CN116600247A

CN116600247A - Information association matching method, device, equipment and storage medium

Info

Publication number: CN116600247A
Application number: CN202310566730.7A
Authority: CN
Inventors: 焦恒建; 尹卜一; 周金辉; 舒鑫; 冯朝阳; 王畔
Original assignee: Tiktok Zhitu Technology Co ltd
Current assignee: Tiktok Zhitu Technology Co ltd
Priority date: 2023-05-18
Filing date: 2023-05-18
Publication date: 2023-08-15

Abstract

The disclosure provides an information association matching method, device, equipment and storage medium, aiming at a network signal set and a POI set to be detected, a wireless network signal and a POI with matched positions can be determined as candidate combinations, then a target combination with a feature distance larger than a first preset threshold value between the wireless network signal and the POI can be screened from a plurality of candidate combinations, then semantic relevance detection processing can be carried out on the wireless network signal and the POI in the target combination to obtain at least one association data set, and therefore, multidimensional association hanging is carried out on the wireless network signal and the POI through the position relation, the feature distance and the semantic relevance, the comprehensiveness and the accuracy of the association matching process are effectively improved, the accuracy of the obtained association data set is facilitated to be improved, and meanwhile, the wireless network signal and the POI can be screened and filtered for a plurality of times through the position relation, the feature distance and the semantic relevance, the data processing amount is facilitated to be reduced, and therefore the processing efficiency is effectively improved.

Description

Information association matching method, device, equipment and storage medium

Technical Field

The disclosure relates to the technical field of data processing, and in particular relates to an information association matching method, an information association matching device, information association matching equipment and a storage medium.

Background

Based on the continuous popularization and development of location services, demands for indoor positioning, scene recognition, network positioning and the like are continuously improved, along with the popularization of the use of WIFI equipment, WIFI signals are often used for assisting in positioning, and the current position of a user can be determined specifically through the positions of WIFI signals of surrounding shops scanned by the user equipment.

In order to determine the current position of the user through the positions of WIFI signals of surrounding shops scanned by the user equipment, the association relationship between the shops and the WIFI signals needs to be determined, and the association relationship between the shops and the WIFI signals is mostly determined through a mode of carrying out semantic hooking on the shops and the WIFI signals, so that the accuracy rate is low and the recall rate is low.

Disclosure of Invention

The embodiment of the disclosure at least provides an information association matching method, an information association matching device, information association matching equipment and a storage medium.

The embodiment of the disclosure provides an information association matching method, which comprises the following steps:

acquiring a network signal set to be detected and a point of interest (POI) set, wherein the network signal set comprises a plurality of wireless network signals, and the POI set comprises a plurality of POIs;

Determining a plurality of candidate combinations from the network signal set and the POI set, wherein each candidate combination comprises a wireless network signal and a POI, and the signal emission position of the wireless network signal is matched with the position of the POI;

screening from the plurality of candidate combinations to obtain a target combination, wherein the characteristic distance between the wireless network signal and the POI in the target combination is larger than a first preset threshold;

and carrying out semantic relevance detection processing on the wireless network signals and POIs in the screened target combinations to obtain at least one associated data set, wherein the associated data set comprises the POIs and at least one wireless network signal semantically related to the POIs.

In an alternative embodiment, before the acquiring the set of network signals to be detected and the set of points of interest POIs, the method includes:

acquiring a plurality of initial wireless network signals acquired in advance and initial signal names of each initial wireless network signal, and a plurality of initial POIs acquired in advance and initial POI names of each initial POI;

for the plurality of initial wireless network signals, performing signal filtering on the plurality of initial wireless network signals based on the initial signal names, and removing the initial wireless network signals with names not conforming to a first preset rule to obtain the wireless network signals with names conforming to the first preset rule;

Determining the obtained set of each wireless network signal as a network signal set to be detected;

for the initial POIs, filtering the initial POIs based on the initial POI names, and removing the initial POIs with names not conforming to a second preset rule to obtain POIs with names conforming to the second preset rule;

and determining the obtained set of the POIs as a POI set to be detected.

In an optional implementation manner, the filtering the signals of the plurality of initial wireless network signals based on the initial signal names, and removing the initial wireless network signals with names not conforming to the first preset rule to obtain the wireless network signals with names conforming to the first preset rule, includes:

adjusting the initial signal names aiming at each initial signal name to obtain the signal names after filtering processing, wherein the signal names accord with the first preset rule;

determining the initial wireless network signal corresponding to the signal name as a wireless network signal in a network signal set to be detected, and obtaining a network signal set comprising a plurality of wireless network signals;

wherein the filtering of the initial signal names includes one or more of:

If the initial signal name contains the content indicating the network attribute, deleting the content; if the characterization of the initial signal name is a messy code, deleting the initial signal name; and deleting the content which does not indicate the name in the initial signal name under the condition that the name length of the initial signal name is larger than a first preset length.

In an optional implementation manner, the filtering the POIs based on the initial POI names, and removing the initial POIs whose names do not conform to the second preset rule to obtain the POIs whose names conform to the second preset rule, includes:

screening the initial POI names aiming at each initial POI name to obtain the POI names left after screening;

determining the initial POI corresponding to the POI name as the POI in the POI set to be detected, and obtaining a POI set comprising a plurality of POIs;

wherein filtering the initial POI name includes one or more of:

if the name length of the initial POI name is smaller than a second preset length, deleting the initial POI name; and deleting the content if the content representing the preset character exists in the initial POI name, and deleting the initial POI name after character deletion if the name length of the initial POI name after character deletion is smaller than the second preset length.

In an alternative embodiment, after the determining that the set including the obtained POIs is the set of POIs to be detected, the method includes:

traversing at least one pre-stored name matched with the POI name of the POI from a preset name database based on the POI name of the POI for each POI in the obtained POI set, wherein the pre-stored name and the POI name are different names referring to the same POI;

and supplementing the POI corresponding to the pre-stored name to the POI set, and taking the supplemented POI set as the POI set to be detected.

In an alternative embodiment, the determining a plurality of candidate combinations from the set of network signals and the set of POIs includes:

for each POI in the POI set, determining a preset range taking the position of the POI as the center based on the position of the POI;

determining, for each wireless network signal in the set of network signals, whether the wireless network signal is located in the preset range based on a signal emission location of the wireless network signal;

if so, the wireless network signal and the POI are determined to be one of the candidate combinations.

