House resource searching method, electronic equipment, storage medium and program product

Abstract

The invention relates to a housing resource searching method, electronic equipment, a storage medium and a program product. The housing resource searching efficiency can be improved. The method comprises the following steps: in response to a housing resource search instruction, obtaining at least one structured field and at least one unstructured field corresponding to the housing resource search instruction; generating a vectorized target query statement based on the at least one structured field, the at least one unstructured field and a target similarity algorithm; based on the target similarity algorithm, multiple target house resources matched with the target query statement are searched from a house resource database, the house resource database comprises multiple house resources and a vectorization index corresponding to each house resource, and the vectorization indexes comprise structured indexes and unstructured indexes of the house resources.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of search, and particularly relates to a house source search method, an electronic device, a storage medium and a program product. BACKGROUND

[0002] The current user house searching dimension is limited, and the user can only select and search through the existing fields (i.e. structured fields) in the search box. If there is other dimension demand (such as a house source near a park, a house source with good lighting, etc.), the user needs to exclude one by one from the house source search result, and then filter out the house source meeting the demand. Thus, the house source search process in the related art is time-consuming, and the search efficiency is low. SUMMARY

[0003] In order to solve the above technical problems or at least partially solve the above technical problems, the present disclosure provides a house source search method, an electronic device, a storage medium and a program product.

[0004] In a first aspect of the embodiments of the present disclosure, a house source search method is provided, which includes: in response to a house source search instruction, acquiring at least one structured field and at least one unstructured field corresponding to the house source search instruction; generating a vectorized target query statement based on the at least one structured field, the at least one unstructured field and a target similarity algorithm; and searching a plurality of target house sources matching the target query statement from a house source database based on the target similarity algorithm, the house source database including a plurality of house sources and a vectorized index corresponding to each house source, the vectorized index including a structured index and an unstructured index of the house source.

[0005] In some embodiments of the present disclosure, the query statement includes a plurality of query conditions, and the searching a plurality of target house sources matching the target query statement from the house source database based on the target similarity algorithm includes: generating a plurality of query statements based on the target query statement, the plurality of query statements including the target query statement and at least one first query statement, the at least one first query statement being respectively less than the target query statement by one query condition; searching a plurality of house source sets matching the plurality of query statements from the house source database based on the target similarity algorithm, each house source set including a plurality of candidate house sources; and rearranging the candidate house sources in the plurality of house source sets to obtain the plurality of target house sources.

[0006] In some embodiments of the present disclosure, the rearranging the candidate house sources in the plurality of house source sets to obtain the plurality of target house sources comprises: determining similarity scores of the candidate house sources in the plurality of house source sets with the plurality of query conditions respectively; determining a weighting value of the similarity score of each candidate house source with the plurality of query conditions respectively; determining the candidate house source with a weighting value of the similarity score greater than or equal to a score threshold as the target house source.

[0007] In some embodiments of the present disclosure, the rearranging the candidate house sources in the plurality of house source sets to obtain the plurality of target house sources comprises: inputting the candidate house sources in the plurality of house source sets into a rearrangement model respectively, and outputting the plurality of target house sources; wherein the rearrangement model is used to filter the candidate house source with a weighting value of the similarity score greater than or equal to a score threshold from the candidate house sources in the plurality of house source sets as the target house source, and the weighting value of the similarity score is the weighting value of the similarity score of each candidate house source with the plurality of query conditions respectively.

[0008] In some embodiments of the present disclosure, in response to the house source search instruction, the at least one structured field and the at least one unstructured field corresponding to the house source search instruction are obtained: in response to the house source search instruction, a target search statement input by a user in a search information input interface or extracted from a target chat record of the user is obtained; based on a plurality of search keywords in the target search statement, the at least one structured field and the at least one unstructured field are generated; wherein the at least one structured field comprises a structured field in the plurality of search keywords, a target unstructured field in the plurality of search keywords, and an unstructured field semantically same as the target unstructured field.

[0009] In some embodiments of the present disclosure, in response to the house source search instruction, the target search statement is extracted from the target chat record of the user, comprising: in response to the house source search instruction, inputting the target chat record into a target search statement generation model to output the target search statement; wherein the target search statement generation model is used to extract key information from the chat record and generate a search statement based on the extracted key information.

