CN113918802B

CN113918802B - Navigation method, device, equipment, medium and product

Info

Publication number: CN113918802B
Application number: CN202111524386.2A
Authority: CN
Inventors: 钟子宏
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2021-12-14
Filing date: 2021-12-14
Publication date: 2022-03-08
Anticipated expiration: 2041-12-14
Also published as: CN113918802A

Abstract

The disclosed embodiment provides a navigation method, a navigation device, a navigation apparatus, a navigation medium and a navigation product, wherein the method comprises the following steps: acquiring a current search semantic text input by a target object; predicting the probability that the current search semantic text belongs to the search semantic text category of the target object; determining the target search semantic text category of the current search semantic text according to the probability that the current search semantic text belongs to the search semantic text category of the target object; acquiring historical destination information of a target object under the category of a target search semantic text; determining target destination information of a current search semantic text from historical destination information of a target object; and generating target navigation planning information of the target object to the target destination information according to the historical navigation information of the target object corresponding to the target destination information. The scheme provided by the embodiment of the disclosure can realize intelligent navigation by utilizing a machine learning technology in the field of artificial intelligence, and can be applied to the fields of intelligent traffic, maps and the like.

Description

Navigation method, device, equipment, medium and product

Technical Field

The present disclosure relates to the field of computer technologies, and in particular, to a navigation method, a navigation apparatus, a computer device, a computer-readable storage medium, and a computer program product.

Background

In a navigation scheme in the related art, a user generally inputs a place name of a destination to be visited in a search box, and a map background generates a navigation route after the place name of the destination input by the user is matched with the place of the map background according to a database system.

However, in some cases, the user does not directly input the accurate destination location name in the search box, but inputs the fuzzy text information, the scheme adopting the related art cannot infer the real meaning behind the fuzzy text information input by the user, and the situation that the destination location matched with the navigation is wrong often occurs, so that the user may need to manually input the destination location name for many times, thereby reducing the user experience.

Disclosure of Invention

The embodiment of the disclosure provides a navigation method, a navigation device, a computer readable storage medium and a computer program product, which can improve the accuracy of navigation.

The embodiment of the disclosure provides a navigation method, which comprises the following steps: acquiring a current search semantic text input by a target object; predicting the probability that the current search semantic text belongs to the search semantic text category of the target object; determining the target search semantic text category of the current search semantic text according to the probability that the current search semantic text belongs to the search semantic text category of the target object; acquiring historical destination information of the target object under the category of the target search semantic text; determining target destination information of the current search semantic text from historical destination information of the target object; and generating target navigation planning information for the target object to reach the target destination information according to the historical navigation information of the target object corresponding to the target destination information.

The embodiment of the disclosure provides a navigation method, which comprises the following steps: displaying a navigation interface, the navigation interface including a destination input area; responding to an input instruction of the destination input area, and acquiring current search information input by a target object, wherein the current search information comprises a current search semantic text; according to the current search information, displaying historical destination information which belongs to the same target search semantic text category as the current search semantic text on the navigation interface, wherein the historical destination information is destination information in the historical navigation information of the target object; determining target destination information from the history destination information in response to an operation instruction for the displayed history destination information; and displaying target navigation planning information of the target object reaching the target destination information according to the target destination information.

The disclosed embodiment provides a navigation device, which comprises: the first acquisition unit is used for acquiring a current search semantic text input by a target object; a prediction unit for predicting a probability that the current search semantic text belongs to a search semantic text category of the target object; the first determining unit is used for determining the target search semantic text category of the current search semantic text according to the probability that the current search semantic text belongs to the search semantic text category of the target object; the first obtaining unit is further used for obtaining historical destination information of the target object under the category of the target search semantic text; the first determining unit is further used for determining target destination information of the current search semantic text from historical destination information of the target object; and the generating unit is used for generating target navigation planning information of the target object reaching the target destination information according to the historical navigation information of the target object corresponding to the target destination information.

In some exemplary embodiments of the disclosure, the prediction unit is further configured to perform the following steps: obtaining a text sequence prediction model of a hidden input parameter matrix, a hidden parameter vector, an input hidden parameter matrix, an output parameter vector and a previous hidden layer vector which are determined by training; inputting the current search semantic text at the current time T +1 into the text sequence prediction model to obtain a current hidden layer vector according to the hidden input parameter matrix, the current search semantic text, the hidden parameter matrix, the previous hidden layer vector and the hidden parameter vector, and obtaining the probability that the current search semantic text belongs to each search semantic text category of the target object according to the input hidden parameter matrix, the current hidden layer vector and the output parameter vector, wherein T is an integer greater than 1.

In some exemplary embodiments of the present disclosure, the navigation device further includes: the obtaining unit is used for obtaining historical search semantic texts of the target object from the T-1 th time to the T-n th time and corresponding search semantic text category labels, wherein n is an integer larger than 0; the obtaining unit is further configured to input the historical search semantic texts of the target object at the T-1 th time to the T-n th time to the text sequence prediction model, and obtain probabilities that the historical search semantic texts of the target object at the T-1 th time to the T-n th time belong to search semantic text categories of the target object respectively; the training unit is used for training the text sequence prediction model according to the probability that the historical search semantic texts of the target object from the T-1 th moment to the T-n th moment respectively belong to the search semantic text categories of the target object and the corresponding search semantic text category labels; the obtaining unit is further used for obtaining a historical search semantic text of the target object at the Tth moment and a corresponding search semantic text category label; the obtaining unit is further configured to input the historical search semantic text of the target object at the tth moment to the trained text sequence prediction model, and obtain a probability that the historical search semantic text of the target object at the tth moment belongs to the search semantic text category of the target object; and the evaluation unit is used for evaluating the text sequence prediction model according to the probability that the historical search semantic text of the target object at the T-th moment belongs to the search semantic text category of the target object and the corresponding search semantic text category label so as to determine the hidden input parameter matrix, the hidden parameter vector, the input hidden parameter matrix, the output parameter vector and the previous hidden layer vector of the text sequence prediction model.

In some exemplary embodiments of the present disclosure, the first obtaining unit is further configured to perform the following steps: obtaining historical navigation information of the target object and a historical search semantic text corresponding to the historical navigation information, wherein the historical navigation information comprises historical destination information of historical navigation, and the historical destination information of the historical navigation comprises a historical destination name of the historical navigation, historical destination interest point information, a parcel of a historical destination position, parcel attributes of the historical destination and historical destination attributes; determining a search semantic text category to which the historical search semantic text belongs and a historical destination name corresponding to the search semantic text category according to the historical destination name of the historical navigation, the historical destination interest point information, a section where the historical destination is located, a section attribute where the historical destination is located and the historical destination attribute; and obtaining the historical destination information of the target object under the target search semantic text type according to the search semantic text type to which the historical search semantic text belongs and the corresponding historical destination name.

In some exemplary embodiments of the present disclosure, the first determining unit is further configured to perform the following steps: obtaining historical operation behavior information of the target object aiming at the historical destination information; inputting the historical operation behavior information into a destination prediction model so as to output a probability corresponding to the historical destination information through the destination prediction model; sequentially displaying the historical destination information in a descending order according to the probability corresponding to the historical destination information; determining the target destination information from the historical destination information in response to an operation instruction on the historical destination information.

In some exemplary embodiments of the disclosure, the generating unit is further configured to perform the following steps: inputting historical navigation information of the target object corresponding to the target destination information into a navigation planning model so as to select the target navigation planning information from the historical navigation information of the target object corresponding to the target destination information through the navigation planning model; the historical navigation information of the target object corresponding to the target destination information comprises historical navigation track information, historical geographic position information and historical destination information thereof.

In some exemplary embodiments of the disclosure, the first obtaining unit is further configured to perform the following steps: acquiring current search information input by the target object; and performing semantic extraction processing on the current search information to obtain the current search semantic text.

The disclosed embodiment provides a navigation device, which comprises: a display unit for displaying a navigation interface including a destination input area; the second acquisition unit is used for responding to an input instruction of the destination input area and acquiring current search information input by a target object, wherein the current search information comprises a current search semantic text; the display unit is further used for displaying historical destination information which belongs to the same target search semantic text category as the current search semantic text on the navigation interface according to the current search information, wherein the historical destination information is destination information in the historical navigation information of the target object; a second determination unit configured to determine target destination information from the history destination information in response to an operation instruction to the displayed history destination information; the display unit is further used for displaying target navigation planning information of the target object reaching the target destination information according to the target destination information.

