A keyword search method, device, and medium for enterprise query functions

Abstract

The present application discloses a keyword search method, device, and medium for enterprise query functions. The method includes: obtaining keywords of the enterprise name to be queried; performing word segmentation on the keywords to obtain multiple query word segments; in a pre-constructed index document, respectively matching the multiple query word segments with index word segments to obtain multiple source table primary key values corresponding to the keywords and the relevance of the source table primary key values; the index document includes each source table primary key value and the index word segments corresponding to each source table primary key value; retrieving multiple rows of target data in the source table database according to each source table primary key value; generating display information of the target enterprise in each row of target data according to the set display fields; determining the display order of each target enterprise according to the relevance of each source table primary key value; and displaying the information to be displayed of each target enterprise on the client according to the display order of each target enterprise. The query efficiency and accuracy are improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a keyword search method, device and medium for an enterprise query function. Background Art

[0002] At present, keyword retrieval technology is based on precise query matching of keywords. It uses query statements to substitute the keywords entered by users, traverses the company name field in the database, and returns the results containing this keyword in the target field. It is displayed to users through an interactive interface. The results returned by the above method will inevitably include the keywords entered by users. Therefore, ideally, the query results are accurate.

[0003] However, in actual use, there is a great deal of uncertainty in the content input by users, that is, the user's input may contain homophones, be out of order, or contain less information (such as only remembering the iconic words contained in the name). At this time, it is necessary to directly traverse all the company names in the database with inaccurate keywords, resulting in low query efficiency and difficulty in outputting accurate results. Summary of the invention

[0004] The embodiments of the present application provide a keyword search method, device and medium for an enterprise query function, which are used to solve the problems of low query efficiency and difficulty in outputting accurate results.

[0005] The present application embodiment adopts the following technical solutions:

[0006] On the one hand, an embodiment of the present application provides a keyword search method for an enterprise query function, the method comprising: obtaining keywords of the enterprise name to be queried; segmenting the keywords to obtain multiple query segmentations; in a pre-constructed index document, matching the multiple query segmentations with the index segmentations respectively to obtain multiple source table primary key values ​​corresponding to the keywords and the relevance of the source table primary key values; the index document includes each source table primary key value and the index segmentation corresponding to each source table primary key value; searching in the source table database according to each source table primary key value to obtain multiple rows of target data; generating display information of the target enterprise in each row of target data according to the set display field; determining the display order of each target enterprise according to the relevance of each source table primary key value; the display order is in reverse order; and displaying the information to be displayed of each target enterprise on the client according to the display order of each target enterprise.

[0007] In one example, before obtaining the keywords of the company name to be queried, the method also includes: extracting all company names from the constructed source database; segmenting all the company names separately to obtain the index document of the source database; and saving the index document to each distributed index node of the distributed search.

[0008] In one example, in the pre - constructed index document, matching the multiple query word segments with the index word segments respectively to obtain multiple source table primary key values corresponding to the keyword and the relevance of the source table primary key values specifically includes: allocating the query tasks of the multiple query word segments to each distributed index node to generate sub - query tasks for each distributed index node, so that each distributed index node matches the query word segments in its respective sub - query task with the index word segments in the index document to obtain the query results of its respective sub - query task; the query results include multiple matching source table primary key values and the matching index word segments for each matching source table primary key value; aggregating the query results of each sub - query task to obtain multiple source table primary key values corresponding to the keyword and the number of target index word segments for each source table primary key value; determining the relevance of each source table primary key value according to the number of target index word segments for each source table primary key value.

[0009] In one example, before segmenting the enterprise names of all the source table primary key values to obtain multiple index documents of the source database, the method further includes: judging whether there is a first enterprise name with missing values, and if so, filtering the first enterprise name; judging whether there is a second enterprise name with duplicates, and if so, filtering the second enterprise name; judging whether there is a third enterprise name with outliers, and if so, filtering the third enterprise name.

[0010] In one example, determining the display order of each target enterprise according to the relevance of each source table primary key value specifically includes: obtaining the geographical location of the IP to which the client belongs; calculating the relative geographical location between the client and each target enterprise; compensating the relevance of each target enterprise according to the relative geographical location of each target enterprise to obtain the location - compensated relevance of each target enterprise; the closer the relative geographical location, the higher the location compensation coefficient, and the location compensation coefficient is greater than 1; determining the display order of each target enterprise according to the location - compensated relevance of each target enterprise.