[0010] In some embodiments of the present disclosure, the generating the vectorized target query statement based on the at least one structured field, the at least one unstructured field, and a target similarity algorithm comprises: inputting the at least one structured field and the at least one unstructured field into a query statement generation model to output the target query statement; wherein the query statement generation model is a model with preset model hyperparameters, and the similarity algorithm is the target similarity algorithm.

[0011] In a second aspect, the present disclosure provides a house source searching device, which comprises: an acquisition module configured to acquire at least one structured field and at least one unstructured field corresponding to a house source searching instruction in response to the house source searching instruction; a generation module configured to generate a vectorized target query statement based on the at least one structured field, the at least one unstructured field and a target similarity algorithm; and a searching module configured to search a plurality of target house sources matching the target query statement from a house source database based on the target similarity algorithm, wherein the house source database comprises a plurality of house sources and a vectorized index corresponding to each house source, and the vectorized index comprises a structured index and an unstructured index of the house source.

[0012] In some embodiments of the present disclosure, the query statement comprises a plurality of query conditions, and the searching module is specifically configured to generate a plurality of query statements based on the target query statement, wherein the plurality of query statements comprises the target query statement and at least one first query statement, and each first query statement is less than the target query statement by one query condition; search a plurality of house source sets matching the plurality of query statements from the house source database based on the target similarity algorithm, wherein each house source set comprises a plurality of candidate house sources; and rearrange the candidate house sources in the plurality of house source sets to obtain the plurality of target house sources.

[0013] In some embodiments of the present disclosure, the searching module is specifically configured to determine similarity scores of the candidate house sources in the plurality of house source sets with respect to the plurality of query conditions respectively; determine a weighting value of the similarity score of each candidate house source with respect to the plurality of query conditions respectively; and determine the candidate house sources with the weighting value of the similarity score greater than or equal to a score threshold as the plurality of target house sources.

[0014] In some embodiments of the present disclosure, the searching module is specifically configured to input the candidate house sources in the plurality of house source sets into a rearrangement model respectively, and output the plurality of target house sources; wherein the rearrangement model is configured to filter the candidate house sources with the weighting value of the similarity score greater than or equal to the score threshold from the candidate house sources in the plurality of house source sets as the target house sources, and the weighting value of the similarity score is the weighting value of the similarity score of each candidate house source with respect to the plurality of query conditions respectively.

[0015] In some embodiments of the present disclosure, the acquisition module is specifically configured to acquire a target search statement input by a user in a search information input interface or extracted from a target chat record of the user in response to the house source searching instruction; generate the at least one structured field and the at least one unstructured field based on a plurality of search keywords in the target search statement; and wherein the at least one structured field comprises a structured field in the plurality of search keywords, a target unstructured field in the plurality of search keywords and an unstructured field with the same semantics as the target unstructured field.

[0016] In some embodiments of the present disclosure, the acquisition module is specifically configured to input the target chat record into a target search sentence generation model in response to the house source search instruction, and output the target search sentence; wherein the target search sentence generation model is used for key information extraction on the chat record, and generates a search sentence based on the extracted key information.

[0017] In some embodiments of the present disclosure, the generation module is specifically configured to input the at least one structured field and the at least one unstructured field into a query sentence generation model, and output the target query sentence; wherein the query sentence generation model is a model with preset model hyperparameters, and the similarity algorithm is the target similarity algorithm.

[0018] In a third aspect of the embodiments of the present disclosure, an electronic device is provided, which includes a processor, a memory, and a computer program stored in the memory and executable on the processor, and when the computer program is executed by the processor, the house source search method according to the first aspect is implemented.

[0019] In a fourth aspect of the embodiments of the present disclosure, a computer readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the house source search method according to the first aspect is implemented.

[0020] In a fifth aspect of the embodiments of the present disclosure, a computer program product is provided, which includes a computer program, and when the computer program product is executed on a processor, the processor executes the computer program to implement the house source search method according to the first aspect.

[0021] In a sixth aspect of the embodiments of the present disclosure, a chip is provided, which includes a processor and a communication interface, the communication interface is coupled with the processor, and the processor is used to run program instructions to implement the house source search method according to the first aspect.

