Data fusion method, search method, device, electronic equipment and storage medium

Abstract

The present disclosure provides a data fusion method, a search method, a device, an electronic device and a storage medium, relates to the technical field of computers, and in particular to the fields of data processing, cloud computing, intelligent search and the like. The specific implementation scheme is as follows: obtaining to-be-fused data from different data sources, wherein the to-be-fused data comprises heterogeneous data with non-uniform data formats; performing data format conversion processing on attribute fields of the to-be-fused data to obtain attribute data with uniform data formats; performing attribute classification processing on the attribute data to obtain attribute classification data; and performing data fusion processing according to the attribute classification data to obtain a data fusion result. By using the present disclosure, massive multi-source heterogeneous data can be fused to improve the data processing speed.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to the fields of data processing, cloud computing, and intelligent search. Background Technology

[0002] In the era of big data, massive amounts of data from different data sources exhibit heterogeneity (e.g., data with different formats). This vast amount of multi-source heterogeneous data, with its enormous volume and diverse formats, inevitably impacts data processing speed. However, most existing data fusion methods are designed for small-scale datasets. To address the challenges of processing massive amounts of multi-source heterogeneous data, an efficient data fusion method is urgently needed to improve processing speed. Summary of the Invention

[0003] This disclosure provides a data fusion method, apparatus, electronic device, and storage medium.

[0004] According to one aspect of this disclosure, a data fusion method is provided, comprising:

[0005] Data to be merged is obtained from different data sources, including heterogeneous data with inconsistent data formats;

[0006] The attribute fields of the data to be merged are converted to obtain attribute data with a unified data format.

[0007] The attribute data is subjected to attribute classification processing to obtain attribute classification data;

[0008] Data fusion processing is performed based on the attribute-level data to obtain the data fusion result.

[0009] According to one aspect of this disclosure, a search method is provided, comprising:

[0010] Extract search keywords from the obtained search request;

[0011] The search keywords are sent to the backend server, wherein the backend server stores the data fusion results obtained according to the above data fusion method;

[0012] Receive search results obtained by matching the data fusion results with the search keywords.

[0013] According to another aspect of this disclosure, a data fusion apparatus is provided, comprising:

[0014] The data acquisition unit is used to acquire data to be merged from different data sources, including heterogeneous data with inconsistent data formats.

[0015] The data conversion unit is used to convert the data format of the attribute fields of the data to be merged, so as to obtain attribute data with a unified data format.

[0016] A data classification unit is used to perform attribute classification processing on the attribute data to obtain attribute classification data.

[0017] The data fusion unit is used to perform data fusion processing based on the attribute-level data to obtain the data fusion result.

[0018] According to another aspect of this disclosure, a search apparatus is provided, comprising:

[0019] The keyword extraction unit is used to extract search keywords from the acquired search request;

[0020] A keyword sending unit is used to send the search keywords to a backend server, wherein the backend server stores the data fusion results obtained according to the above data fusion method;

[0021] The search result receiving unit is used to receive the search results obtained by matching the data fusion results with the search keywords.

[0022] According to another aspect of this disclosure, an electronic device is provided, comprising:

[0023] At least one processor; and

[0024] The memory is communicatively connected to the at least one processor; wherein,

[0025] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method provided in any embodiment of this disclosure.

[0026] According to another aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions for causing the computer to perform the methods provided in any embodiment of this disclosure.

[0027] According to another aspect of this disclosure, a computer program product is provided, including computer instructions that, when executed by a processor, implement the methods provided in any embodiment of this disclosure.

[0028] By employing this disclosure, data to be merged can be obtained from different data sources, including heterogeneous data with inconsistent data formats. By converting the data formats of the attribute fields of this data, attribute data with a unified data format can be obtained. Attribute classification processing of this attribute data yields attribute-level data. Data fusion processing is then performed based on this attribute-level data to obtain the data fusion result. This allows for the fusion of massive amounts of multi-source heterogeneous data, thereby improving data processing speed.

