TW201805839A - Data processing method, device and system - Google Patents

Abstract

Provided are a data processing method, device and system, the method comprising: a query terminal receiving a query request of a user, wherein the query request comprises search keywords; acquiring dimension keywords, index keywords and time granularity keywords among the search keywords, and first data corresponding to dimension features matching the dimension keywords, second data corresponding to index features matching the index keywords, and third data corresponding to time granularity features matching the time granularity keywords; and the query terminal determining target data fed back to the user according to the first data, the second data and the third data. In the present embodiment, the user only need to input the search keywords once without traversing each data outlet to search data, and a search engine database can then find data related to the search keywords in all the data outlets, thereby improving the efficiency of searching data.

Description

Data processing method, equipment and system

The present invention relates to Internet technology, and in particular, to a data processing method, device, and system.

With the rapid development of the Internet, data has exploded. At present, all companies with huge data assets have a large amount of data. The above-mentioned companies generally use four types of data export. The data export refers to the storage space where the data is stored or the software application that can generate data, and the storage space or software application can provide a data source for the database and store a large amount of data. Presented to all employees of the company, the four kinds of data exports are data application program exports (such as Alibaba's Taobao Business and Baidu's Baidu Index, etc.), report exports (such as the company's salary report), and knowledge base platform exports. (Such as Baidu Encyclopedia of Baidu Company) and export of cluster physical tables (such as personal information of company users).

For the non-technical employees of the above companies, they generally need to search the above four data outlets in order to obtain the required data. For example, a non-technical employee of the company needs to obtain the company's "amount of home improvement products in a certain day", then the non-technical staff must search the company's data application program export, report export, knowledge base platform export, and cluster physical table export in order. ,straight Until you find the company's "amount of home improvement products in a certain day".

Because in actual applications, the amount of data presented by each of the above data outlets is large, so non-technical employees searching for each data outlet in turn to find the data will inevitably cause inefficient data query.

The invention provides a data processing method, equipment and system to improve the efficiency of searching data.

In one aspect, the present invention provides a data processing system, including: a query terminal and a search engine database; the query terminal is configured to receive a query request from a user, the query request includes a search keyword; Describing the dimensional keywords, index keywords, and time fineness keywords in the search keywords, and sending the dimensional keywords, the indicator keywords, and the time fineness keywords to the search engine database The search engine database pre-stores the data in the data exit and the characteristic information of the data, the characteristic information includes at least one of the following: dimensional characteristics, index characteristics, and temporal nuance characteristics; the search engine data A library for obtaining first data corresponding to a dimensional feature matching the dimensional keyword, second data corresponding to an index feature matching the index keyword, and time fineness matching the time fineness keyword The third material corresponding to the sexual characteristics, and send the first material, the second material, and the third material Said terminal inquiry; the query terminal, for further information according to the first, the second The data and the third data, determine target data to be fed back to the user, and display the target data to the user.

In another aspect, the present invention provides a data processing method, including: a query terminal receiving a user's query request, the query request including a search keyword; and the query terminal obtaining a dimensional keyword and an index key in the search keyword Word and time fineness keywords; the query terminal sends the dimensional keywords, the index keywords, and the time fineness keywords to a search engine database to enable the search engine database to obtain and The first data corresponding to the dimensional feature matched by the dimensional keyword, the second data corresponding to the index feature matched with the index keyword, and the third data corresponding to the temporal nuance feature matched with the temporal nuance key Data, the search engine database pre-stores the data in the data exit, and the characteristic information of the data, the characteristic information includes at least one of the following: dimensional characteristics, index characteristics, and temporal nuance characteristics; the query terminal Receive the first data, the second data, and the third data sent by the search engine database The query according to the first data terminal, the second information and the third information to determine the user back to the target data.

In another aspect, the present invention provides a data processing method, including: a query terminal receiving a user's query request, the query request including a search keyword; and the query terminal obtaining at least two types of keys in the search keyword Word; the query terminal sends at least two types of keywords to a search engine database, so that the search engine database obtains source data corresponding to the at least two types of keywords respectively; the query terminal receives the search The source data sent by the engine database; the query terminal determines the target data to be returned to the user according to the source data.

In another aspect, the present invention provides a data processing method, which includes: a search engine database receives a dimensional keyword, an index keyword, and a time fineness keyword sent by a query terminal, the dimensional keyword, the index keyword, And the time fineness keyword is received by the query terminal from the query request of the user and obtained from the search keywords included in the query request; the search engine database stores the data in the data exit in advance, And characteristic information of the data, the characteristic information includes at least one of the following: dimensional characteristics, index characteristics, and temporal nuance characteristics; the search engine data obtains first data corresponding to the dimensional characteristics matching the dimensional keywords , Second data corresponding to the index characteristic matching the index keyword, and third data corresponding to the temporal nuance characteristic matching the temporal nuance keyword; the search engine data combines the first data, Sending the second data and the third data to the query terminal, so that the query terminal First data, said second data and said third data, return OK Target data for the user.

In yet another aspect, the present invention provides a data processing method, including: searching a database of a search engine to obtain first data in a data application, and dimensional characteristics, index characteristics, and time nuance characteristics of the first data; the search engine The database obtains the second data in the report, the knowledge base platform, the cluster physical table, and the dimensional characteristics of the second data; the search engine database stores the first data and the dimensions of the first data Characteristics, index characteristics, temporal nuance characteristics; the search engine database stores the second data, and the dimensional characteristics of the second data.

In another aspect, the present invention provides a query terminal, including: a receiving unit, a processing unit, and a sending unit; the receiving unit is configured to receive a user's query request, the query request includes a search keyword; and the processing unit Is coupled to the receiving unit, and is configured to obtain a dimension keyword, an index keyword, and a time nuance keyword in the search keyword; and the sending unit is coupled to the processing unit, and is configured to couple the dimension The keywords, the index keywords, and the time-specificity keywords are sent to a search engine database, so that the search engine database obtains the first data corresponding to the dimensional characteristics matching the dimensional keywords, and The second data corresponding to the index characteristics of the index keyword matching, and the third data corresponding to the time nuance characteristics matching the time nuance keyword matching Data, the search engine database pre-stores the data in the data export, and the characteristic information of the data, the data export includes at least one of the following: data applications, reports, knowledge base platforms, and cluster physical tables. The feature information includes at least one of the following: a dimensional feature, an index feature, and a temporal nuance feature; the receiving unit is further configured to receive the first data, the second data, and the sent data from the search engine database Third data; the processing unit is further configured to determine target data to be returned to the user according to the first data, the second data, and the third data.