[0022] Compared with the prior art, the technical scheme provided by the embodiments of the present disclosure has the following advantages: in response to a house source search instruction, at least one structured field and at least one unstructured field corresponding to the house source search instruction are obtained; a vectorized target query statement is generated based on the at least one structured field, the at least one unstructured field, and a target similarity algorithm; based on the target similarity algorithm, a plurality of target house sources matching the target query statement are searched from a house source database, the house source database including a plurality of house sources and a vectorized index corresponding to each house source, the vectorized index including a structured index and an unstructured index of the house source. In this way, in the embodiments of the present disclosure, the query statement generated by combining the structured field and the unstructured field can quickly search for house sources that meet the user's query requirements for the unstructured field, thereby improving the house source search efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0023] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and serve to explain the principles of the present disclosure together with the specification.

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the accompanying drawings required to be used in the embodiments or prior art description will be briefly introduced. Obviously, for those skilled in the art, other drawings can also be obtained based on these drawings without creative labor.

[0025] Figure 1 One of the flowcharts of the house source search method provided by the embodiments of the present disclosure;

[0026] Figure 2 The second flowchart of the house source search method provided by the embodiments of the present disclosure;

[0027] Figure 3 The third flowchart of the house source search method provided by the embodiments of the present disclosure;

[0028] Figure 4 The structural block diagram of a house source search device provided by the embodiments of the present disclosure;

[0029] Figure 5 The structural block diagram of an electronic device provided by the embodiments of the present disclosure. DETAILED DESCRIPTION

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the accompanying drawings required to be used in the embodiments or prior art description will be briefly introduced. Obviously, for those skilled in the art, other drawings can also be obtained based on these drawings without creative labor.

[0031] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present disclosure. However, the present disclosure can be practiced without the specific details. In other instances, well-known methods, procedures, components, and circuits have not been described in detail so as not to obscure the present disclosure.

[0032] The terms "first", "second", etc. in the description and claims of the present disclosure are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present disclosure can be implemented in an order other than that illustrated or described herein, and the objects distinguished by "first", "second", etc. are generally a category, and do not limit the number of objects, for example, the first object can be one or more. In addition, "and / or" in the specification and claims indicates at least one of the connected objects, and the character " / ", generally indicates that the objects before and after are in a "or" relationship.

[0033] The electronic device in the embodiments of the present disclosure can be a mobile electronic device or a non-mobile electronic device. The mobile electronic device can be a mobile phone, a tablet computer, a notebook computer, a palm computer, a vehicle-mounted electronic device, a wearable device, an ultra-mobile personal computer (UMPC), a netbook, or a personal digital assistant (PDA), etc. The non-mobile electronic device can be a personal computer (PC), a television (TV), a cashier machine, or a self-service machine, etc. The embodiments of the present disclosure are not limited specifically.

[0034] The execution subject of the house source search method provided by the embodiments of the present disclosure can be the electronic device (including the mobile electronic device and the non-mobile electronic device) described above, or the functional module and / or functional entity capable of implementing the house source search method in the electronic device. The specific implementation can be determined according to the actual use demand, and the embodiments of the present disclosure are not limited.

[0035] The house source search method provided by the embodiments of the present disclosure will be described in detail below in combination with the drawings, specific embodiments and application scenarios.

[0036] As shown in FIG. 1, the embodiments of the present disclosure provide a house source search method, which can include the following steps 101 to 103. Figure 1

[0037] 101, in response to a house source search instruction, at least one structured field and at least one unstructured field corresponding to the house source search instruction are acquired.

[0038] ​Among them, the structured field refers to the part of the house information with a clear format and fixed rules, which can usually be clearly represented by the rows and columns of the table or database. The data type of these fields is clear, which is convenient for accurate search, sorting and statistical analysis. Such as house type, house structure, building area, floor, etc.

[0039] The unstructured field refers to the part of the house information with relatively free content form and no fixed format requirement. These information are usually text description, which focuses more on the detailed and personalized introduction of the house, and can provide users with more rich details, but relatively difficult to be accurately and automatically processed. Such as decoration condition, shop equipment condition and house feature highlights (such as good lighting, close to subway, etc.), house description information, and surrounding supporting facilities such as traffic condition and education resources.

[0040] In some embodiments of the present disclosure, the search instruction can carry a search statement, in which case at least one structured field and at least one unstructured field can be extracted from the search instruction.

[0041] In some embodiments of the present disclosure, the search instruction can not carry a search statement.