The embodiment of the disclosure provides computer equipment, which comprises a processor, a memory and an input/output interface; the processor is respectively connected with the memory and the input/output interface, wherein the input/output interface is used for receiving data and outputting data, the memory is used for storing a computer program, and the processor is used for calling the computer program so as to enable the computer equipment comprising the processor to execute the navigation method in the embodiment of the disclosure.

The disclosed embodiments provide a computer-readable storage medium storing a computer program adapted to be loaded and executed by a processor to cause a computer device having the processor to perform the navigation method in the disclosed embodiments.

Embodiments of the present disclosure provide a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, causing the computer device to perform the method provided in the various alternatives in the embodiments of the present disclosure.

In the disclosed embodiment, on one hand, by acquiring the current search semantic text input by the target object, predicting the probability that the current search semantic text belongs to the search semantic text category of the target object, and determining the target search semantic text category of the current search semantic text according to the probability that the current search semantic text belongs to the search semantic text category of the target object, semantic reasoning can be performed on the current search semantic text input by the target object, and the semantic similarity degree between the current search semantic text category and each search semantic text category of the target object is predicted (represented by the probability between the current search semantic text category and the search semantic text category), so that navigation through the semantic reasoning can be realized, and the real intention behind the current search semantic text input by the target object can be deduced even if the destination place name input by the target object is not the destination place name to be visited, therefore, the navigation accuracy can be improved, multiple times of manual input or modification of input information of the target object are not needed, and the user experience is improved; on one hand, the historical destination information of the target object under the category of the target search semantic text is obtained to determine the target destination information of the current search semantic text from the historical destination information of the target object, so that the target navigation planning information of the target object reaching the target destination information can be generated according to the historical navigation information of the target object corresponding to the target destination information, namely, the semantic reasoning and the historical navigation information of the target object are combined, the target navigation planning information which is most suitable for the target object can be planned, and the navigation accuracy is further improved.

Drawings

In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present disclosure, and other drawings can be obtained by those skilled in the art without creative efforts.

Fig. 1 is a network interaction architecture diagram of a navigation method according to an embodiment of the present disclosure.

Fig. 2 is a scene diagram of a navigation method in the related art.

Fig. 3 is a flowchart of a navigation method provided by an embodiment of the present disclosure.

Fig. 4 is a flowchart of another navigation method provided by the embodiments of the present disclosure.

Fig. 5 is a flowchart of another navigation method provided by the embodiments of the present disclosure.

Fig. 6 is a scene schematic diagram of a navigation method provided by an embodiment of the present disclosure.

Fig. 7 is a schematic frame diagram of a navigation device according to an embodiment of the present disclosure.

Fig. 8 is a schematic frame diagram of another navigation device provided in the embodiments of the present disclosure.

Fig. 9 is a schematic structural diagram of a computer device according to an embodiment of the present disclosure.

Detailed Description

The technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the drawings in the embodiments of the present disclosure, and it is obvious that the described embodiments are only a part of the embodiments of the present disclosure, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments disclosed herein without making any creative effort, shall fall within the protection scope of the present disclosure.

In the embodiment of the disclosure, a current search semantic text of current search information input by a target object may be acquired based on a machine learning technique and the like in the field of artificial intelligence, a probability that the current search semantic text belongs to a search semantic text category of the target object is predicted, target destination information of the current search semantic text is determined from historical destination information of the target object, and target navigation planning information for the target object to reach the target destination information is generated.

Among them, Artificial Intelligence (AI) is a theory, method, technique and application system that simulates, extends and expands human Intelligence using a digital computer or a machine controlled by a digital computer, senses the environment, acquires knowledge and uses the knowledge to obtain the best result. In other words, artificial intelligence is a comprehensive technique in computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can react in a manner similar to human intelligence, thereby reducing the cost of artificial resources. Artificial intelligence is the research of the design principle and the realization method of various intelligent machines, so that the machines have the functions of perception, reasoning and decision making.

The artificial intelligence technology is a comprehensive subject and relates to the field of extensive technology, namely the technology of a hardware level and the technology of a software level. The artificial intelligence infrastructure generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a voice processing technology, a natural language processing technology, machine learning/deep learning, automatic driving, intelligent traffic and the like.

Key technologies of Speech Technology (Speech Technology) are automatic Speech recognition Technology and Speech synthesis Technology, as well as voiceprint recognition Technology. The computer can listen, see, speak and feel, and the development direction of the future human-computer interaction is provided, wherein the voice becomes one of the best viewed human-computer interaction modes in the future.

Natural Language Processing (NLP) is an important direction in the fields of computer science and artificial intelligence. It studies various theories and methods that enable efficient communication between humans and computers using natural language. Natural language processing is a science integrating linguistics, computer science and mathematics. Therefore, the research in this field will involve natural language, i.e. the language that people use everyday, so it is closely related to the research of linguistics. Natural language processing techniques typically include text processing, semantic understanding, machine translation, robotic question and answer, knowledge mapping, and the like.

Machine Learning (ML) is a multi-domain cross discipline, and relates to a plurality of disciplines such as probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory and the like. The special research on how a computer simulates or realizes the learning behavior of human beings so as to acquire new knowledge or skills and reorganize the existing knowledge structure to continuously improve the performance of the computer. Machine learning is the core of artificial intelligence, is the fundamental approach for computers to have intelligence, and is applied to all fields of artificial intelligence. Machine learning and deep learning generally include techniques such as artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and formal education learning.

With the research and progress of artificial intelligence technology, the artificial intelligence technology develops research and application in a plurality of fields, such as common smart homes, smart wearable devices, virtual assistants, smart speakers, smart marketing, unmanned driving, automatic driving, unmanned aerial vehicles, robots, smart medical care, smart customer service, internet of vehicles, smart traffic, and the like.

The scheme provided by the embodiment of the disclosure relates to the technologies such as artificial intelligence voice technology, NLP, machine learning and the like, and is specifically explained by the following embodiments:

in the embodiment of the present disclosure, please refer to fig. 1, where fig. 1 is a network interaction architecture diagram of a navigation method provided in the embodiment of the present disclosure, and the embodiment of the present disclosure may be implemented by a user equipment. The user equipment may acquire data from the computer equipment 101 and display the data, where the computer equipment 101 may perform data interaction with the user equipment, and the computer equipment 101 may be a server where an application program is located, may also belong to the user equipment (i.e., a background of the user equipment), and the like, which is not limited herein.

The user equipment may be the user equipment 102a, the user equipment 102b, or the user equipment 102c, and the like, and the embodiment of the present disclosure may be implemented by any one of the user equipment 102a, the user equipment 102b, or the user equipment 102c, and the like.

Specifically, taking the user equipment 102b as an example, the computer device 101 may obtain a current search semantic text input by the target object from the user equipment 102 b; predicting the probability that the current search semantic text belongs to the search semantic text category of the target object; determining the target search semantic text category of the current search semantic text according to the probability that the current search semantic text belongs to the search semantic text category of the target object; acquiring historical destination information of the target object under the category of the target search semantic text; the computer device 101 may return the acquired historical destination information of the target object in the category of the target search semantic text to the user device 102b, so that the target object determines the target destination information of the current search semantic text from the historical destination information of the target object through the user device 102b, the user device 102b returns the target destination information of the current search semantic text to the computer device 101, and the computer device 101 may generate target navigation planning information for the target object to reach the target destination information according to the historical navigation information of the target object corresponding to the target destination information.

Specifically, also taking the user device 102b as an example, a navigation interface may be displayed on the screen of the user device 102b, and the navigation interface may include a destination input area; the user equipment 102b responds to an input instruction of the destination input area, and obtains current search information input by the target object, wherein the current search information comprises a current search semantic text; according to the current search information, displaying historical destination information which belongs to the same target search semantic text category as the current search semantic text on the navigation interface, wherein the historical destination information is destination information in the historical navigation information of the target object; determining target destination information from the history destination information in response to an operation instruction for the displayed history destination information; and displaying target navigation planning information of the target object reaching the target destination information according to the target destination information.

