CN-120780787-B - Geographic information retrieval method, device, equipment and medium based on large language model

CN120780787BCN 120780787 BCN120780787 BCN 120780787BCN-120780787-B

Abstract

The embodiment of the disclosure discloses a geographic information retrieval method, a geographic information retrieval device, geographic information retrieval equipment and a geographic information retrieval medium based on a large language model. The method comprises the steps of obtaining target geographic text information input by a target user terminal, carrying out recognition processing on the target geographic text information based on a semantic analyzer to generate a text feature information set, obtaining preset generation rules, generating a space operation sequence based on the preset generation rules, carrying out search pretreatment on the space operation sequence based on the space operation sequence to generate a pretreated space operation sequence, carrying out stage search processing based on the pretreated space operation sequence to generate geographic entity information, generating a geographic entity page corresponding to the geographic entity information, and sending the geographic entity page to the target user terminal for display. The embodiment accurately generates the user query result and avoids the waste of transmission resources.

Inventors

HE FENGYI
GAO SHIHU
YANG YAN
LU YANAN
LI QIYU
LIU MENG
GAO ZHEN
ZHENG BINYAN
LI PEI

Assignees

南方电网新能设计研究院(广东)有限公司

Dates

Publication Date: 20260505
Application Date: 20250530

Claims (8)

1. A geographic information retrieval method based on a large language model comprises the following steps: Acquiring target geographic text information input by a target user terminal; Based on a semantic analyzer, carrying out recognition processing on the target geographic text information to generate a text feature information set, wherein the text feature information in the text feature information set is one of a spatial relationship entity, geographic attribute information and logic relationship information; The semantic analyzer is used for identifying and processing the target geographic text information to generate a text characteristic information set, and the semantic analyzer comprises the following steps: performing geographic word pre-segmentation processing on the target geographic text information to generate pre-segmented geographic words so as to obtain a pre-segmented geographic word set; for each pre-segmentation geographic word in the pre-segmentation geographic word set, responding to the pre-segmentation geographic word meeting a first preset condition, and performing secondary segmentation processing on the pre-segmentation geographic word to obtain a segmentation geographic word; combining the rest pre-divided geographic words and the divided geographic words into a geographic word set; screening out each geographic word representing the logical connection from the geographic word set, and generating a corresponding virtual segmenter according to the logical connection represented by each selected geographic word; According to each generated virtual segmenter, carrying out replacement processing on the geographic word set to generate a geographic word set containing a nested logic structure, and obtaining a nested geographic word set; for each nested geographic word in the nested geographic word set, classifying the nested geographic word to generate a classified geographic word; Clustering the generated classified geographic words to generate a text feature information set; Acquiring a preset generation rule and generating a space operation sequence based on the preset generation rule, wherein the space operation sequence is used for comprising the arrangement sequence of each space operation, and the space operation in the space operation sequence comprises space range definition and attribute screening; Based on the space operation sequence, carrying out retrieval pretreatment on the space operation sequence to generate a pretreated space operation sequence; Based on the preprocessed space operation sequence, performing stage search processing to generate geographic entity information; generating a geographic entity page corresponding to the geographic entity information, and sending the geographic entity page to the target user terminal for display.
2. The method of claim 1, wherein after the semantic parser-based recognition processing of the target geographic text information to generate a set of text feature information, the method further comprises: Selecting text feature information representing a spatial relationship entity from the text feature information set as target feature information; and performing context error correction processing on the target characteristic information to generate the corrected target characteristic information.
3. The method of claim 2, wherein the performing a contextual error correction process on the target feature information to generate error corrected target feature information comprises: Responding to the classified geographic words included in the target feature information, wherein the classified geographic words have the geographic words with the characteristic directions, and matching the classified geographic words with the characteristic directions with pre-stored geographic entity topological relation data; Supplementing the classified geographic words of the characteristic direction based on the pre-stored geographic entity topological relation data obtained by matching to obtain the supplemented geographic words; Responding to the classified geographic words included in the target feature information, wherein the classified geographic words represent fuzzy distances, and for each classified geographic word representing the fuzzy distance, a distance threshold corresponding to the classified geographic word is generated based on a pre-trained regional density model; based on the generated distance threshold, carrying out replacement processing on the classified geographic words which represent the fuzzy distance; The target feature information after the addition and replacement processing is performed is determined as the error-corrected target feature information.
4. The method of claim 1, wherein the performing a phase retrieval process based on the pre-processed spatial operation sequence to generate geographic entity information comprises: performing an index retrieval operation based on the spatial semantic index to generate at least one retrieval result; And performing stage screening on the at least one search result to generate a screened search result sequence serving as geographic entity information.
5. The method of claim 4, wherein the performing phase screening on the at least one search result to generate a screened search result sequence as geographic entity information comprises: performing spatial screening processing on the at least one search result to generate a set of spatially screened search results; Performing time filtering processing on the space-screened search result set to generate a filtered search result; and sequencing the filtered search results according to a preset sequencing rule to generate a filtered search result sequence serving as geographic entity information.
6. A geographic information retrieval device based on a large language model, comprising: the first acquisition unit is configured to acquire target geographic text information input by a target user terminal; The identification unit is configured to perform identification processing on the target geographic text information based on a semantic analyzer to generate a text feature information set, wherein the text feature information in the text feature information set is one of a spatial relationship entity, geographic attribute information and logical relationship information, and the identification unit is further configured to: performing geographic word pre-segmentation processing on the target geographic text information to generate pre-segmented geographic words so as to obtain a pre-segmented geographic word set; for each pre-segmentation geographic word in the pre-segmentation geographic word set, responding to the pre-segmentation geographic word meeting a first preset condition, and performing secondary segmentation processing on the pre-segmentation geographic word to obtain a segmentation geographic word; combining the rest pre-divided geographic words and the divided geographic words into a geographic word set; screening out each geographic word representing the logical connection from the geographic word set, and generating a corresponding virtual segmenter according to the logical connection represented by each selected geographic word; According to each generated virtual segmenter, carrying out replacement processing on the geographic word set to generate a geographic word set containing a nested logic structure, and obtaining a nested geographic word set; for each nested geographic word in the nested geographic word set, classifying the nested geographic word to generate a classified geographic word; Clustering the generated classified geographic words to generate a text feature information set; A second obtaining unit configured to obtain a preset generation rule and generate a spatial operation sequence based on the preset generation rule, wherein the spatial operation sequence is used for including an arrangement sequence of each spatial operation, and the spatial operations in the spatial operation sequence include spatial range definition and attribute screening; a search preprocessing unit configured to perform search preprocessing on the spatial operation sequence based on the spatial operation sequence to generate a preprocessed spatial operation sequence; a stage retrieval unit configured to perform stage retrieval processing based on the preprocessed spatial operation sequence to generate geographical entity information; and the generation unit is configured to generate a geographic entity page corresponding to the geographic entity information and send the geographic entity page to the target user terminal for display.
7. An electronic device, comprising: One or more processors; A storage device having one or more programs stored thereon; when executed by the one or more processors, causes the one or more processors to implement the method of any of claims 1 to 5.
8. A computer readable medium having stored thereon a computer program, wherein the program when executed by a processor implements the method of any of claims 1 to 5.