In an optional implementation manner, the feature distance includes a name similarity, and the screening from the plurality of candidate combinations to obtain the target combination includes:

For each candidate combination, determining a signal feature vector of a wireless network signal based on a signal name of the wireless network signal in the candidate combination, and determining a POI feature vector of a POI based on a POI name of the POI in the candidate combination;

calculating the name similarity between the wireless network signal and the POI based on the signal feature vector and the POI feature vector;

and if the name similarity is larger than a first preset threshold, determining the candidate combination as a target combination to be analyzed.

In an optional implementation manner, the performing semantic relevance detection processing on the wireless network signals and POIs in each screened target combination to obtain at least one associated data set includes:

determining the signal name of a wireless network signal and the POI name of a POI in each screened target combination;

carrying out semantic relevance detection on the wireless network signal and the POI through the signal name and the POI name to obtain a semantic relevance detection result of the target combination, wherein the semantic relevance detection result comprises that the wireless network signal is related to the POI semanteme or the wireless network signal is not related to the POI semanteme;

And obtaining at least one associated data set based on the semantic relevance detection result of each target combination.

In an optional implementation manner, the performing semantic relevance detection on the wireless network signal and the POI through the signal name and the POI name to obtain a semantic relevance detection result of the target combination includes:

inputting the signal names and the POI names into a pre-trained semantic relevance detection model, and extracting features of the signal names and the POI names through a feature extraction layer of the semantic relevance detection model to obtain fusion feature vectors output by the feature extraction layer;

inputting the fusion feature vector to a plurality of semantic correlation classification layers which are sequentially connected in the semantic correlation detection model, and aiming at any semantic correlation classification layer except the last semantic correlation classification layer in the plurality of semantic correlation classification layers, obtaining an intermediate feature vector and a classification result output by the semantic correlation classification layer, wherein the classification result comprises the wireless network signal and the POI semantic correlation or the wireless network signal and the POI semantic uncorrelation;

Aiming at the classification result, if the classification result is determined to be reliable, determining the classification result as a semantic relevance detection result of the target combination;

if the classification result is determined to be unreliable, the intermediate feature vector is input to a semantic correlation classification layer next to the semantic correlation classification layer, the above process is repeated until the obtained classification result is reliable, or until the classification result output by the final semantic correlation classification layer is obtained, and the classification result is used as a semantic correlation detection result of the target combination.

In an optional implementation manner, the inputting the fused feature vector into the multiple semantic correlation classification layers sequentially connected in the semantic correlation detection model, for any one semantic correlation classification layer except the last semantic correlation classification layer in the multiple semantic correlation classification layers, obtains an intermediate feature vector and a classification result output by the semantic correlation classification layer, and includes:

inputting the fusion feature vector into a plurality of feature processing layers which are sequentially connected in the semantic relevance detection model, and aiming at any one feature processing layer except the last feature processing layer in the plurality of feature processing layers, carrying out feature processing on the fusion feature vector through the feature processing layer to obtain a processed intermediate feature vector;

Inputting the intermediate feature vector to a relevance classifier connected with the feature processing layer in the semantic relevance detection model to obtain a classification result which is output after the relevance classifier processes the intermediate feature vector, wherein the semantic relevance classification layer comprises the feature processing layer and the relevance classifier.

In an alternative embodiment, before the determining that the classification result is reliable, the method further includes, before determining the classification result as the semantic relevance detection result of the target combination:

calculating uncertainty of the classification result based on the classification result;

if the uncertainty is smaller than a second preset threshold, determining that the classification result is credible;

and if the uncertainty is not smaller than the second preset threshold, determining that the classification result is not credible.

The embodiment of the disclosure also provides an information association matching device, which comprises:

the information acquisition module is used for acquiring a network signal set to be detected and a POI set, wherein the network signal set comprises a plurality of wireless network signals, and the POI set comprises a plurality of POIs;

a combination determining module for determining a plurality of candidate combinations from the network signal set and the POI set, each of the candidate combinations including a wireless network signal and a POI, the signal emission location of the wireless network signal and the location of the POI matching;

The combination screening module is used for screening a target combination from the plurality of candidate combinations, and the characteristic distance between the wireless network signal and the POI in the target combination is larger than a first preset threshold;

the data association module is used for carrying out semantic relevance detection processing on the wireless network signals and POIs in the screened target combinations to obtain at least one association data set, and the association data set comprises the POIs and at least one wireless network signal semantically related to the POIs.

In an alternative embodiment, the apparatus further comprises an information generating module, where the information generating module is configured to:

and determining the obtained set of the POIs as a POI set to be detected.

In an optional implementation manner, the information generating module is configured to, when configured to perform signal filtering on the plurality of initial wireless network signals based on the initial signal names, remove the initial wireless network signals whose names do not meet a first preset rule, and obtain each wireless network signal whose name meets the first preset rule, specifically be:

Wherein the filtering of the initial signal names includes one or more of:

In an optional implementation manner, the information generating module is configured to, when configured to filter the plurality of initial POIs based on the initial POI names, remove the initial POIs whose names do not meet a second preset rule, and obtain each POI whose names meet the second preset rule, specifically configured to:

wherein filtering the initial POI name includes one or more of:

In an alternative embodiment, the information generating module is further configured to:

In an alternative embodiment, the combination determining module is specifically configured to:

In an optional implementation manner, the feature distance includes a name similarity, and the combination screening module is specifically configured to:

In an alternative embodiment, the data association module is specifically configured to:

In an optional implementation manner, the data association module is specifically configured to, when being configured to perform semantic relevance detection on the wireless network signal and the POI through the signal name and the POI name, obtain a semantic relevance detection result of the target combination:

In an optional implementation manner, when the data association module is used for inputting the fusion feature vector into the multi-layer semantic correlation classification layer sequentially connected in the semantic correlation detection model, for any one semantic correlation classification layer except for the last semantic correlation classification layer in the multi-layer semantic correlation classification layer, the data association module is specifically used for obtaining an intermediate feature vector and a classification result output by the semantic correlation classification layer:

In an alternative embodiment, the data association module is further configured to:

The embodiment of the disclosure also provides an electronic device, including: a processor, a memory and a bus, the memory storing machine-readable instructions executable by the processor, the processor and the memory in communication via the bus when the electronic device is running, the machine-readable instructions when executed by the processor performing the steps of the information correlation matching method of any one of the possible embodiments described above.

The disclosed embodiments also provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the information correlation matching method in any of the possible implementations.