[0042] In some embodiments of the present disclosure, the above step 101 can be implemented by the following steps 101a to 101b.

[0043] 101a, in response to the house search instruction, obtaining the target search statement input by the user in the search information input interface.

[0044] Among them, the search information input interface is used for inputting the search statement or selecting the structured field and unstructured field included in the search statement, which is not limited here.

[0045] 101b, generating at least one structured field and at least one unstructured field based on a plurality of search keywords in the target search statement.

[0046] In some embodiments of the present disclosure, the at least one structured field includes the structured field in the plurality of search keywords, the target unstructured field in the plurality of search keywords and the unstructured field with the same semantics as the target unstructured field.

[0047] In the embodiment of the present disclosure, in addition to the target structured field in the plurality of search keywords, the at least one unstructured field can also include the unstructured field with the same semantics as the target unstructured field, that is, rewriting and expanding the unstructured field in the plurality of search keywords to obtain at least one unstructured field, which greatly improves the accuracy of house matching and search effect, and more house sources accurately matched with the house search instruction can be searched more quickly.

[0048] In some embodiments of the present disclosure, the unstructured fields in the plurality of search keywords can not be rewritten or expanded.

[0049] In some embodiments of the present disclosure, the step 101 can be implemented by the following steps 101c to 101d.

[0050] 101c, in response to the house source search instruction, extracting a target search statement from the target chat record of the user.

[0051] The target search statement is a house purchase appeal or a house rental appeal.

[0052] 101d, generating at least one structured field and at least one unstructured field based on the plurality of search keywords in the target search statement.

[0053] The at least one structured field includes the structured field in the plurality of search keywords, the target unstructured field in the plurality of search keywords, and the unstructured field with the same semantics as the target unstructured field.

[0054] In the embodiments of the present disclosure, the target search statement is extracted from the chat record, and at least one structured field and at least one unstructured field are generated based on the plurality of search keywords in the target search statement, so that the user can find a house through natural language description, and the interaction flexibility and convenience are improved.

[0055] In some embodiments of the present disclosure, the step 101c can be implemented by the following step 101c1.

[0056] 101c1, in response to the house source search instruction, inputting the target chat record into a target search statement generation model to output the target search statement.

[0057] The target search statement generation model is used for key information extraction on the chat record and generates a search statement based on the extracted key information.

[0058] In the embodiments of the present disclosure, the target search statement generation model can be a large language model (LLM), or a model obtained by optimizing and training based on the LLM model, which is not limited here. The LLM model is a model based on machine learning and natural language processing technology. Through large-scale unsupervised training on massive text data, the pattern and structure of natural language are learned, so as to simulate human language cognition and generation process, and have strong language understanding and generation ability.

[0059] In the embodiments of the present disclosure, the target search statement generation model can quickly extract the target search statement from the target chat record, that is, the target search statement generation model can extract the target search statement from the natural language description, and then generate a vectorized query statement including a structured field and an unstructured field based on the target search statement, which can improve the house search efficiency.

[0060] 102. generating a vectorized target query statement based on the at least one structured field, the at least one unstructured field, and a target similarity algorithm.

[0061] The vectorized target query statement is a query content in natural language form (such as search words or questions input by a user) converted into a vector form through specific technical means, and relevant operations and comparisons are performed in a vector space to achieve more accurate and semantic information retrieval or matching. The vector can capture the semantic information of the query statement to some extent, so that the computer system can find the most relevant house source according to the similarity between vectors.

[0062] The target similarity algorithm is a vector-based similarity algorithm, which can be Euclidean distance or Manhattan distance.

[0063] For example, the target similarity algorithm can be a K-Nearest Neighbors (KNN) algorithm.

[0064] In the embodiments of the present disclosure, the vectorization processing of the unstructured field converts the unstructured description information of the house source into a vector for storage, and improves the intelligent degree of the search system through similarity matching, so that the system can more flexibly process complex house source descriptions and customer demands.

[0065] In some embodiments of the present disclosure, the above step 102 can be implemented by the following step 102a.

[0066] 102a. inputting the at least one structured field and the at least one unstructured field into a query statement generation model, and outputting the target query statement.

[0067] The query statement generation model has preset model hyperparameters, and the similarity algorithm is the target similarity algorithm.