The user device may be a mobile phone (e.g., the user device 102 c) or a laptop (e.g., the user device 102 b), or may also be a playback device in a vehicle (e.g., the user device 102 a), and the like, which is not limited herein. The user device 102a may be regarded as a playing device in the vehicle 103, and an application program may be displayed in the user device 102a, where the application program may be a map-type application program, an internet taxi-taking application program, a navigation-type application program, an internet game-type application program, or the like. The user equipment in fig. 1 is only an exemplary part of the equipment, and in the present disclosure, the user equipment is not limited to the equipment illustrated in fig. 1.

It is understood that the user equipment mentioned in the embodiments of the present disclosure may be a computer device, and the computer device in the embodiments of the present disclosure includes, but is not limited to, a terminal device or a server. In other words, the computer device may be a server or a terminal device, or may be a system of a server and a terminal device. The above-mentioned terminal device may be an electronic device, including but not limited to a mobile phone, a tablet computer, an intelligent voice interaction device, an intelligent appliance, a vehicle-mounted terminal, a desktop computer, a notebook computer, a palm computer, a vehicle-mounted device, an Augmented Reality/Virtual Reality (AR/VR) device, a helmet display, an intelligent television, a wearable device, an intelligent speaker, a digital camera, a camera, and other Mobile Internet Devices (MID) with network access capability, or a terminal device in a scene such as a train, a ship, or a flight.

The above-mentioned server may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as cloud service, a cloud database, cloud computing, a cloud function, cloud storage, Network service, cloud communication, middleware service, domain name service, security service, vehicle-road cooperation, a Content Delivery Network (CDN), a big data and artificial intelligence platform, and the like.

Optionally, the data related to the embodiments of the present disclosure may be stored in a computer device, or the data may be stored based on a cloud storage technology and a block chain technology, which is not limited herein.

The navigation scheme in the related art does not have the capability of performing semantic reasoning according to the text information of the context input by the user in the search box of the map application and the historical navigation information of the user, so that the real place information behind the fuzzy text information input by the user cannot be deduced, and the place name containing the same text content can be searched simply according to the text input by the user.

For example, as shown in fig. 2, in a search of a related art map-like application, a user inputs fuzzy text information in a search box, such as: when the user goes home, the user clicks the search virtual key, and then the place name with the word "go home" appears (for example, the place names of the house, the restaurant, the Hunan cuisine, and the like). However, the user's essence needs to express that it means "go back to his home", and thus the map search should be for the specific address of the user's own home. However, the existing map searching method cannot achieve the purpose, so that the accuracy of navigation and positioning is reduced, a user needs to input the name of the own home of the user again in a search box, and the user experience is reduced.

Fig. 3 is a flowchart of a navigation method provided by an embodiment of the present disclosure. The method provided by the embodiment of the present disclosure may be executed by the computer device and/or the user equipment in the embodiment of fig. 1.

As shown in fig. 3, the method provided by the embodiment of the present disclosure may include the following steps.

In step S310, the current search semantic text input by the target object is acquired.

In an exemplary embodiment, obtaining the current search semantic text input by the target object may include: acquiring current search information input by the target object; and performing semantic extraction processing on the current search information to obtain the current search semantic text.

In the embodiment of the present disclosure, the target object may be any user who obtains a navigation service through an application (for example, a map application, hereinafter, simply referred to as a map).

In the embodiment of the present disclosure, the current search information refers to a search word/query word (query) input by a user through a search box provided by the application, and is different from a manner of a destination place name directly input in the related art, the current search information in the embodiment of the present disclosure is fuzzy text information, where the fuzzy text information refers to that after the application acquires the current search information input by the user, the current search information cannot directly obtain destination information that the user currently desires to go from the current search information, but after semantic reasoning is performed on the current search information, the current search information can be inferred to obtain destination information (referred to as target destination information) that the user really desires to go behind the current search information, that is, the current search information is not target destination information, for example, a target destination name.

It should be noted that the search box provided by the application program may be a text input box, for example, the user inputs text information such as "i want to go home", "i want to go to and shop" and the like in the text input box as the current search information; the search box can also be a voice input control, a user can input similar voice information such as 'i want to go home', 'i want to go to shop' and the like through the voice input control, and the voice information input by the user can be converted into corresponding text information to serve as current search information by utilizing a voice recognition technology.

In the embodiment of the present disclosure, the current search semantic text refers to a semantic text obtained by performing semantic extraction processing on current search information input by a user in a search box, and the semantic extraction processing may include word segmentation, part-of-speech tagging, semantic feature extraction, and other NLP technologies. For example, if the current search information is "i want to go home", the extracted current search semantic text is "go home". For another example, if the current search information is "i want to go and shop", the extracted current search semantic text is "shop".

In step S320, a probability that the current search semantic text belongs to the search semantic text category of the target object is predicted.

In the embodiment of the present disclosure, the search semantic text category refers to a category corresponding to fuzzy text information included in a search semantic text (including a current search semantic text and a historical search semantic text), and may include, for example, "go home", "shopping", "business trip", "work trip", and the like, which do not include specific destination information. The search semantic text is not destination information but may be used to determine destination information, and therefore, the search semantic text may also be referred to as "place-related semantic text" such as "go home", "shopping", "business", "work", and the like.

In the embodiment of the disclosure, one or more search semantic text categories may be pre-constructed according to the history navigation information of the target object and the history search semantic text corresponding to the history navigation information, the history navigation information may include history destination information corresponding to the history navigation, the history destination information of the history navigation may include a history destination name of the history navigation, history destination interest point information, a fragment of the history destination location, and history destination attributes, and the search semantic text category may be determined based on the history destination information of the history navigation including the history destination name of the history navigation, the history destination interest point information, the fragment of the history destination location, and the history destination attributes.

For example, if the historical destination name corresponding to a certain historical navigation is "XX building in cell XX", and it is known that "XX building in cell XX" is a residential building in a residential cell by combining the property of the parcel in which the historical destination location is located and the property of the historical destination, the search semantic text category corresponding to the historical search semantic text corresponding to the historical navigation can be determined as "going home".

For another example, if the historical destination name corresponding to a certain historical navigation is "XX mall", and it is known that "XX mall" is located in a commercial building by combining the parcel attribute of the historical destination location and the historical destination attribute, the search semantic text category corresponding to the historical search semantic text corresponding to the historical navigation can be determined as "shopping".

In an exemplary embodiment, predicting a probability that the current search semantic text belongs to the search semantic text category of the target object includes: obtaining a matrix of hidden input parameters determined by training

Hidden parameter matrix

Hidden parameter vector

Inputting the hidden parameter matrix

Outputting the parameter vector

And previous hidden layer vector

The text sequence prediction model of (1); the current search semantic text of the current time T +1

Inputting the current hidden layer vector into the text sequence prediction model to obtain a current hidden layer vector according to the hidden input parameter matrix, the current search semantic text, the hidden parameter matrix, the previous hidden layer vector and the hidden parameter vector

And obtaining the probability that the current search semantic text belongs to each search semantic text category of the target object according to the input hidden parameter matrix, the current hidden layer vector and the output parameter vector

And T is an integer greater than 1.

In an exemplary embodiment, the method may further include: obtaining historical search semantic texts of the target object from the T-1 th time to the T-n th time (assumed to be respectively expressed as

，…，

) And corresponding search semantic text category labels, wherein n is an integer greater than 0, and T is an integer greater than n; history search words of the target object from the T-1 th time to the T-n th timeSemantic texts are respectively input into the text sequence prediction model, and the probability that the historical search semantic texts of the target object from the T-1 th moment to the T-n th moment respectively belong to the search semantic text category of the target object is obtained; training a text sequence prediction model according to the probability that the historical search semantic texts of the target object from the T-1 th moment to the T-n th moment respectively belong to the search semantic text categories of the target object and the corresponding search semantic text category labels; obtaining historical search semantic texts of the target object at the Tth time (assumed to be respectively expressed as

) And a corresponding search semantic text category label; inputting the historical search semantic text of the target object at the Tth moment into the trained text sequence prediction model, and obtaining the probability that the historical search semantic text of the target object at the Tth moment belongs to the search semantic text category of the target object; and evaluating the text sequence prediction model according to the probability that the historical search semantic text of the target object at the T-th moment belongs to the search semantic text category of the target object and the corresponding search semantic text category label thereof so as to determine the hidden input parameter matrix, the hidden parameter vector, the input hidden parameter matrix, the output parameter vector and the previous hidden layer vector of the text sequence prediction model.