According to the information association matching method, device and equipment and storage medium, a network signal set to be detected and a POI set of interest points can be obtained, the network signal set comprises a plurality of wireless network signals, the POI set comprises a plurality of POIs, a plurality of candidate combinations are determined from the network signal set and the POI set, each candidate combination comprises a wireless network signal and a POI, the signal sending position of the wireless network signal is matched with the position of the POI, a target combination is obtained through screening from the plurality of candidate combinations, the characteristic distance between the wireless network signal and the POI in the target combination is larger than a first preset threshold, semantic relevance detection processing is conducted on the wireless network signal and the POI in each screened target combination, and at least one association data set is obtained, wherein the association data set comprises the POI and at least one wireless network signal semantically related to the POI.

In this way, for the network signal set and the POI set to be detected, the wireless network signal and the POI with matched positions can be determined as candidate combinations, so that the wireless network signal and the POI can be associated based on the position relation, then a target combination with the feature distance between the wireless network signal and the POI larger than a first preset threshold value can be screened from a plurality of candidate combinations, so that the wireless network signal and the POI can be further associated based on the feature distance, then the wireless network signal and the POI in the target combination can be subjected to semantic relevance detection processing to obtain at least one associated data set, so that the wireless network signal and the POI can be further associated based on the semantic relevance, and the wireless network signal and the POI can be subjected to multidimensional association hanging through the position relation, the feature distance and the semantic relevance, so that the comprehensiveness and the accuracy of the association matching process are effectively increased, the accuracy of the associated data set obtained by matching the wireless network signal and the POI is facilitated to be improved, and the accuracy of the information relevance between the wireless network signal and the POI is improved. Meanwhile, wireless network signals and POIs can be screened and filtered for many times through the position relation, the feature distance and the semantic relativity, so that the data processing amount is reduced, the processing efficiency is effectively improved, and the processing time is shortened.

Furthermore, when the semantic relevance detection processing is performed on the wireless network signals and the POIs in the target combination, the reliability of the classification result output by the semantic relevance classification layer in the semantic relevance detection model can be judged, and the classification result is determined to be the semantic relevance detection result under the condition that the classification result is reliable by judging whether the classification result is reliable or not, so that the semantic relevance detection result can be obtained in advance without processing of all the semantic relevance classification layers, the data processing speed is improved, the calculation cost is reduced, the calculation resource consumption is avoided, and the data processing performance is improved.

The foregoing objects, features and advantages of the disclosure will be more readily apparent from the following detailed description of the preferred embodiments taken in conjunction with the accompanying drawings.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the aspects of the disclosure.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings required for the embodiments are briefly described below, which are incorporated in and constitute a part of the specification, these drawings showing embodiments consistent with the present disclosure and together with the description serve to illustrate the technical solutions of the present disclosure. It is to be understood that the following drawings illustrate only certain embodiments of the present disclosure and are therefore not to be considered limiting of its scope, for the person of ordinary skill in the art may admit to other equally relevant drawings without inventive effort.

Fig. 1 shows a flowchart of an information association matching method provided by an embodiment of the present disclosure;

FIG. 2 is a schematic structural diagram of a semantic relevance detection model according to an embodiment of the present disclosure;

FIG. 3 is a schematic diagram of an information correlation matching process provided by an embodiment of the present disclosure;

FIG. 4 shows one of the schematic diagrams of an information-correlation matching apparatus provided by the embodiments of the present disclosure;

FIG. 5 is a second schematic diagram of an information-related matching device according to an embodiment of the disclosure;

fig. 6 shows a schematic structural diagram of an electronic device according to an embodiment of the disclosure.

Detailed Description

For the purposes of making the objects, technical solutions and advantages of the embodiments of the present disclosure more apparent, 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 apparent that the described embodiments are only some embodiments of the present disclosure, but not all embodiments. The components of the embodiments of the present disclosure, which are generally described and illustrated in the figures herein, may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present disclosure provided in the accompanying drawings is not intended to limit the scope of the disclosure, as claimed, but is merely representative of selected embodiments of the disclosure. All other embodiments, which can be made by those skilled in the art based on the embodiments of this disclosure without making any inventive effort, are intended to be within the scope of this disclosure.

It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures.

The term "and/or" is used herein to describe only one relationship, meaning that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist together, and B exists alone. In addition, the term "at least one" herein means any one of a plurality or any combination of at least two of a plurality, for example, including at least one of A, B, C, and may mean including any one or more elements selected from the group consisting of A, B and C.

According to research, when the association relation between a store and a WIFI signal is determined, an extreme gradient lifting (eXtreme Gradient Boosting, XGB) model or a pre-training language (Bidirectional Encoder Representation from Transformers, BERT) model is mostly used for semantic association, however, the XGB model is mostly used by manually designed characteristics, such as the duty ratio of the longest public subsequence, the duty ratio of an editing distance and the like, and has the advantages of poor fitting effect, low generalization effect, low BERT model operation speed, high calculation cost, low accuracy of matching results of the store and the WIFI signal and low recall rate.

Based on the above study, the disclosure provides an information association matching method, for a network signal set and a POI set to be detected, a wireless network signal and a POI with matched positions can be determined as candidate combinations, then a target combination with a feature distance between the wireless network signal and the POI larger than a first preset threshold value can be screened from a plurality of candidate combinations, then semantic relevance detection processing can be performed on the wireless network signal and the POI in the target combination to obtain at least one association data set, in this way, multidimensional association hanging is performed on the wireless network signal and the POI through the position relationship, the feature distance and the semantic relevance, the comprehensiveness and the accuracy of the association matching process are effectively increased, and meanwhile, multiple screening and filtering can be performed on the wireless network signal and the POI through the position relationship, the feature distance and the semantic relevance, so that the data processing amount is reduced, the processing efficiency is effectively improved, and the processing time is reduced.

For the sake of understanding the present embodiment, first, a detailed description will be given of an information association matching method disclosed in the present embodiment, and an execution main body of the information association matching method provided in the present embodiment is generally an electronic device with a certain computing capability, and in this embodiment, the electronic device may be a server. The 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 services, cloud databases, cloud computing, cloud storage, big data, artificial intelligent platforms and the like.

In other embodiments, the electronic device may be a terminal device or other processing device, where the terminal device may be a User Equipment (UE), a User terminal, a computing device, and the other processing device may be a device including a processor and a memory, and is not limited herein. In some possible implementations, the information association matching method may be implemented by a processor invoking computer readable instructions stored in a memory.

An information association matching method provided by an embodiment of the present disclosure is described below.

Referring to fig. 1, a flowchart of an information association matching method according to an embodiment of the present disclosure is shown, where the method includes steps S101 to S104, where:

s101: a network signal set to be detected and a point of interest (POI) set are obtained, wherein the network signal set comprises a plurality of wireless network signals, and the POI set comprises a plurality of POIs.

