target named entity.
In a possible embodiment, the method further comprises: a detecting module 65, configured to determine whether the named entities identified include other named entities except the target named entity before the target search area is determined according to the target named entity;
the determining module 62 is specifically configured to:
When the named entities which are identified comprise other named entities except the target named entity, determining the target retrieval area according to the target named entity;
and when the identified named entities do not comprise other named entities except the target named entity, searching according to the target named entity to obtain at least one piece of target POI information corresponding to the search information.
The description of the processing flow of each module in the device and the interaction flow between the modules may refer to the related description in the above method embodiments, and will not be described in detail here.
An embodiment of the present application further provides a computer device 70, as shown in fig. 7, which is a schematic structural diagram of the computer device 70 provided in the embodiment of the present application, and includes: a processor 71, a memory 72, and a bus 73. The memory 72 stores machine-readable instructions (e.g., corresponding execution instructions of the identification module 61, the determination module 62, and the retrieval module 63 in the apparatus in fig. 6) executable by the processor 71, when the computer device 70 is running, the processor 71 and the memory 72 communicate via the bus 73, and when the processor 71 executes the following processes:
Acquiring retrieval information input by a user through a client, inputting the retrieval information into a pre-trained named entity recognition model, and recognizing a named entity in the retrieval information;
if the identified named entities comprise target named entities used for representing geographic positions, determining a target retrieval area according to the target named entities; the target retrieval area takes POI information corresponding to the target named entity as a center;
and retrieving and acquiring at least one piece of target POI information corresponding to the retrieval information from the target retrieval area according to other named entities except the target named entity in the named entities.
In one possible embodiment, processor 71 executes instructions that train the named entity recognition model in the following manner:
obtaining a plurality of sample texts and a label labeling sequence corresponding to each sample text; the label labeling sequence comprises a plurality of entity type labeling labels which are in one-to-one correspondence with a plurality of character strings forming the sample text;
for each sample text, inputting the sample text into a conditional random field model, and determining a score corresponding to each label prediction sequence in a plurality of label prediction sequences corresponding to the sample text; the tag annotation sequence is one of a plurality of the tag prediction sequences;
And training the conditional random field model according to the score to obtain the named entity recognition model.
In one possible embodiment, the instructions executed by processor 71 for determining a score corresponding to each predicted sequence of tags in the plurality of predicted sequences of tags corresponding to the sample text includes:
based on the conditional random field model, aiming at each label prediction sequence, determining the transition probability corresponding to each two adjacent prediction labels in the label prediction sequence according to a probability transition matrix;
and determining the score corresponding to the label prediction sequence according to the transition probability corresponding to the prediction labels adjacent to each two positions in the label prediction sequence.
In one possible embodiment, the instructions executed by processor 71 for training the conditional random field model based on the score include:
determining model loss according to the score corresponding to each label prediction sequence and the score of the label prediction sequence corresponding to the label labeling sequence;
adjusting parameters of the conditional random field model according to the model loss to complete the current round of training of the conditional random field model;
And determining the conditional random field model after multiple rounds of training as the named entity recognition model.
In one possible embodiment, the determining a target search area according to the target named entity in the instructions executed by the processor 71 includes:
retrieving according to the target named entity and acquiring at least one piece of POI information corresponding to the target named entity; the POI information comprises a POI name and geographical position information;
aiming at each piece of POI information, determining an area range corresponding to the POI information by taking geographical position information in the POI information as a center;
and determining the area range corresponding to each POI information as the target retrieval area.
In a possible implementation manner, in the instructions executed by the processor 71, the retrieving and acquiring at least one piece of POI information corresponding to the target named entity according to the target named entity includes:
acquiring a plurality of candidate POI information related to the target named entity from a POI database by taking the target named entity as a retrieval key word;
determining the relevance of each alternative POI information and the target named entity;
and selecting at least one piece of POI information corresponding to the target named entity from the candidate POI information according to the degree of correlation between each piece of candidate POI information and the target named entity.
In a possible embodiment, the determining the relevance of each candidate POI information to the target named entity in the instructions executed by the processor 71 includes:
determining the relevance of each candidate POI information and the target named entity based on the current position information of the client and the geographic position information in each candidate POI information;
alternatively, the first and second electrodes may be,
and determining the relevance of each piece of candidate POI information and the target named entity based on the target named entity and the number of the same characters in the POI names in each piece of candidate POI.
In a possible implementation manner, in the instructions executed by the processor 71, the retrieving and acquiring at least one piece of POI information corresponding to the target named entity according to the target named entity includes:
acquiring a plurality of candidate POI information related to the target named entity from a POI database by taking the target named entity as a retrieval key word;
clustering the alternative POI information according to the geographical position information in the alternative POI information to form a plurality of POI groups; the distance between any one piece of alternative POI information in each POI group and at least one piece of other POI information in the POI group is not more than a preset distance threshold value;
And aiming at each POI group, determining POI information corresponding to the target named entity according to each alternative information in the POI group.
In a possible embodiment, the determining, by the processor 71, POI information corresponding to the target named entity for each candidate POI information included in each of the candidate POI groups includes:
for each alternative POI group, if the number of alternative POI information in the alternative POI group is 1, determining the alternative POI information in the alternative POI group as the POI information corresponding to the target named entity;
if the number of the alternative POI information in the alternative POI group is larger than 1, determining the central POI information corresponding to the alternative POI group according to the geographical location information in each alternative POI information in the alternative POI group, and determining the central POI information as the POI information corresponding to the target named entity.
In a possible implementation manner, the instructions executed by processor 71, before determining the target search area according to the target named entity, further include:
judging whether other named entities except the target named entity are included in the identified named entities;
the determining a target retrieval region according to the target named entity includes:
When the named entities which are identified comprise other named entities except the target named entity, determining the target retrieval area according to the target named entity;
and when the identified named entities do not comprise other named entities except the target named entity, searching according to the target named entity to obtain at least one piece of target POI information corresponding to the search information.
An embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the POI information acquisition method are executed.
Specifically, the storage medium can be a general storage medium, such as a mobile disk, a hard disk, or the like, and when a computer program on the storage medium is run, the POI information acquisition method can be executed, and the accuracy of POI information acquisition can be improved by identifying the category of the entity in the retrieval information, and performing step-by-step retrieval on the identified target named entity and other named entities, which can represent the geographic location, to acquire target POI information.
It can be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the system and the apparatus described above may refer to corresponding processes in the method embodiments, and are not described in detail in this application. In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. The above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is merely a logical division, and there may be other divisions in actual implementation, and for example, a plurality of modules or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of devices or modules through some communication interfaces, and may be in an electrical, mechanical or other form.
The modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a non-volatile computer-readable storage medium executable by a processor. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing program codes, such as a U disk, a removable hard disk, a ROM, a RAM, a magnetic disk, or an optical disk.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims (13)