In the embodiment of the present disclosure, the text sequence prediction model is generated by training an RNN (Recurrent Neural Network/Recurrent Neural Network) model using a sequence composed of historical search semantic texts and a training sample composed of search semantic text category labels corresponding to each historical search semantic text in the sequence, and therefore, may also be referred to as an RNN semantic reasoning model.

The RNN semantic inference model is a recurrent neural network that takes sequence data as input, recurses in the evolution direction of the sequence, and all nodes (cyclic units) are connected in a chain, and aims to process time series data.

In the disclosed embodiment, the historical search semantic text is obtained by performing semantic extraction processing on historical search information input by the target object at historical time (for example, time T-1 to time T-n), the historical search information corresponds to historical navigation information, the text sequence prediction model is obtained by training the historical search semantic text, therefore, the trained text sequence prediction model can take the historical navigation information and the input historical search information of the target object as the context information of the current search information input by the target object at the current moment, perform semantic reasoning or fuzzy matching on the current search semantic text corresponding to the current search information based on the context information, the fuzzy inference search navigation based on the RNN is realized, namely, the inference search and the fuzzy search can be realized by utilizing the method provided by the embodiment of the disclosure.

In the embodiment of the disclosure, the inference search is a method for performing semantic recognition to obtain extracted key information (for example, a current search semantic text and a historical search semantic text), and searching the extracted key information by using an RNN technology.

In the embodiment of the present disclosure, the fuzzy search refers to that when the search intention of the user is ambiguous, the search engine performs fuzzy matching on a query term (query, for example, current search information) of the user and content (doc, for example, target destination information) to be retrieved to find out content related to the query.

It should be noted that the method for predicting the probability that the current search semantic text belongs to the search semantic text category of the target object is not limited to the above-mentioned RNN semantic reasoning model, in other embodiments, the semantic similarity between the current search semantic text and each pre-constructed search semantic text category may be calculated, and the probability that the current search semantic text belongs to the search semantic text category of the target object is generated according to the semantic similarity, that is, the greater the semantic similarity is, the greater the corresponding probability is, and conversely, the smaller the semantic similarity is, the smaller the corresponding probability is.

In step S330, a target search semantic text category of the current search semantic text is determined according to a probability that the current search semantic text belongs to the search semantic text category of the target object.

For example, the search semantic text category with the highest probability may be used as the target search semantic text category of the current search semantic text, but the disclosure is not limited thereto.

In step S340, historical destination information of the target object under the target search semantic text category is acquired.

In an exemplary embodiment, obtaining historical destination information of the target object in the target search semantic text category may include: obtaining historical navigation information of the target object and a historical search semantic text corresponding to the historical navigation information, wherein the historical navigation information comprises historical destination information of historical navigation, and the historical destination information of the historical navigation comprises a historical destination name of the historical navigation, historical destination interest point information, a parcel of a historical destination position, parcel attributes of the historical destination and historical destination attributes; determining a search semantic text category to which the historical search semantic text belongs and a historical destination name corresponding to the search semantic text category according to the historical destination name of the historical navigation, the historical destination interest point information, a section where the historical destination is located, a section attribute where the historical destination is located and the historical destination attribute; and obtaining the historical destination information of the target object under the target search semantic text type according to the search semantic text type to which the historical search semantic text belongs and the corresponding historical destination name.

In the embodiment of the present disclosure, the property of the section where the history destination is located refers to a property of the section where the history destination is located, and refers to a functional property of a certain large section, and the property of the history destination refers to a property of the history destination, and refers to a functional property of a certain small section. In the embodiment of the disclosure, the search semantic text category to which the history search semantic text belongs and the history destination name corresponding to the search semantic text category are determined by combining the segment attribute to which the history destination belongs and the history destination attribute, so that the accuracy of determining the search semantic text category to which the history search semantic text belongs can be improved, and furthermore, after the target destination information is subsequently determined, the accuracy of navigation positioning can be improved based on the segment attribute to which the history destination belongs and the history destination attribute which are contained in the history navigation information.

In step S350, target destination information of the current search semantic text is determined from the historical destination information of the target object.

In an exemplary embodiment, determining the target destination information of the current search semantic text from the historical destination information of the target object may include: obtaining historical operation behavior information of the target object aiming at the historical destination information; inputting the historical operation behavior information into a destination prediction model so as to output a probability corresponding to the historical destination information through the destination prediction model; sequentially displaying the historical destination information in a descending order according to the probability corresponding to the historical destination information; determining the target destination information from the historical destination information in response to an operation instruction on the historical destination information.

In the embodiment of the disclosure, the destination prediction model is obtained by training a machine learning model based on historical operation behavior information for historical destination information in historical navigation information of a target object, and can be used for predicting the probability of a historical destination navigated by the target object before. The historical operation behavior information may include relevant operation behavior information of the target object in the historical navigation, such as clicking behavior of the user on the historical destination name, payment behavior of the historical navigation, and the like. For example, the LR (Logistic Regression) algorithm may be used as the machine learning model, but the present disclosure is not limited thereto.

For example, because when each search semantic text category is constructed, the historical destination information corresponding to each search semantic text category can be determined, and the search semantic text category includes the target search semantic text category, after the target search semantic text category of the current search semantic text is determined, the historical destination information under the target search semantic text category can be found, the probability of each piece of historical destination information under the target search semantic text category can be determined according to the information such as the number of times, time and the like when the target object navigates to reach the historical destination information in the historical navigation, and then each piece of historical destination information under the target search semantic text category can be sequentially displayed in the navigation interface according to the sequence from large to small of the probability of each piece of historical destination information under the target search semantic text category, for selection by the user, the historical destination information selected by the user is referred to as target destination information.

In step S360, according to the historical navigation information of the target object corresponding to the target destination information, target navigation planning information for the target object to reach the target destination information is generated.

In an exemplary embodiment, generating target navigation planning information for the target object to reach the target destination information according to the historical navigation information of the target object corresponding to the target destination information may include: inputting historical navigation information of the target object corresponding to the target destination information into a navigation planning model so as to select the target navigation planning information from the historical navigation information of the target object corresponding to the target destination information through the navigation planning model; the historical navigation information of the target object corresponding to the target destination information comprises historical navigation track information, historical geographic position information and historical destination information thereof.

In the embodiment of the disclosure, the navigation planning model is obtained by training a machine learning model based on historical navigation information of a target object, and the navigation planning model can be used for predicting target navigation planning information of target destination information. Based on the historical navigation information of the target object, an optimal route is selected as a target navigation route for the target destination information, and the target navigation planning information is obtained by planning the target navigation route.

The navigation method provided by the disclosed embodiment, on one hand, by obtaining the current search semantic text input by the target object, predicting the probability that the current search semantic text belongs to the search semantic text category of the target object, determining the target search semantic text category of the current search semantic text according to the probability that the current search semantic text belongs to the search semantic text category of the target object, performing semantic reasoning on the current search semantic text input by the target object, predicting the semantic similarity degree between the current search semantic text category and each search semantic text category of the target object (represented by the probability between the current search semantic text category and the search semantic text category), so that the navigation can be realized through the semantic reasoning, and the real intention behind the current search semantic text input by the target object can be deduced even if the destination place name input by the target object is not the destination place name to be visited, therefore, the navigation accuracy can be improved, multiple times of manual input or modification of input information of the target object are not needed, and the user experience is improved; on one hand, the historical destination information of the target object under the category of the target search semantic text is obtained to determine the target destination information of the current search semantic text from the historical destination information of the target object, so that the target navigation planning information of the target object reaching the target destination information can be generated according to the historical navigation information of the target object corresponding to the target destination information, namely, the semantic reasoning and the historical navigation information of the target object are combined, the target navigation planning information which is most suitable for the target object can be planned, and the navigation accuracy is further improved.