[0029] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0030] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0031] Figure 1 This is a schematic diagram of an application scenario according to an embodiment of the present disclosure;

[0032] Figure 2 This is a flowchart illustrating a data fusion method according to an embodiment of the present disclosure;

[0033] Figure 3 This is a flowchart illustrating the search method according to an embodiment of the present disclosure;

[0034] Figure 4 This is a schematic diagram of hierarchical fusion in an application example according to an embodiment of the present disclosure;

[0035] Figure 5 This is a schematic diagram illustrating the parsing of a search request in an application example according to an embodiment of this disclosure;

[0036] Figure 6 This is a schematic diagram of a search scenario in an application example according to an embodiment of this disclosure;

[0037] Figures 7-8 This is a schematic diagram showing search results in an application example according to an embodiment of this disclosure;

[0038] Figure 9 This is a schematic diagram of the composition structure of a data fusion apparatus according to an embodiment of the present disclosure;

[0039] Figure 10 This is a schematic diagram of the composition of a search device according to an embodiment of the present disclosure;

[0040] Figure 11 This is a block diagram of an electronic device used to implement the data fusion method or search method of the embodiments of this disclosure. Detailed Implementation

[0041] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0042] In this document, the term "and / or" merely describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. The term "at least one" in this document indicates any combination of at least two of a plurality of elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C. The terms "first" and "second" in this document refer to and distinguish between multiple similar technical terms, not to restrict the order or to limit there to only two. For example, "first feature" and "second feature" refer to two categories / two features; the first feature can be one or more, and the second feature can also be one or more.

[0043] Furthermore, to better illustrate this disclosure, numerous specific details are set forth in the following detailed description. Those skilled in the art will understand that this disclosure can be practiced without certain specific details. In some instances, methods, means, components, and circuits well known to those skilled in the art have not been described in detail in order to highlight the main points of this disclosure.

[0044] Figure 1 This is a schematic diagram of an application scenario according to an embodiment of the present disclosure, such as... Figure 1As shown, the process includes two parts: data fusion processing performed by the backend server 101 and search processing performed by the client 105. The search processing performed by the client 105 is only one example of data operation; the client can also perform other data operations based on the data fusion processing performed by the backend server. During this data fusion processing, the backend server 101 can obtain data to be fused from different data sources (such as multiple platforms), perform data preprocessing 102 (such as data format standardization) on the data to be fused, and then perform data fusion processing 103 to obtain the data fusion result. The data fusion result is then stored in database 104, which supports online or offline search. During this search processing, the user can enter a search request on the display interface of the client 105, carrying the search keyword "xx hotel". The database 104 on the backend server is then matched with the data fusion result based on the search keyword to obtain the search result, which is then displayed on the display interface of the client 105. The search results can include two parts: first information is displayed in the first area 106 of the client display interface. This first information is obtained by aggregating multiple basic information provided by different platforms (such as different hotel platforms) for the same point of interest (POI); and second information is displayed in the second area 107 of the client display interface. This second information is obtained by aggregating multiple price comparison information provided by different platforms (such as different hotel platforms) for the same POI.

[0045] According to embodiments of this disclosure, a data fusion method is provided. Figure 2 This is a flowchart illustrating a data fusion method according to an embodiment of the present disclosure. This method can be applied to a data fusion device, for example, deployed in a single-machine, multi-machine, or cluster system as a terminal, server, or other processing device. It can achieve the integration and processing of massive amounts of heterogeneous data from multiple sources. The terminal can be a user equipment (UE), mobile device, personal digital assistant (PDA), handheld device, computing device, in-vehicle device, wearable device, etc. In some possible implementations, the method can also be implemented by a processor calling computer-readable instructions stored in memory. Figure 2 As shown, it includes:

[0046] S201. Obtain data to be merged from different data sources, including heterogeneous data with inconsistent data formats.

[0047] S202. Perform data format conversion on the attribute fields of the data to be merged to obtain attribute data with a unified data format.

[0048] S203. Perform attribute classification processing on the attribute data to obtain attribute classification data.

[0049] S204. Perform data fusion processing based on the attribute-level data to obtain the data fusion result.

[0050] In one example from S201-S204, data to be merged is obtained from different data sources, such as different ticketing platforms, clients or websites of different lifestyle applications (e.g., hotel applications, movie ticket applications, concert applications, etc.). Considering that the data to be merged belongs to different data sources, the data format definitions are different. In other words, the data to be merged includes heterogeneous data with inconsistent data formats. Before performing data operations such as search processing, data fusion processing is required. Specifically, the attribute fields of the data to be merged (e.g., the fields to be processed to describe POI attributes) can be converted to obtain attribute data with a unified data format. Attribute classification processing can be performed on the attribute data to obtain attribute classification data (e.g., attribute classification data to describe POI attributes). Data fusion processing can then be performed based on this attribute classification data to obtain the data fusion result (e.g., the data fusion result obtained based on POI identity).

[0051] By employing the embodiments of this disclosure, data to be merged can be obtained from different data sources. This data includes heterogeneous data with inconsistent data formats. By converting the data format of the attribute fields of the data to be merged, attribute data with a unified data format can be obtained. By performing attribute classification processing on the attribute data, attribute-classified data can be obtained. Data fusion processing is then performed based on the attribute-classified data to obtain the data fusion result. This enables the fusion of massive amounts of multi-source heterogeneous data, thereby improving data processing speed.