Here, when association detection needs to be performed on the wireless network signal and the POI, a network signal set to be detected and a POI set need to be acquired first.

Optionally, for clearly determining the wireless network signal and the POI, the network signal set and the POI set may present the wireless network signal and the POI, respectively, in the form of a list.

Here, the wireless network signal may be a WIFI signal, and the POI may be a place of a store level.

For example, a mall may be used as one POI, and each store in the mall may be used as one POI.

Optionally, the network signal set may further include information such as a name and a location of each of the wireless network signals, and the POI set may further include information such as a name and a location of each of the POIs.

In order to improve accuracy and richness of data, the network signal set and the POI set may be obtained by preprocessing a plurality of initial wireless network signals and a plurality of initial POIs.

Specifically, in some possible embodiments, before the acquiring the network signal set to be detected and the POI set, the method includes:

and determining the obtained set of the POIs as a POI set to be detected.

The initial wireless network signals and the initial signal names of the initial wireless network signals can be acquired through a signal scanning device, and the initial POIs and the initial POI names of the initial POIs can be obtained by grabbing from a network.

The store owner can upload information of the store to each platform for propaganda after the store is opened, and the customer can upload information of the experienced store to each platform for sharing after visiting the store, so that a large number of initial POIs exist on the network, and the initial POIs can be obtained by grabbing from the network.

In practical applications, the initial signal names and the initial POI names mostly have non-standard or non-semantic content, the content is not related to the subsequent process of performing association detection between wireless network signals and POIs, in order to reduce the subsequent data processing amount, the plurality of initial wireless network signals may be filtered based on each initial signal name, and the plurality of initial POIs may be filtered based on each initial POI name, so as to obtain a network signal set and a POI set to be detected, in this process, data that are not used for the process of performing association detection between wireless network signals and POIs may be effectively removed, and the practicality and usability of the network signal set and the POI set may be improved.

In order to obtain each wireless network signal with a name conforming to the first preset rule, in some possible implementation manners, for each initial signal name, the initial signal name is adjusted to obtain a filtered signal name, where the signal name conforms to the first preset rule; and determining the initial wireless network signal corresponding to the signal name as the wireless network signal in the network signal set to be detected, and obtaining the network signal set comprising a plurality of wireless network signals.

Wherein the filtering of the initial signal names includes one or more of:

In some possible ways, the initial signal name often includes content indicating a network attribute, where the content indicating a network attribute includes at least content indicating a frequency band adopted by the network signal, a corresponding carrier, and the like, where the content indicating a network attribute is not relevant for a subsequent process of detecting association between the wireless network signal and the POI, and is therefore deleted.

In other possible ways, the original signal name may be characterized as a scrambling code due to naming irregularities, etc., at which point the original signal name may be considered invalid and thus deleted.

In still other possible embodiments, the initial signal name may include a detailed address of the initial POI, which results in a name length of the initial signal name being greater than a first preset length, where a content of the initial signal name that is irrelevant to the name is too high in proportion to facilitate subsequent resolution and identification of the name, so that content of the initial signal name that is not indicated by the name is deleted.

In order to obtain the POIs with the names conforming to the second preset rule, in some possible implementation manners, for each initial POI name, screening the initial POI name to obtain the POI names left after screening; and determining the initial POI corresponding to the POI name as the POI in the POI set to be detected, and obtaining the POI set comprising a plurality of POIs.

Wherein filtering the initial POI name includes one or more of:

And the name length of the POI names left after screening is greater than or equal to the second preset length.

Here, the initial POI name may include a part of preset characters, and optionally, the preset characters include, but are not limited to, emoticons, numerals, and punctuation marks such as brackets, and the content represented as the preset characters does not include actual semantic information, so that the part of the content may be deleted. If the name length of the initial POI name after character deletion is smaller than the second preset length, deleting the initial POI name after character deletion.

For example, if the second preset rule indicates that the name length of the POI name left after screening needs to be greater than or equal to 2, both the initial POI name with the name length less than 2 and the initial POI name after character deletion are deleted.

Here, since the plurality of initial POIs and the initial POI name of each initial POI may be obtained by capturing from a network, and information on the network may have situations of incomplete information, non-updated information, etc., in order to improve the comprehensiveness and richness of the POI set, after the determining that the obtained set including each POI is the POI set to be detected, expansion processing may be performed on the obtained POI set.

Accordingly, in some possible embodiments, for each POI in the obtained POI set, traversing at least one pre-stored name matched with the POI name of the POI from a preset name database based on the POI name of the POI, wherein the pre-stored name and the POI name are different names referring to the same POI; and supplementing the POI corresponding to the pre-stored name to the POI set, and taking the supplemented POI set as the POI set to be detected.

The preset name database stores a plurality of prestored POIs and at least one prestored name corresponding to each prestored POI.

Optionally, the preset name database may be generated based on a brand library and an alias library, where the brand library includes a plurality of first POIs and a plurality of names of each of the first POIs, where the first POIs are chained brand POIs, so that the first POIs and the plurality of names of the first POIs may be conveniently obtained through a network; the alias library comprises a plurality of second POIs and a plurality of names of each second POI, wherein the second POIs are POIs with a plurality of names, and the brand library and the alias library are integrated to obtain the preset name database.

For example, for a certain barbecue in the alias library, the name of the record is determined as Zhang Jun barbecue based on business registration information of the barbecue, and the name of the store is acquired as Zhang Sanbarbecue based on map information corresponding to the barbecue, so the barbecue itself has two names.

S102: a plurality of candidate combinations are determined from the set of network signals and the set of POIs, each of the candidate combinations including a wireless network signal and a POI whose signal emitting location matches the location of the POI.

In this step, if each wireless network signal and each POI are respectively matched, the amount of processing data is extremely large, and the possible positions between many wireless network signals and the POI are far apart, so that unnecessary processing is caused, therefore, full calculation is not required, but a plurality of candidate combinations are determined based on the signal sending positions of the wireless network signals and the positions of the POI, the wireless network signals with matched positions and the POI are matched, and therefore, the wireless network signals and the POI are effectively associated based on the position relation.

Here, the signal sending position of the wireless network signal and the position of the POI are matched, which may be the same position as the signal sending position of the wireless network signal and the position of the POI, or may be the signal sending position of the wireless network signal and the position of the POI are located in the same range, and may be specifically set according to the accuracy requirement of the candidate combination.