In the following, a text sequence prediction model is taken as an RNN semantic reasoning model, and the probability that the current search semantic text belongs to the search semantic text category of the user is predicted based on the RNN semantic reasoning model for illustration. The method provided by the embodiment of the disclosure can be a fuzzy inference navigation method based on an RNN semantic reasoning model, can perform semantic reasoning by using the RNN semantic reasoning model, acquires historical destination information which the user has gone to by combining a historical track of the user stored in a map, constructs at least one search semantic text category, trains the constructed RNN semantic reasoning model based on the historical search semantic text and the search semantic text categories corresponding to the historical search semantic text, is used for calculating the probability that the current search semantic text of the user belongs to each search semantic text category, is used for determining the target search semantic text category to which the current search semantic text belongs, and then calculates the probability of the historical destination information which the user has gone to under the target search semantic text category. And sorting according to the probability of the historical destination information from large to small, and displaying on a map mobile client (map application program) of the user according to the sequence for the user to select. And after the user selects the target destination information at the map mobile client, the map background generates an optimal navigation route for navigation according to the geographical position information matched with the target destination information.

The method provided by the embodiment of the disclosure can mainly comprise the following nine stages: the method comprises a data acquisition stage, a place-related semantic reasoning text word bank construction stage, a training sample and test sample construction stage, an RNN semantic reasoning model training and testing stage, a search word input and prediction sample extraction stage, an RNN semantic reasoning model prediction stage, a geographic position information matching stage and a target destination determination and navigation stage. The overall flow is shown in fig. 4.

Step S401 is a data acquisition phase. Historical navigation information of a user is obtained from a map background management system, and the historical navigation information can comprise historical navigation track information, historical geographic position information and historical destination information. In the subsequent steps, historical navigation track information, historical geographic position information and historical destination information can be used as text information data, and training samples and test samples are constructed according to the text information data and are used for constructing an RNN semantic reasoning model.

Wherein the historical navigation track information may include, but is not limited to: the navigation method includes the steps of passing by the location name in each navigation, the area (province, city, county, town, street, village and the like) of the location, the name of the navigation starting location, the POI (Point of Interest) information of the navigation starting location, the name of the navigation ending location (which can also be called history destination name), the POI information of the navigation ending location (which can also be called history destination POI information), the total mileage of the navigation track in each navigation, the time consumption of each navigation, the average speed per hour of each navigation, the highest speed per hour of each navigation, the lowest speed per hour of each navigation and the like.

Historical geographic location information may include, but is not limited to: and point POI information, point longitude and latitude information and the like of each navigation track.

Historical destination information may include, but is not limited to: historical destination point names, historical destination point POI information, a parcel of historical destination locations, parcel attributes of historical destinations, historical destination attributes, and the like.

Voice or text information input by a user in a map search box during each historical navigation is collected, and the user can input text in the map search box, such as: a "go home" text is input in the search box, or a "go home" voice is input as history search information through the map voice robot. The map background collects texts through a search box or collects voice contents through a voice robot and converts the voice contents into text information, and after semantic extraction processing is carried out, historical search semantic texts can be obtained.

Step S402 is a semantic reasoning text word bank construction stage related to the place.

Specifically, the text information data of the history navigation in step S401 is input, the text content related to the place in the text information data is marked, and the property category label of the text content is marked, so as to determine the search semantic text category label corresponding to the history search semantic text at each history time.

For example, if the historical destination name corresponds to a cell name, the corresponding label is "go home" to search for a semantic text category label; if the historical destination name corresponds to a mall name, the historical destination name is correspondingly marked as a 'shopping' search semantic text category label.

After the search semantic text category label corresponding to the historical search semantic text at each historical time is determined, a word bank comprising mapping relations among the historical search semantic text (text content related to the place), the search semantic text category label and the historical destination name can be constructed, and the word bank is called a semantic reasoning text word bank related to the place.

Step S403 is a training sample and test sample construction stage.

Specifically, the history search semantic text for each history time in step S402 is input and recorded as

Selecting one of them

Obtaining a time series sample set of T-1 phase

And the method is used for constructing a training sample and a testing sample.

For the

Randomly cut into training samples (e.g., assuming a ratio a, a being a real number greater than 0 and less than 1) and test samples (ratio 1-a) at a certain ratio, e.g., will

Randomly cutting into training samples: test sample = 8:2, i.e. training samples were obtained by random slicing at a ratio of 8:2

And a test specimen

. Semantic text search for t-period history

And the evaluation sample is used for evaluating the trained and tested RNN semantic reasoning model.

Step S404 is an RNN semantic reasoning model construction phase. An RNN semantic reasoning model is constructed and expressed by the following formula:

（1）

wherein the content of the first and second substances,

representing the input vector at time T-1,

represents the output vector at time T-1,

a hidden layer vector representing the T-1 time;

、

、

respectively representing a hidden input parameter matrix, a hidden parameter matrix and an input hidden parameter matrix at the T-1 th moment;

，

respectively representing a hidden parameter vector and an output parameter vector at the T-1 th moment;

and

respectively, represent activation functions.

In the embodiments of the present disclosure, it is,

it is possible to use a tanh (hyperbolic tangent function) or a ReLU (Rectified Linear Unit) function,

sigmoid functions may be employed. Initial state, hidden layer vector of hidden layer

Using the true input vector of the first phase (time T-n)

. In the embodiment of the present disclosure, it is assumed that n periods are used, T-1 represents the last period of the T-th period, and T-2 represents the last period of the T-th period.

Wherein an activation function is a function added to an artificial neural network intended to help the network learn complex patterns in the data. Similar to neuron-based models in the human brain, the activation function ultimately determines what is to be transmitted to the next neuron. In an artificial neural network, the activation function of a node defines the output of the node at a given input or set of inputs.

Wherein the hyperbolic sine function (sinh) is expressed as:

（2）

its definition domain and value domain are respectively

。

The hyperbolic cosine function (cosh) is expressed as:

（3）

it has the definition domain of

Value range of

When x =0, the minimum value 1 is taken.

The hyperbolic tangent function (tanh) is computationally equal to the ratio of hyperbolic sine (sinh) to hyperbolic cosine (cosh), i.e., tanh (x) = sinh (x)/cosh (x).

Step S405 is a RNN semantic reasoning model training and testing stage.

Specifically, input training samples

And a test specimen

To RNN semantic reasoning model to obtain RNN semantic reasoning model

、

、

、

，

And

。

inputting evaluation samples in a trained and tested RNN semantic reasoning model

And the method is used for evaluating the trained and tested RNN semantic reasoning model. If the model accuracy reaches more than 90 percent (the threshold value of the confidence interval), the model is passed, otherwise, the sample is put back to random sampling for continuous sampling training until the model reaches the standard. Using the evaluation sample

The following can be obtained by calculation:

（4）

wherein the Confidence interval (Confidence interval) refers to the estimated interval of the overall parameter constructed by the sample statistics. In statistics, the confidence interval for a probability sample is an interval estimate for some overall parameter of the sample.

Step S406 is a map search box search term input stage. And the user inputs a search word in the map search box at the current moment as current search information.

Step S407 is a prediction sample extraction stage.

And extracting a subject, a predicate, an object and the like of the current search information by adopting an NLP technology. And representing the extracted text content as a current search semantic text as a prediction sample

。

For example: assuming that the user inputs 'I want to go home' in the map search box at the current moment, relevant key information 'go home' is extracted as a prediction sample through an NLP technology

。

Step S408 is the RNN semantic inference model prediction phase.

The prediction sample of step S407 is input

Parameter matrix of step S405

、

、

、

，

And

using the following formula:

（5）

obtaining the probability that the current search semantic text belongs to each search semantic text category of the target object

。

Step S409 is a stage of calculating probabilities of prediction samples and classes (abbreviation of search semantic text class).

According to the semantic reasoning text word bank related to the place constructed in the step S402, the probability of each search semantic text category to which the current search semantic text corresponding to the current search information input by the user at the current time belongs is obtained, and is expressed as: { Userid, search semantic text category, probability }. Userid represents a unique identification of the user.

Step S410 is a stage of matching the geographical location information by using the geographical location information database.

The disclosed embodiment can construct a geographic position information database based on historical navigation information, and the geographic position information database can include but is not limited to: historical destination names, areas where historical destinations are located, area attributes where historical destinations are located, historical destination attributes, historical destination POI information, and semantic text related to places (i.e., historical search semantic text).