[0011] In one example, determining the display order of each target enterprise according to the location - compensated relevance of each target enterprise specifically includes: determining multiple target enterprise sets in the same organizational structure; determining the hierarchical compensation coefficient of each target enterprise according to the level of each target enterprise in the target enterprise set; the higher the level, the higher the hierarchical compensation coefficient, and the hierarchical compensation coefficient is greater than 1; compensating the relevance of each target enterprise according to the hierarchical compensation coefficient of each target enterprise to obtain the hierarchical - compensated relevance; determining the display order of each target enterprise according to the hierarchical - compensated relevance and the location - compensated relevance.

[0012] In one example, determining the display order of each target enterprise based on the level compensation correlation and the position compensation correlation specifically includes: retrieving the level compensation weight and the position compensation weight of the target enterprise in the constructed mapping relationship table; taking weighted sum of the level compensation correlation and the position compensation correlation according to the level compensation weight and the position compensation weight to obtain a final correlation of each target enterprise; and determining the display order of each target enterprise according to the final correlation of each target enterprise.

[0013] In one example, the information to be displayed of each target enterprise is displayed on the client according to the display order of each target enterprise, specifically including: obtaining the click-view information of the user on the client; if the click-view information includes viewing the detailed information page of the target enterprise and exiting the query, determining that the display information meets the query needs of the user; if the click-view information includes multiple page turns and exiting the query, determining that the display information does not meet the query needs of the user.

[0014] On the other hand, an embodiment of the present application provides a keyword search device for an enterprise query function, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute any one of the above-described keyword search methods for an enterprise query function.

[0015] On the other hand, an embodiment of the present application provides a non-volatile computer storage medium for keyword search for an enterprise query function, which stores computer executable instructions, and the computer executable instructions can execute any of the above-mentioned keyword search methods for an enterprise query function.

[0016] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects:

[0017] With keyword segmentation as a prerequisite, it is possible to establish an index document of the company name, and reflect the content of the company name and the position of the company name in the source database through the source table primary key value and the index segmentation, so as to match the keyword with the index segmentation, and obtain the source table primary key value of the keyword and the correlation of the source table primary key value, and then directly and quickly locate the row target data of the source database based on the source table primary key value, that is, the company information that the user may want to query, and display the highly relevant ones more prominently in the front end, thereby improving the query efficiency and query accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] To more clearly illustrate the technical solutions of this application, the following will, with reference to the accompanying drawings, elaborate on some embodiments of this application. In the drawings:

[0019] Figure 1 It is a schematic flowchart of a keyword search method for an enterprise query function provided by an embodiment of this application;

[0020] Figure 2 It is a schematic structural diagram of a keyword search device for an enterprise query function provided by an embodiment of this application. Detailed implementation manners

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the following will, in conjunction with specific embodiments and corresponding accompanying drawings, clearly and completely describe the technical solutions of this application. Obviously, the described embodiments are only some embodiments of this application, rather than all embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this application.

[0022] The following will refer to the accompanying drawings to elaborate on some embodiments of this application in detail.

[0023] Figure 1 It is a schematic flowchart of a keyword search method for an enterprise query function provided by an embodiment of this application. This method can be applied to different business fields, such as the Internet finance business field, the e-commerce business field, the instant messaging business field, the game business field, the official business field, etc. Some input parameters or intermediate results in this process allow manual intervention and adjustment to help improve accuracy.

[0024] The implementation of the analysis method involved in the embodiments of this application can be a terminal device or a server, and this application does not make special restrictions on this. For the convenience of understanding and description, the following embodiments will be described in detail taking the server as an example.

[0025] It should be noted that this server can be a single device or a system composed of multiple devices, that is, a distributed server, and this application does not make specific limitations on this.

[0026] Figure 1 The process in

[0027] S101: Obtain the keyword of the enterprise name to be queried.

[0028] Among them, the user inputs keywords of the enterprise name to be queried in the client interface. It should be noted that the user needs to be guided to make the input content more readable. For example, a prompt word for entering the enterprise name is displayed, and when the information is ambiguous, the user is guided to separate the input keywords with specific delimiters, etc.