In some possible embodiments, the determining a plurality of candidate combinations from the network signal set and the POI set may be determining, for each POI in the POI set, a preset range centered at a location of the POI based on the location of the POI; determining, for each wireless network signal in the set of network signals, whether the wireless network signal is located in the preset range based on a signal emission location of the wireless network signal; if so, the wireless network signal and the POI are determined to be one of the candidate combinations.

Specifically, the interval distance between the signal sending position of the wireless network signal and the position of the POI may be calculated, and whether the wireless network signal is located in the preset range may be determined according to the interval distance.

Here, the position of the POI and the signal emitting position of the wireless network signal may be expressed in terms of longitude and latitude.

For example, for a business's business store, there may be 700 wireless network signals within 200 meters centered at that store location, so 700 candidate combinations may be available.

In this way, the position of the POI and the signal sending position of the wireless network signal can be used for performing distance calculation, so that the wireless network signal in a preset range centered on the position of the POI is screened out, and a candidate combination is formed.

S103: and screening the plurality of candidate combinations to obtain a target combination, wherein the characteristic distance between the wireless network signal and the POI in the target combination is larger than a first preset threshold value.

Here, based on the candidate combination, the wireless network signal and the POI may be further associated based on the feature distance, and the target combination may be obtained by screening.

The feature distance may include a similarity of names between the wireless network signal and the POI in the candidate combination, a co-occurrence of traffic between the wireless network signal and the POI in the candidate combination, a signal strength between the wireless network signal and the POI in the candidate combination, and the like.

Specifically, in some possible embodiments, the feature distance includes the name similarity, for each of the candidate combinations, determining a signal feature vector of the wireless network signal based on a signal name of the wireless network signal in the candidate combination, and determining a POI feature vector of the POI based on a POI name of the POI in the candidate combination; calculating the name similarity between the wireless network signal and the POI based on the signal feature vector and the POI feature vector; and if the name similarity is larger than a first preset threshold, determining the candidate combination as a target combination to be analyzed.

In the above step, the respective embedded vectors may be extracted from the signal name and the POI name, respectively, and the embedded vectors may be used as signal feature vectors and POI feature vectors.

Optionally, when the signal names of the wireless network signals in the plurality of candidate combinations are the same, the signal feature vector determination is only performed once on the signal names, and correspondingly, when the POI names of the POIs in the plurality of candidate combinations are the same, the POI feature vector determination is only performed once on the POI names.

For example, a sense-bert model may be used to determine the vector, and the signal name and the POI name are input into the sense-bert model respectively, so as to obtain an Embedding vector with 768 dimensions output by the sense-bert model. The sense-bert model may perform incremental calculations during the extraction of the Embedding vector, i.e., the same name is extracted only once.

Then, two Embedding vectors corresponding to the candidate combination can be directly adopted to calculate the name similarity between the wireless network signal and the POI.

Alternatively, a similarity algorithm may be used to calculate cosine similarity, pearson correlation coefficient, etc., and a spatial distance algorithm may be used to calculate euclidean distance, hamming distance, etc., where the smaller the distance, the higher the characterization similarity.

The value of the first preset threshold can be set according to actual needs, and the filtering rate and recall rate of the target combination can be guaranteed through the value adjustment of the first preset threshold.

S104: and carrying out semantic relevance detection processing on the wireless network signals and POIs in the screened target combinations to obtain at least one associated data set, wherein the associated data set comprises the POIs and at least one wireless network signal semantically related to the POIs.

Here, the wireless network signal and the POI may be further associated based on the semantic relevance, so as to obtain a semantic relevance detection result of the target combination, and then at least one associated data set may be obtained based on the semantic relevance detection result of each target combination.

The detection precision of the semantic relativity detection is higher than that of the name similarity.

Specifically, in some possible embodiments, for each screened target combination, determining a signal name of a wireless network signal and a POI name of a POI in the target combination; carrying out semantic relevance detection on the wireless network signal and the POI through the signal name and the POI name to obtain a semantic relevance detection result of the target combination, wherein the semantic relevance detection result comprises that the wireless network signal is related to the POI semanteme or the wireless network signal is not related to the POI semanteme; and obtaining at least one associated data set based on the semantic relevance detection result of each target combination.

Here, semantic relevance detection can be performed on the wireless network signal and the POI through names, so that comprehensiveness and richness of association matching are improved.

Optionally, before the semantic relevance detection is performed on the wireless network signal and the POI through the signal name and the POI name, secondary filtering may be performed on the signal name and the POI name to obtain a filtered signal name and a filtered POI name, and then the semantic relevance detection is performed on the wireless network signal and the POI through the filtered signal name and the filtered POI name.

Specifically, for the target combination, in the case that the POI name is composed of a pure number, if the pure number is completely contained in the signal name, the POI name is considered to be actually composed of a number, the POI name is reserved, and if the pure number is not completely contained in the signal name, the POI name is deleted. Similarly, in the case where the signal name is composed of a pure number, for the target combination, if the pure number is completely contained in the POI name, the signal name is held if the signal name itself is considered to be composed of a number, and if the pure number is not completely contained in the POI name, the signal name is deleted.

The semantic relevance detection can be performed by adopting a pre-trained semantic relevance detection model.

Therefore, in some possible embodiments, the signal name and the POI name may be input to a pre-trained semantic relevance detection model, and feature extraction is performed on the signal name and the POI name by a feature extraction layer of the semantic relevance detection model, so as to obtain a fused feature vector output by the feature extraction layer; inputting the fusion feature vector to a plurality of semantic correlation classification layers which are sequentially connected in the semantic correlation detection model, and aiming at any semantic correlation classification layer except the last semantic correlation classification layer in the plurality of semantic correlation classification layers, obtaining an intermediate feature vector and a classification result output by the semantic correlation classification layer, wherein the classification result comprises the wireless network signal and the POI semantic correlation or the wireless network signal and the POI semantic uncorrelation; aiming at the classification result, if the classification result is determined to be reliable, determining the classification result as a semantic relevance detection result of the target combination; if the classification result is determined to be unreliable, the intermediate feature vector is input to a semantic correlation classification layer next to the semantic correlation classification layer, the above process is repeated until the obtained classification result is reliable, or until the classification result output by the final semantic correlation classification layer is obtained, and the classification result is used as a semantic correlation detection result of the target combination.

Here, by judging whether the classification result is credible or not, under the condition that the classification result is credible, the semantic relativity detection result is obtained in advance, so that all semantic relativity classification layers are not needed to be passed, the data processing speed is improved, the calculation cost is reduced, the calculation resource consumption is avoided, and the data processing performance is improved.