Inputting the { Userid, search semantic text category, probability } of step S409 and the semantic reasoning text lexicon related to the place constructed in step S402, determining the search semantic text category with the highest probability in the { Userid, search semantic text category, probability } of step S409 as a target search semantic text category, matching based on the semantic reasoning text lexicon related to the place constructed in step S402, and obtaining each historical destination information, such as a historical destination name, under the target search semantic text category, namely: { Userid, target search semantic text category, historical destination name }.

And matching the { Userid, the target search semantic text category and the historical destination name } with a geographic position information database to obtain { Userid, the target search semantic text category, the historical destination name, the fragment of the historical destination, the fragment attribute of the historical destination, the historical destination attribute history and the destination POI information }.

Step S411, obtaining the probability of the historical destination under the corresponding category, and matching the historical destination under the corresponding category with the geographic position information database to obtain the underground corresponding probability and the geographic position information of each historical destination.

Specifically, after the category of the current search semantic text is determined by the RNN semantic reasoning model according to the probability of the category, a destination prediction model can be constructed by using LR through the historical operation behavior information such as click behavior, payment behavior, and the like of the user for each historical destination name, so as to perform probability prediction of each historical destination name in the target search semantic text category, predict the probability of each historical destination name in the target search semantic text category, and sort the probabilities in a descending order. If there is only one historical destination name under the target search semantic text category, the destination prediction model may not be constructed.

Step S412, sequentially recommending the historical destinations and the POI information corresponding to the search words to the user map client according to the sequence of the probabilities of the historical destination names under the target search semantic text category from large to small for the user to select.

Step S413 is a target destination determination and navigation phase.

And when the user selects a target destination from the recommended historical destination names, namely the target destination information is determined, the map background generates an optimal navigation route for navigation according to the geographic position information of the matched target destination.

Specifically, after the target destination name is determined, historical navigation track information, historical geographic position information and historical destination information corresponding to the target destination name are input into a trained navigation planning model, a softmax algorithm is constructed, a route with the maximum probability is selected from multiple routes, so that target navigation planning information is generated, and navigation is performed.

The navigation method provided by the embodiment of the invention can obtain the historical destination information under the category of the target search semantic text and predict the probability of each piece of historical destination information under the category of the target search semantic text based on the category of the target search semantic text and the pre-constructed word stock by acquiring the historical navigation track information, the historical geographic position information and the historical destination information of a user as text information data, constructing a training sample and a testing sample according to the text information for constructing an RNN semantic reasoning model, then calculating the probability of the current search semantic text corresponding to the input current search information and the category of the search semantic text in the word stock according to the current search information input by a map search box and the RNN semantic reasoning model for determining the category of the target search semantic text to which the current search semantic text belongs, the method is used for determining the target destination information and generating the target navigation planning information based on the target destination information, so that accurate navigation can be realized through semantic reasoning. The embodiment of the disclosure uses the RNN semantic reasoning model in semantic reasoning, so that semantic understanding can be performed according to the context information to reason the real target destination information behind the fuzzy text information input by the user, that is, the embodiment of the disclosure has the capability of semantic similarity reasoning by combining the historical navigation information of the user, can improve the accuracy of navigation positioning, and improves the user experience.

As shown in fig. 5, the method provided by the embodiment of the present disclosure may include the following steps.

In step S510, a navigation interface including a destination input area is displayed.

In step S520, in response to the input instruction to the destination input area, current search information input by the target object is acquired, where the current search information includes a current search semantic text.

In step S530, according to the current search information, displaying, on the navigation interface, historical destination information belonging to the same target search semantic text category as the current search semantic text, where the historical destination information is destination information in the historical navigation information of the target object.

In step S540, in response to an operation instruction to the displayed history destination information, target destination information is determined from the history destination information.

In step S550, target navigation planning information for the target object to reach the target destination information is displayed according to the target destination information.

The specific implementation of the embodiment of fig. 5 can refer to the contents of the other embodiments described above.

For example, as shown in fig. 6, when the user inputs "go home", the navigation interface displays the historical destination information that the user has historically navigated: and the user can select one of the XX building, the YY building and the ZZ building to be used as target destination information to realize the accurate navigation.

Further, please refer to fig. 7, fig. 7 is a schematic view of a navigation device according to an embodiment of the disclosure. The navigation means may be a computer program (comprising program code etc.) running on a computer device, for example the navigation means may be an application software; the apparatus may be used to perform the corresponding steps in the methods provided by the embodiments of the present disclosure. As shown in fig. 7, the navigation device 700 may be used in a computer device, and specifically, the navigation device 700 may include: a first obtaining unit 710, a prediction unit 720, a first determining unit 730, and a generating unit 740.

The first obtaining unit 710 may be configured to obtain a current search semantic text input by the target object.

The prediction unit 720 may be configured to predict a probability that the current search semantic text belongs to the search semantic text category of the target object.

The first determining unit 730 may be configured to determine a target search semantic text category of the current search semantic text according to a probability that the current search semantic text belongs to the search semantic text category of the target object.

The first obtaining unit 710 may be further configured to obtain historical destination information of the target object in the target search semantic text category.

The first determining unit 730 may be further configured to determine target destination information of the current search semantic text from the historical destination information of the target object.

The generating unit 740 may be configured to generate target navigation planning information for the target object to reach the target destination information according to the historical navigation information of the target object corresponding to the target destination information.

In an exemplary embodiment, the prediction unit 720 may be further configured to perform the following steps: obtaining a text sequence prediction model of a hidden input parameter matrix, a hidden parameter vector, an input hidden parameter matrix, an output parameter vector and a previous hidden layer vector which are determined by training; inputting the current search semantic text at the current time T +1 into the text sequence prediction model, so as to obtain a current hidden layer vector according to the hidden input parameter matrix, the current search semantic text, the hidden parameter matrix, the previous hidden layer vector and the hidden parameter vector, and obtain the probability that the current search semantic text belongs to each search semantic text category of the target object according to the input hidden parameter matrix, the current hidden layer vector and the output parameter vector.

In an exemplary embodiment, the navigation device 700 may further include: the obtaining unit can be used for obtaining historical search semantic texts of the target object from the T-1 th time to the T-n th time and corresponding search semantic text category labels; the obtaining unit may be further configured to input the historical search semantic texts of the target object at times T-1 to T-n to the text sequence prediction model, and obtain probabilities that the historical search semantic texts of the target object at times T-1 to T-n belong to search semantic text categories of the target object, respectively; the training unit can be used for training the text sequence prediction model according to the probabilities that the historical search semantic texts of the target object from the T-1 th moment to the T-n th moment respectively belong to the search semantic text categories of the target object and the corresponding search semantic text category labels; the obtaining unit may be further configured to obtain a historical search semantic text of the target object at a tth time and a search semantic text category tag corresponding to the historical search semantic text; the obtaining unit may be further configured to input the historical search semantic text of the target object at the tth time into the trained text sequence prediction model, and obtain a probability that the historical search semantic text of the target object at the tth time belongs to the search semantic text category of the target object; and the evaluation unit may be configured to evaluate the text sequence prediction model according to the probability that the historical search semantic text of the target object at the T-th time belongs to the search semantic text category of the target object and the search semantic text category label corresponding to the probability, so as to determine the hidden input parameter matrix, the hidden parameter vector, the input hidden parameter matrix, the output parameter vector, and the previous hidden layer vector of the text sequence prediction model.

In an exemplary embodiment, the first obtaining unit 710 may be further configured to perform the following steps: obtaining historical navigation information of the target object and a historical search semantic text corresponding to the historical navigation information, wherein the historical navigation information comprises historical destination information of historical navigation, and the historical destination information of the historical navigation comprises a historical destination name of the historical navigation, historical destination interest point information, a parcel of a historical destination position, parcel attributes of the historical destination and historical destination attributes; determining a search semantic text category to which the historical search semantic text belongs and a historical destination name corresponding to the search semantic text category according to the historical destination name of the historical navigation, the historical destination interest point information, a section where the historical destination is located, a section attribute where the historical destination is located and the historical destination attribute; and obtaining the historical destination information of the target object under the target search semantic text type according to the search semantic text type to which the historical search semantic text belongs and the corresponding historical destination name.

In an exemplary embodiment, the first determining unit 730 may be further configured to perform the following steps: obtaining historical operation behavior information of the target object aiming at the historical destination information; inputting the historical operation behavior information into a destination prediction model so as to output a probability corresponding to the historical destination information through the destination prediction model; sequentially displaying the historical destination information in a descending order according to the probability corresponding to the historical destination information; determining the target destination information from the historical destination information in response to an operation instruction on the historical destination information.