[0068] The preset model hyperparameter is a parameter that needs to be preset before training of the query statement generation model, is not learned through a training process of the model, and is used for controlling aspects such as a structure, a training process, or an optimization strategy of the model. A value of the hyperparameter has a far-reaching influence on performance, training efficiency, and final performance of the model, like some key "experimental conditions" set before the "experiment" of training the model.

[0069] In the embodiments of the present disclosure, the LLM model is used to convert natural language text into a domain specific language (Domain Specific Language, DSL), that is, text2DSL. The LLM can learn the patterns and relationships between natural language and corresponding DSL expressions in a large amount of text, thereby analyzing and understanding the input natural language text, and automatically generating a DSL statement meeting the requirements. The DSL statement is a structured query language (Structured Query Language, SQL).

[0070] In the embodiments of the present disclosure, the query statement generation model can quickly output the target query statement, thereby improving the efficiency of searching for a house source.

[0071] 103. Based on the target similarity algorithm, a plurality of target house sources matching the target query statement are searched from a house source database.

[0072] The house source database includes a plurality of house sources and a vectorized index corresponding to each house source, and the vectorized index includes a structured index and an unstructured index of the house source.

[0073] In the data storage stage, existing house source description information including broker house evaluation, owner self-recommendation, and house source point of interest (Point of Interest, POI) information is aggregated into a table, is chunked according to different dimensions, and is then embedded and stored in the house source database, thereby facilitating subsequent approximate query. The house source database can be a relational database, such as an Elasticsearch (ES) index system.

[0074] In some embodiments of the present disclosure, a hybrid retrieval mode combining text retrieval and semantic retrieval can be used to implement house source recall. The text retrieval can include accurate search, range search, and matching search, without limitation.

[0075] In some embodiments of the present disclosure, the query statement includes a plurality of query conditions, and the hybrid retrieval mode is combined with Figure 1 As shown in FIG. 10, the step 103 can be implemented through the following steps 103a to 103c. Figure 2 As shown in FIG. 10, the step 103 can be implemented through the following steps 103a to 103c.

[0076] 103a. Generate multiple query statements based on the target query statement. The multiple query statements include the target query statement and at least one first query statement. Each of the at least one first query statement has one less query condition than the target query statement.

[0077] If the multiple query conditions include N (N is an integer greater than 1) query conditions, the number of the multiple query statements may be any one of 2, 3, ..., (N+1), which is not limited here.

[0078] 103b. Based on the target similarity algorithm, multiple property resource sets matching the multiple query statements are searched from the property resource database, each property resource set including multiple candidate properties.

[0079] 103c. Rearrange the candidate properties in the plurality of property resource sets to obtain the plurality of target properties.

[0080] For example, for each query condition, the similarity is calculated based on the knn algorithm to retrieve the k most matching properties. The retrieved properties are then reordered and the final result is output.

[0081] In the disclosed embodiment, according to the target query statement, the query conditions are appropriately deleted to form at least one downgraded first query statement, thereby realizing multi-way retrieval based on multiple query statements, and then the multiple search results that can be retrieved are rearranged to obtain multiple target properties, which greatly improves the accuracy of property matching and the search effect, and can more quickly search for more properties that accurately match the property search instructions.

[0082] In some embodiments of the present disclosure, Figure 2 ,like Figure 3 As shown, the above step 103c can be specifically implemented through the following steps 103c1 to 103c3.

[0083] 103c1. Determine similarity scores between each of the candidate properties in the plurality of property resource sets and the plurality of query conditions.

[0084] 103c2. Determine a weighted value of the similarity score between each candidate property and the multiple query conditions.

[0085] The weight of the similarity score of the query condition can be set according to actual conditions and is not limited here.

[0086] 103c3. Determine candidate properties whose weighted similarity scores are greater than or equal to the score threshold as the multiple target properties.

[0087] In the embodiments of the present disclosure, different query conditions are set with corresponding weighting values, and then the candidate house sources are rearranged based on the weighting values of the similarity scores of each candidate house source with the plurality of query conditions, to obtain a plurality of target house sources, thereby greatly improving the accuracy and search effect of house source matching.

[0088] In some embodiments of the present disclosure, the step 103c can be implemented by the following step 103c4.

[0089] 103c4, input the candidate house sources in the plurality of house source sets into the rearrangement model respectively, and output the plurality of target house sources.