In yet another aspect, the present invention provides a search engine database including a receiver, a memory, a processor, and a transmitter; the receiver is configured to receive a dimensional keyword, an index keyword, and a time sent by a query terminal; Subtlety keywords, the dimension keywords, the index keywords, and the time fineness keywords are the query request received by the query terminal from the user, and are obtained from the search keywords included in the query request Said memory is used to store data in a data export and characteristic information of said data, said data export includes at least one of the following: a data application, a report, a knowledge base platform, and a cluster physical table, said The feature information includes at least one of the following: a dimensional feature, an index feature, and a temporal nuance feature; the processor is coupled to the receiver and the memory, and is configured to obtain a dimensional feature correspondence that matches the dimensional keyword The first data of, the second data corresponding to the index characteristics matching the index keywords, and A third material corresponding to the time nuance feature matching the time nuance keyword; the transmitter is coupled to the processor and is configured to combine the first material, the second material, and the first material Three data are sent to the query terminal, so that the query terminal determines the target data to be returned to the user according to the first data, the second data, and the third data.

In the present invention, data in a data application, a report, a knowledge base platform, and a cluster physical table are collected in advance into a search engine database, and dimensional features, index features, and time nuance features are added to each piece of collected data. At least one of: when a search engine receives a search keyword entered by a user, the search keyword is first split to obtain a dimensional keyword, an index keyword, and a time fineness keyword; then, in a pre-built search In the engine database, find the data that matches the dimensional keywords, index keywords, and time-specific keywords, and display the matched data to the user; the user does not need to go through each data exit to find the data, only Enter a search keyword once, and the search engine database can find all the data related to the search keyword in all data outlets, thereby improving the efficiency of finding data.

1‧‧‧Query terminal

2‧‧‧Search Engine Database

10‧‧‧ users

11‧‧‧Query terminal

12‧‧‧ Semantic Recognition Module

13‧‧‧Search Engine Database

14‧‧‧Sequencer

15‧‧‧ Data Application

16‧‧‧statement

17‧‧‧ Knowledge Base Platform

18‧‧‧ cluster physical table

19‧‧‧ Syntax Parser

1900‧‧‧Query terminal

1922‧‧‧Processing element

1926‧‧‧Power Components

1932‧‧‧Memory

1950‧‧‧Interface

1958‧‧‧Input and output interface

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are Some embodiments of the invention, for those of ordinary skill in the art, On the premise of paying creative labor, other schemes can be obtained based on these schemes.

FIG. 1 is a schematic diagram of an optional application scenario of the present invention; FIG. 2 is a schematic structural diagram of a data processing system provided by an embodiment of the present invention; FIG. 3 is a flowchart of a data processing method provided by Embodiment 1 of the present invention; FIG. 5 is a flowchart of a data processing method provided by Embodiment 2 of the present invention; FIG. 5 is a flowchart of a data processing method provided by Embodiment 3 of the present invention; FIG. 6 is a flowchart of a data processing method provided by Embodiment 4 of the present invention; 7 is a flowchart of a data processing method provided in Embodiment 5 of the present invention; FIG. 8 is a flowchart of a data processing method provided in Embodiment 6 of the present invention; FIG. 9 is a flowchart of a data processing method provided in Embodiment 7 of the present invention; FIG. 10 is a flowchart of a data processing method provided by Embodiment 8 of the present invention; FIG. 11 is a flowchart of a data processing method provided by Embodiment 9 of the present invention; FIG. 12 is a schematic structural diagram of a query terminal provided by Embodiment 1 of the present invention FIG. 13 is a schematic structural diagram of a query terminal provided in Embodiment 2 of the present invention; FIG. 14 is a schematic structural diagram of a query terminal provided in Embodiment 3 of the present invention; and FIG. 15 is a structural schematic diagram of a search engine database provided by an embodiment of the present invention .

Exemplary embodiments will be described in detail here, examples of which are illustrated in the drawings. When the following description refers to drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present invention. Rather, they are merely examples of devices and methods consistent with aspects of the invention as detailed in the scope of the patent application.

In the prior art, when a non-technical employee of a company needs to obtain the company's "amount of home improvement products in a certain day", he needs to search the company's data application program export, report export, knowledge base platform export, and cluster physical table export in order. Until you find the company's "amount of home improvement products in a certain day", this will lead to a decline in the efficiency of data search. In response to this problem, a data processing method is proposed in this case. The specific process of the data processing method provided in this case will now be described with reference to FIG. 1.

As shown in FIG. 1, the user 10 queries the data through the query terminal 11. The user 10 may be a non-technical person in the company or a consumer. Alternatively, the inquiry terminal 11 may be a terminal device in the company to which the user 10 belongs, or may be a personal computer, a notebook computer, or other device of the user 10. The search terminal 11 is equipped with a search engine, and the user 10 can enter a search keyword in the search box of the search engine through the keyboard of the query terminal 11, for example, the search keyword is "the transaction amount of the home improvement category in the last day", and the semantic recognition module 12 This search keyword is split into dimensional keywords, index keywords, and time nuance keywords in the field of big data. Specifically, the dimensional keywords are "home improvement category", the index keywords are "transaction amount", time The subtlety keyword is "last day". The method for the semantic recognition module 12 to split the search keyword into a dimension keyword, an index keyword, and a time nuance keyword will be described in detail in the following embodiments.