The semantic relativity detection model can be trained by a Chinese corpus, so that the recognition effect of Chinese names is improved. Moreover, the semantic relativity detection model can be pre-trained by using a task based on similar word replacement and whole word masking (Masked language modeling, MLM) and a task based on sentence sequential prediction (Sentence Order Prediction, SOP), so that the capability of the semantic relativity detection model for Chinese word segmentation is improved.

Optionally, the semantic relevance detection model may output a classification result in the form of a probability score, and may determine whether the wireless network signal is semantically relevant to the POI according to a comparison result of the probability score and a third preset threshold.

The value of the third preset threshold can be set according to actual needs, and the accuracy of detection can be guaranteed through the value adjustment of the third preset threshold.

For example, if the third preset threshold is 0.5, if the classification result is a probability score greater than 0.5, the wireless network signal is related to the POI semantic, and if the classification result is a probability score less than or equal to 0.5, the wireless network signal is not related to the POI semantic.

In some possible embodiments, the semantic correlation classification layer includes a feature processing layer and a correlation classifier, specifically, the fused feature vector may be input to a plurality of feature processing layers sequentially connected in the semantic correlation detection model, for any one feature processing layer except for a last feature processing layer in the plurality of feature processing layers, feature processing is performed on the fused feature vector by using the feature processing layer to obtain a processed intermediate feature vector, and then the intermediate feature vector is input to a correlation classifier connected with the feature processing layer in the semantic correlation detection model, so as to obtain a classification result output by the correlation classifier after processing the intermediate feature vector, where the semantic correlation classification layer includes the feature processing layer and the correlation classifier.

Alternatively, the feature processing layer may output intermediate feature vectors in a matrix form, and the relevance classifier may output classification results.

To determine whether the classification is authentic, in some possible embodiments, an uncertainty of the classification result may be calculated based on the classification result before the classification result is determined to be the semantic relevance detection result of the target combination if the classification result is determined to be authentic; if the uncertainty is smaller than a second preset threshold, determining that the classification result is credible; and if the uncertainty is not smaller than the second preset threshold, determining that the classification result is not credible.

Here, the uncertainty of the classification result may be calculated based on the output probability of the correlation classifier and the number of the correlation classifiers.

The value of the second preset threshold can be set according to actual needs, and the data processing speed can be guaranteed through the value adjustment of the second preset threshold.

Referring to fig. 2, fig. 2 is a schematic structural diagram of a semantic relevance detection model provided by the implementation of the present disclosure. As shown in fig. 2, this embodiment is illustrated by taking a 12-layer semantic correlation classification layer as an example, where the semantic correlation detection model includes a feature extraction layer and a 12-layer semantic correlation classification layer, and each of the semantic correlation classification layers includes a feature processing layer and a correlation classifier that are connected.

Inputting the signal names and the POI names into a feature extraction layer of the semantic relevance detection model, and carrying out feature extraction on the signal names and the POI names through the feature extraction layer to obtain fusion feature vectors output by the feature extraction layer.

And then inputting the fusion feature vector to a plurality of layers of semantic relevance classifying layers which are sequentially connected in the semantic relevance detecting model, specifically inputting the fusion feature vector to a plurality of layers of feature processing layers which are sequentially connected in the semantic relevance detecting model, aiming at any one of the plurality of layers of feature processing layers except the last layer of feature processing layer, carrying out feature processing on the fusion feature vector through the feature processing layer to obtain a processed intermediate feature vector, and then inputting the intermediate feature vector to a relevance classifier which is connected with the feature processing layer in the semantic relevance detecting model to obtain a classification result which is output by the relevance classifier after processing the intermediate feature vector, wherein the classification result comprises a wireless network signal and the POI semantic relevance or the wireless network signal and the POI semantic relevance.

Calculating uncertainty of the classification result aiming at the classification result, if the uncertainty is smaller than a second preset threshold, determining that the classification result is reliable, and determining the classification result as a semantic relevance detection result of the target combination; if the uncertainty is not less than the second preset threshold, determining that the classification result is not credible, inputting the intermediate feature vector to a semantic correlation classification layer of a next layer of the semantic correlation classification layer, repeating the above processes until the obtained classification result is credible, or until the classification result output by the final layer of semantic correlation classification layer is obtained, and taking the classification result as a semantic correlation detection result of the target combination.

Illustratively, the semantic relevance detection model may adopt an adaptive inference model (Fastbert) of a distilled Chinese natural language pre-training model (MLM as correction BERT, macbert), which may help to reduce the calculation cost of using a depth model and increase the calculation speed of the model without reducing the performance.

Optionally, for each POI, a wireless network signal related to its semantic meaning may be retained, and a wireless network signal not related to its semantic meaning may be removed, thereby obtaining the association data set, where the association data set includes the POI and at least one wireless network signal related to its semantic meaning.

Specifically, the target combination indicating the semantic relevance of the wireless network signal and the POI according to the semantic relevance detection result can be used as an associated data set and stored, so that at least one wireless network signal associated with each POI is determined.

Optionally, when the wireless network signals and the POIs form the semantic-related target combination during merging and storing, the wireless network signals can be subjected to scoring processing from the wireless network signal dimension to the corresponding POIs based on a preset scoring rule, the POI with the highest score is determined to be the associated POI associated with the wireless network signals, the wireless network signals and the corresponding associated POI are further stored, each wireless network signal is ensured to be only associated with one POI, and each POI can be associated with at least one wireless network signal.

To clearly demonstrate the information-related matching process, fig. 3 shows a schematic diagram of an information-related matching process provided by an embodiment of the present disclosure. As shown in fig. 3, a plurality of initial wireless network signals collected in advance and initial signal names of each initial wireless network signal may be acquired first, and a plurality of initial POIs collected in advance and initial POI names of each initial POI may be obtained by performing signal filtering on the plurality of initial wireless network signals based on the respective initial signal names for the plurality of initial wireless network signals, to obtain a network signal set to be detected, filtering on the plurality of initial POIs based on the respective initial POI names for the plurality of initial POIs, and supplementing the filtered POIs to obtain a POI set to be detected.

A plurality of candidate combinations may then be determined from the set of network signals and the set of POIs by calculating a separation distance between a signaling location of the wireless network signal and a location of the POI, each of the candidate combinations comprising a wireless network signal and a POI, the signaling location of the wireless network signal matching the location of the POI.

Then, in the case that the feature distance includes a name similarity, a signal feature vector of the wireless network signal may be determined based on a signal name of the wireless network signal in the candidate combination, and a POI feature vector of the POI may be determined based on a POI name of the POI in the candidate combination, and then, based on the signal feature vector and the POI feature vector, a name similarity between the wireless network signal and the POI may be calculated, and the candidate combination with the name similarity greater than a first preset threshold may be determined as a target combination.

And further, carrying out semantic relevance detection processing on the wireless network signals and the POIs in the target combination to obtain a semantic relevance detection result of the target combination, wherein the semantic relevance detection result comprises semantic relevance of the wireless network signals and the POIs or semantic irrelevance of the wireless network signals and the POIs, and the target combination which is related to the semantic relevance detection result and indicates the wireless network signals and the POIs according to the semantic relevance detection result of each target combination can be used as a correlation data set to be stored in a merging mode, so that at least one correlation data set is obtained, and the correlation data set comprises the POIs and at least one wireless network signal related to the POIs.

According to the information association matching method provided by the embodiment of the disclosure, for a network signal set and a POI set to be detected, a wireless network signal and a POI which are matched in position can be determined as candidate combinations, so that the wireless network signal and the POI can be associated based on the position relationship, then a target combination with the feature distance between the wireless network signal and the POI larger than a first preset threshold value can be screened out from a plurality of candidate combinations, so that the wireless network signal and the POI can be further associated based on the feature distance, then the wireless network signal and the POI in the target combination can be subjected to semantic relevance detection processing to obtain at least one associated data set, and therefore, the wireless network signal and the POI can be further associated based on the semantic relevance, the multidimensional association hooking is carried out on the wireless network signal and the POI based on the position relationship, the feature distance and the semantic relevance, the comprehensiveness and the accuracy of the association matching process are effectively increased, the accuracy of the association data set obtained for the wireless network signal and the POI is facilitated to be improved, and the accuracy of information relevance between the wireless network signal and the POI is improved. Meanwhile, wireless network signals and POIs can be screened and filtered for many times through the position relation, the feature distance and the semantic relativity, so that the data processing amount is reduced, the processing efficiency is effectively improved, and the processing time is shortened.

It will be appreciated by those skilled in the art that in the above-described method of the specific embodiments, the written order of steps is not meant to imply a strict order of execution but rather should be construed according to the function and possibly inherent logic of the steps.

Based on the same inventive concept, the embodiment of the disclosure further provides an information association matching device corresponding to the information association matching method, and since the principle of solving the problem by the device in the embodiment of the disclosure is similar to that of the information association matching method in the embodiment of the disclosure, the implementation of the device can refer to the implementation of the method, and the repetition is omitted.

Referring to fig. 4 and fig. 5, fig. 4 is a schematic diagram of a first information correlation matching device provided in an embodiment of the disclosure, and fig. 5 is a schematic diagram of a second information correlation matching device provided in an embodiment of the disclosure. As shown in fig. 4, an information association matching apparatus 400 provided by an embodiment of the present disclosure includes:

an information obtaining module 410, configured to obtain a network signal set to be detected and a POI set, where the network signal set includes a plurality of wireless network signals, and the POI set includes a plurality of POIs;

a combination determining module 420 configured to determine a plurality of candidate combinations from the set of network signals and the set of POIs, each of the candidate combinations including a wireless network signal and a POI, a signal emission location of the wireless network signal and a location of the POI matching;

A combination screening module 430, configured to screen a target combination from the plurality of candidate combinations, where a feature distance between a wireless network signal and a POI in the target combination is greater than a first preset threshold;

the data association module 440 is configured to perform semantic relevance detection processing on the wireless network signals and POIs in each of the screened target combinations, so as to obtain at least one associated data set, where the associated data set includes a POI and at least one wireless network signal semantically related to the POI.

In an alternative embodiment, as shown in fig. 5, the apparatus further includes an information generating module 450, where the information generating module 450 is configured to:

and determining the obtained set of the POIs as a POI set to be detected.

In an optional implementation manner, the information generating module 450 is configured to, when used in the filtering the signals of the plurality of initial wireless network signals based on the initial signal names, remove the initial wireless network signals whose names do not meet the first preset rule, and obtain each wireless network signal whose name meets the first preset rule, specifically be:

Wherein the filtering of the initial signal names includes one or more of:

In an optional implementation manner, the information generating module 450 is configured to, when configured to perform POI filtering on the plurality of initial POIs based on the respective initial POI names, remove initial POIs whose names do not meet a second preset rule, and obtain each POI whose names meet the second preset rule, specifically configured to:

wherein filtering the initial POI name includes one or more of:

In an alternative embodiment, the information generating module 450 is further configured to:

In an alternative embodiment, the combination determination module 420 is specifically configured to:

for each POI in the POI set, determining a preset range taking the POI position as a center based on the POI position of the POI;

In an alternative embodiment, the feature distance includes a name similarity, and the combination filtering module 430 is specifically configured to:

In an alternative embodiment, the data association module 440 is specifically configured to:

In an optional implementation manner, the data association module 440 is specifically configured to, when configured to perform semantic relevance detection on the wireless network signal and the POI by using the signal name and the POI name, obtain a semantic relevance detection result of the target combination:

In an optional implementation manner, when the data association module 440 is configured to input the fused feature vector to the multiple semantic correlation classification layers sequentially connected in the semantic correlation detection model, for any semantic correlation classification layer except for the last semantic correlation classification layer in the multiple semantic correlation classification layers, the data association module is specifically configured to:

In an alternative embodiment, the data association module 440 is further configured to:

The process flow of each module in the apparatus and the interaction flow between the modules may be described with reference to the related descriptions in the above method embodiments, which are not described in detail herein.

According to the information association matching device provided by the embodiment of the disclosure, for a network signal set and a POI set to be detected, a wireless network signal and a POI which are matched in position can be determined as candidate combinations, so that the wireless network signal and the POI can be associated based on the position relationship, then a target combination with the feature distance between the wireless network signal and the POI larger than a first preset threshold value can be screened out from a plurality of candidate combinations, so that the wireless network signal and the POI can be further associated based on the feature distance, then the wireless network signal and the POI in the target combination can be subjected to semantic relevance detection processing to obtain at least one associated data set, and therefore, the wireless network signal and the POI can be further associated based on the semantic relevance, the multidimensional association hooking is carried out on the wireless network signal and the POI based on the position relationship, the feature distance and the semantic relevance, the comprehensiveness and the accuracy of the association matching process are effectively increased, the accuracy of the association data set obtained for the wireless network signal and the POI is facilitated to be improved, and the accuracy of information relevance between the wireless network signal and the POI is improved. Meanwhile, wireless network signals and POIs can be screened and filtered for many times through the position relation, the feature distance and the semantic relativity, so that the data processing amount is reduced, the processing efficiency is effectively improved, and the processing time is shortened.