[0090] The rearrangement model is used to filter out, from the candidate house sources in the plurality of house source sets, the candidate house sources with a weighting value of the similarity score greater than or equal to a score threshold as the target house sources, and the weighting value of the similarity score is a weighting value of the similarity score of each candidate house source with the plurality of query conditions.

[0091] In the embodiments of the present disclosure, the rearrangement model can be an LLM model, or the rearrangement model is a model combining RAG and LLM capabilities. The rearrangement efficiency and accuracy of the house source can be improved by the rearrangement model.

[0092] The combination of the knowledge retrieval advantage of RAG and the language generation capability of LLM can utilize the accurate knowledge retrieved by RAG to correct the possible knowledge deviation of LLM when answering the user's question, thereby improving the accuracy and reliability of the answer.

[0093] In the embodiments of the present disclosure, the query statement generated by combining the structured field and the unstructured field can quickly search for house sources that meet the user's query requirements for the unstructured field, thereby improving the house source search efficiency.

[0094] In the embodiments of the present disclosure, based on the capabilities of RAG and LLM, the capabilities are constructed in the two stages of house source database construction and house source search, so that more accurate house sources can be matched by using natural language description of house sources, and the similarity matching of POI, school district, house evaluation, and self-recommendation information can be supported.

[0095] Figure 4 The structural block diagram of a house source search device according to an embodiment of the present disclosure is shown in FIG. 1. Figure 4As shown, the method comprises: acquiring, in response to a house search instruction, at least one structured field and at least one unstructured field corresponding to the house search instruction; generating, based on the at least one structured field, the at least one unstructured field, and a target similarity algorithm, a vectorized target query statement; and searching, based on the target similarity algorithm, a plurality of target houses matching the target query statement from a house database, wherein the house database comprises a plurality of houses and a vectorized index corresponding to each house, and the vectorized index comprises a structured index and an unstructured index of the house.

[0096] In some embodiments of the present disclosure, the query statement comprises a plurality of query conditions, and the search module 403 is specifically configured to generate, based on the target query statement, a plurality of query statements, wherein the plurality of query statements comprises the target query statement and at least one first query statement, and each first query statement is less than the target query statement by one query condition; search, based on the target similarity algorithm, a plurality of house sets matching the plurality of query statements from the house database, wherein each house set comprises a plurality of candidate houses; and rearrange the candidate houses in the plurality of house sets to obtain the plurality of target houses.

[0097] In some embodiments of the present disclosure, the search module 403 is specifically configured to determine similarity scores of the candidate houses in the plurality of house sets with respect to the plurality of query conditions; determine a weighting value of the similarity score of each candidate house with respect to the plurality of query conditions; and determine, as the plurality of target houses, the candidate houses whose weighting values of the similarity scores are greater than or equal to a score threshold.

[0098] In some embodiments of the present disclosure, the search module 403 is specifically configured to input the candidate houses in the plurality of house sets into a rearrangement model to output the plurality of target houses, wherein the rearrangement model is configured to filter, from the candidate houses in the plurality of house sets, the candidate houses whose weighting values of the similarity scores are greater than or equal to a score threshold as the target houses, and the weighting value of the similarity score is a weighting value of the similarity score of each candidate house with respect to the plurality of query conditions.

[0099] In some embodiments of the present disclosure, the acquisition module 401 is specifically configured to acquire, in response to the house search instruction, a target search statement input by a user in a search information input interface or extracted from a target chat record of the user; generate, based on a plurality of search keywords in the target search statement, the at least one structured field and the at least one unstructured field; and wherein the at least one structured field comprises a structured field in the plurality of search keywords, a target unstructured field in the plurality of search keywords, and an unstructured field semantically identical to the target unstructured field.

[0100] In some embodiments of the present disclosure, the acquisition module 401 is specifically used to respond to the housing search instruction, input the target chat record into the target search statement generation model, and output the target search statement; wherein, the target search statement generation model is used to extract key information from the chat record and generate a search statement based on the extracted key information.

[0101] In some embodiments of the present disclosure, the generation module 402 is specifically used to input the at least one structured field and the at least one unstructured field into a query statement generation model and output the target query statement; wherein, the query statement generation model has model hyperparameters that are preset model hyperparameters, and the similarity algorithm is the target similarity algorithm.