[0029] In some embodiments of the present application, a source database needs to be constructed in advance.

[0030] First, construct a table template for the source table. The fields of the table template include labels such as enterprise name, enterprise type, establishment time, business status, registered capital, unified social credit code, and legal representative. It should be noted that there may be multiple source tables.

[0031] Then, determine the unique primary key of the source table. For example, the unique primary keys of source table 1 and source table 2 are the same. Source table 2 and source table 1 can be considered as related tables. For example, the unique primary key is the enterprise name. The fields of source table 1 include labels such as enterprise name, enterprise type, and establishment time, and the fields of source table 2 include labels such as enterprise name, business status, and legal representative.

[0032] Finally, fill the obtained source enterprise information into the corresponding table template to generate the source table of the source database.

[0033] S102: Segment the keywords to obtain multiple query segments.

[0034] Among them, ik_max_word can be used for the finest-grained segmentation. Different from ik_smart, the segmentation result of ik_max_word is more detailed, which is more effective for short-word matching.

[0035] For example, if the keyword is ABCD, the query segments after segmentation are AB and CD.

[0036] S103: In the pre-constructed index document, match the multiple query segments with the index segments respectively to obtain the multiple source table primary key values corresponding to the keywords and the relevance of the source table primary key values; the index document includes each source table primary key value and the index segments corresponding to each source table primary key value.

[0037] In some embodiments of the present application, an index document needs to be constructed. The purpose of establishing the index is to improve the query efficiency and shorten the query time.

[0038] Specifically, first, in the constructed source database, extract all enterprise names. Then, segment all enterprise names respectively to obtain the index document of the source database. Then, save the index document to each distributed index node of the distributed search.

[0039] It should be noted that after extracting all enterprise names, data governance needs to be carried out on the enterprise names. The purpose is to distinguish available data from invalid data. Invalid data refers to unavailable data such as missing values (null values), duplicates, and outliers (such as disordered numbers, letters, symbols, etc. in the enterprise name). Generally, these data will be processed, transformed, or logically deleted.

[0040] Specifically, judge whether there is a missing value in the first enterprise name. If so, filter the first enterprise name. Judge whether there is a duplicate in the second enterprise name. If so, filter the second enterprise name. Judge whether there is an outlier in the third enterprise name. If so, filter the third enterprise name.

[0041] Among them, during the process of word segmentation for all enterprise names, the distributed search engine Elasticsearch (abbreviated as ES) is introduced. The generated index documents are inserted into the specified index nodes using the index API. This process will parse the documents into the form of an inverted index. Among them, ik_max_word can be used for the finest-grained division.

[0042] It should be noted that the table structure fields in the index document include the source table primary key value and the index word segmentation. For example, if the source table primary key value is xx, and it includes three enterprise names ABCD, EFGH, and ABJK, the index word segmentation of this source table primary key value includes AB, CD, EF, GH, and JK.

[0043] In some embodiments of the present application, the process of matching multiple query word segmentations with the index word segmentations is as follows:

[0044] First, distribute the query tasks of multiple query word segmentations to each distributed index node to generate sub-query tasks for each distributed index node, so that each distributed index node can match the query word segmentations in its respective sub-query tasks with the index word segmentations in the index document to obtain the query results of its respective sub-query tasks. Among them, the query results include multiple matching source table primary key values and the matching index word segmentations for each matching source table primary key value.

[0045] It should be noted that during the process of index word segmentation matching, the Match query method can be used. It has multiple parameters that can be adjusted, such as the query term, word segmenter, operator, and fuzziness, etc. By setting these parameters, the final query result can be affected, but these parameters are not fixed and need to be adjusted repeatedly according to the data situation of the query samples.

[0046] Then, aggregate the query results of each sub-query task to obtain multiple source table primary key values corresponding to the keywords, and the number of target index word segmentations for each source table primary key value.

[0047] Finally, determine the relevance of each source table primary key value based on the number of target index terms segmented for each source table primary key value.

[0048] For example, assign the AB query task to index node a and the CD query task to index node b. Index node a matches AB with the index terms in the index document, and index node b matches CD with the index terms in the index document. As a result, the matching source table primary key values for index node a are xx and yy, with the matching index terms being AB, and the matching source table primary key value for index node b is xx, with the matching index term being CD. Then, aggregate the query results of index node a and index node b to obtain the source table primary key values xx and yy, and the index terms AB and CD corresponding to the source table primary key value xx, so the number of target index terms is 2, and the index term corresponding to the source table primary key value yy is AB, so the number of target index terms is 1.