1. A method for obtaining POI information is characterized in that the method for obtaining the POI information comprises the following steps:
acquiring retrieval information input by a user through a client, inputting the retrieval information into a pre-trained named entity recognition model, and recognizing a named entity in the retrieval information;
if the identified named entities comprise target named entities used for representing geographic positions, determining a target retrieval area according to the target named entities; the target retrieval area takes POI information corresponding to the target named entity as a center;
and retrieving and acquiring at least one piece of target POI information corresponding to the retrieval information from the target retrieval area according to other named entities except the target named entity in the named entities.
2. The method of claim 1, wherein the named entity recognition model is trained by:
Obtaining a plurality of sample texts and a label labeling sequence corresponding to each sample text; the label labeling sequence comprises a plurality of entity type labeling labels which are in one-to-one correspondence with a plurality of character strings forming the sample text;
for each sample text, inputting the sample text into a conditional random field model, and determining a score corresponding to each label prediction sequence in a plurality of label prediction sequences corresponding to the sample text; the tag annotation sequence is one of a plurality of the tag prediction sequences;
and training the conditional random field model according to the score to obtain the named entity recognition model.
3. The method of claim 2, wherein determining a score corresponding to each of the plurality of predicted sequences of tags corresponding to the sample text comprises:
based on the conditional random field model, aiming at each label prediction sequence, determining the transition probability corresponding to each two adjacent prediction labels in the label prediction sequence according to a probability transition matrix;
and determining the score corresponding to the label prediction sequence according to the transition probability corresponding to the prediction labels adjacent to each two positions in the label prediction sequence.
4. The method of claim 2 wherein training the conditional random field model based on the score comprises:
determining model loss according to the score corresponding to each label prediction sequence and the score of the label prediction sequence corresponding to the label labeling sequence;
adjusting parameters of the conditional random field model according to the model loss to complete the current round of training of the conditional random field model;
and determining the conditional random field model after multiple rounds of training as the named entity recognition model.
5. The method of claim 1, wherein determining a target search area based on the target named entity comprises:
retrieving according to the target named entity and acquiring at least one piece of POI information corresponding to the target named entity; the POI information comprises a POI name and geographical position information;
aiming at each piece of POI information, determining an area range corresponding to the POI information by taking geographical position information in the POI information as a center;
and determining the area range corresponding to each POI information as the target retrieval area.
6. The method according to claim 5, wherein the retrieving and obtaining at least one piece of POI information corresponding to the target named entity according to the target named entity comprises:
Acquiring a plurality of candidate POI information related to the target named entity from a POI database by taking the target named entity as a retrieval key word;
determining the relevance of each alternative POI information and the target named entity;
and selecting at least one piece of POI information corresponding to the target named entity from the candidate POI information according to the degree of correlation between each piece of candidate POI information and the target named entity.
7. The method of claim 6, wherein the determining the relevance of each candidate POI information to the target named entity comprises:
determining the relevance of each candidate POI information and the target named entity based on the current position information of the client and the geographic position information in each candidate POI information;
alternatively, the first and second electrodes may be,
and determining the relevance of each piece of candidate POI information and the target named entity based on the target named entity and the number of the same characters in the POI names in each piece of candidate POI.
8. The method according to claim 5, wherein the retrieving and obtaining at least one piece of POI information corresponding to the target named entity according to the target named entity comprises:
Acquiring a plurality of candidate POI information related to the target named entity from a POI database by taking the target named entity as a retrieval key word;
clustering the alternative POI information according to the geographical position information in the alternative POI information to form a plurality of POI groups; the distance between any one piece of alternative POI information in each POI group and at least one piece of other POI information in the POI group is not more than a preset distance threshold value;
and aiming at each POI group, determining POI information corresponding to the target named entity according to each alternative information in the POI group.
9. The method according to claim 8, wherein the determining, for each alternative POI information included in each alternative POI group, POI information corresponding to the target named entity comprises:
for each alternative POI group, if the number of alternative POI information in the alternative POI group is 1, determining the alternative POI information in the alternative POI group as the POI information corresponding to the target named entity;
if the number of the alternative POI information in the alternative POI group is larger than 1, determining the central POI information corresponding to the alternative POI group according to the geographical location information in each alternative POI information in the alternative POI group, and determining the central POI information as the POI information corresponding to the target named entity.
10. The method of claim 1, wherein prior to determining a target search area based on the target named entity, further comprising:
judging whether other named entities except the target named entity are included in the identified named entities;
the determining a target retrieval region according to the target named entity includes:
when the named entities which are identified comprise other named entities except the target named entity, determining the target retrieval area according to the target named entity;
and when the identified named entities do not comprise other named entities except the target named entity, searching according to the target named entity to obtain at least one piece of target POI information corresponding to the search information.
11. An apparatus for obtaining POI information, the apparatus comprising:
the system comprises an identification module, a search module and a search module, wherein the identification module is used for acquiring search information input by a user through a client, inputting the search information into a pre-trained named entity identification model and identifying a named entity in the search information;
the determining module is used for determining a target retrieval area according to a target named entity when the identified named entity comprises the target named entity for representing the geographic position; the target retrieval area takes POI information corresponding to the target named entity as a center;
And the retrieval module is used for retrieving and acquiring at least one piece of target POI information corresponding to the retrieval information from the target retrieval area according to other named entities except the target named entity in the named entities.
12. A computer device, comprising: a processor, a storage medium and a bus, the storage medium storing machine-readable instructions executable by the processor, the processor and the storage medium communicating via the bus when the computer device is running, the processor executing the machine-readable instructions to perform the steps of the method according to any one of claims 1 to 10.
13. A computer-readable storage medium, having stored thereon a computer program which, when being executed by a processor, is adapted to carry out the steps of the method according to any one of claims 1 to 10.
CN201910907458.8A 2019-09-24 2019-09-24 POI information acquisition method and device Pending CN111858787A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910907458.8A CN111858787A (en) 2019-09-24 2019-09-24 POI information acquisition method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910907458.8A CN111858787A (en) 2019-09-24 2019-09-24 POI information acquisition method and device