The semantic recognition module 12 sends the split dimension keyword “home improvement category”, the index keyword “transaction amount”, and the time detail keyword “last day” to the search engine database 13 and the search engine database 13 The data sources include data application 15, report 16, knowledge base platform 17, and cluster physical table 18. Among them, data application 15 can be data products, such as Alibaba's Taobao Business and Baidu's Baidu Index. A data product is a web product in the form of a web page. The biggest difference between a data product and a common web product is that the data product carries a large amount of data and needs to interact frequently with background data sources. The background data source specifically stores the data application. 15 Operational information device. In this embodiment, the data in the data application 15 and the report 16 can be stored in the search engine database 13 through the parser 19. Taking the data application 15 as an example, since the data application 15 is It is developed by the Software Development Kit (SDK), so the data in the data application 15 can be obtained into the parser 19 through the SDK. Syntax parser 19 can parse a structured query language (Structured Query Language, SQL for short) of dimensional characteristics, index characteristics, time nuance characteristics and read table names. For example, a piece of SQL is as follows: SELECT stat_date AS date, user_type AS user type, se_lpv_pc_1d_001 AS Pv, se_uv_pc_1d_001 AS Uv

FROM tbbi.ads_tb_log_1d

WHERE ds = '20151026'

The parser 19 can parse the SQL dimension of this segment as "user type", the index feature is "Pv, Uv", the time nuance feature is "last day", and the table name read is "tbbi.ads_tb_log_1d" . Through the foregoing method, the syntax parser 19 can parse out the dimensional characteristics, index characteristics, and time nuance characteristics of each data in the data application 15 and the report 16.

The grammar parser 19 sends the parsed data to the search engine database 13. The search engine database 13 not only stores the data itself, but also stores the dimensional characteristics, index characteristics, and temporal nuance characteristics of the data. In addition, the search engine database 13 can also store the data in the knowledge base platform 17 and the cluster physical table 18. The storage process is as follows: each data in the knowledge base platform 17 and the cluster physical table 18 is split and extracted. Dismantle The dimensional characteristics of each of the divided materials are stored in the search engine database 13 for each of the materials in the knowledge base platform 17 and the cluster physical table 18 and the dimensional characteristics of each of the materials. As such, each piece of data stored in the search engine database 13 has at least dimensional characteristics.

When the search engine database 13 receives the dimensional keyword "home improvement category", the index keyword "transaction amount", and the time nuanced keyword "last day" sent by the semantic recognition module 12, it finds the key related to the dimension The data matching the word "home improvement category", the data matching the index keyword "transaction amount", and the data matching the time granularity keyword "last day", the search engine database 13 sends the found matching data to Sorter 14, if there is only one matching data found by the search engine database 13, the sorter 14 sends the matching data to the query terminal 11, and the query terminal 11 displays the matching data; if the search engine database 13 finds a match If there are multiple data, the sorter 14 sorts the multiple matching data according to a preset algorithm, and sends the sorted multiple matching data to the query terminal 11. The query terminal 11 displays the multiple matches in the order of sorting. data. In this embodiment, the preset algorithm that the sorter 14 uses to sort the multiple matching materials includes at least one of the following: Pagerank algorithm, CUS distance algorithm, and Latent Dirichlet Allocation (LDA) algorithm. Method, breadth first search (BFS) algorithm, etc.

In this embodiment, data in application programs, reports, knowledge base platforms, and cluster physical tables are collected in advance into the search engine database, and dimensional features, index features, and time are added for each piece of collected data. At least one of the subtle features; when the search engine receives the search keywords entered by the user, it first splits the search keywords to obtain the dimensional keywords, index keywords, and time fineness keywords; then, in advance The search engine database is created to find the data that matches the dimensional keywords, index keywords, and time-specific keywords, and displays the matched data to users; users do not need to go through each data exit to search for data You only need to enter a search keyword once, and the search engine database can find all the data related to the search keyword in all data outlets, thereby improving the efficiency of finding data.

FIG. 2 is a schematic structural diagram of a data processing system according to an embodiment of the present invention. As shown in FIG. 2, the data processing system includes a query terminal 1 and a search engine database 2, where the query terminal 1 is configured to receive a query request from a user. The query request includes a search keyword; the query terminal obtains a dimensional keyword, an index keyword, and a time fineness keyword in the search keyword, and sets the dimension keyword, the index keyword, and The time granularity keywords are sent to the search engine database.

As shown in FIG. 1, the query terminal 11 receives a query request from the user 10. There are various ways for the query request. For example, the user 10 enters text or voice on the search engine of the query terminal 11, and the text or voice includes the user 10 pre-retrieved keywords. As shown in FIG. 1, the semantic recognition module 12 and the sorter 14 may be modules belonging to the query terminal 11. The semantic recognition module 12 splits the search keyword into dimensional keywords and index keywords in the big data field. And time subtlety keywords, specifically, the dimension keyword is "home improvement category", the index keyword is "transaction amount", The time nuance keyword is "last day". The semantic recognition module 12 also sends the dimension keyword “home improvement category”, the index keyword “transaction amount”, and the time detail keyword “last day” to the search engine database 2.

The search engine database 2 previously stores data in the data outlet and characteristic information of the data, and the characteristic information includes at least one of the following: dimensional characteristics, index characteristics, and temporal nuance characteristics.

Optionally, in this embodiment, the data export includes: data applications, reports, knowledge base platforms, and cluster physical tables. The search engine database 13 stores data applications, reports, knowledge base platforms, and cluster physical tables. Data, and the characteristic information of each data, each data in the data application has dimensional characteristics, index characteristics and time nuance characteristics, and the data in reports, knowledge base platforms and cluster physical tables all have dimensional characteristics.

The search engine database 2 is configured to obtain first data corresponding to a dimensional feature matching the dimensional keyword, second data corresponding to an index feature matching the index keyword, and matching with the temporal nuance keyword The third material corresponding to the time nuance characteristic, and sending the first material, the second material, and the third material to the query terminal.

When the search engine database 13 receives the dimensional keyword "home improvement category", the index keyword "transaction amount", and the time nuanced keyword "last day" sent by the semantic recognition module 12, it can find the dimensions and dimensions respectively. The data matching the keyword "home improvement category", the data matching the indicator keyword "transaction amount", and the time detail keyword "recent "Day" match. The search engine database 13 can match the dimensional keyword "home improvement category" identified by the semantic recognition module 12 with the dimensional characteristics of the stored data to obtain the dimensional characteristics that match the dimensional keywords. Corresponding first data, the first data may be multiple data, and the first data may be data derived from a data application 15, a report 16, a knowledge base platform 17, or a cluster physical table 18.