[0052] In one embodiment, the attribute fields of the data to be fused are converted to obtain attribute data with a unified data format. This includes: if the attribute field includes a field to be processed in the data to be fused that describes POI attributes, extracting a first template for describing POI attribute standardization, and converting the data format of the field to be processed according to the first template to obtain the attribute data. Using this embodiment, the field to be processed that describes POI attributes can be converted using a first template (such as a tabular chart) for describing POI attribute standardization. Since the first template includes a unified data format satisfied by POI attribute standardization, unified attribute data can be obtained through data format conversion. After the attribute data format is unified, it facilitates more efficient subsequent attribute classification and data fusion processing.

[0053] In one embodiment, attribute data is subjected to attribute classification processing to obtain attribute-classified data. This includes: extracting a second template for describing POI attribute classification; and performing attribute classification processing on the attribute data according to the second template to obtain attribute-classified data. Using this embodiment, for attribute-classified data (such as attribute-classified data describing POI attributes), attribute classification processing can be performed using a second template (such as a tabular chart) for describing POI attribute classification. Since the second template includes different data types satisfied by POI attribute classification, attribute-classified data can be obtained through attribute classification processing. After classification, data fusion processing can be performed according to the classification fusion priority, facilitating more efficient subsequent data fusion processing.

[0054] It should be noted that the first template and the second template can be integrated into the same template, or they can be separate templates. When they are separate templates, the processing speed is faster.

[0055] In one embodiment, the method further includes: obtaining the POI attribute classification based on at least one of the following: the administrative region to which the POI belongs, the geographic coordinates corresponding to the POI, and rich media information including POI attribute descriptions. This embodiment allows for diverse methods of POI attribute classification, thereby meeting different design requirements for data classification accuracy or data fusion accuracy according to actual user needs.

[0056] In one embodiment, data fusion processing is performed based on attribute-level data to obtain a data fusion result. This includes: performing data fusion processing on the attribute-level data based on the POI identity identifier, mapping the attribute-level data onto a map containing the POI identity identifier, and obtaining the data fusion result. Using this embodiment, the attribute-level data can be mapped onto a map based on the POI identity identifier (i.e., a POI identity system with unique identifiers, or a baseline identity system). This achieves geographical uniformity of POI addresses and provides globally unique POI identity identifiers, facilitating various client applications to match the data fusion result based on these unique POI identity identifiers.

[0057] In one embodiment, the method further includes performing data fusion processing based on the hierarchical fusion priority corresponding to the attribute-level data. In this embodiment, for example, the hierarchical fusion priority may include: first-level attributes, second-level attributes, and third-level attributes, with first-level attributes > second-level attributes > third-level attributes. When a first-level attribute can be obtained, data fusion processing is performed using the first-level attribute; otherwise, if the first-level attribute cannot be found, data fusion processing is performed using the second-level attribute; if neither can be found, finally, data fusion processing is performed using the third-level attribute. This can improve the data fusion speed and match different data fusion accuracies.

[0058] According to embodiments of this disclosure, a search method is provided. Figure 3 This is a flowchart illustrating a search method according to an embodiment of the present disclosure. This method can be applied to a search device, for example, deployed in a single-machine, multi-machine, or cluster system as a terminal, server, or other processing device. It can perform processing such as searching massive amounts of multi-source heterogeneous data. The terminal can be a user equipment (UE), mobile device, personal digital assistant (PDA), handheld device, computing device, in-vehicle device, wearable device, etc. In some possible implementations, the method can also be implemented by a processor calling computer-readable instructions stored in memory. Figure 3 As shown, it includes:

[0059] S301. Extract search keywords from the obtained search request.

[0060] S302. The search keyword is sent to the backend server, wherein the backend server stores the data fusion result obtained according to the data fusion method of the above embodiment.

[0061] S303. Receive the search results obtained by fusion of matching data based on the search keywords.

[0062] In one example of S301-S303, search keywords, such as "xx hotel" or "xx movie", can be extracted from the obtained search request. The search keywords, such as "xx hotel" or "xx movie", are sent to the backend server. Since the backend server stores the data fusion results after data fusion processing, such as "fusion results of multiple platforms for xx hotel" or "fusion results of multiple platforms for xx movie", the search keyword is matched with the data fusion results to obtain the search results, such as price comparison information for "xx hotel" or "xx movie".