Publications (1)

Publication Number Publication Date
CN111858787A true CN111858787A (en) 2020-10-30

Family

ID=72970614

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910907458.8A Pending CN111858787A (en) 2019-09-24 2019-09-24 POI information acquisition method and device

Country Status (1)

Country Link
CN (1) CN111858787A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113794808A (en) * 2021-09-01 2021-12-14 北京亿心宜行汽车技术开发服务有限公司 Ordering method and system for designated driving telephone

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2013105282A (en) * 2011-11-11 2013-05-30 Nippon Telegr & Teleph Corp <Ntt> Information retrieval device and method and program
KR20150057472A (en) * 2013-11-19 2015-05-28 한국과학기술연구원 Method for solving ambiguity for extraction of a POI, Method for extracting a POI from a document and Apparatus for extracting a POI
CN108363698A (en) * 2018-03-13 2018-08-03 腾讯大地通途（北京）科技有限公司 Point of interest relation recognition method and device
CN108388559A (en) * 2018-02-26 2018-08-10 中译语通科技股份有限公司 Name entity recognition method and system, computer program of the geographical space under
CN109145219A (en) * 2018-09-10 2019-01-04 百度在线网络技术（北京）有限公司 Point of interest Effective judgement method and apparatus based on internet text mining
CN109190130A (en) * 2018-08-30 2019-01-11 昆明理工大学 A kind of research method matching proposed algorithm with machine translator based on POI similarity
CN109460509A (en) * 2018-10-12 2019-03-12 平安科技（深圳）有限公司 User interest point appraisal procedure, device, computer equipment and storage medium
CN109492066A (en) * 2018-10-30 2019-03-19 百度在线网络技术（北京）有限公司 A kind of determination method, apparatus, equipment and the storage medium of point of interest branch name
CN110019616A (en) * 2017-12-04 2019-07-16 腾讯科技（深圳）有限公司 A kind of POI trend of the times state acquiring method and its equipment, storage medium, server
CN110114790A (en) * 2016-12-07 2019-08-09 谷歌有限责任公司 For showing the graphic user interface for the entity sorted out jointly