[0049] It should be noted that when determining the relevance of each source table primary key value based on the number of target index terms segmented for each source table primary key value, for example, the number of target index terms can be retrieved in the quantity mapping relation table to obtain the corresponding relevance.

[0050] Obviously, the position of the query terms in the source database can be located through the source table primary key value, which can achieve the effect of traversing and retrieving the source database. The process of locating the source table primary key value is relatively more efficient than traversing the query.

[0051] In addition, the relevance of each source table primary key value can also be determined by combining the number of target index terms, the position of the target index terms in the index document, and the number of occurrences of the target index terms in the entire document.

[0052] For example, Elasticsearch calculates the relevance of each term based on the TF-IDF (Term Frequency-Inverse Document Frequency) algorithm. TF-IDF is a classic information retrieval algorithm that takes into account two factors: term frequency and inverse document frequency. Term frequency represents the number of times a term appears in the index document, and inverse document frequency represents the prevalence of a term in the entire index document collection. By comprehensively considering these two factors, the TF-IDF algorithm can calculate the importance of each term for the search condition.

[0053] S104: Retrieve multiple rows of target data in the source table database according to each source table primary key value.

[0054] S105: Generate the information to be displayed for the target enterprise in each row of target data according to the set display fields.

[0055] S106: Determine the display order of each target enterprise according to the relevance of each source table primary key value, and the display order is in reverse order.

[0056] In some embodiments of the present application, tuning parameters can be set with the aim of optimizing the query results to make them more in line with the user's expectations, rather than returning a large amount of unordered relevant data to the user and having the user select the values that meet the expectations from the query results. The entire tuning process is to use the program to select more accurate values from these query results on behalf of the user.

[0057] There can be many tuning parameters, which need to be set by software designers according to actual business needs. It should be noted that the tuning result is not to clear the non-conforming data, but to conduct a general macro-control on the entire query method. By setting the influence weights of these parameter fields, it can ultimately be reflected in the relevance.

[0058] For example, regarding the business status of an enterprise. During the query process, the business preferably shows the enterprises with normal operations to the user. Therefore, for abnormal operations such as cancellation and revocation, the display priority will be lower.

[0059] Regarding the enterprise type. Due to data quality issues, for some types of entities such as individual industrial and commercial households, the data is relatively less. Therefore, the business preferably gives a lower display priority to individual industrial and commercial households in the query results; on the contrary, for enterprises such as limited liability companies and joint stock limited companies, the display order is more forward.

[0060] Regarding the number of external investments and the number of branches. When the number of query keywords given by the user is small, the user's perception of the full name of the enterprise is relatively vague. Therefore, from the perspective of function design, the parent company and the main company of the group in these query results will be preferentially shown to the user. Using the number of external investments (i.e., the number of holding subsidiaries) and the number of branches as a parameter to affect the query results, so as to achieve as much as possible that in the query results, for enterprises under the same organizational structure, the higher-level enterprises are shown in a more forward position.

[0061] Regarding the region. Before the query, the system will retrieve the region where the client IP is located and pass it as a parameter to the background. The region attribute will be a factor affecting the query results. For example, when searching for "xx Technology" in Shanghai, the relative position of Shanghai xx Technology is more forward than that of Guangxi xx Technology.

[0062] For relevant fields. When querying the enterprise name, the enterprise information mastered by the user may be relatively fragmented. For example, it is known that the enterprise name contains "Tianyuan", and it is known that the legal person of this enterprise is "Zhang San", as well as the former names of this enterprise. Therefore, similar queries and matches will be performed on other different fields according to the results of word segmentation. Of course, the impact of these fields on the overall relevance is lower than that of the enterprise name field.