[0063] Using the embodiments of this disclosure, for data operations in search scenarios, the search keywords in the search request are matched from the data fusion results of the above embodiments. Since the data fusion results are obtained by converting, classifying, and fusing multi-source heterogeneous data from different platforms, the data format is unified. According to the attribute classification, it can also meet the classification of different search types and search accuracy requirements. Not only can relevant search results from different platforms be obtained in the same search result, but the search is also faster and more accurate.

[0064] In one embodiment, the method further includes displaying search results. Where the search keyword includes a POI, the search results represent the fusion results of different resource providers for the same POI. Using this embodiment, since the data fusion result can be obtained by performing data fusion processing on the fields to be processed to describe the attributes of the POI, for example, the data fusion result can be obtained based on POI identity identifiers. Therefore, the search keyword can be matched against the data fusion result based on the POI identity identifiers to obtain the fusion results of different resource providers for the same POI, facilitating price comparison for users of the same POI.

[0065] In one embodiment, displaying search results includes: in a first area of ​​the client display interface, a first piece of information is obtained by aggregating multiple basic information sources provided by different resource providers for the same POI; in a second area of ​​the client display interface, multiple second pieces of information are obtained by aggregating multiple price comparison information sources provided by different resource providers for the same POI. The search results are then displayed using graphic rendering based on the first and second pieces of information. Using this embodiment, the search results include: the first piece of information obtained from the basic information, and the multiple second pieces of information obtained from the multiple price comparison information sources, satisfying users' needs for price comparison queries for the same POI on the same client, resulting in faster and more accurate searches.

[0066] The data fusion method and search method provided in the above-described embodiments of this disclosure are illustrated below.

[0067] In search scenarios, data varies across different ticketing platforms. For example, different platforms may offer different prices and discounts for the same hotel. Therefore, users need to access data from different platforms and compare prices. Taking online ticketing as an example, websites providing online ticketing services are called online ticketing service providers. Each ticketing platform operates independently, and users typically only have access to ticketing information from one platform. To compare ticketing services offered by different platforms, multiple clients need to be downloaded. The ability to integrate resources from different ticketing platforms urgently needs improvement. There is a need to provide users with a unified entry point for searching ticketing information and accessing ticketing services based on a search engine, integrating data from different industries and ticketing platforms. The following application example illustrates this with hotel and movie tickets, but it is not limited to hotels and movie tickets; it can also integrate resources from different ticketing platforms for other ticketing services such as attraction tickets.

[0068] The technical terms used in the application examples are explained below:

[0069] a. Point of Interest (POI): A POI is a basic entity in geospatial space. A POI can be a building, a shop, a city, a bus stop, a street, etc. Many commercial and consumer entities are POIs, such as hotels in accommodation consumption and cinemas and amusement parks in leisure and entertainment consumption.

[0070] b. Multi-source heterogeneous POIs: The ticketing services provided by various ticketing platforms have different requirements. The attribute descriptions of the same POI have their own professional terminology. The quality levels of service resource providers accessed through various ticketing platforms are uneven, and the data standards are not uniform. To integrate and manage such multi-source heterogeneous POIs, preprocessing with standardized data formats is required.

[0071] c. POI Price Comparison: For the same commercial POI, different ticketing platforms will have different prices. It is necessary to integrate the resources of each ticketing platform to finally display the price comparison information of each ticketing platform for the same POI.

[0072] By directly connecting various ticketing platforms to the Points of Interest (POIs) of service providers at different levels, each ticketing platform can directly provide pricing information. Alternatively, intermediaries can connect to the POIs of service providers at different levels, with the intermediaries controlling pricing resources. Each ticketing platform then connects with the intermediary to display pricing information. Both methods offer exclusivity in data fusion and processing, meaning that users are shielded from information about inconsistent prices across different channels. This means that users who want to obtain and compare prices from multiple channels need to download multiple clients corresponding to each ticketing platform or intermediary. This is especially problematic for intermediaries that do not have independent clients, preventing users from obtaining pricing information and comparing prices.

[0073] This application example mainly includes the following: data fusion of fields to be processed to describe POI attributes, data fusion results being stored in a database to build an offline system, and data fusion results from the offline system being used to implement search in the online system. The details are as follows:

[0074] I. Data Fusion for Fields to be Processed that Describe POI Attributes

[0075] 1. POI Attribute Standardization & Attribute Classification:

[0076] The POI attribute data of different ticketing platforms under different search scenarios are different, and the ways in which the fields used to describe POI attributes are expressed by different ticketing platforms are also diverse. A unified attribute standard is needed. Considering that different attributes have different impacts on the accuracy of data fusion, the attribute data needs to be used in a graded manner according to different requirements of data fusion accuracy.