In an exemplary embodiment, the generating unit 740 may be further configured to perform the following steps: inputting historical navigation information of the target object corresponding to the target destination information into a navigation planning model so as to select the target navigation planning information from the historical navigation information of the target object corresponding to the target destination information through the navigation planning model; the historical navigation information of the target object corresponding to the target destination information comprises historical navigation track information, historical geographic position information and historical destination information thereof.

In an exemplary embodiment, the first obtaining unit 710 may be further configured to perform the following steps: acquiring current search information input by the target object; and performing semantic extraction processing on the current search information to obtain the current search semantic text.

The embodiment of the disclosure provides a navigation device, which can be operated in user equipment.

Further, please refer to fig. 8, fig. 8 is a schematic view of a navigation device according to an embodiment of the disclosure. The navigation means may be a computer program (comprising program code etc.) running on a computer device, for example the navigation means may be an application software; the apparatus may be used to perform the corresponding steps in the methods provided by the embodiments of the present disclosure. As shown in fig. 8, the navigation apparatus 800 may be used in a computer device, and specifically, the navigation apparatus 800 may include: a display unit 810, a second acquisition unit 820, and a second determination unit 830.

The display unit 810 may be used to display a navigation interface including a destination input area.

The second obtaining unit 820 may be configured to obtain current search information input by the target object in response to an input instruction to the destination input area, where the current search information includes a current search semantic text.

The display unit 810 may be further configured to display, on the navigation interface, historical destination information that belongs to the same target search semantic text category as the current search semantic text according to the current search information, where the historical destination information is destination information in the historical navigation information of the target object.

The second determining unit 830 may be configured to determine target destination information from the history destination information in response to an operation instruction to the displayed history destination information.

The display unit 810 may be further configured to display target navigation planning information for the target object to reach the target destination information according to the target destination information.

Referring to fig. 9, fig. 9 is a schematic structural diagram of a computer device according to an embodiment of the present disclosure. As shown in fig. 9, the computer device in the embodiment of the present disclosure may include: one or more processors 901, memory 902, and input-output interface 903. The processor 901, the memory 902, and the input/output interface 903 are connected by a bus 904. The memory 902 is used for storing a computer program, which includes program instructions, and the input/output interface 903 is used for receiving data and outputting data, for example, for data interaction between a host and a computer device, or for data interaction between virtual machines in the host; the processor 901 is configured to execute program instructions stored in the memory 902.

The processor 901 may perform the following operations:

acquiring a current search semantic text input by a target object;

predicting the probability that the current search semantic text belongs to the search semantic text category of the target object;

determining the target search semantic text category of the current search semantic text according to the probability that the current search semantic text belongs to the search semantic text category of the target object;

acquiring historical destination information of the target object under the category of the target search semantic text;

determining target destination information of the current search semantic text from historical destination information of the target object;

and generating target navigation planning information for the target object to reach the target destination information according to the historical navigation information of the target object corresponding to the target destination information.

Alternatively, the processor 901 may perform the following operations:

displaying a navigation interface, the navigation interface including a destination input area;

responding to an input instruction of the destination input area, and acquiring current search information input by a target object, wherein the current search information comprises a current search semantic text;

according to the current search information, displaying historical destination information which belongs to the same target search semantic text category as the current search semantic text on the navigation interface, wherein the historical destination information is destination information in the historical navigation information of the target object;

determining target destination information from the history destination information in response to an operation instruction for the displayed history destination information;

and displaying target navigation planning information of the target object reaching the target destination information according to the target destination information.

In some possible embodiments, the processor 901 may be a Central Processing Unit (CPU), and the processor may also be other general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, and the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The memory 902 may include a read-only memory and a random access memory, and provides instructions and data to the processor 901 and the input/output interface 903. A portion of the memory 902 may also include non-volatile random access memory. For example, memory 902 may also store device type information.

In a specific implementation, the computer device may execute, through each built-in functional module, an implementation manner provided in each step in the foregoing embodiments, which may be referred to specifically for the implementation manner provided in each step in the foregoing embodiments, and details are not described herein again.

The disclosed embodiments provide a computer device, including: the navigation device comprises a processor, an input/output interface and a memory, wherein the processor acquires a computer program in the memory, and executes each step of the method shown in the embodiment to perform navigation operation.

The embodiments of the present disclosure further provide a computer-readable storage medium, where a computer program is stored, where the computer program is suitable for being loaded by the processor and executing the navigation method provided in each step in the foregoing embodiments, and specific reference may be made to implementation manners provided in each step in the foregoing embodiments, which are not described herein again. In addition, the beneficial effects of the same method are not described in detail. For technical details not disclosed in embodiments of the computer-readable storage medium to which the present disclosure relates, refer to the description of embodiments of the method of the present disclosure. By way of example, a computer program can be deployed to be executed on one computer device or on multiple computer devices at one site or distributed across multiple sites and interconnected by a communication network.

The computer readable storage medium may be the navigation device provided in any of the foregoing embodiments or an internal storage unit of the computer device, such as a hard disk or a memory of the computer device. The computer readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a Smart Memory Card (SMC), a Secure Digital (SD) card, a flash card (flash card), and the like, provided on the computer device. Further, the computer-readable storage medium may also include both an internal storage unit and an external storage device of the computer device. The computer-readable storage medium is used for storing the computer program and other programs and data required by the computer device. The computer readable storage medium may also be used to temporarily store data that has been output or is to be output.

Embodiments of the present disclosure also provide a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions to cause the computer device to perform the method provided in the various alternatives in the embodiments described above.

The terms "first," "second," and the like in the description and in the claims and the drawings of the embodiments of the present disclosure are used for distinguishing between different objects and not for describing a particular order. Furthermore, the terms "comprises" and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, product, or apparatus that comprises a list of steps or elements is not limited to the listed steps or modules, but may alternatively include other steps or modules not listed or inherent to such process, method, apparatus, product, or apparatus.

Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the examples described in connection with the embodiments disclosed herein may be embodied in electronic hardware, computer software, or combinations of both, and that the components and steps of the examples have been described in a functional general in the specification for the purpose of clearly illustrating the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present disclosure.

The method and the related apparatus provided by the embodiments of the present disclosure are described with reference to the flowchart and/or the structural diagram of the method provided by the embodiments of the present disclosure, and each flow and/or block of the flowchart and/or the structural diagram of the method, and the combination of the flow and/or block in the flowchart and/or the block diagram can be specifically implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable navigation device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable navigation device, create means for implementing the functions specified in the flowchart flow or flows and/or block or blocks of the block diagram. These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable navigation device to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block or blocks of the block diagram. These computer program instructions may also be loaded onto a computer or other programmable navigation device to cause a series of operational steps to be performed on the computer or other programmable device to produce a computer implemented process such that the instructions which execute on the computer or other programmable device provide steps for implementing the functions specified in the flowchart flow or flows and/or block or blocks.

The disclosure of the present invention is not intended to be limited to the particular embodiments disclosed, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A navigation method, comprising:

acquiring historical navigation information of a target object and a corresponding historical search semantic text, wherein the historical navigation information comprises historical destination information of historical navigation, and the historical destination information of the historical navigation comprises a historical destination name of the historical navigation, historical destination interest point information, a parcel of a historical destination position, parcel attributes of the historical destination and historical destination attributes;

determining a search semantic text category to which the historical search semantic text belongs and a historical destination name corresponding to the search semantic text category according to the historical destination name of the historical navigation, the historical destination interest point information, a section where the historical destination is located, a section attribute where the historical destination is located and the historical destination attribute;

acquiring a current search semantic text input by the target object;

2. The method of claim 1, wherein predicting a probability that the current search semantic text belongs to a search semantic text category of the target object comprises:

obtaining a text sequence prediction model of a hidden input parameter matrix, a hidden parameter vector, an input hidden parameter matrix, an output parameter vector and a previous hidden layer vector which are determined by training;

inputting the current search semantic text at the current time T +1 into the text sequence prediction model to obtain a current hidden layer vector according to the hidden input parameter matrix, the current search semantic text, the hidden parameter matrix, the previous hidden layer vector and the hidden parameter vector, and obtaining the probability that the current search semantic text belongs to each search semantic text category of the target object according to the input hidden parameter matrix, the current hidden layer vector and the output parameter vector, wherein T is an integer greater than 1.