[0063] Based on this, the process of determining the display order of each target enterprise according to the relevance of each source table primary key value is as follows:

[0064] On the one hand, first, obtain the geographical location of the IP address of the client. Then, calculate the relative geographical location between the client and each target enterprise. Then, compensate the relevance of each target enterprise according to its relative geographical location to obtain the location-compensated relevance of each target enterprise. Among them, the closer the relative geographical location, the higher the location compensation coefficient, and the location compensation coefficient is greater than 1. Finally, determine the display order of each target enterprise according to the location-compensated relevance of each target enterprise. For example, in the location compensation mapping relationship table, retrieve the relative geographical location to obtain the location compensation coefficient value, and multiply the location compensation coefficient value by the relevance to obtain the location-compensated relevance.

[0065] On the other hand, determine multiple target enterprise sets in the same organizational structure. Then, determine the hierarchical compensation coefficient of each target enterprise according to the level of each target enterprise in the target enterprise set. The higher the level, the higher the hierarchical compensation coefficient, and the hierarchical compensation coefficient is greater than 1. Then, compensate the relevance of each target enterprise according to the hierarchical compensation coefficient of each target enterprise to obtain the hierarchical-compensated relevance. Finally, determine the display order of each target enterprise according to the hierarchical-compensated relevance and the location-compensated relevance.

[0066] Among them, in the process of determining the display order of each target enterprise according to the hierarchical-compensated relevance and the location-compensated relevance, the hierarchical compensation weight and the location compensation weight of the target enterprise can be retrieved in the constructed mapping relationship table. Then, according to the hierarchical compensation weight and the location compensation weight, perform a weighted sum of the hierarchical-compensated relevance and the location-compensated relevance to obtain the final relevance of each target enterprise. Finally, determine the display order of each target enterprise according to the final relevance of each target enterprise.

[0067] S107: According to the display order of each target enterprise, display the information to be displayed of each target enterprise on the client.

[0068] That is, after the entire query function receives the keywords submitted by the user and executes the above analysis and query logic, it will sort the results in reverse order according to the relevance of each result, and output and display them to the front end through the API interface.

[0069] In some embodiments of the present application, the recording and application of behavior data are reflected at the business function design level. When designing the query function of the application system, the actions and behaviors of user queries will be recorded, and corresponding feedback mechanisms and analysis methods will be designed. At the business function level, the success of this query will be judged based on the user's behavior, and through the analysis model, a basis will be provided for the adjustment of various parameters in the query logic. For example, if the user enters the detailed information page to view after a query and then exits the query function, it can be considered that this query is successful. Or if the user makes multiple page turns after the query and finally exits the query function, it can be considered that this query is failed.

[0070] Based on this, obtain the click and view information of the user on the client. Then, if the click and view information includes viewing the detailed information page of the target enterprise and exiting the query, it is determined that the displayed information meets the user's query requirements. If the click and view information includes multiple page turns and exiting the query, it is determined that the displayed information does not meet the user's query requirements.

[0071] In summary, use an inverted index to establish the association between the index tokenized content and the source table, so as to realize finding the primary key value of the source table and the target enterprise content in the row data corresponding to the primary key value of the source table through the index tokenized content, as well as the position of the target enterprise content in the source table. In addition, establish a matching model, perform word segmentation on the input keywords, match them with the tokens in the index, and based on the TF-IDF algorithm, find the results matching the keywords. In addition, model tuning. By introducing business parameters and setting the influence weights of the fields mapped by the business parameters on the relevance, the query results can be made more in line with user expectations. In addition, result output. The query results are output in reverse order according to the relevance, and those with higher relevance are more prominent in the front-end display. In addition, the ES distributed search engine is applied. ES usually uses an inverted index, which can greatly improve the query efficiency. In addition, the matching model is mainly a basic matching model built based on the TF-IDF algorithm. The basic matching model defines the core logic of the query. The key point of model adjustment lies in the business parameters mentioned. The basic model only defines the query idea, and through the compensation adjustment based on the business parameters, the query is made more in line with the application scenario.

[0072] It should be noted that although the embodiments of the present application are introduced and described in sequence for steps S101 to S107 with reference to Figure 1 this does not mean that steps S101 to S107 must be executed in a strict order. The reason why the embodiments of the present application are in accordance with Figure 1The order shown in the figure introduces and explains steps S101 to S107 in sequence, which is to facilitate those skilled in the art to understand the technical solution of the embodiment of the present application. In other words, in the embodiment of the present application, the sequence between steps S101 to S107 can be appropriately adjusted according to actual needs.