In addition, the search engine database 13 can also match the index keyword "transaction amount" identified by the semantic recognition module 12 with the index characteristics of the stored data to obtain a second index corresponding to the index characteristics matching the index keywords. Data, the second data may be a plurality of data derived from the data application 15.

In addition, the search engine database 13 can also match the temporal nuance keywords “last day” identified by the semantic recognition module 12 with the temporal nuance characteristics of the stored data to obtain a match with the temporal nuance keywords. The third data corresponding to the temporal nuance feature may be a plurality of data derived from the data application program 15.

The search engine database 13 sends the first data, the second data, and the third data obtained to the query terminal 11, and may specifically send it to the sequencer 14 in the query terminal 11.

The query terminal 1 is further configured to determine, according to the first data, the second data, and the third data, target data that is returned to the user, and display the target data to the user.

If there is only one matching data found by the search engine database 13, the first data, the second data, and the third data are the same information. Data, the sorter 14 sends the matching data to the display of the query terminal 11, and the display of the query terminal 11 displays the matching data.

If there are multiple matching data found by the search engine database 13, that is, the first data, the second data, and the third data are not the same data, the sorter 14 performs a preset algorithm on the multiple data. The matching materials are sorted, and the sorted matching materials are sent to the display of the query terminal 11. The display of the query terminal 11 displays the multiple matching materials in the order of sorting.

In this embodiment, data in a data application, a report, a knowledge base platform, and a cluster physical table are collected in advance into a search engine database, and dimensional features, index features, and time nuance features are added for each piece of collected data. At least one of: when a search engine receives a search keyword entered by a user, the search keyword is first split to obtain a dimensional keyword, an index keyword, and a time fineness keyword; then, in a pre-built search In the engine database, find the data that matches the dimensional keywords, index keywords, and time-specific keywords, and display the matched data to the user; the user does not need to go through each data exit to find the data, only Enter a search keyword once, and the search engine database can find all the data related to the search keyword in all data outlets, thereby improving the efficiency of finding data.

FIG. 3 is a flowchart of a data processing method provided in Embodiment 1 of the present invention. As shown in FIG. 3, the method includes the following steps:

Step S201: The query terminal receives a query request from a user, where the query request includes a search keyword.

As shown in FIG. 1, the query terminal 11 receives a query request from the user 10. There are various ways for the query request. For example, the user 10 enters text or voice on the search engine of the query terminal 11. The text or voice includes the user 10 pre-retrieved keywords; or, the search engine of the query terminal 11 is provided with a drop-down list in which keywords are stored in advance, and the user can enter the pre-retrieved keywords by selecting keywords in the list and clicking ; Or alternatively, the user 10 previews the text information on the query terminal 11, and the user 10 selects a keyword from the previewed text information, and searches for the keyword by dragging, sliding, and clicking a function key.

The user 10 queries the data through the query terminal 11. The user 10 may be a non-technical person in the company or a consumer. The query terminal 11 may be a terminal device in the company to which the user 10 belongs, or the user 10 Personal computers, laptops, and more. The search terminal 11 is equipped with a search engine, and the user 10 can enter a search keyword in the search box of the search engine through the keyboard of the search terminal 11, for example, the search keyword is "the transaction amount of home improvement category in the last day".

Step S202: The query terminal obtains a dimensional keyword, an index keyword, and a time fineness keyword among the search keywords.

As shown in FIG. 1, the semantic recognition module 12 and the sorter 14 may be modules belonging to the query terminal 11 or modules belonging to the search engine database 13. The query terminal 11 and the search engine database 13 may be Connect directly or indirectly through other devices. In this embodiment, the semantic recognition module 12 and the sorter 14 belong to the query terminal 11, the query The terminal 11 and the search engine database 13 are directly connected as an example.

The semantic recognition module 12 splits the search keywords into dimensional keywords, index keywords, and time nuance keywords in the field of big data. Specifically, the dimensional keywords are "home improvement category" and the index keywords are "transactions." Amount ", time nuance keyword is" last day ".

Step S203: The query terminal sends the dimensional keywords, the index keywords, and the time fineness keywords to a search engine database, so that the search engine database obtains the dimensional keywords. The first data corresponding to the matched dimensional feature, the second data corresponding to the index feature matching the index key, and the third data corresponding to the time subtle feature matching the time subtle key.

In an embodiment, the search engine database pre-stores data in the data outlet and characteristic information of the data, and the characteristic information includes at least one of the following: dimensional characteristics, index characteristics, and temporal nuance characteristics.

The query terminal sends the dimensional keyword, the index keyword, and the time fineness keyword to a search engine database, the search engine database stores data in a data outlet in advance, and the Characteristic information of the data, the characteristic information includes at least one of the following: dimensional characteristics, index characteristics, and temporal nuance characteristics, and the data outlet includes at least one of the following: a data application, a report, a knowledge base platform, and a cluster physical table. Optionally, in this embodiment, the data export includes: data applications, reports, knowledge base platforms, and cluster physical tables, and the search engine database 13 stores data applications, reports, knowledge base platforms, and collections. The data in the group physical table.

As shown in FIG. 1, the data sources of the search engine database 13 include a data application 15, a report 16, a knowledge base platform 17, and a cluster physical table 18. Among them, the data application 15 may specifically be a data product, such as Alibaba ’s Taobao Business and Baidu's Baidu Index, etc., data products are web products in the form of web pages. The biggest difference between data products and ordinary web products is that data products carry a large amount of data and need to frequently interact with background data sources. The data source is specifically a device storing data operable by the data application program 15. In this embodiment, the data in the data application program 15 can be stored in the search engine database 13 through the parser 19. Specifically, the data in the data application program 15 is acquired into the parser 19 through the SDK. Syntax parser 19 can parse a structured query language (Structured Query Language, SQL for short) of dimensional characteristics, index characteristics, time nuance characteristics and read table names. For example, a piece of SQL is as follows: SELECT stat_date AS date, user_type AS user type, se_lpv_pc_1d_001 AS Pv, se_uv_pc_1d_001 AS Uv

FROM tbbi.ads_tb_log_1d

WHERE ds = '20151026'

The parser 19 can parse the SQL dimension of this segment as "user type", the index feature is "Pv, Uv", the time nuance feature is "last day", and the table name read is "tbbi.ads_tb_log_1d" . Through the foregoing method, the grammar parser 19 can parse out the dimensional characteristics, index characteristics, and temporal nuance characteristics of each data in the data application program 15. The grammar parser 19 sends the parsed data to the search engine database 13. The search engine database 13 not only stores the data itself, but also stores the dimensional characteristics, index characteristics, and temporal nuance characteristics of the data.

In addition, the search engine database 13 can also store data in the report 16, the knowledge base platform 17, and the cluster physical table 18. The storage process is as follows: For each data in the report 16, the knowledge base platform 17, and the cluster physical table 18 Perform splitting to extract the dimensional characteristics of each data after the split, and store each data in the report 16, the knowledge base platform 17, and the cluster physical table 18, and the dimensional characteristics of each data in the search engine database 13. As such, each piece of data stored in the search engine database 13 has at least dimensional characteristics.

When the search engine database 13 receives the dimensional keyword "home improvement category", the index keyword "transaction amount", and the time nuanced keyword "last day" sent by the semantic recognition module 12, it can find the dimensions and dimensions respectively. Data matching the keyword "home improvement category", data matching the index keyword "transaction amount", and data matching the time nuanced keyword "last day".

In this embodiment, the search engine database 13 stores the data in the data application 15 and the dimensional characteristics, index characteristics, and time nuance characteristics of each data in the data application 15. In addition, the search engine database 13 also stores the data in the report 16, the knowledge base platform 17, and the cluster physical table 18, as well as the report 16, the knowledge base platform 17, and the cluster properties. Manage the dimensional characteristics of each data in Table 18. In addition, the dimensional characteristics of each data in the search engine database 13 may be different and may be the same; the index characteristics of each data may be different and may be the same; the temporal nuance characteristics of each data may be different and may be the same.

The search engine database 13 in this embodiment can match the dimensional keyword “home improvement category” identified by the semantic recognition module 12 with the dimensional characteristics of the stored data to obtain a corresponding dimensional characteristic that matches the dimensional keyword. The first data may be multiple data, and the first data may be data derived from a data application 15, a report 16, a knowledge base platform 17, or a cluster physical table 18.

In addition, the search engine database 13 can also match the index keyword "transaction amount" identified by the semantic recognition module 12 with the index characteristics of the stored data to obtain a second index corresponding to the index characteristics matching the index keywords. Data, the second data may be a plurality of data derived from the data application 15.

In addition, the search engine database 13 can also match the temporal nuance keywords “last day” identified by the semantic recognition module 12 with the temporal nuance characteristics of the stored data to obtain a match with the temporal nuance keywords. The third data corresponding to the temporal nuance feature may be a plurality of data derived from the data application program 15.

Step S204: The query terminal receives the first data, the second data, and the third data sent by the search engine database.

The search engine database 13 sends the first data, the second data, and the third data obtained to the query terminal 11. It is sent to the sequencer 14 in the query terminal 11.

Step S205: The query terminal determines target data to be returned to the user according to the first data, the second data, and the third data.

If there is only one matching data found by the search engine database 13, that is, the first data, the second data, and the third data are the same data, the sorter 14 sends the matching data to the query terminal 11. Display, the display of the query terminal 11 displays the matching data.

If there are multiple matching data found by the search engine database 13, that is, the first data, the second data, and the third data are not the same data, the sorter 14 performs a preset algorithm on the multiple data. The matching materials are sorted, and the sorted matching materials are sent to the display of the query terminal 11. The display of the query terminal 11 displays the multiple matching materials in the order of sorting.

In this embodiment, data in a data application, a report, a knowledge base platform, and a cluster physical table are collected in advance into a search engine database, and dimensional features, index features, and time nuance features are added for each piece of collected data. At least one of: when a search engine receives a search keyword entered by a user, the search keyword is first split to obtain a dimensional keyword, an index keyword, and a time fineness keyword; then, in a pre-built search In the engine database, find the data that matches the dimensional keywords, index keywords, and time-specific keywords, and display the matched data to the user; the user does not need to go through each data exit to find the data, only Enter your search keywords once and the search engine database will be It can find all the data related to the search keyword in all data outlets, thereby improving the efficiency of finding data.

FIG. 4 is a flowchart of a data processing method provided in Embodiment 2 of the present invention. As shown in FIG. 4, based on the embodiment shown in FIG. 3, the query terminal obtains a dimensional keyword, The method for index keywords and time-specific keywords can include the following steps:

Step S301: The query terminal performs segmentation processing on the search keywords to obtain multiple target segmentations.

For example, as described in step S2 () 1, the search keyword entered by the user is "the transaction amount of the home improvement category in the last day". The query terminal 11 can also use TF-idf algorithm to split the search keywords entered by the user to obtain multiple target participles. The multiple target participles are "last day", "home improvement category", and "transaction amount". .

Step S302: The query terminal queries a preset mapping table according to each target segmentation, and the mapping table includes dimensional segmentation, index segmentation, and time subtle segmentation.

In this embodiment, the query terminal 11 establishes a mapping table in advance, and the mapping table includes dimension segmentation, index segmentation, and time subtle segmentation. The dimension segmentation may be a plurality of segmentations having dimensional characteristics, and the index segmentation may be a plurality of segmentations Feature word segmentation, time subtlety segmentation can be multiple segmentation with time subtleness features. According to the multiple target segmentations after step S301, the query terminal 11 queries the mapping table respectively, and for each target segment, determines whether there is a segmentation matching the target segmentation in the mapping table.

Step S303: The query terminal determines, as the dimension keyword, a target segmentation that matches the dimensional segmentation among the plurality of target segmentations.

For example, if the "home improvement category" in the multiple target participles matches the dimensional participle in the mapping table, the "home improvement category" is used as the dimensional keyword in the search key.

Step S304: The query terminal determines, as the index key, a target segmentation that matches the index segmentation among the plurality of target segmentations.

For example, if the "transaction amount" in the above multiple target segmentations matches the index segmentation in the mapping table, then the "transaction amount" is used as the index keyword in the search keywords.

Step S305: The query terminal determines a target segmentation that matches the temporal subtle segmentation among the multiple target segmentations as the temporal subtleness keyword.

For example, if the "last day" in the multiple target participles matches the time subtlety participle in the mapping table, then the "last day" is used as the time subtlety key in the search key.

In this embodiment, a plurality of target segmentations are obtained by performing segmentation processing on the search keywords, and the dimension keywords, index keywords, and time nuance keywords in the multiple target segmentations are queried according to a pre-established mapping table, thereby improving the determination Retrieve the efficiency of dimensional keywords, index keywords, and time nuance keywords in keywords.

FIG. 5 is a flowchart of a data processing method provided in Embodiment 3 of the present invention. As shown in FIG. 5, based on any one of the foregoing embodiments, and based on Embodiment 2, the specifics of the data processing method provided in this embodiment are specific. Steps as under:

Step S401: The query terminal receives a query request from a user, where the query request includes a search keyword.

Step S402: The query terminal obtains a dimensional keyword, an index keyword, and a time fineness keyword among the search keywords.

Step S403: The query terminal sends the dimensional keyword, the index keyword, and the time fineness keyword to a search engine database, so that the search engine database obtains the dimensional keyword. The first data corresponding to the matched dimensional feature, the second data corresponding to the index feature matching the index key, and the third data corresponding to the time subtle feature matching the time subtle key.

Step S404: The query terminal receives the first data, the second data, and the third data sent by the search engine database.

Steps S401 to S404 are consistent with steps S201 to S204, respectively, and specific methods are not described herein again.

Step S405: The query terminal determines whether the first data, the second data, and the third data are the same data. If yes, step S406 is performed; otherwise, step S407 is performed.

Step S406: The query terminal determines the same data as target data to be returned to the user.

As shown in FIG. 1, if there is only one matching data found by the search engine database 13, that is, the first data, the second data, and the third data are the same data, the sorter 14 matches the matching data. Sent to the display of the query terminal 11, and the display of the query terminal 11 displays the matching information material.

Step S407: The query terminal sorts the first data, the second data, and the third data, and determines the sorted data as target data to be returned to the user.

If there are multiple matching data found by the search engine database 13, that is, the first data, the second data, and the third data are not the same data, the sorter 14 performs a preset algorithm on the multiple data. The matching materials are sorted, and the sorted matching materials are sent to the display of the query terminal 11. The display of the query terminal 11 displays the multiple matching materials in the order of sorting.

In step S407, the method by which the query terminal sorts the first data, the second data, and the third data may specifically include the following steps:

Step S51: The query terminal calculates a weight value of each of the first data, the second data, and the third data.

Specifically, the weight value of each data can be calculated through the Pagerank algorithm.

Step S52: The query terminal calculates a similarity between each of the first data, the second data, and the third data and the search keyword.

Specifically, a CUS distance algorithm can be used to calculate the similarity between each data and the search keywords entered by the user.

Step S53: The query terminal calculates the ranking value of each material according to the weight value and similarity of each material.

Specifically, the value obtained by adding the weight value and the similarity of each material may be used as the ranking value of the material.

Step S54: The query terminal sorts each of the first material, the second material, and the third material according to the ranking value of each material.

Specifically, each of the first material, the second material, and the third material may be sorted according to a sort value of each material in a descending order.

Optionally, the query terminal determines, according to the ranking value of each of the materials, the materials in which the ranking value of the first material, the second material, and the third material is greater than a first threshold; the query terminal Sort the data whose ranking value is greater than the first threshold according to the size of the ranking value.

In addition, after calculating a ranking value of each of the first data, the second data, and the third data, the first data, the second data, and the third data may be determined. The data with a middle ranking value greater than the first threshold is sorted according to the size of the ranking value.

In this embodiment, a plurality of materials matching the search keywords found by the search engine database are sorted, and the sorting is based on the sorted value of each data, the sorted value and the weight value of each data, and the data and The similarity of the search keywords is related. The larger the ranking value, the stronger the correlation between the data and the search keywords. The sorted data is returned to the user, and the user can easily view the search keywords. The most relevant data improves the user experience.

FIG. 6 is a flowchart of a data processing method provided in Embodiment 4 of the present invention. As shown in FIG. 6, on the basis of any of the foregoing embodiments, and based on Embodiment 2, the specifics of the data processing method provided by this embodiment are specific. Proceed as follows:

Step S601: The query terminal receives a query request from a user, where the query request includes a search keyword.

Step S602: The query terminal obtains a dimensional keyword, an index keyword, and a time fineness keyword among the search keywords.

Step S603: The query terminal sends the dimensional keyword, the index keyword, and the time fineness keyword to a search engine database, so that the search engine database obtains the dimensional keyword. The first data corresponding to the matched dimensional feature, the second data corresponding to the index feature matching the index key, and the third data corresponding to the time subtle feature matching the time subtle key.

Step S604: The query terminal receives the first data, the second data, and the third data sent by the search engine database.

Step S605: The query terminal determines target data to be returned to the user according to the first data, the second data, and the third data.

Steps S601 to S605 are consistent with steps S201 to S205, respectively, and specific methods are not described herein again.

Step S606: The query terminal receives the user's click operation on the target data.

After step S407, multiple sorted target data can be displayed In the query terminal, the user can click to view the multiple target data through the query terminal. When a user clicks on some target data, the query terminal can receive the user's click operation on the target data.

Step S607: The query terminal establishes an association relationship between the user and the target data according to the click operation.

The association relationship includes a degree of association that identifies a degree of association between the user and the target material.

In this embodiment, the query terminal establishes an association relationship between the user and the target data according to a click operation generated when the user clicks on a certain target data. In addition, the user and his click can be calculated according to association rules and collaborative filtering rules. The relevance of the target data, the number of target data clicked by the user may be multiple.

Step S608: When the user does not enter the search keyword, the query terminal displays the target data according to the association relationship.

When the user does not enter a search keyword in the query terminal 11, the query terminal 11 may display the target data according to the association relationship between the user and the target data that the user clicked, that is, the query terminal 11 may display the target data that the user clicked. Visible to users.

Specifically, the association relationship includes a degree of association that identifies a degree of association between the user and the target data. The displaying, by the query terminal, the target material according to the association relationship includes: displaying, by the query terminal, the target material with a correlation degree greater than a second threshold.

Optionally, the query terminal displays the target data whose relevance is greater than a second threshold. The user ’s association with each target Including the degree of association between the user and the target data, the query terminal 11 may also display the target data with the degree of relevance clicked by the user greater than the second threshold.

In this embodiment, by establishing the association between the user and the target data that they clicked on, when the user does not enter a search keyword, the user ’s clicked information can be displayed according to the association relationship between the user and the target data. The target data improves the convenience of user query data.

FIG. 7 is a flowchart of a data processing method provided by Embodiment 5 of the present invention. As shown in FIG. 7, the specific steps of the data processing method provided by this embodiment are as follows:

Step S501: The query terminal receives a query request from a user, where the query request includes a search keyword.

As shown in FIG. 1, the query terminal 11 receives a query request from the user 10. There are various ways for the query request. For example, the user 10 enters text or voice on the search engine of the query terminal 11, and the text or voice includes the user 10 pre-retrieved keywords.

Step S502: The query terminal obtains at least two types of keywords in the search keywords.

In this embodiment, when the query terminal categorizes the search keywords requested by the user, the query keywords may not be limited to the three types of keywords: dimension keywords, index keywords, and time nuance keywords, because they are not user requests. Each search keyword of the query includes three types of keywords: a dimension keyword, an index keyword, and a time nuance keyword. Therefore, the semantic recognition module 12 corresponding to the query terminal 11 shown in FIG. 1 can also use The search keywords that the user requested the query are split into at least two types of keywords, for example, the user is Seller, the search keyword that the seller requests to query is "Are there any customers evaluating my products?" The verb "evaluation" and the noun "product" can be split.

Step S503: The query terminal sends at least two types of keywords to a search engine database, so that the search engine database obtains source data corresponding to the at least two types of keywords respectively.

The query terminal sends the verb "evaluation" and the noun "commodity" to the search engine database. The search engine database stores the product information of all the products of the seller and the evaluation information of each product. The search engine database obtains the product information of all the products of the seller according to the "product". The product information specifically includes the name, the origin, the material, etc., and obtains the evaluation information of all products according to the "evaluation".

Step S504: The query terminal receives the source data sent by the search engine database.

The search engine database sends product information and evaluation information to the query terminal. Since there can be multiple product information here, there can also be multiple evaluation information.

Step S505: The query terminal determines target data to be returned to the user according to the source data.

The query terminal may determine the product information that is returned to the user with the most evaluation information according to the number of evaluation information for each product, or may return the first several pieces of evaluation information to each user. The embodiment does not limit the specific implementation manner of the query terminal determining the target data to be returned to the user.

In this embodiment, by classifying the search keywords, the classification results The results are not limited to dimensional keywords, index keywords, and time nuance keywords, which improves the flexibility of classification of search keywords, increases the flexibility of search keywords, and expands the scope of search.

FIG. 8 is a flowchart of a data processing method provided in Embodiment 6 of the present invention. As shown in FIG. 8, the specific steps of the data processing method provided in this embodiment are as follows:

Step S701: The search engine database receives the dimensional keywords, index keywords, and time fineness keywords sent by the query terminal.

The dimension keyword, the index keyword, and the time fineness keyword are obtained by the query terminal from a query request of a user and are obtained from the search keywords included in the query request.

In this embodiment, the search engine database stores the data in the data exit and the characteristic information of the data in advance. The characteristic information includes at least one of the following: dimensional characteristics, index characteristics, and temporal nuance characteristics.

The data export includes at least one of the following: a data application, a report, a knowledge base platform, and a cluster physical table.

Step S702, the search engine data obtains first data corresponding to a dimensional feature matching the dimensional keyword, second data corresponding to an index feature matching the index keyword, and a time nuanced keyword The third data corresponding to the temporal nuance feature of the match.

Step S703: The search engine data sends the first data, the second data, and the third data to the query terminal, so that the query terminal uses the first data, the second data Information and said third Data to determine the target data that is returned to the user.

The principle of the method in this embodiment is consistent with the principle of the method in the embodiment shown in FIG. 3, and the specific process is not repeated here.

In this embodiment, data in a data application, a report, a knowledge base platform, and a cluster physical table are collected in advance into a search engine database, and dimensional features, index features, and time nuance features are added for each piece of collected data. At least one of: when a search engine receives a search keyword entered by a user, the search keyword is first split to obtain a dimensional keyword, an index keyword, and a time fineness keyword; then, in a pre-built search In the engine database, find the data that matches the dimensional keywords, index keywords, and time-specific keywords, and display the matched data to the user; the user does not need to go through each data exit to find the data, only Enter a search keyword once, and the search engine database can find all the data related to the search keyword in all data outlets, thereby improving the efficiency of finding data.

FIG. 9 is a flowchart of a data processing method provided in Embodiment 7 of the present invention. As shown in FIG. 9, specific steps of the data processing method provided in this embodiment are as follows:

Step S801: The search engine database stores data in the data application program, the report, the knowledge base platform, and the cluster physical table.

Based on the embodiment shown in FIG. 3, before receiving the search keywords input by the user, the search engine database 13 previously stores data in data applications, reports, knowledge base platforms, and cluster physical tables.