[0077] Taking the integration of the first template for describing POI attribute standardization and the second template for describing POI attribute hierarchy into a single template (as shown in Table 1), we can perform data format conversion on the fields to be processed that describe POI attributes using the data format in the "Example" section of this template, thereby obtaining unified attribute data. Then, for this attribute data, we can perform attribute hierarchy processing using the "hierarchy" corresponding to the "attribute" in this template, resulting in attribute hierarchy data. For example, we can divide the POI according to its administrative region to obtain secondary attributes, and so on, without further elaboration.

[0078]

[0079] Table 1

[0080] 2. Data fusion processing based on POI identity identifiers:

[0081] When fusing data from heterogeneous POIs from various ticketing platforms, a POI identity system (or baseline identity system) with unique identifiers is required. This baseline identity system can be a map with the POI identifier, so that each POI of different types (such as cinema merchant POIs, hotel merchant POIs, scenic spot merchant POIs, etc.) has a globally unique POI identifier, so that the data fusion results can be matched based on the POI identifier in the future.

[0082] Based on the attribute data of POIs, the data from various ticketing platforms is unified and integrated into a baseline identity system. For example, it is mapped onto a map with the POI identity identifier, and the coordinates of the map with the POI identity identifier are used as a unified coordinate system to obtain the final data fusion result. This ensures that in the data fusion result, the addresses of the same POI for each ticketing platform are geographically unified and have a globally unique POI identity identifier.

[0083] Furthermore, based on different attribute field types (number, text, rich text, etc.), the attribute data based on POI points can be filtered using various similarity algorithms to improve data fusion efficiency. The following templates (such as the charts shown in Table 2) can also be used for data fusion processing.

[0084]

[0085] Table 2

[0086] Furthermore, data fusion processing can also be performed using the attribute hierarchy (such as first-level attributes, second-level attributes, third-level attributes, etc.) in the above template (as shown in the chart in Table 1) to match the needs of different fusion precision. Figure 4This is a schematic diagram of hierarchical fusion in an application example according to an embodiment of this disclosure, such as... Figure 4 As shown, heterogeneous POIs from various ticketing platforms are unified and integrated into a baseline identity system. After attribute preprocessing, hierarchical fusion processing can be performed. For example, the hierarchical fusion priority can include: first-level attributes, second-level attributes, and third-level attributes, with first-level attributes > second-level attributes > third-level attributes. When a first-level attribute can be obtained, it is used for data fusion processing; otherwise, if a first-level attribute cannot be found, second-level attributes are used; if neither is found, third-level attributes are used for data fusion processing. This can improve the data fusion speed and match different data fusion accuracies. The fusion results obtained after hierarchical fusion are stored in a database to obtain a fused POI database, which can also be stored in an offline system.

[0087] II. Input the data fusion results into the database to build an offline system.

[0088] The offline system can store the data fusion results (the fused POI library). The offline system can be an offline database, running independently of the data fusion processing. Alternatively, the offline system can include a data submission module and a POI fusion module. That is, in the offline system, the data submission module receives multi-source heterogeneous POIs from various ticketing platforms (supporting multiple data submission modes, including offline full data and real-time batch data), and the POI fusion module performs data fusion processing to obtain the data fusion results.

[0089] The multi-source heterogeneous POIs received by the data submission module can be submitted by various ticketing platforms in a data schema format according to a pre-agreed data format, or they can be original data submitted by various ticketing platforms. If the original data does not meet the data format, the data submission module will perform data format conversion preprocessing before use.

[0090] III. Implementing online system search based on data fusion results from offline systems.

[0091] 1. Analyze the search requirements of the search request.

[0092] After the online system receives a user's search request (e.g., a search request sent by a user through different entry points such as a mini-program, client, or H5 webpage platform), it parses the search request to obtain the search keywords. To improve search accuracy, it can utilize the prior experience of neural network training. That is, it can use consumption data such as hotel consumption data, cinema consumption data, and concert consumption data as training samples to pre-train a classification template or classification model. The trained classification template or classification model is then used to parse the search request to identify which specific category (e.g., hotel, cinema, concert, etc.) the search keywords in the search request satisfy.

[0093] Furthermore, the search requirements identified through trained classification templates or models can be provided to other search services, not limited to entity search or key-value (KV) search, to better identify the search requirements in the search request.

[0094] Figure 5 This is a schematic diagram illustrating the parsing of a search request according to an application example of an embodiment of this disclosure, such as... Figure 5 As shown, taking a classification model as an example, after training the classification model with training data for hotel demand entity recognition, the search request is input into the classification model (e.g., an NLP named entity recognition model). The classification model parses the search request's search requirements and determines whether it is a hotel POI entity / hotel POI demand entity. If so, the hotel entity word is output. Table 3 shows the input / output examples of parsing based on POI entities using this classification model. The input example is the following search request, and the output example is the entity corresponding to the search requirement.