Description

Geographic information retrieval method, device, equipment and medium based on large language model Technical Field The embodiment of the disclosure relates to the technical field of computers, in particular to a geographic information retrieval method, a geographic information retrieval device, geographic information retrieval equipment and geographic information retrieval media based on a large language model. Background When a user searches a geographic position, a relatively fuzzy statement is usually input to search. At present, when geographic position retrieval is carried out through fuzzy terms, the mode of fuzzy query or similarity query is carried out through fuzzy terms and a preset word stock, and the geographic position retrieval is generally carried out according to the probability. However, when the above manner is adopted for geographic location retrieval, there are often the following technical problems: The fuzzy statement input by the user cannot be processed, so that more results which are inconsistent with the retrieval purpose of the user are retrieved, more transmission resources are required to be consumed when the results are transmitted to the user, and the transmission resources are wasted. The above information disclosed in this background section is only for enhancement of understanding of the background of the inventive concept and, therefore, may contain information that does not form the prior art that is already known to those of ordinary skill in the art in this country. Disclosure of Invention The disclosure is in part intended to introduce concepts in a simplified form that are further described below in the detailed description. The disclosure is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. Some embodiments of the present disclosure propose geographic information retrieval methods, apparatus, devices, and computer-readable media based on a large language model to solve one or more of the technical problems mentioned in the background section above. In a first aspect, some embodiments of the present disclosure provide a geographic information retrieval method based on a large language model, where the method includes obtaining target geographic text information input by a target user terminal, performing recognition processing on the target geographic text information based on a semantic analyzer to generate a text feature information set, where the text feature information in the text feature information set is one of a spatial relationship entity, geographic attribute information, and logical relationship information, obtaining a preset generation rule, and generating a spatial operation sequence based on the preset generation rule, performing retrieval preprocessing on the spatial operation sequence based on the spatial operation sequence to generate a preprocessed spatial operation sequence, performing stage retrieval processing based on the preprocessed spatial operation sequence to generate geographic entity information, generating a geographic entity page corresponding to the geographic entity information, and transmitting the geographic entity page to the target user terminal for display. In a second aspect, some embodiments of the present disclosure provide a geographic information retrieval device based on a large language model, the device including a first acquisition unit configured to acquire target geographic text information input by a target user terminal, a recognition unit configured to perform recognition processing on the target geographic text information based on a semantic analyzer to generate a text feature information set, wherein the text feature information in the text feature information set is one of a spatial relationship entity, geographic attribute information, and logical relationship information, a second acquisition unit configured to acquire a preset generation rule, and generate a spatial operation sequence based on the preset generation rule, a retrieval preprocessing unit configured to perform retrieval preprocessing on the spatial operation sequence to generate a preprocessed spatial operation sequence based on the spatial operation sequence, a stage retrieval unit configured to perform stage retrieval processing based on the preprocessed spatial operation sequence to generate geographic entity information, and a generation unit configured to generate a geographic page corresponding to the geographic entity information, and transmit the geographic entity page to the target user terminal for display. In a third aspect, some embodiments of the present disclosure provide an electronic device comprising one or more processors, and storage means having one or more programs stored thereon, which when executed by the one or more processors, cause the one or more processors to implement the method described in any o