[0073] By Figure 1 the method, on the premise of segmenting keywords, can establish an index document of enterprise names, and reflect the content of the enterprise name and the location of the enterprise name in the source database through the source table primary key value and index segmentation. Thus, the keywords can be matched with the index segmentation to obtain the source table primary key value of the keywords and the relevance of the source table primary key value, and then directly and quickly locate the row target data in the source database based on the source table primary key value, that is, the enterprise information that the user may want to query, and display the ones with higher relevance more prominently in the front-end display, improving the query efficiency and query accuracy.

[0074] Furthermore, it improves the query accuracy under non-ideal conditions, and has a higher query error tolerance rate. In the case where the user input content is not standardized (such as disordered, less information, etc.), it can also return some results that may meet the user's expectations, thereby enhancing the user experience; while traditional search methods usually return null or irrelevant data in such cases.

[0075] Furthermore, it improves the query efficiency. By applying the ES distributed search engine, the query request will be sliced and the query tasks will be distributed to multiple query points. After the query is completed, the results will be aggregated and output. In terms of query efficiency, compared with the traditional query method, it is improved by dozens of times.

[0076] Furthermore, the application scope is wider. In the scenario of a large amount of data, keyword queries can also be realized, and the query efficiency is relatively high, and the query response time can be controlled within 3 seconds.

[0077] Thus, it solves the problems of poor user experience such as the retrieval result being empty and there being no user's expected value in the retrieval result in the use of traditional retrieval methods under non-ideal conditions.

[0078] Among them, non-ideal conditions may include homophones, disordered order, less input information, or the keywords containing symbols such as spaces and commas due to input habits, and the keywords containing multiple types of information (such as the user only remembers that the enterprise name contains "Tianyuan", but knows that the legal person's name is "Zhang San") and other special situations. This method can solve the problem that in the above situations, corresponding candidate units can still be provided for the user, and the unit name that the user wants to see can be matched as much as possible. Thus, in the enterprise retrieval function, when the user input content is not standardized, relatively more accurate and more in line with the user's query target query results can also be provided.

[0079] Based on the same idea, some embodiments of the present application also provide the corresponding devices and non-volatile computer storage media for the above method.

[0080] Figure 2 The following is a schematic structural diagram of a keyword search device for enterprise query functions provided by an embodiment of the present application, including:

[0081] At least one processor; and,

[0082] A memory communicatively connected to the at least one processor; wherein,

[0083] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a keyword search method for enterprise query functions described in any one of the above.

[0084] A non-volatile computer storage medium for keyword search for enterprise query functions provided by some embodiments of the present application stores computer-executable instructions, and the computer-executable instructions can execute a keyword search method for enterprise query functions described in any one of the above.

[0085] Each embodiment in the present application is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other, and the key points of each embodiment are the differences from other embodiments. In particular, for the device and medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0086] The devices and media provided by the embodiments of the present application correspond one-to-one with the methods. Therefore, the devices and media also have beneficial technical effects similar to those of the corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be elaborated here.

[0087] Those skilled in the art should understand that the embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program code.

[0088] The present invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each flow and / or block in the flowchart illustrations and / or block diagrams, and combinations of flows and / or blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to the processors of a general purpose computer, special purpose computer, embedded processor or other programmable data processing device to produce a machine, such that the instructions executed by the processors of the computer or other programmable data processing device create means for implementing the functions specified in the flow Figure 1 one or more flows and / or blocks Figure 1 one or more blocks.

[0089] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means that implement the functions specified in the flow Figure 1 one or more flows and / or blocks Figure 1 one or more blocks.

[0090] These computer program instructions may also be loaded onto a computer or other programmable data processing device, such that a series of operational steps are performed on the computer or other programmable device to produce a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in the flow Figure 1 one or more flows and / or blocks Figure 1 one or more blocks.

[0091] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0092] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM) and / or non-volatile memory such as read only memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.

[0093] A computer-readable medium includes both permanent and non-permanent, removable and non-removable media and can implement information storage by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transitory medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer-readable medium does not include transitory computer-readable media such as modulated data signals and carrier waves.

[0094] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or apparatus comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or apparatus. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or apparatus comprising the element.

[0095] The above description is only for the embodiments of the present application and is not intended to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the technical principles of the present application shall fall within the protection scope of the present application.

