A method and device for managing business data

Abstract

The invention relates to the field of business data management methods, and discloses a business data management method and device, including an information terminal, a storage repository module, a user terminal, a processing module, a search module, and an expansion module. The information terminal can copy and store the business information required by users on various APPs, and according to the technical solution in the present invention, enable the users of the user terminal to obtain more accurate information content after searching. Moreover, by uploading redundant information with low click-through rate and long recording time to the cloud database for storage, it can effectively reduce the file occupancy of local information, enabling users to extract information faster when searching. Additionally, while the user is searching for information content, information that is closer in time within the same information group as the searched content can be recommended to the user for reference.

Description

Technical Field

[0001] The present invention relates to the field of business data management methods, and specifically to a business data management method and device. Background Art

[0002] With the rapid development of e-commerce, the sources of enterprise business data have become increasingly diversified, covering multiple shopping platforms, mobile applications, and Internet of Things terminal devices. During the operation process, merchants will collect and understand the information of relevant sold categories of different software by using keys (data keys), and sort and collect the currently sold data to facilitate the understanding of the information contained in the products. However, the information that can be stored in the storage repository of the existing data keys will gradually increase and become complex as the information is continuously loaded, resulting in the storage of some useless information in the storage end, occupying the storage space. This will interfere with the main search results with invalid information during the user's information search process, increasing the search difficulty. Summary of the Invention

[0003] The present invention provides a business data management method and device, which overcomes the deficiencies described in the background art.

[0004] The technical solution adopted by the present invention to solve its technical problems is as follows:

[0005] A business data management method and device, including

[0006] An information terminal for collecting information;

[0007] A storage repository module, which is signal-connected to the information terminal and stores information from the information terminal through manual loading;

[0008] A user terminal for obtaining the information stored in the storage repository module through searching;

[0009] A processing module, and the storage repository module is signal-connected to the user terminal through the processing module;

[0010] A search module, which is signal-connected to the processing module. During the search, the user terminal notifies the processing module of the search requirements, and the processing module commands the search module to search for the information stored in the storage repository module and feedback the main search results to the user terminal;

[0011] An expansion module, which is respectively signal-connected to the user terminal and the processing module. When the user terminal obtains the main search results from the processing module, the expansion module obtains secondary search results similar to the main search results according to the search requirements and the main search results, and feeds back the secondary search results to the user terminal through the processing module;

[0012] The repository module further includes a cloud database and a plurality of information units. The number of information terminals of the cloud database is set to be plural, and each information unit corresponds to one information terminal respectively, so as to store the information in different information terminals through each information unit; and all information units are signal-connected to the cloud database.

[0013] In a preferred technical solution, the search module includes a plurality of retrieval units, and the number of retrieval units is the same as the number of information units. Each retrieval unit is respectively signal-connected to each corresponding information unit.

[0014] A business data management method is applied to the described business data management device. The information stored in the repository module is actively collected and stored by the user through the data key at the information terminal. When storing information, the information is stored in different information units according to videos, pictures, and texts respectively, and the time stamp when the information is stored is recorded;

[0015] The information unit counts the occurrence frequency of the same entries in the different stored information, and groups the information with the same entries to form information groups. The information unit sets the information with the latest time stamp in all information groups as the information node, and counts the access times of all information;

[0016] The information storage method of the repository module is rolling storage. The rolling storage method is set with a time period threshold. When the time stamp of the information node in the information group exceeds the set time period threshold from the current time, the information unit uploads the information in the information group except the information node to the cloud database, and deletes the same information content stored in the information unit.

[0017] In a preferred technical solution, when the storage module obtains search results according to the search requirements of the user terminal, each retrieval unit searches for the information node with the corresponding entry in the corresponding information unit according to the search entry proposed by the user terminal, and calculates the access volume of the information nodes in the top 10% of the time stamps in the information database module;

[0018] Calculate the weight of each information node according to the weighted algorithm, and feedback the information node with the highest weight to the user terminal. The weighted formula is as follows:

[0019] ;

[0020] Wherein, is the weight of the corresponding information node;

[0021] is the frequency of the entry appearing in the information group;

[0022] is the time period of rolling storage;

[0023] is the time decay coefficient;

[0024] is the number of visits to the corresponding information node within the most recent 7 days;

[0025] is the median of the access volume of the information nodes in the top 10% of the timestamps in the information library module.

[0026] In a preferred technical solution, the actual click-through rate and the average click-through rate of all information nodes in the repository module are statistically calculated. The time decay coefficient of is dynamically adjusted according to the click-through rate of the corresponding information node, and the adjustment rule formula is as follows:

[0027] ;

[0028] where The reference value of is 0.05;

[0029] is the average click-through rate of all information nodes in the information library module;

[0030] is the actual click-through rate of the information node for which the current calculation weight is being calculated.

[0031] In a preferred technical solution, when the processing end feeds back the corresponding search results according to the search requirements of the user end, two pieces of information within the information group corresponding to the search result in the cloud database are provided and fed back to the user end;

[0032] The two pieces of information provided in the cloud database corresponding to the search result are the two pieces of information with timestamps closest to the timestamp of the time node corresponding to the information group.

[0033] Compared with the prior art, this technical solution has the following advantages:

[0034] Through the technical solution in the present invention, the user of the user end can obtain more accurate information content after searching. And by uploading redundant, low-click-through-rate, and long-recorded-time information to the cloud database for storage, the file occupancy of local information can be effectively reduced, enabling the user to extract information faster during searching. Also, when the user searches for information content, information with a time closer to that of the same information group as the search content can be recommended to the user for reference. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The present invention will be further described below in conjunction with the drawings and embodiments.

[0036] Figure 1This is the overall logic diagram of the present invention. Detailed implementation manners

[0037] As Figure 1 shown, the present invention proposes a business data management method and device, including

[0038] An information terminal for collecting information;

[0039] A repository module, which is signal-connected to the information terminal and stores information from the information terminal through manual loading;

[0040] A user terminal that obtains the information stored in the repository module through searching;

[0041] A processing module, and the repository module is signal-connected to the user terminal through the processing module;

[0042] A search module, which is signal-connected to the processing module. When searching, the user terminal tells the processing module the search requirements, and the processing module commands the search module to search for the information stored in the repository module and feedback the main search results to the user terminal;

[0043] An expansion module, which is respectively signal-connected to the user terminal and the processing module. When the user terminal obtains the main search results from the processing module, the expansion module obtains secondary search results similar to the main search results according to the search requirements and the main search results, and feeds back the secondary search results to the user terminal through the processing module;

[0044] It should be noted that the repository module in the present invention can be carried on a key (data key), and the corresponding information copying program is started by inserting it into a computer through SUB, and the information on the corresponding software is obtained by docking with the API of the corresponding APP, so that users can understand, copy and read the required information and materials.

[0045] The repository module further includes a cloud database and a plurality of information units. The number of information terminals of the cloud database is set to be plural, and each information unit corresponds to an information terminal respectively, so as to store the information in different information terminals through each information unit; and all information units are signal-connected to the cloud database.

[0046] Among them, the search module includes a plurality of retrieval units, and the number of retrieval units is the same as the number of information units. Each retrieval unit is respectively signal-connected to each corresponding information unit.

[0047] A business data management method is applied to the described business data management device. The information stored in the repository module is actively collected and stored by the user at the information terminal through a data key. When storing information, the information is stored in different information units according to videos, pictures, and texts respectively, and the timestamp when the information is stored is recorded.

[0048] The information unit counts the occurrence frequency of the same entries in the stored different information, and groups the information with the same entries to form an information group. The information unit sets the information with the latest timestamp in all information groups as the information node, and counts the access times of all information. The information storage method of the repository module is rolling storage, and a time period threshold is set for the rolling storage method. When the timestamp of the information node in the information group exceeds the set time period threshold from the current time, the information unit uploads the information in the information group except the information node to the cloud database, and deletes the same information content stored in the information unit.

[0049] Based on the above content, the general usage method of the business data management method in the present invention is as follows: deposit and record new information, update entries, reset the time difference, the user initiates a search, record the time of the first access from the user side, obtain the information node according to the entry, record and increase the access times of the information node, calculate the weight of the information node, return the sorted result to check whether the information exceeds the time threshold. If it exceeds the time threshold, upload the non-node data to the cloud database. If it does not exceed the threshold, record the access time of the information node.

[0050] In a preferred technical solution, when the storage module obtains a search result according to the search requirements of the user side, each retrieval unit searches for the information node with the corresponding entry in the corresponding information unit according to the search entry proposed by the user side, and calculates the access volume of the information nodes in the top 10% of the timestamps in the information library module.

[0051] Calculate the weight of each information node according to the weighted algorithm, and feedback the information node with the highest weight to the user side. The weighted formula is as follows:

[0052] ;

[0053] Among them, is the weight of the corresponding information node;

[0054] is the frequency of the entry appearing in the information group;

[0055] is the time period of rolling storage;

[0056] is the time decay coefficient;

[0057] is the number of visits to the corresponding information node within the most recent 7 days;

[0058] is the median of the access volume of the information nodes in the top 10% of the information repository module before the timestamp.

[0059] It should be noted that in this formula, the main reason for using + 1 in the word frequency factor is as follows: To avoid the zero-value problem. If log2(F) is directly used, when the word frequency F = 0 (for example, a new entry is not counted or there are residues after data cleaning), the logarithmic value log_2(0) will tend to negative infinity, resulting in calculation errors or abnormal weights.

[0060] In order to more clearly understand the technical effects achieved by the technical solution, the parameters of a certain information node are set according to the above formula as follows:

[0061] = 10, = 5 days, = 12 times, = 50 times. Based on this data, the following results are obtained:

[0062]

[0063] The characteristic of this technical solution is to use logarithmic compression to avoid the monopoly of high-frequency entries. According to the above formula, when the frequency of an entry appearing in the information group is 10 times, the weight is 3.32;

[0064] When the frequency of an entry appearing in the information group is 100 times, the weight is 6.66;

[0065] When the frequency of an entry appearing in the information group is 1000 times, the weight is 9.97.

[0066] And in order to make the information have a higher popularity weight in implementation, so that merchants can have a certain understanding of the popularity of the product, in the above formula is a coefficient value obtained by judging and calculating according to the click-through rate of the information nodes in the information repository, and is set as the median of the access volume of the top 10% of the information nodes in order to balance the impact of the calculation resources and the fluctuations of individual information hotspots in the storage repository module. By covering 10% of the partial information, users can search for the information they want among the information nodes with relatively prominent current weights, so as to improve the accuracy of information search.

[0067] Among them, the business data management method in the present invention needs to count the actual click-through rate and the average click-through rate of all the information nodes in the storage repository module, The time decay coefficient is dynamically adjusted according to the click-through rate of the corresponding information node, and the adjustment rule formula is as follows:

[0068] ;

[0069] Among them, The reference value of is 0.05;

[0070] is the average click-through rate of all information nodes in the information library module;

[0071] is the actual click-through rate of the information node for which the current weight is calculated;

[0072] The high-quality content and low-quality content in the information library module are calculated and obtained through the above formula. Because the actual click-through rate of high-quality content is often higher than that of low-quality content. For example, when the high-quality content information is 2%, is 3%, The decay coefficient of is 0.03. When the low-quality content information is 2%, is 0.5%, The decay coefficient of is 0.08. According to the weight calculation formula, the higher the decay coefficient, the higher the decay speed of this piece of information content, and the weight result drops rapidly. While the lower the decay coefficient, the slower the decay speed of this piece of information, and the weight result decays slowly, and the weight retention time is added. Thus, it can be seen that when The high-quality content with 3% is searched by the user side, and its weight coefficient can be retained for a longer time, so that the user side can retrieve more accurate content subsequently.

[0073] And, when the processing side feeds back the corresponding search results according to the search requirements of the user side, two pieces of information in the information group corresponding to the search results in the cloud database are provided and fed back to the user side;

[0074] The two pieces of information provided in the information group corresponding to the search results in the cloud database are the two pieces of information with timestamps closest to the time node corresponding to the information group.

[0075] As described above, it is only a preferred embodiment of the present invention, so the scope of implementation of the present invention cannot be limited thereby. That is, equivalent changes and modifications made according to the scope of the present invention patent and the content of the specification should still fall within the scope covered by the present invention.

Claims

1. An operation data management device, characterized in that, including an information terminal for collecting information; a repository module, signal - connected to the information terminal, storing information from the information terminal through manual loading; a user terminal, obtaining the information stored in the repository module through searching; a processing module, the repository module is signal - connected to the user terminal through the processing module; a search module, signal - connected to the processing module. When searching, the user terminal informs the processing module of the search requirement, and the processing module commands the search module to search for the information stored in the repository module and feedback the main search result to the user terminal; an expansion module, signal - connected to the user terminal and the processing module respectively. When the user terminal obtains the main search result from the processing module, the expansion module obtains secondary search results similar to the main search result according to the search requirement and the main search result, and feeds the secondary search result back to the user terminal through the processing module; the repository module further includes a cloud database and multiple information units. The number of information terminals of the cloud database is set to be plural, and each information unit corresponds to an information terminal respectively to store the information in different information terminals respectively; and all information units are signal - connected to the cloud database; the search module includes multiple retrieval units, and the number of retrieval units is the same as the number of information units. Each retrieval unit is signal - connected to each corresponding information unit respectively; The business data management method based on the above - mentioned business data management device is as follows: The information stored in the repository module is actively collected and stored by the user at the information terminal through a data key. When storing information, the information is stored in different information units according to video, picture, and text respectively, and the timestamp when the information is stored is recorded; The information unit counts the occurrence frequency of the same entries in the stored different information, groups the information with the same entries to form information groups, sets the information with the latest timestamp in all information groups as information nodes, and counts the access times of all information; The information storage method of the repository module is rolling storage. The rolling storage method is set with a time - period threshold. When the timestamp of the information node in the information group exceeds the set time - period threshold from the current time, the information unit uploads the information in the information group except the information node to the cloud database and deletes the same information content stored in the information unit; When the storage module obtains a search result according to the search requirement of the user terminal, each retrieval unit searches for the information node with the corresponding entry in the corresponding information unit according to the search term proposed by the user terminal, and calculates the access volume of the information nodes in the top 10% of the timestamps in the information database module; Calculate the weight of each information node according to the weighted algorithm, and feedback the information node with the highest weight to the user terminal. The weighted - algorithm formula is as follows: ； Among them, is the weight of the corresponding information node; is the frequency of the entry's appearance in the information group; is the time period for rolling storage; is the time decay coefficient; is the number of accesses to the corresponding information node within the most recent 7 days; It is the median of the access volumes of the information nodes in the information repository module that are in the first 10% before the timestamp.

2. The operation data management device according to claim 1, characterized in that Statistically analyze the actual click-through rate and average click-through rate of all information nodes in the repository module. The time decay coefficient of is dynamically adjusted according to the click-through rate of the corresponding information node, and the adjustment rule formula is as follows: ； Among them, The reference value is 0.05; is the average click-through rate of all information nodes in the information repository module; The actual click-through rate of the information node for which the weight is currently being calculated.

3. The operation data management device according to claim 2, wherein When the processing terminal feedbacks the corresponding search result according to the search requirement of the user terminal, provide two pieces of information in the information group corresponding to the search result in the cloud database and feedback them to the user terminal; The two pieces of information provided in the cloud database in the information group corresponding to the search result are the two pieces of information with timestamps closest to the timestamp of the time node corresponding to the information group.

Images