[0095]

[0096] Table 3

[0097] Specifically, to improve the parsing accuracy of this classification model, processing techniques such as data mining and generalization can be performed, including the following:

[0098] 1) Demand Vector (pattern) Mining & Generalization

[0099] During model training, for users with POI ticket purchase needs, the actual search requests of users are analyzed to extract entities other than POI entities from the search requests. These entities are the demand patterns. For example, a search request might include: "How much does it cost to stay at xx Hotel in Beijing for one night?", where "xx Hotel in Beijing" is the POI entity, and "How much does it cost to stay for one night" is the demand pattern. Generalizing demand patterns involves finding similar expressions, such as "How much does it cost to stay once" or "What's the cheapest price to stay for one night?". Specifically, demand pattern mining can be done by subtracting all POI entity names from the search requests of all users with ticket purchase needs; generalizing demand patterns involves similarity matching from NLP short text search requests to obtain similar descriptions.

[0100] 2) Training of the classification model

[0101] Based on the extracted POI entities, a POI entity dictionary can be obtained; based on the mining and generalization of the extracted demand patterns, a demand pattern dictionary (or ticket purchase demand dictionary) can be obtained. Using the POI entity dictionary and the ticket purchase demand dictionary as training data, a classification model can be trained. A small amount of training data can be used to train the classification model. In online ticket purchase, this classification model is specifically a text multi-classification model that determines whether a user's query has a certain type of ticket purchase demand.

[0102] Furthermore, a demand template can be constructed using the [POI entity dictionary] and [demand pattern dictionary]. If the demand vector pattern obtained from the search request matches the demand pattern dictionary, the demand template is invoked. Then, the classification template is used to parse the POI entities in the search request to clarify what specific search demand exists for that POI entity.

[0103] 2. POI online search

[0104] Figure 6 This is a schematic diagram of a search scenario in an application example according to an embodiment of this disclosure, such as... Figure 6 As shown, the POI online search includes: identifying the search requirements (such as ticket purchase requirements) based on the user's input search request, extracting the POI entity name (such as "xx hotel") from the search request, searching the POI database in the offline system using a key-value search method, obtaining the target POI entity corresponding to the POI entity name and the corresponding POI identity identifier, and obtaining the search results based on the target POI entity and the corresponding POI identity identifier. The search results can be price comparison information provided by various ticketing platforms, and the price comparison information can be displayed in a rich media format.

[0105] Figure 7 This is a schematic diagram showing search results in an application example according to an embodiment of this disclosure, such as... Figure 7 As shown, a price comparison page 701 for the same hotel is displayed on the client's screen after data fusion from various ticketing platforms. This price comparison page 701 is the search result returned by the backend server after receiving the client's search request (the search request carries the search keyword "xx hotel"). Specifically, in one example of the backend server, the backend server may include an online system and an offline system. After receiving the client's search request through the online system, the offline system matches the stored data fusion results according to the search keyword. The data fusion result matching the search keyword is the search result. The search result includes two parts: the first area 702 displays first information, which is the information obtained by fusion of multiple basic information provided by different hotel platforms for the same POI; the second area 703 displays multiple second pieces of information, which are multiple pieces of information obtained by fusion of multiple price comparison information provided by different hotel platforms for the same POI.

[0106] Figure 8 This is a schematic diagram showing search results in an application example according to an embodiment of this disclosure, such as... Figure 8 As shown, the client's screen displays a price comparison page 801 for the same movie, which is the result of data fusion from various ticketing platforms. This price comparison page 801 is the search result returned by the backend server after receiving the client's search request (which carries the search keyword "xx movie") in the search box 802. Specifically, in one example of the backend server, the backend server may include an online system and an offline system. After receiving the client's search request through the online system, the offline system matches the stored data fusion results according to the search keyword. The data fusion result that matches the search keyword is the search result. The search result includes two parts: the first area 803 displays first information, which is the information obtained by fusion of multiple basic information provided by different cinema platforms for the same POI; the second area 804 displays multiple second pieces of information, which are multiple pieces of information obtained by fusion of multiple price comparison information provided by different cinema platforms for the same POI.

[0107] According to embodiments of this disclosure, a data fusion apparatus is provided. Figure 9 This is a schematic diagram of the composition structure of a data fusion apparatus according to an embodiment of the present disclosure, such as... Figure 9As shown, the data fusion device 900 includes: a data acquisition unit 901, used to acquire data to be fused from different data sources, the data to be fused including heterogeneous data with inconsistent data formats; a data conversion unit 902, used to convert the data format of the attribute fields of the data to be fused to obtain attribute data with a unified data format; a data classification unit 903, used to perform attribute classification processing on the attribute data to obtain attribute-classified data; and a data fusion unit 904, used to perform data fusion processing based on the attribute-classified data to obtain a data fusion result.