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2013105282A (en) * 2011-11-11 2013-05-30 Nippon Telegr & Teleph Corp <Ntt> Information retrieval device and method and program
KR20150057472A (en) * 2013-11-19 2015-05-28 한국과학기술연구원 Method for solving ambiguity for extraction of a POI, Method for extracting a POI from a document and Apparatus for extracting a POI
CN110114790A (en) * 2016-12-07 2019-08-09 谷歌有限责任公司 For showing the graphic user interface for the entity sorted out jointly
CN110019616A (en) * 2017-12-04 2019-07-16 腾讯科技（深圳）有限公司 A kind of POI trend of the times state acquiring method and its equipment, storage medium, server
CN108388559A (en) * 2018-02-26 2018-08-10 中译语通科技股份有限公司 Name entity recognition method and system, computer program of the geographical space under
CN108363698A (en) * 2018-03-13 2018-08-03 腾讯大地通途（北京）科技有限公司 Point of interest relation recognition method and device
CN109190130A (en) * 2018-08-30 2019-01-11 昆明理工大学 A kind of research method matching proposed algorithm with machine translator based on POI similarity
CN109145219A (en) * 2018-09-10 2019-01-04 百度在线网络技术（北京）有限公司 Point of interest Effective judgement method and apparatus based on internet text mining
CN109460509A (en) * 2018-10-12 2019-03-12 平安科技（深圳）有限公司 User interest point appraisal procedure, device, computer equipment and storage medium
CN109492066A (en) * 2018-10-30 2019-03-19 百度在线网络技术（北京）有限公司 A kind of determination method, apparatus, equipment and the storage medium of point of interest branch name

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
陈丹双: "POI(Point of Interest)名称识别及其在对话导航系统中的应用", 中国优秀硕士论文信息科技, no. 1, 15 January 2016 (2016-01-15) *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113794808A (en) * 2021-09-01 2021-12-14 北京亿心宜行汽车技术开发服务有限公司 Ordering method and system for designated driving telephone
CN113794808B (en) * 2021-09-01 2024-01-30 北京亿心宜行汽车技术开发服务有限公司 Method and system for ordering representative driving telephone

Similar Documents

Publication Publication Date Title
JP6846469B2 (en) Method and device for determining the effectiveness of points of interest based on Internet text mining
Vargas-Munoz et al. OpenStreetMap: Challenges and opportunities in machine learning and remote sensing
CN108304423B (en) Information identification method and device
CN112069276B (en) Address coding method, address coding device, computer equipment and computer readable storage medium
CN110390054A (en) Point of interest recalls method, apparatus, server and storage medium
US8429204B2 (en) Short point-of-interest title generation
RU2598165C1 (en) Non-deterministic disambiguation and comparison of data of location of commercial enterprise
CN111141301B (en) Navigation end point determining method, device, storage medium and computer equipment
WO2018223331A1 (en) Systems and methods for text attribute determination using conditional random field model
CN111460248A (en) System and method for online-to-offline services
CN110717010A (en) Text processing method and system
CN111859174A (en) Method and system for determining recommended boarding point
CN111831929B (en) Method and device for acquiring POI information
CN111931077A (en) Data processing method and device, electronic equipment and storage medium
CN112711645B (en) Method and device for expanding position point information, storage medium and electronic equipment
KR20170035694A (en) Method and system for recommending course for travel related query
CN111858787A (en) POI information acquisition method and device
CN110647595B (en) Method, device, equipment and medium for determining newly-added interest points
US20190318014A1 (en) Facilitating identification of an intended country associated with a query
US20220248170A1 (en) Methods and systems for recommending pick-up points
CN109918583B (en) Task information processing method and device
CN110633370B (en) OTA hotel label generation method, system, electronic device and medium
Kim et al. Personalized POI embedding for successive POI recommendation with large-scale smart card data
CN113175940A (en) Data processing method, device, equipment and storage medium
CN112650791A (en) Method and device for processing fields, computer equipment and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination