Entity archive query method based on PC terminal and mobile terminal integrated management device

Abstract

The application relates to the technical field of file recommendation, in particular to an entity file query method based on PC terminal and mobile terminal integrated management devices, which comprises the following steps: obtaining a current search term, and screening out to-be-queried entity files containing all the current search terms from all to-be-queried entity files; determining an initial recommendation index corresponding to each candidate file according to a historical calling condition of each candidate file; determining a target matching degree corresponding to each candidate file based on a matching condition between the current search term and feature words in each candidate file and an important degree of the feature words for the candidate file; thus determining a target recommendation index corresponding to each candidate file; and carrying out entity file recommendation page display based on the target recommendation indexes corresponding to all the candidate files, so as to realize entity file query. The application realizes entity file recommendation and entity file query, and improves the rationality of entity file recommendation and the efficiency of entity file query.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of archive recommendation, and particularly relates to an entity archive query method based on PC terminal and mobile terminal integrated management device. BACKGROUND

[0002] In the process of entity archive query, entity archive recommendation is often made for staff so that the staff can quickly query the required entity archive. At present, when making archive recommendation, the method usually adopted is: making archive recommendation based on the similarity between the search terms input by the staff and the archives.

[0003] However, when making entity archive recommendation based on the similarity between the search terms input by the staff and each entity archive in the archive storage room, the following technical problems often exist:

[0004] When making entity archive recommendation, if only the similarity between the search terms and the entity archive is considered, the rationality of the entity archive recommendation may be poor due to the single consideration factor, thereby leading to poor efficiency of entity archive query. SUMMARY

[0005] In order to solve the technical problem of poor efficiency of entity archive query due to poor rationality of entity archive recommendation, the present application provides an entity archive query method based on PC terminal and mobile terminal integrated management device.

[0006] In a first aspect, the present application provides an entity archive query method based on PC terminal and mobile terminal integrated management device, which comprises:

[0007] obtaining a current search term and screening a to-be-queried entity archive containing all the current search terms from all to-be-queried entity archives as a candidate archive;

[0008] determining an initial recommendation index corresponding to each candidate archive according to the historical calling condition of each candidate archive;

[0009] obtaining a feature word in each candidate archive and determining a target matching degree corresponding to each candidate archive based on the matching condition between the current search term and the feature word in each candidate archive and the importance of the feature word to the candidate archive;

[0010] determining a target recommendation index corresponding to each candidate archive according to the initial recommendation index and the target matching degree corresponding to each candidate archive;

[0011] displaying an entity archive recommendation page based on the target recommendation index corresponding to all candidate archives to realize entity archive query.

[0012] In a possible implementation manner of the first aspect, the initial recommendation index corresponding to each candidate archive is determined according to historical calling conditions of each candidate archive, and the initial recommendation index corresponding to each candidate archive is determined according to the following steps.

[0013] An arbitrary candidate archive is determined as a marker archive, and a total number of historical calls of the marker archive is determined as a total number of historical calls corresponding to the marker archive;

[0014] A time of each historical call of the marker archive is determined as a historical calling time of the marker archive, and a sequence of historical calling times corresponding to the marker archive is obtained;

[0015] The initial recommendation index corresponding to the marker archive is determined according to the total number of historical calls and the sequence of historical calling times corresponding to the marker archive.

[0016] In a possible implementation manner of the first aspect, the initial recommendation index corresponding to the marker archive is determined according to the total number of historical calls and the sequence of historical calling times corresponding to the marker archive, and the initial recommendation index corresponding to the marker archive is determined according to the following steps.

[0017] A time length between each historical calling time in the sequence of historical calling times corresponding to the marker archive and a current time is determined as a target time length, and a sequence of target time lengths corresponding to the marker archive is obtained;

[0018] The initial recommendation index corresponding to the marker archive is determined according to the total number of historical calls and the sequence of target time lengths corresponding to the marker archive, wherein the total number of historical calls is positively correlated with the initial recommendation index, and a target time length in the sequence of target time lengths is negatively correlated with the initial recommendation index.

[0019] In a possible implementation manner of the first aspect, the feature word in each candidate archive is obtained according to the following steps.

[0020] Each candidate archive is subjected to word segmentation processing by using a JieBa word segmentation tool, and a target word segmentation is obtained;

[0021] If the target word segmentation is not a stop word, the target word segmentation is determined as a feature word.

[0022] In a possible implementation manner of the first aspect, the target matching degree corresponding to each candidate archive is determined according to a matching condition between the current search word and the feature word in each candidate archive and an importance degree of the feature word for the candidate archive, and the target matching degree corresponding to each candidate archive is determined according to the following steps.

[0023] A word vector corresponding to each current search word and a word vector corresponding to each feature word are obtained by using a World2Vec model;

[0024] According to the TF-IDF value corresponding to each feature word in each candidate profile, and the matching between the word vector corresponding to all current search words and the word vector corresponding to all feature words in each candidate profile, the target matching degree corresponding to each candidate profile is determined, wherein the TF-IDF value corresponding to the feature word represents the importance of the feature word to the candidate profile.

[0025] In combination with the first aspect, in a possible implementation, the target matching degree corresponding to each candidate profile is determined according to the TF-IDF value corresponding to each feature word in each candidate profile, and the matching between the word vector corresponding to all current search words and the word vector corresponding to all feature words in each candidate profile, including:

[0026] The cosine similarity between the word vector corresponding to each current search word and the word vector corresponding to each feature word is normalized to determine the word reference similarity between each current search word and each feature word;

[0027] According to the word reference similarity between all current search words and all feature words in each candidate profile, and the TF-IDF value corresponding to all feature words in each candidate profile, the target matching degree corresponding to each candidate profile is determined, wherein the word reference similarity and the TF-IDF value are positively correlated with the target matching degree.

[0028] In combination with the first aspect, in a possible implementation, the target recommendation index corresponding to each candidate profile is determined according to the initial recommendation index corresponding to each candidate profile and the target matching degree, including:

[0029] The product of the initial recommendation index corresponding to each candidate profile and the target matching degree is determined as the target recommendation index corresponding to each candidate profile.

[0030] In combination with the first aspect, in a possible implementation, the entity profile recommendation page display based on the target recommendation index corresponding to all candidate profiles, including:

[0031] Based on the target recommendation index corresponding to all candidate profiles, all candidate profiles are sorted in descending order to obtain a candidate profile sequence.

[0032] The candidate profile sequence is used to form an entity profile recommendation page for display to the staff.

[0033] In combination with the first aspect, in a possible implementation, the method further includes:

[0034] Based on the displayed entity file recommendation page, staff members select the required file from all candidate files as the target file, and identify the entity files to be queried other than the target file as reference files.

[0035] The time when the target file is accessed in history is determined as the target access time, and the union of the preset time periods corresponding to all target access times is determined as the target time range, wherein the target access time is the time at the center of its corresponding preset time period;

[0036] Reference files that have been accessed within the target time range will be identified as candidate files;

[0037] Based on the access history of the target file and each candidate file within the target time range, determine the degree of target association between the target file and each candidate file;

[0038] From all candidate files, select a predetermined number of candidate files that have the highest degree of target relevance with the target file, and use them as the target related files;

[0039] Recommend related files based on all target-related files.

[0040] In conjunction with the first aspect above, in one possible implementation, the formula corresponding to the degree of target association between the target file and the candidate files is:

[0041] ;

[0042] in, It represents the degree of relevance between the target file and the h-th candidate file; h is the sequence number of the candidate file. G is the total number of times the h-th candidate file is called within the target time range; F is the number of preset time periods during which the h-th candidate file has been called; and f is the sequence number of the preset time periods during which the h-th candidate file has been called. It is an exponential function with the natural constant as its base; yes and The distance between them; It is the file query location to which the h-th alternative file needs to be sent when the h-th alternative file is called within the f-th preset time period; It is the file query location to which the target file needs to be sent when the target file is retrieved within the f-th preset time period.

[0043] Secondly, the present invention provides a physical file query system based on an integrated management device for PC and mobile terminals, the system comprising:

[0044] The acquisition and screening module is configured to acquire the current search term and screen, from all the to-be-queried entity archives, a to-be-queried entity archive containing all the current search terms as a candidate archive;

[0045] The initial recommendation index determination module is configured to determine, according to the historical calling condition of each candidate archive, an initial recommendation index corresponding to each candidate archive;

[0046] The acquisition and determination module is configured to acquire a feature word in each candidate archive and determine, based on a matching condition between the current search term and the feature word in each candidate archive and an importance degree of the feature word to the candidate archive, a target matching degree corresponding to each candidate archive;

[0047] The target recommendation index determination module is configured to determine, according to the initial recommendation index and the target matching degree corresponding to each candidate archive, a target recommendation index corresponding to each candidate archive;

[0048] The page display module is configured to display an entity archive recommendation page based on the target recommendation index corresponding to all the candidate archives, so as to realize entity archive query.

[0049] In a third aspect, a server is provided, including a memory and a processor. The memory is configured to store executable program code, and the processor is configured to call and run the executable program code from the memory, so that the device executes the method in the first aspect or any possible implementation manner of the first aspect.

[0050] In a fourth aspect, a computer program product is provided, including computer program code. When the computer program code runs on a computer, the computer program code causes the computer to execute the method in the first aspect or any possible implementation manner of the first aspect.

[0051] In a fifth aspect, a computer readable storage medium is provided, which stores computer program code. When the computer program code runs on a computer, the computer program code causes the computer to execute the method in the first aspect or any possible implementation manner of the first aspect.

[0052] The present application has the following beneficial effects:

[0053] The entity archive query method based on the PC terminal and mobile terminal integrated management device of the application realizes entity archive recommendation and entity archive query, solves the technical problem of poor efficiency of entity archive query caused by poor rationality of entity archive recommendation, and improves the rationality of entity archive recommendation and the efficiency of entity archive query. Specifically, when the entity archive recommendation is performed, the entity archive with relatively high similarity to the current search term, that is, the candidate archive, is screened out; and the historical calling condition of the candidate archive, the matching condition between the current search term and the feature word in the candidate archive, and the importance of the feature word to the candidate archive are comprehensively considered, the entity archive recommendation is realized, and the entity archive query is realized. Therefore, the factors considered when the entity archive recommendation is performed are relatively rich, so that the rationality of the entity archive recommendation is improved, and the efficiency of the entity archive query is improved. BRIEF DESCRIPTION OF DRAWINGS

[0054] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and for those skilled in the art, other drawings can be obtained without creative labor based on these drawings.

[0055] Figure 1 The flow chart of the entity archive query method based on the PC terminal and mobile terminal integrated management device of the application;

[0056] Figure 2 The composition structure schematic diagram of the entity archive query system based on the PC terminal and mobile terminal integrated management device of the application;

[0057] Figure 3 The structure schematic diagram of the computer equipment of the application. DETAILED DESCRIPTION

[0058] In order to further illustrate the technical means and effects adopted by the application to achieve the predetermined application purpose, the specific implementation, structure, features and effects of the technical solutions according to the application are described in detail as follows by combining the drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0059] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the application belongs.

[0060] REFERENCE Figure 1This document illustrates the flowchart of some embodiments of a physical file query method based on an integrated PC and mobile management device according to the present invention. The physical file query method based on the integrated PC and mobile management device includes the following steps:

[0061] Step S1: Obtain the current search term and filter out the entity files containing all the current search terms from all the entity files to be queried as candidate files.

[0062] Here, "current search term" can be a search term entered in the search box for the current archive query. "Current archive query" refers to the archive query to be performed at the current moment. "Entity archive to be queried" can be any physical archive participating in the archive query. Physical archives can be paper archives. For example, when querying a specific physical archive from a particular archive storage room, all physical archives stored in that archive storage room can be recorded as entity archives to be queried.

[0063] As an example, when staff members search for files, they can enter one or more search terms in the search window on their mobile devices. After the search terms are entered, each search term displayed in the search window is recorded as the current search term. If all current search terms exist in the file to be searched, then the file to be searched is identified as a candidate file. The staff member searching for the file can be an administrator.

[0064] It should be noted that the entity files to be queried often contain the current search terms. Therefore, entity files that do not contain the current search terms are often not the entity files to be queried. Thus, filtering out entity files that contain all the current search terms can reduce interference from other files to a certain extent.

[0065] Step S2: Determine the initial recommendation index for each candidate file based on its historical retrieval history.

[0066] Among them, the historical access to archives can also be referred to as the historical retrieval of archives.

[0067] As an example, this step may include the following steps:

[0068] The first step is to designate any candidate file as a marked file, and to determine the total number of historical calls to the marked file as the total number of historical calls to the marked file.

[0069] In this context, retrieving a specific file can also be referred to as accessing a specific file. The total number of times a file has been accessed in history is the total number of times that file has been accessed within a historical time period. The historical time period can be a period ending at the current time, and its duration can be six months. It should be noted that in this embodiment of the invention, the process of accessing different entity files in history can be the process of accessing different entity files within a historical time period.

[0070] In practice, the process of retrieving and returning files from the archives storage room is often referred to as the file retrieval process. Specifically, during the physical file retrieval process, staff members typically need to locate the required file, retrieve it from the archives storage room, place it at the front desk for relevant personnel to consult or take away for handling related matters, and finally return the physical file to the archives storage room.

[0071] The second step is to determine the time of each historical access to the marked file as the historical access time of the marked file, thus obtaining the historical access time sequence corresponding to the marked file.

[0072] The time of a specific historical retrieval of a marked file can be represented by the moment when the marked file was retrieved from the file storage during that historical retrieval process. The historical retrieval time sequence can be a time series.

[0073] The third step, based on the total number of historical calls and the historical call time sequence corresponding to the aforementioned tagged profiles, determines the initial recommendation metrics corresponding to the aforementioned tagged profiles, which may include the following sub-steps:

[0074] The first sub-step involves determining the duration between each historical call time and the current time in the historical call time sequence corresponding to the above-mentioned marked file as the target duration, thereby obtaining the target duration sequence corresponding to the above-mentioned marked file.

[0075] Among them, the historical call time in the historical call time sequence can correspond one-to-one with the target duration in the target duration sequence.

[0076] The second sub-step involves determining the initial recommendation metrics corresponding to the aforementioned tagged archives based on the total number of historical calls and the target duration sequence.

[0077] The total number of historical calls can be positively correlated with the initial recommendation metric. The target duration in the target duration sequence can be negatively correlated with the initial recommendation metric.

[0078] For example, the formula for determining the initial recommendation metric corresponding to the candidate profile can be:

[0079] ;

[0080] ;

[0081] in, is the initial recommendation metric corresponding to the i-th candidate file. i is the index of the candidate file. This represents the total number of historical calls corresponding to the i-th candidate file, which is the total number of historical calls to the i-th candidate file. n is the maximum value among all the total number of historical calls corresponding to the entity files to be queried. The total number of historical calls corresponding to the entity files to be queried is the total number of historical calls to the entity files to be queried. It is an exponential function with the natural constant as its base. It can characterize the overall difference between the current moment and the moment when the i-th candidate file was called in history. It is the number of target durations in the target duration sequence corresponding to the i-th candidate file, which can be equal to j is the index of the target duration in the target duration sequence corresponding to the i-th candidate file. Since Furthermore, the number of target durations in the target duration sequence corresponding to the i-th candidate file is equal to the number of historical call times in its corresponding historical call time sequence. Therefore, j can also be used as the order of historical calls to the i-th candidate file, and j can also be used as the sequence number of historical call times in the historical call time sequence corresponding to the i-th candidate file. This is the function for taking the absolute value. It is the j-th target duration in the target duration sequence corresponding to the i-th candidate file. It is the time of the j-th historical call to the i-th candidate file, which is the j-th historical call time in the historical call time sequence corresponding to the i-th candidate file. t is the current time.

[0082] It should be noted that when A larger value generally indicates that the i-th candidate file is called more often, meaning it is more frequently accessed and has a relatively higher probability of being called in this instance. The smaller the value, the closer the time when the history of the i-th candidate file was accessed is to the current time. This generally means the i-th candidate file is more likely to be accessed at this stage, and thus has a relatively high probability of being accessed in this instance. Therefore, when... The larger the value, the higher the probability that the i-th candidate file will be used in this call.

[0083] Step S3: Obtain the feature words in each candidate file, and determine the target matching degree for each candidate file based on the matching between the current search term and the feature words in each candidate file, as well as the importance of the feature words to the candidate files.

[0084] As an example, this step may include the following steps:

[0085] In the first step, use the JieBa tokenization tool to tokenize each candidate file, and record each resulting token as a target token.

[0086] Among them, JieBa tokenization is also called结巴分词.

[0087] In the second step, if the target token is not a stop word, then determine the target token as a feature word.

[0088] Among them, stop words refer to common high-frequency words that are excluded in natural language processing, such as function words, conjunctions, and auxiliary words like "de", "le", "he", etc. Their main function is to improve the efficiency of text processing and eliminate interference with semantic analysis.

[0089] It should be noted that a stop word list can be used to determine whether a target token is a stop word.

[0090] In the third step, through the World2Vec (Word to Vector) model, obtain the word vectors corresponding to each current search term and the word vectors corresponding to each feature word.

[0091] Among them, each different current search term can represent a current search term. Each different feature word can represent a feature word.

[0092] In the fourth step, according to the TF-IDF (Term Frequency–Inverse Document Frequency) value corresponding to each feature word in each candidate file, and the matching situation between the word vectors corresponding to all current search terms and the word vectors corresponding to all feature words in each candidate file, determine the target matching degree corresponding to each candidate file.

[0093] Among them, the TF-IDF value corresponding to a feature word can characterize the importance of this feature word for the candidate file.

[0094] For example, determining the target matching degree corresponding to each candidate file may include the following sub-steps:

[0095] In the first sub-step, determine the normalized value of the cosine similarity between the word vectors corresponding to each current search term and the word vectors corresponding to each feature word as the word reference similarity between each current search term and each feature word.

[0096] In the second sub-step, according to the word reference similarity between all current search terms and all feature words in each candidate file, and the TF-IDF values corresponding to all feature words in each candidate file, determine the target matching degree corresponding to each candidate file. It should be noted that "结巴分词" in the original text seems to be a Chinese term that needs to be translated accurately. Here, I translated it directly as "结巴分词" because there is no specific English equivalent provided. If there is a correct English name for it, it should be used instead.

[0097] Among them, both word reference similarity and TF-IDF value are positively correlated with target matching degree.

[0098] For example, the formula for determining the target matching degree of the candidate file can be:

[0099] ;

[0100] in, is the target match degree corresponding to the i-th candidate file. i is the index of the candidate file. M is the number of current search terms. a is the category index of the current search term. is the number of feature words in the i-th candidate file. b is the index of the feature word category in the i-th candidate file. It is the word reference similarity between the current search term of type a and the feature term of type b in the i-th candidate file. It is the TF-IDF value corresponding to the b-th feature word in the i-th candidate file.

[0101] It should be noted that when A larger value generally indicates a greater similarity between the current search term (type a) and the feature term (type b) in the i-th candidate file. A larger value generally indicates that the b-th feature word is more important to the i-th candidate file, and that the b-th feature word is more representative of the i-th candidate file to a certain extent. Therefore, when A larger value usually indicates that the current search term is more similar to the important feature words in the i-th candidate file, and that the i-th candidate file is more likely to be called in this operation.

[0102] Step S4: Determine the target recommendation index for each candidate file based on the initial recommendation index and target matching degree corresponding to each candidate file.

[0103] As an example, the product of the initial recommendation metric and the target matching degree corresponding to each candidate file can be used to determine the target recommendation metric for each candidate file.

[0104] For example, the formula for determining the target recommendation metric corresponding to the candidate profile can be:

[0105] ;in, is the target recommendation metric corresponding to the i-th candidate file. i is the index of the candidate file. It is the initial recommendation metric corresponding to the i-th candidate file. It is the target matching degree corresponding to the i-th candidate file.

[0106] It should be noted that when A larger value generally indicates that the i-th candidate file is being called more frequently in the current stage, and that the i-th candidate file is more likely to be called in this instance. A larger value generally indicates a greater similarity between the current search term and the important features in the i-th candidate file, suggesting a higher probability that the i-th candidate file will be retrieved in this search. Therefore, when... The larger the value, the higher the probability that the i-th candidate file will be used in this instance, and the more likely the i-th candidate file should be ranked higher when making this file recommendation.

[0107] Step S5: Display the entity file recommendation page based on the target recommendation indicators corresponding to all candidate files to enable entity file query.

[0108] As an example, this step may include the following steps:

[0109] The first step is to sort all candidate files according to the target recommendation indicators corresponding to all candidate files, in descending order, to obtain a candidate file sequence.

[0110] It should be noted that the higher a candidate file appears in the candidate file sequence, the higher the target recommendation index for that candidate file tends to be.

[0111] The second step is to compile the above candidate file sequence into a physical file recommendation page for display to staff.

[0112] The physical archive recommendation page can be the page containing the candidate archive sequence, which can be used to assist staff in searching for archives.

[0113] It should be noted that after finding the required file, sometimes it is also possible to search for related files. Therefore, in order to facilitate the relevant personnel to find the required files, related file recommendations can be made after the required file is found.

[0114] Optionally, recommending related profiles may include the following steps:

[0115] The first step is to use the displayed entity file recommendation page to select the required files from all candidate files as target files, and to identify the entity files to be queried that are not the target files as reference files.

[0116] The second step is to determine the time when the target file is accessed in history each time as the target access time, and to determine the target time range by the union of the preset time periods corresponding to all target access times.

[0117] The target call time can be the time at the center of its corresponding preset time period. The preset time period can be a pre-set time period surrounding the target call time, and its corresponding duration can be a pre-set duration. For example, the duration of the preset time period can be 5 minutes.

[0118] The third step is to identify the reference files that have been accessed within the aforementioned target time frame as candidate files.

[0119] The fourth step is to determine the degree of target association between the target files and each candidate file based on their access patterns within the target time frame.

[0120] For example, the formula for determining the degree of relevance between the target file and the candidate files can be:

[0121] ;

[0122] in, This represents the degree of relevance between the target file and the h-th candidate file. h is the sequence number of the candidate file. G is the total number of times the h-th candidate file is called within the target time range. F is the number of preset time periods during which the h-th candidate file was called. f is the sequence number of the preset time periods during which the h-th candidate file was called. It is an exponential function with the natural constant as its base. yes and The distance between them. It is the file query location to which the h-th alternative file needs to be sent when the h-th alternative file is called within the f-th preset time period. This refers to the file retrieval location where the target file needs to be sent when retrieving the target file within the f-th preset time period. The file retrieval location where the physical file needs to be sent when retrieving it can be the location where relevant personnel open and query the physical file during the file retrieval process. For example, if a file retrieval process involves taking the physical file from the archive storage room and placing it at the front desk for relevant personnel to query, then the file retrieval location where the physical file needs to be sent during that retrieval is the front desk. Conversely, if a file retrieval process involves taking the physical file from the archive storage room and placing it at the front desk, and then relevant personnel taking the physical file from the front desk to a designated location to handle related tasks, then the location where the physical file is opened and queried during the handling of those tasks is the file retrieval location where the physical file needs to be sent during that retrieval.

[0123] It should be noted that when A larger value generally indicates that the h-th alternative file is called more frequently in the vicinity of the target file, and that the h-th alternative file may be called again after the target file is called. The smaller the value, the closer the file query locations need to be when calling the target file and the h-th alternative file within the same time period; the more likely the target file and the h-th alternative file are to be used to process the same task; and the more related the target file and the h-th alternative file are. Therefore, when The larger the value, the more likely it is that the h-th alternative file will be called after the target file is called, and the more likely the h-th alternative file is to be associated with the target file.

[0124] The fifth step is to select a predetermined number of candidate files from all the candidate files that have the highest degree of relevance to the target file mentioned above, and use them as the target-related files.

[0125] The preset quantity can be the number of associated files that are pre-set to be displayed to staff. For example, the preset quantity could be 10.

[0126] Step 6: Recommend related files based on all target related files.

[0127] For example, based on the degree of relevance between all target-related files and target files, all target-related files are sorted in descending order to obtain a target-related file sequence. This sequence is then displayed to staff to facilitate file recommendations. Generally, the earlier a target-related file appears in the target-related file sequence, the greater the degree of relevance between it and other target files.

[0128] It should be noted that the integrated PC (Personal Computer) and mobile management device includes both PC and mobile terminals, primarily for integrated management across both platforms. Specifically, the integrated PC and mobile management device may include: PC-based physical file management software, a mobile operation and query platform, and a mobile workstation for file management, with a unique RFID (Radio Frequency Identification) electronic tag pre-attached to each file.

[0129] PC-based physical archive management software is primarily used for the digital storage, classification, retrieval, and statistical analysis of archives. It provides access control functions to ensure data security and privacy; it also features RFID data management capabilities, enabling tag reading and writing via RFID devices; furthermore, the software has big data analytics capabilities, generating visual reports to help managers monitor archive usage and trends in real time.

[0130] The mobile operation and query platform is mainly designed as a flexible and convenient operating platform for archive management personnel. It has functions such as archive inventory, shelving, and de-shelving, can quickly locate the physical location of archives, and can scan RFID tags to query archive information, thus simplifying the archive operation process.

[0131] The mobile workstation for document management integrates an RFID reader / writer module, which can quickly identify documents and pinpoint their exact location.

[0132] The specific process for querying archives in this embodiment of the invention can be as follows: the administrator inputs the relevant information of the archives to be queried into the mobile terminal operation query platform as search terms; the PC terminal physical archive management software receives the current search terms input by the administrator on the mobile terminal operation query platform, and based on the received current search terms, executes steps S1 to S5 to obtain a candidate archive sequence and a target associated archive sequence; the candidate archive sequence and the target associated archive sequence are fed back to the mobile terminal operation query platform to realize archive recommendation, thereby assisting staff in querying archives.

[0133] refer to Figure 2 Based on the same inventive concept as the above-described method embodiments, this invention provides a physical file query system based on an integrated PC and mobile management device. The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When executed by the processor, the computer program implements the steps of a physical file query method based on an integrated PC and mobile management device, specifically including:

[0134] The acquisition and filtering module 201 is used to acquire the current search term and filter out the entity files containing all the current search terms from all the entity files to be queried as candidate files;

[0135] The initial recommendation metric determination module 202 is used to determine the initial recommendation metric for each candidate file based on the historical call history of each candidate file.

[0136] The acquisition and determination module 203 is used to acquire the feature words in each candidate file, and determine the target matching degree corresponding to each candidate file based on the matching between the current search term and the feature words in each candidate file, as well as the importance of the feature words to the candidate file;

[0137] The target recommendation index determination module 204 is used to determine the target recommendation index for each candidate file based on the initial recommendation index and target matching degree corresponding to each candidate file.

[0138] The page display module 205 is used to display the entity file recommendation page based on the target recommendation indicators corresponding to all candidate files, so as to realize the entity file query.

[0139] Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. For example, as shown... Figure 3 As shown, the computer device 300 includes: a memory 301, a processor 302, and a computer program 303 stored in the memory 301 and running on the processor 302. When the processor 302 executes the computer program 303, the computer device can execute any of the aforementioned physical file query methods based on an integrated PC and mobile terminal management device.

[0140] Based on the same inventive concept as the above-described method embodiments, the present invention provides a server, including a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, enabling the device to execute any of the above-described entity file query methods based on an integrated PC and mobile management device.

[0141] Based on the same inventive concept as the above-described method embodiments, the present invention provides a computer program product comprising: computer program code, which, when executed on a computer, causes the computer to execute any of the above-described entity file query methods based on an integrated PC and mobile management device.

[0142] Based on the same inventive concept as the above-described method embodiments, the present invention provides a computer-readable storage medium storing computer program code, which, when executed on a computer, causes the computer to perform any of the above-described entity file query methods based on an integrated PC and mobile management device.

[0143] In summary, when recommending entity files, this invention filters out entity files with relatively high similarity to the current search term, i.e., candidate files. It comprehensively considers factors such as the historical retrieval status of candidate files, the matching between the current search term and the feature words in the candidate files, and the importance of the feature words to the candidate files, thus achieving entity file recommendation and enabling entity file querying. Therefore, this invention considers relatively rich factors when recommending entity files, thereby improving the rationality of entity file recommendations and ultimately improving the efficiency of entity file querying.

[0144] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for querying entity files based on an integrated management device for PC and mobile terminals, characterized in that, Includes the following steps: Get the current search term, and filter out the entity files containing all the current search terms from all the entity files to be queried, as candidate files; Based on the historical access history of each candidate file, determine the initial recommendation metrics for each candidate file; The feature words in each candidate file are obtained, and the target matching degree for each candidate file is determined based on the matching between the current search term and the feature words in each candidate file, as well as the importance of the feature words to the candidate file. Based on the initial recommendation metrics and target matching degree corresponding to each candidate file, the target recommendation metrics corresponding to each candidate file are determined. The entity file recommendation page is displayed based on the target recommendation indicators corresponding to all candidate files, so as to realize the entity file query. The method further includes: based on the displayed entity file recommendation page, staff members select the required files from all candidate files as target files, and identify the entity files to be queried other than the target files as reference files; the time of each historical call to the target file is identified as the target call time, and the union of the preset time periods corresponding to all target call times is identified as the target time range, wherein the target call time is the time at the center of its corresponding preset time period; the reference files called within the target time range are identified as candidate files; the degree of target association between the target file and each candidate file is determined based on the call history of the target file and each candidate file within the target time range; a preset number of candidate files with the highest degree of target association with the target file are selected from all candidate files as target associated files; and associated files are recommended based on all target associated files. The formula for the degree of relevance between the target file and the candidate files is: ； in, It represents the degree of relevance between the target file and the h-th candidate file; h is the sequence number of the candidate file. G is the total number of times the h-th candidate file is called within the target time range; F is the number of preset time periods during which the h-th candidate file has been called; and f is the sequence number of the preset time periods during which the h-th candidate file has been called. It is an exponential function with the natural constant as its base; yes and The distance between them; It is the file query location to which the h-th alternative file needs to be sent when the h-th alternative file is called within the f-th preset time period; It is the file query location to which the target file needs to be sent when the target file is retrieved within the f-th preset time period.

2. The method for querying entity files based on an integrated PC and mobile management device according to claim 1, characterized in that, The step of determining the initial recommendation metrics for each candidate file based on its historical retrieval history includes: Any candidate file is designated as a marked file, and the total number of times the marked file has been accessed in history is determined as the total number of times the marked file has been accessed in history. The time when the marked file is accessed in history is determined as the historical access time of the marked file, thus obtaining the historical access time sequence corresponding to the marked file; Based on the total number of historical calls and the historical call time sequence corresponding to the tagged profile, the initial recommendation index corresponding to the tagged profile is determined.

3. The method for querying entity files based on an integrated PC and mobile management device according to claim 2, characterized in that, The step of determining the initial recommendation metric corresponding to the tagged profile based on the total number of historical calls and the historical call time sequence includes: The duration between each historical call time and the current time in the historical call time sequence corresponding to the marked file is determined as the target duration, thus obtaining the target duration sequence corresponding to the marked file; Based on the total number of historical calls and the target duration sequence corresponding to the marked archive, the initial recommendation index corresponding to the marked archive is determined. The total number of historical calls is positively correlated with the initial recommendation index, and the target duration in the target duration sequence is negatively correlated with the initial recommendation index.

4. The method for querying entity files based on an integrated PC and mobile management device according to claim 1, characterized in that, The step of obtaining the feature words in each candidate file includes: The JieBa word segmentation tool is used to segment each candidate file to obtain the target word. If the target word segment is not a stop word, then the target word segment is identified as a feature word.

5. The method for querying entity files based on an integrated PC and mobile management device according to claim 1, characterized in that, The determination of the target matching degree for each candidate file based on the matching between the current search term and the feature words in each candidate file, as well as the importance of the feature words to the candidate files, includes: The World2Vec model is used to obtain the word vectors corresponding to each current search term and each feature word. Based on the TF-IDF value of each feature word in each candidate file, and the matching between the word vectors of all current search terms and the word vectors of all feature words in each candidate file, the target matching degree of each candidate file is determined. The TF-IDF value of the feature word represents the importance of the feature word to the candidate file.

6. The method for querying entity files based on an integrated PC and mobile management device according to claim 5, characterized in that, The step of determining the target matching degree for each candidate file based on the TF-IDF value corresponding to each feature word in each candidate file, and the matching results between the word vectors corresponding to all current search terms and the word vectors corresponding to all feature words in each candidate file, includes: The normalized value of the cosine similarity between the word vector corresponding to each current search term and the word vector corresponding to each feature word is determined as the word reference similarity between each current search term and each feature word. The target matching degree for each candidate file is determined based on the word reference similarity between all current search terms and all feature words in each candidate file, as well as the TF-IDF value of all feature words in each candidate file. Both word reference similarity and TF-IDF value are positively correlated with the target matching degree.

7. The method for querying entity files based on an integrated PC and mobile management device according to claim 1, characterized in that, The step of determining the target recommendation metric for each candidate file based on the initial recommendation metric and target matching degree for each candidate file includes: The product of the initial recommendation metric and the target matching degree for each candidate file is used to determine the target recommendation metric for each candidate file.

8. The method for querying entity files based on an integrated PC and mobile management device according to claim 1, characterized in that, The method of displaying entity file recommendation pages based on target recommendation metrics corresponding to all candidate files includes: Based on the target recommendation metrics corresponding to all candidate files, all candidate files are sorted in descending order to obtain a candidate file sequence. The candidate file sequence is used to create an entity file recommendation page for display to staff.

Images