[0108] In one embodiment, the data conversion unit is configured to, when the attribute field includes a field to be processed in the data to be merged that describes the POI attribute, extract a first template for describing the standardization of the POI attribute; and perform data format conversion processing on the field to be processed according to the first template to obtain the attribute data.

[0109] In one embodiment, the data classification unit is used to extract a second template for describing the classification of POI attributes; and to perform attribute classification processing on the attribute data according to the second template to obtain attribute classification data.

[0110] In one embodiment, the system further includes an attribute classification generation unit, which is used to obtain the POI attribute classification based on at least one of the following: the administrative region to which the POI belongs, the geographic coordinates corresponding to the POI, and rich media information including the POI attribute description.

[0111] In one embodiment, the data fusion unit is used to perform data fusion processing on the attribute-level data according to the POI identity identifier, and map the attribute-level data onto a map with the POI identity identifier to obtain a data fusion result.

[0112] In one embodiment, the system further includes a hierarchical fusion unit, which performs the data fusion processing according to the hierarchical fusion priority corresponding to the attribute hierarchical data.

[0113] According to embodiments of this disclosure, a search device is provided. Figure 10 This is a schematic diagram of the composition structure of a search device according to an embodiment of the present disclosure, such as... Figure 10 The search device 1000 shown includes: a keyword extraction unit 1001, used to extract search keywords from the acquired search request; a keyword sending unit 1002, used to send the search keywords to a backend server, wherein the backend server stores data fusion results obtained according to the above data fusion method; and a search result receiving unit 1003, used to receive search results obtained by matching the search keywords with the data fusion results.

[0114] In one embodiment, a display unit is further included for displaying the search results; wherein, when the search keywords include POI, the search results are used to characterize the fusion results of different resource providers for the same point of interest POI.

[0115] In one embodiment, the display unit is configured to: merge multiple basic information provided by different resource providers for the same POI in a first area of ​​the client display interface to obtain first information; merge multiple price comparison information provided by different resource providers for the same POI in a second area of ​​the client display interface to obtain multiple second information; and perform graphic rendering based on the first information and the multiple second information to display the search results.

[0116] The acquisition, storage, and application of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0117] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0118] Figure 11 A schematic block diagram of an example electronic device 1100 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0119] like Figure 11 As shown, the electronic device 1100 includes a computing unit 1101, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 1102 or a computer program loaded into a random access memory (RAM) 1103 from a storage unit 1108. The RAM 1103 may also store various programs and data required for the operation of the electronic device 1100. The computing unit 1101, ROM 1102, and RAM 1103 are interconnected via a bus 1104. An input / output (I / O) interface 1105 is also connected to the bus 1104.

[0120] Multiple components in electronic device 1100 are connected to I / O interface 1105, including: input unit 1106, such as keyboard, mouse, etc.; output unit 1107, such as various types of displays, speakers, etc.; storage unit 1108, such as disk, optical disk, etc.; and communication unit 1109, such as network card, modem, wireless transceiver, etc. Communication unit 1109 allows electronic device 1100 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0121] The computing unit 1101 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1101 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 1101 performs the various methods and processes described above, such as data fusion methods or search methods. For example, in some embodiments, the data fusion method or search method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 1108. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 1100 via ROM 1102 and / or communication unit 1109. When the computer program is loaded into RAM 1103 and executed by the computing unit 1101, one or more steps of the data fusion method or search method described above may be performed. Alternatively, in other embodiments, the computing unit 1101 may be configured to perform a data fusion method or a search method by any other suitable means (e.g., by means of firmware).

[0122] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0123] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0124] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0125] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0126] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0127] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0128] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this disclosure can be achieved, and this is not limited herein.

[0129] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A data fusion method, comprising: Data to be merged is obtained from different data sources, including heterogeneous data with inconsistent data formats; The attribute fields of the data to be merged are converted to obtain attribute data with a unified data format; wherein, the attribute fields include fields to be processed to describe the attributes of POI; The attribute data is subjected to attribute classification processing to obtain attribute classification data; wherein, the attribute classification data includes attribute classification data used to describe POI attributes; the attribute classification data has a corresponding classification fusion priority; Based on the hierarchical fusion priority corresponding to the attribute-level data, data fusion processing is performed to obtain the data fusion result; The step of performing data fusion processing based on the attribute-level data to obtain a data fusion result includes: performing data fusion processing on the attribute-level data based on POI identity identifiers, mapping the attribute-level data to a baseline identity system, and obtaining a data fusion result; wherein, the POI identity identifier is a unique identifier for different POIs in the baseline identity system; the baseline identity system is a map containing the POI identity identifiers; in the data fusion result, the addresses of the same POI from different data sources are geographically unified and have globally unique POI identity identifiers.