[0102] In the embodiments of the present disclosure, each module can implement the housing search method provided by the above method embodiments and can achieve the same technical effect. To avoid repetition, it will not be described here.

[0103] Figure 5 This is a structural diagram of an electronic device provided in an embodiment of the present disclosure, which is used to exemplify an electronic device that implements any housing search method in an embodiment of the present disclosure and should not be understood as a specific limitation on the embodiments of the present disclosure.

[0104] like Figure 5 As shown, the electronic device 500 may include a processor (e.g., a central processing unit, a graphics processing unit, etc.) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage device 508 into a random access memory (RAM) 503. Various programs and data required for the operation of the electronic device 500 are also stored in the RAM 503. The processor 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0105] Typically, the following devices may be connected to the I / O interface 505: an input device 506 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 507 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 508 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 509. The communication device 509 may allow the electronic device 500 to communicate with other devices wirelessly or by wire to exchange data. Although the electronic device 500 is shown as having various devices, it should be understood that it is not required to implement or have all of the devices shown. More or fewer devices may be implemented or have alternatively.

[0106] In particular, according to embodiments of the present disclosure, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, embodiments of the present disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for executing the methods illustrated by the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network through the communication apparatus 509, or installed from the storage apparatus 508, or installed from the ROM 502. When the computer program is executed by the processor 501, the functions defined in any of the house searching methods provided by embodiments of the present disclosure can be performed.

[0107] It should be noted that the computer-readable medium described above in the present disclosure can be a computer-readable signal medium or a computer-readable storage medium or any combination thereof. The computer-readable storage medium may, for example, be, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or any suitable combination thereof. More specific examples of the computer-readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus or device. In the present disclosure, the computer-readable signal medium can include a data signal propagated in a baseband or as a carrier wave in a propagated data signal, which bears computer-readable program code. Such propagated data signal can take many forms, including but not limited to, an electromagnetic signal, an optical signal or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable medium that can send, propagate or transfer a program for use by or in connection with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to, a wire, cable, optical fiber, RF (radio frequency), or any suitable combination thereof.

[0108] In some embodiments, the client, server can communicate using any known or future developed network protocols, such as HTTP (HyperText Transfer Protocol), and can be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), the Internet, and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any current known or future developed networks.

[0109] The computer readable medium described above can be included in the electronic device described above; or can exist independently of the electronic device and be not assembled into the electronic device.

[0110] The computer readable medium described above carries one or more programs, when the one or more programs are executed by the electronic device, cause the electronic device to: in response to a house source search instruction, acquire at least one structured field and at least one unstructured field corresponding to the house source search instruction; generate a vectorized target query statement based on the at least one structured field, the at least one unstructured field, and a target similarity algorithm; search for a plurality of target house sources matching the target query statement from a house source database based on the target similarity algorithm, the house source database including a plurality of house sources and a vectorized index corresponding to each house source, the vectorized index including a structured index and an unstructured index of the house source.

[0111] In the embodiments of the present disclosure, the computer program code for performing the operations of the present disclosure can be written in one or more programming languages or combinations of languages including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" programming language or similar programming languages. The program code can be executed entirely on the computer, partially on the computer, as an independent software package, partially on the computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, connected through the Internet using an Internet service provider).

[0112] The computer readable program instructions can also be loaded onto a computer, other programmable data processing apparatus, or computer readable storage devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other computer readable storage devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0113] The units described in the embodiments of the present disclosure can be implemented by software, or by hardware. In some cases, the names of the units do not constitute a limitation on the units themselves.

[0114] The functions described above in the detailed description can be performed by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that can be used include Field-programmable Gate Arrays (FPGAs), Program-specific Integrated Circuits (ASICs), Program-specific Standard Products (ASSPs), System-on-a-chip systems (SOCs), Complex Programmable Logic Devices (CPLDs), etc.

[0115] In the context of the present disclosure, a computer readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. The computer readable medium can be a computer readable signal medium or a computer readable storage medium. The computer readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the computer readable storage medium can include one or more lines of electrical connections, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0116] The above description merely illustrates the preferred embodiments of the disclosure and a principle for applying the technologies. It is understood by those skilled in the art that the disclosed scope of the disclosure is not limited to the technical solutions formed by the specific combinations of the technical features described above, and should also cover other technical solutions formed by the combinations of the technical features described above or their equivalent features without departing from the disclosed concept. For example, the technical solutions formed by the mutual replacement of the above-described features and the technical features with similar functions disclosed in the disclosure (but not limited to) can be formed.