3. The method of claim 2, further comprising:

obtaining a historical search semantic text of the target object from the T-1 th moment to the T-n th moment and a corresponding search semantic text category label, wherein n is an integer larger than 0;

respectively inputting the historical search semantic texts of the target object from the T-1 th moment to the T-n th moment to the text sequence prediction model to obtain the probabilities that the historical search semantic texts of the target object from the T-1 th moment to the T-n th moment respectively belong to the search semantic text categories of the target object;

training the text sequence prediction model according to the probability that the historical search semantic texts of the target object from the T-1 th moment to the T-n th moment respectively belong to the search semantic text categories of the target object and the corresponding search semantic text category labels;

obtaining a historical search semantic text of the target object at the Tth moment and a corresponding search semantic text category label;

inputting the historical search semantic text of the target object at the Tth moment into the trained text sequence prediction model, and obtaining the probability that the historical search semantic text of the target object at the Tth moment belongs to the search semantic text category of the target object;

and evaluating the text sequence prediction model according to the probability that the historical search semantic text of the target object at the T-th moment belongs to the search semantic text category of the target object and the corresponding search semantic text category label thereof so as to determine the hidden input parameter matrix, the hidden parameter vector, the input hidden parameter matrix, the output parameter vector and the previous hidden layer vector of the text sequence prediction model.

4. The method of claim 1, wherein obtaining historical destination information for the target object in the target search semantic text category comprises:

obtaining historical navigation information of the target object and a historical search semantic text corresponding to the historical navigation information, wherein the historical navigation information comprises historical destination information of historical navigation, and the historical destination information of the historical navigation comprises a historical destination name of the historical navigation, historical destination interest point information, a parcel of a historical destination position, parcel attributes of the historical destination and historical destination attributes;

and obtaining the historical destination information of the target object under the target search semantic text type according to the search semantic text type to which the historical search semantic text belongs and the corresponding historical destination name.

5. The method of claim 1, wherein determining target destination information for the current search semantic text from historical destination information for the target object comprises:

obtaining historical operation behavior information of the target object aiming at the historical destination information;

inputting the historical operation behavior information into a destination prediction model so as to output a probability corresponding to the historical destination information through the destination prediction model;

sequentially displaying the historical destination information in a descending order according to the probability corresponding to the historical destination information;

determining the target destination information from the historical destination information in response to an operation instruction on the historical destination information.

6. The method of claim 1, wherein generating target navigation planning information for the target object to reach the target destination information according to historical navigation information of the target object corresponding to the target destination information comprises:

inputting historical navigation information of the target object corresponding to the target destination information into a navigation planning model so as to select the target navigation planning information from the historical navigation information of the target object corresponding to the target destination information through the navigation planning model;

the historical navigation information of the target object corresponding to the target destination information comprises historical navigation track information, historical geographic position information and historical destination information thereof.

7. The method of claim 1, wherein obtaining current search semantic text entered by a target object comprises:

acquiring current search information input by the target object;

and performing semantic extraction processing on the current search information to obtain the current search semantic text.

8. A navigation method, comprising:

according to the current search information, displaying historical destination information which belongs to the same target search semantic text category as the current search semantic text on the navigation interface, wherein the historical destination information is destination information in the historical navigation information of the target object, the historical navigation information of the target object comprises historical destination information of historical navigation, and the historical destination information of the historical navigation comprises historical destination names of the historical navigation, historical destination interest point information, a section where a historical destination position is located, a section attribute where the historical destination is located and historical destination attributes; the search semantic text category to which the history search semantic text of the target object belongs is determined according to the history destination name of the history navigation, the history destination interest point information, the fragment of the history destination position, the fragment attribute of the history destination and the history destination attribute, and the search semantic text category of the target object comprises the target search semantic text category;

9. A navigation device, comprising:

the device comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring historical navigation information of a target object and a historical search semantic text corresponding to the historical navigation information, the historical navigation information comprises historical destination information of historical navigation, and the historical destination information of the historical navigation comprises a historical destination name of the historical navigation, historical destination interest point information, a section where a historical destination position is located, a section attribute where the historical destination is located and a historical destination attribute, and acquiring a current search semantic text input by the target object;

the prediction unit is used for determining a search semantic text category to which the historical search semantic text belongs and a corresponding historical destination name thereof according to the historical destination name of the historical navigation, the historical destination interest point information, the segment where the historical destination is located, the segment attribute of the historical destination and the historical destination attribute, and predicting the probability that the current search semantic text belongs to the search semantic text category of the target object;

the first determining unit is used for determining the target search semantic text category of the current search semantic text according to the probability that the current search semantic text belongs to the search semantic text category of the target object;

the first obtaining unit is further used for obtaining historical destination information of the target object under the category of the target search semantic text;

the first determining unit is further used for determining target destination information of the current search semantic text from historical destination information of the target object;

and the generating unit is used for generating target navigation planning information of the target object reaching the target destination information according to the historical navigation information of the target object corresponding to the target destination information.

10. The apparatus of claim 9, wherein the prediction unit is to perform the steps of:

11. The apparatus of claim 10, further comprising:

the obtaining unit is used for obtaining historical search semantic texts of the target object from the T-1 th time to the T-n th time and corresponding search semantic text category labels, wherein n is an integer larger than 0;

the training unit is used for training the text sequence prediction model according to the probability that the historical search semantic texts of the target object from the T-1 th moment to the T-n th moment respectively belong to the search semantic text categories of the target object and the corresponding search semantic text category labels;

the obtaining unit is further configured to obtain a historical search semantic text of the target object at the tth moment and a search semantic text category label corresponding to the historical search semantic text;

the obtaining unit is further configured to input the historical search semantic text of the target object at the tth moment to the trained text sequence prediction model, and obtain a probability that the historical search semantic text of the target object at the tth moment belongs to the search semantic text category of the target object;

and the evaluation unit is used for evaluating the text sequence prediction model according to the probability that the historical search semantic text of the target object at the T-th moment belongs to the search semantic text category of the target object and the corresponding search semantic text category label so as to determine the hidden input parameter matrix, the hidden parameter vector, the input hidden parameter matrix, the output parameter vector and the previous hidden layer vector of the text sequence prediction model.

12. The apparatus as claimed in claim 9, wherein said first obtaining unit is further configured to perform the steps of:

13. The apparatus as recited in claim 9, wherein said first determining unit is further configured to:

14. The apparatus of claim 9, wherein the generating unit is further configured to perform the steps of:

15. The apparatus as claimed in claim 9, wherein said first obtaining unit is further configured to perform the steps of:

acquiring current search information input by the target object;

16. A navigation device, comprising:

a display unit for displaying a navigation interface including a destination input area;

the second acquisition unit is used for responding to an input instruction of the destination input area and acquiring current search information input by a target object, wherein the current search information comprises a current search semantic text;

the display unit is further used for displaying historical destination information which belongs to the same target search semantic text category as the current search semantic text on the navigation interface according to the current search information, the historical destination information is destination information in the historical navigation information of the target object, the historical navigation information of the target object comprises historical destination information of historical navigation, and the historical destination information of the historical navigation comprises historical destination names of the historical navigation, historical destination interest point information, a section where a historical destination position is located, a section attribute where the historical destination is located and a historical destination attribute; the search semantic text category to which the history search semantic text of the target object belongs is determined according to the history destination name of the history navigation, the history destination interest point information, the fragment of the history destination position, the fragment attribute of the history destination and the history destination attribute, and the search semantic text category of the target object comprises the target search semantic text category;

a second determination unit configured to determine target destination information from the history destination information in response to an operation instruction to the displayed history destination information;

the display unit is further used for displaying target navigation planning information of the target object reaching the target destination information according to the target destination information.

17. A computer device comprising a processor, a memory, an input output interface;

the processor is connected to the memory and the input/output interface, respectively, wherein the input/output interface is configured to receive data and output data, the memory is configured to store a computer program, and the processor is configured to call the computer program to enable the computer device to execute the method according to any one of claims 1 to 7 or the method according to claim 8.

18. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program adapted to be loaded and executed by a processor to cause a computer device having the processor to perform the method of any of claims 1-7 or the method of claim 8.