Claims

1. A keyword search method for enterprise query function, characterized in that: The method comprises: Obtain keywords for the company name to be queried; Segmenting the keywords to obtain multiple query segmentations; In the pre-built index document, the multiple query participles are matched with the index participles respectively to obtain the multiple source table primary key values ​​corresponding to the keyword and the relevance of the source table primary key values; the index document includes each source table primary key value and the index participle corresponding to each source table primary key value; Search the source table database according to each source table primary key value to obtain multiple rows of target data; Generate display information of the target enterprise in each row of target data according to the set display fields; Determine the display order of each target enterprise according to the relevance of each source table primary key value; the display order is in descending order; According to the display order of each target enterprise, the information to be displayed of each target enterprise is displayed on the client; Determining the display order of each target enterprise according to the relevance of each source table primary key value specifically includes: Get the geographical location of the client's IP; Calculating the relative geographic location between the client and each target enterprise; Compensating the respective correlations of each target enterprise according to the relative geographical location of each target enterprise to obtain the location compensation correlation of each target enterprise; the closer the relative geographical location, the higher the location compensation coefficient, and the location compensation coefficient is greater than 1; Identify multiple target enterprise groups in the same organizational structure; Determine the level compensation coefficient of each target enterprise according to the level of each target enterprise in the target enterprise set; the higher the level, the higher the level compensation coefficient, and the level compensation coefficient is greater than 1; According to the hierarchical compensation coefficient of each target enterprise, the relevance of each target enterprise is compensated to obtain the hierarchical compensation relevance; The display order of each target enterprise is determined according to the level compensation relevance and the position compensation relevance.

2. The method according to claim 1, characterized in that Before obtaining the keywords of the company name to be queried, the method further includes: In the constructed source database, all enterprise names are extracted; Segmenting all the company names respectively to obtain index documents of the source database; The index document is saved to each distributed index node of the distributed search.

3. The method according to claim 2, characterized in that The step of matching the multiple query words with the index words in the pre-built index document to obtain the multiple source table primary key values ​​corresponding to the keywords and the correlation between the source table primary key values ​​specifically includes: Assigning query tasks of the multiple query participles to each distributed index node, generating sub-query tasks for each distributed index node, so that each distributed index node matches the query participles in the respective sub-query tasks with the index participles in the index document, and obtaining query results of the respective sub-query tasks; the query results include multiple matching source table primary key values ​​and matching index participles for each matching source table primary key value; Aggregate the query results of each subquery task to obtain multiple source table primary key values ​​corresponding to the keyword and the number of target index word segments for each source table primary key value; The relevance of each source table primary key value is determined according to the target index word quantity of each source table primary key value.

4. The method according to claim 2, characterized in that: Before segmenting the enterprise names of all source table primary key values ​​to obtain multiple index documents of the source database, the method further includes: Determine whether there is a first enterprise name with missing values, and if so, filter the first enterprise name; Determine whether there is a duplicate second enterprise name, and if so, filter the second enterprise name; Determine whether there is an abnormal third enterprise name, and if so, filter the third enterprise name.

5. The method according to claim 1, characterized in that The determining the display order of each target enterprise according to the level compensation relevance and the position compensation relevance specifically includes: In the constructed mapping relationship table, retrieve the hierarchical compensation weight and position compensation weight of the target enterprise; According to the level compensation weight and the position compensation weight, weighted summing the level compensation relevance and the position compensation relevance is performed to obtain a final relevance of each target enterprise; The display order of each target enterprise is determined according to the final relevance of each target enterprise.

6. The method according to claim 1, characterized in that The displaying of the information to be displayed of each target enterprise on the client according to the display order of each target enterprise specifically includes: Obtaining click-view information of the user on the client; If the click-to-view information includes viewing the target enterprise's detailed information page and exiting the query, it is determined that the displayed information meets the user's query needs; If the click to view information includes multiple page turning and exiting the query, it is determined that the displayed information does not meet the user's query needs.

7. A keyword search device for enterprise query function, characterized in that: include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the keyword search method for the enterprise query function as described in any one of claims 1-6 above.

8. A non-volatile computer storage medium for keyword search of enterprise query functions, storing computer executable instructions, characterized in that: The computer executable instructions can execute a keyword search method for an enterprise query function as described in any one of claims 1 to 6 above.

Images