[0117] Further, although operations are depicted in a particular, sequential order, this should not be understood as requiring or implying that the operations are performed in the order illustrated or sequentially. In certain circumstances, multitasking and parallel processing can be advantageous. Likewise, although specific implementation details are included for the purpose of providing a thorough disclosure, these should not be construed as limitations on the scope of the disclosure. Certain features that are described in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable sub-combination.

[0118] Although the subject matter has been described in language specific to structural features and / or methodological acts, it is to be understood that the subject defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

Claims

1. A housing search method, characterized in that: The method comprises: In response to a property search instruction, obtaining at least one structured field and at least one unstructured field corresponding to the property search instruction; generating a vectorized target query statement based on the at least one structured field, the at least one unstructured field, and a target similarity algorithm; Based on the target similarity algorithm, multiple target properties that match the target query statement are searched from a property database, where the property database includes multiple properties and a vectorized index corresponding to each property, and the vectorized index includes a structured index and an unstructured index of the property.

2. The method according to claim 1, characterized in that The query statement includes multiple query conditions, and the target similarity algorithm is used to search a property database for multiple target properties that match the target query statement, including: Based on the target query statement, generating multiple query statements, the multiple query statements including the target query statement and at least one first query statement, each of the at least one first query statement having one less query condition than the target query statement; Based on the target similarity algorithm, searching the property database for multiple property sets that match the multiple query statements, each property set including multiple candidate properties; The candidate properties in the plurality of property sets are rearranged to obtain the plurality of target properties.

3. The method according to claim 2, characterized in that The step of rearranging the candidate properties in the plurality of property sets to obtain the plurality of target properties includes: Determining similarity scores between candidate properties in the plurality of property sets and the plurality of query conditions; Determining weighted values ​​of similarity scores between each candidate property and the plurality of query conditions; Candidate listings whose weighted similarity scores are greater than or equal to the score threshold are determined as the multiple target listings.

4. The method according to claim 2, characterized in that The step of rearranging the candidate properties in the plurality of property sets to obtain the plurality of target properties includes: Inputting candidate properties from the plurality of property sets into a rearrangement model respectively, and outputting the plurality of target properties; In which, the rearrangement model is used to filter out candidate properties from the multiple property sets, and the candidate properties whose weighted similarity scores are greater than or equal to the score threshold are used as the target properties, and the weighted similarity scores are the weighted values ​​of the similarity scores of each candidate property and the multiple query conditions.

5. The method according to claim 1, wherein The step of obtaining, in response to a housing search instruction, at least one structured field and at least one unstructured field corresponding to the housing search instruction includes: In response to the housing search instruction, obtaining a target search statement input by the user in the search information input interface, or extracting the target search statement from the user's target chat history; generating the at least one structured field and the at least one unstructured field based on a plurality of search keywords in the target search statement; The at least one structured field includes structured fields in the multiple search keywords, target unstructured fields in the multiple search keywords, and unstructured fields with the same semantics as the target unstructured fields.

6. The method according to claim 5, characterized in that In response to the housing search instruction, extracting a target search statement from the user's target chat history includes: In response to the housing search instruction, inputting the target chat record into a target search statement generation model and outputting the target search statement; The target search statement generation model is used to extract key information from chat records and generate search statements based on the extracted key information.

7. The method according to claim 1, characterized in that The generating a vectorized target query statement based on the at least one structured field, the at least one unstructured field, and a target similarity algorithm includes: Inputting the at least one structured field and the at least one unstructured field into a query statement generation model, and outputting the target query statement; Among them, the query statement generation model is a model hyperparameter that is a preset model hyperparameter, and the similarity algorithm is the target similarity algorithm.

8. An electronic device, characterized in that: include: A memory and a processor, the memory is used to store a computer program; the processor is used to execute the housing search method according to any one of claims 1 to 7 when calling the computer program.

9. A computer-readable storage medium, characterized in that A computer program is stored thereon, and when the computer program is executed by a processor, the housing search method according to any one of claims 1 to 7 is implemented.

10. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, the housing search method according to any one of claims 1 to 7 is implemented.

Images