Corresponding to the information association matching method in fig. 1, the embodiment of the present disclosure further provides an electronic device 600, as shown in fig. 6, which is a schematic structural diagram of the electronic device 600 provided in the embodiment of the present disclosure, including:

a processor 610, a memory 620, and a bus 630. Wherein the memory 620 is used for storing execution instructions, including a memory 621 and an external memory 622; the memory 621 is also referred to as an internal memory, and is used for temporarily storing operation data in the processor 610 and data exchanged with the external memory 622 such as a hard disk, and the processor 610 exchanges data with the external memory 622 via the memory 621.

In an embodiment of the present application, the memory 620 is specifically configured to store application program codes for executing the solution of the present application, and the processor 610 controls the execution. That is, when the electronic device 600 is operating, communication between the processor 610 and the memory 620 via the bus 630 causes the processor 610 to execute the application code stored in the memory 620, thereby performing the information association matching method described in any of the foregoing embodiments.

The Memory 620 may be, but is not limited to, a random access Memory (Random Access Memory, RAM), a Read Only Memory (ROM), a programmable Read Only Memory (Programmable Read-Only Memory, PROM), an erasable Read Only Memory (Erasable Programmable Read-Only Memory, EPROM), an electrically erasable Read Only Memory (Electric Erasable Programmable Read-Only Memory, EEPROM), etc.

The processor 610 may be an integrated circuit chip having signal processing capabilities. The processor may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; but also digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field programmable gate arrays (Field Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The disclosed methods, steps, and logic blocks in the embodiments of the present application may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

It should be understood that the illustrated structure of the embodiment of the present application does not constitute a specific limitation on the electronic device 600. In other embodiments of the application, electronic device 600 may include more or fewer components than shown, or certain components may be combined, or certain components may be split, or different arrangements of components. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

The disclosed embodiments also provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the information correlation matching method described in the above method embodiments. Wherein the storage medium may be a volatile or nonvolatile computer readable storage medium.

The embodiment of the disclosure further provides a computer program product, which includes computer instructions, where the computer instructions, when executed by a processor, may perform the steps of the information association matching method described in the foregoing method embodiment, and specifically, reference the foregoing method embodiment is omitted herein.

Wherein the above-mentioned computer program product may be realized in particular by means of hardware, software or a combination thereof. In an alternative embodiment, the computer program product is embodied as a computer storage medium, and in another alternative embodiment, the computer program product is embodied as a software product, such as a software development kit (Software Development Kit, SDK), or the like.

It will be clearly understood by those skilled in the art that, for convenience and brevity of description, specific working procedures of the apparatus and device described above may refer to corresponding procedures in the foregoing method embodiments, which are not described herein again. In several embodiments provided in the present disclosure, it should be understood that the disclosed apparatus, device, and method may be implemented in other manners. The apparatus embodiments described above are merely illustrative, e.g., the units described are merely one logical function, and may be implemented in other ways, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some communication interface, device or unit indirect coupling or communication connection, which may be in electrical, mechanical or other form.

The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present disclosure may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.

The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a non-volatile computer readable storage medium executable by a processor. Based on such understanding, the technical solution of the present disclosure may be embodied in essence or a part contributing to the prior art or a part of the technical solution, or in the form of a software product stored in a storage medium, including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in the embodiments of the present disclosure. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

Finally, it should be noted that: the foregoing examples are merely specific embodiments of the present disclosure, and are not intended to limit the scope of the disclosure, but the present disclosure is not limited thereto, and those skilled in the art will appreciate that while the foregoing examples are described in detail, it is not limited to the disclosure: any person skilled in the art, within the technical scope of the disclosure of the present disclosure, may modify or easily conceive changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features thereof; such modifications, changes or substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the disclosure, and are intended to be included within the scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be subject to the protection scope of the claims.

Claims

1. An information association matching method, characterized in that the method comprises the following steps:

2. The method according to claim 1, characterized in that before the acquisition of the set of network signals to be detected and the set of point of interest POIs, the method comprises:

and determining the obtained set of the POIs as a POI set to be detected.

3. The method according to claim 2, wherein the performing signal filtering on the plurality of initial wireless network signals based on the initial signal names, and removing the initial wireless network signals whose names do not meet the first preset rule, to obtain the wireless network signals whose names meet the first preset rule, includes:

wherein the filtering of the initial signal names includes one or more of:

4. The method according to claim 2, wherein filtering the plurality of initial POIs based on the respective initial POI names, and removing initial POIs whose names do not conform to a second preset rule, to obtain respective POIs whose names conform to the second preset rule, includes:

wherein filtering the initial POI name includes one or more of:

5. The method according to claim 2, wherein after said determining comprises the resulting set of individual POIs as the set of POIs to be detected, the method comprises:

6. The method of claim 1, wherein said determining a plurality of candidate combinations from said set of network signals and said set of POIs comprises:

7. The method of claim 1, wherein the feature distance comprises a name similarity, and wherein the screening from the plurality of candidate combinations for the target combination comprises:

8. The method of claim 1, wherein the performing semantic relevance detection processing on the wireless network signals and POIs in each of the screened target combinations to obtain at least one associated data set includes:

9. The method of claim 8, wherein the performing semantic relevance detection on the wireless network signal and the POI by using the signal name and the POI name to obtain a semantic relevance detection result of the target combination includes:

10. The method according to claim 9, wherein the inputting the fused feature vector into the multiple semantic correlation classification layers sequentially connected in the semantic correlation detection model, for any one of the multiple semantic correlation classification layers except for the last semantic correlation classification layer, obtains an intermediate feature vector and a classification result output by the semantic correlation classification layer, includes:

11. The method of claim 9, wherein prior to said determining said classification result as a semantic relevance detection result for said target combination if said classification result is determined to be authentic, said method further comprises:

12. An information association matching apparatus, the apparatus comprising:

13. An electronic device, comprising: a processor, a memory and a bus, the memory storing machine-readable instructions executable by the processor, the processor and the memory in communication over the bus when the electronic device is running, the machine-readable instructions when executed by the processor performing the steps of the information association matching method of any of claims 1 to 11.

14. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a computer program which, when executed by a processor, performs the steps of the information correlation matching method as claimed in any one of claims 1 to 11.