2. The method according to claim 1, wherein, The step of converting the attribute fields of the data to be merged to obtain attribute data with a unified data format includes: Extract the first template used to describe the normalization of POI attributes; The data format of the field to be processed is converted according to the first template to obtain the attribute data.

3. The method according to claim 2, wherein, The attribute data is subjected to attribute classification processing to obtain attribute classification data, including: Extract the second template used to describe the POI attribute hierarchy; The attribute data is processed by attribute classification according to the second template to obtain attribute classification data.

4. The method according to claim 3, further comprising: The POI attribute classification is obtained based on at least one of the following: the administrative region to which the POI belongs, the geographic coordinates corresponding to the POI, and rich media information including the POI attribute description.

5. A search method, comprising: Extract search keywords from the obtained search request; The search keywords are sent to the backend server, wherein the backend server stores the data fusion results obtained according to the data fusion method described in claims 1-4 above; Receive search results obtained by matching the data fusion results with the search keywords.

6. The method according to claim 5, further comprising: The search results are displayed; When the search keywords include Points of Interest (POIs), the search results are used to represent the fusion results of different resource providers for the same POI.

7. The method according to claim 6, wherein, The display of the search results includes: In the first area of ​​the client display interface, the first information is obtained by merging multiple basic information provided by different resource providers for the same POI; In the second area of ​​the client display interface, multiple price comparison information provided by different resource providers for the same POI is merged to obtain multiple second information; Based on the first information and the plurality of second information, image and text rendering is performed to display the search results.

8. A data fusion apparatus, comprising: The data acquisition unit is used to acquire data to be merged from different data sources, including heterogeneous data with inconsistent data formats. The data conversion unit is used to convert the data format of the attribute fields of the data to be merged to obtain attribute data with a unified data format; wherein, the attribute fields include fields to be processed for describing POI attributes; A data grading unit is used to perform attribute grading processing on the attribute data to obtain attribute grading data; wherein, the attribute grading data includes attribute grading data for describing POI attributes; the attribute grading data has a corresponding grading fusion priority. The data fusion unit is used to perform data fusion processing according to the hierarchical fusion priority corresponding to the attribute hierarchical data to obtain the data fusion result; The data fusion unit is further configured to perform data fusion processing on the attribute-level data based on the POI identity identifier, mapping the attribute-level data onto a baseline identity system to obtain a data fusion result; wherein the POI identity identifier is a unique identifier for different POIs in the baseline identity system; the baseline identity system is a map containing the POI identity identifier; in the data fusion result, addresses of the same POI from different data sources are geographically unified and have globally unique POI identity identifiers.

9. The apparatus according to claim 8, wherein, The data conversion unit is used for: Extract the first template used to describe the normalization of POI attributes; The data format of the field to be processed is converted according to the first template to obtain the attribute data.

10. The apparatus according to claim 9, wherein, The data hierarchical unit is used for: Extract the second template used to describe the POI attribute hierarchy; The attribute data is processed by attribute classification according to the second template to obtain attribute classification data.

11. The apparatus according to claim 10, further comprising an attribute hierarchy generation unit, configured to: The POI attribute classification is obtained based on at least one of the following: the administrative region to which the POI belongs, the geographic coordinates corresponding to the POI, and rich media information including the POI attribute description.

12. A search device, comprising: The keyword extraction unit is used to extract search keywords from the acquired search request; A keyword sending unit is used to send the search keywords to a backend server, wherein the backend server stores the data fusion results obtained according to the data fusion method described in claims 1-4. The search result receiving unit is used to receive the search results obtained by matching the data fusion results with the search keywords.

13. The apparatus of claim 12, further comprising a display unit, configured to: The search results are displayed; When the search keywords include Points of Interest (POIs), the search results are used to represent the fusion results of different resource providers for the same POI.

14. The apparatus according to claim 13, wherein, The display unit is used for: In the first area of ​​the client display interface, the first information is obtained by merging multiple basic information provided by different resource providers for the same POI; In the second area of ​​the client display interface, multiple price comparison information provided by different resource providers for the same POI is merged to obtain multiple second information; Based on the first information and the plurality of second information, image and text rendering is performed to display the search results.

15. An electronic device comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

16. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-7.

17. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-7.

Images