Specifically, the data in the data application program 15 can be stored in the search engine database 13 through the parser 19. Specifically, the data in the data application program 15 is acquired into the parser 19 through the SDK. Syntax parser 19 can parse a structured query language (Structured Query Language, SQL for short) of dimensional characteristics, index characteristics, time nuance characteristics and read table names. For example, a piece of SQL is as follows: SELECT stat_date AS date, user_type AS user type, se_lpv_pc_1d_001 AS Pv, se_uv_pc_1d_001 AS Uv

FROM tbbi.ads_tb_log_1d

WHERE ds = '20151026'

The parser 19 can parse the SQL dimension of this segment as "user type", the index feature is "Pv, Uv", the time nuance feature is "last day", and the table name read is "tbbi.ads_tb_log_1d" . Through the foregoing method, the grammar parser 19 can parse out the dimensional characteristics, index characteristics, and temporal nuance characteristics of each data in the data application program 15. The grammar parser 19 sends the parsed data to the search engine database 13. The search engine database 13 not only stores the data itself, but also stores the dimensional characteristics, index characteristics, and temporal nuance characteristics of the data.

In addition, the search engine database 13 can also store data in the report 16, the knowledge base platform 17, and the cluster physical table 18. The storage process is as follows: For each data in the report 16, the knowledge base platform 17, and the cluster physical table 18 Perform a split to extract the dimensional characteristics of each data after the split, Each data in the report 16, the knowledge base platform 17, and the cluster physical table 18, and the dimensional characteristics of each data are stored in the search engine database 13. As such, each piece of data stored in the search engine database 13 has at least dimensional characteristics.

Step S802: The search engine database receives the dimensional keyword, the index keyword, and the time fineness keyword sent by the query terminal, and the dimensional keyword, the index keyword, and the time fineness keyword are the The query terminal receives a query request from a user, and obtains the query request from a search keyword included in the query request.

In this embodiment, the search engine database stores the data in the data exit and the characteristic information of the data in advance. The characteristic information includes at least one of the following: dimensional characteristics, index characteristics, and temporal nuance characteristics.

The data export includes at least one of the following: a data application, a report, a knowledge base platform, and a cluster physical table.

Step S803, the search engine data obtains the first data corresponding to the dimensional feature matching the dimensional keyword, the second data corresponding to the index feature matching the index keyword, and the time fineness keyword The third data corresponding to the temporal nuance feature of the match.

Step S804: The search engine data sends the first data, the second data, and the third data to the query terminal, so that the query terminal can use the first data, the second data, and the second data The data and the third data determine the target data to be fed back to the user.

Step S802-Step S804 Method Principle and Step S701-Step The method described in step S703 has the same principle and will not be repeated here.

In this embodiment, data in a data application, a report, a knowledge base platform, and a cluster physical table are collected in advance into a search engine database, and dimensional features, index features, and time nuance features are added for each piece of collected data. At least one of: when a search engine receives a search keyword entered by a user, the search keyword is first split to obtain a dimensional keyword, an index keyword, and a time fineness keyword; then, in a pre-built search In the engine database, find the data that matches the dimensional keywords, index keywords, and time-specific keywords, and display the matched data to the user; the user does not need to go through each data exit to find the data, only Enter a search keyword once, and the search engine database can find all the data related to the search keyword in all data outlets, thereby improving the efficiency of finding data.

10 is a flowchart of a data processing method provided in Embodiment 8 of the present invention. As shown in FIG. 10, the search engine database stores the data application program, the report, the knowledge base platform, and the cluster physics. The information in the table may specifically include the following steps S901 and S902:

Step S901: The search engine database stores data in the data application.

Step S901 can be implemented through the following steps S11-S13:

Step S11: The search engine database acquires the access logic of the data application to access the data source.

The access logic includes data in the data application, and the data source stores output logic of the data.

This embodiment describes a method for storing data in the data application program into the search engine database, and the method described in this embodiment is different from the data application program 15 through the parser 19 described in the above embodiment. The data in the search engine database 13 is stored.

In this embodiment, the data application program may be specifically in the form of a Web page, which needs to frequently interact with background data sources; the background data source may specifically be a device that stores the data application operation data. Because the data application is developed according to the SDK, the SDK has the largest operation permission for the data application, so the SDK can capture the first access logic of the data application to the background data source through the SDK, and the first access logic includes data The time when the application accesses the background data source, the user ’s second access logic to the data application, and other fields. Therefore, through the existing analysis method, the second access logic of the user to the data application can be obtained. The second access logic of the user to the data application includes field information such as the time when the user accesses the data application, the data in the data application currently accessed by the user, and the like. Therefore, through the existing parsing method, the data in the data application currently accessed by the user can be obtained.

Suppose the user's second access logic to the data application is: SELECT stat_date AS date, user_type AS user type, se_lpv_pc_1d_001 AS Pv, se_uv_pc_1d_001 AS Uv

FROM tbbi.ads_tb_log_1d

WHERE ds = '20151026'

By analyzing the second access logic, the data in the data application currently accessed by the user can be obtained as "tbbi.ads_tb_log_1d", that is, the information behind the FROM field.

In addition, the output logic of each data is stored in the background data source. Therefore, in the background data source, the output logic of the data in the data application currently accessed by the user can be directly found.

Step S12: The search engine database determines characteristic information of data in the data application program according to the output logic.

Specifically, the output logic includes information of the polymer pieces of the data, index information participating in the calculation during the aggregation process, and time information of the index calculation.

The implementation of step S12 is specifically: the search engine database determines that the polymer piece information of the data is a dimensional feature of the data; the search engine database determines that the index information of the data participating in the calculation during the aggregation process is The index characteristics of the data; the search engine database determines the temporal nuance characteristics of the data according to the time information calculated by the index.

Analyze the output logic of the data in the data application currently accessed by the user, and obtain the polymer piece information of the current data, the index information participating in the calculation during the aggregation process, and the time information of the index calculation.

In this embodiment, it is assumed that the output logic of the data in the data application currently accessed by the user is as follows: Select stat_date, user_type, count (1) se_lpv_pc_1d_001, count (distinct uid) se_uv_pc_1d_001

From tbcdm.dwd_tb_log_1d where ds = ’20160119’

Group by user_type, stat_date

By analyzing the above output logic, the polymer piece information of the data in the data application that the user is currently accessing is stat_date, user_type, that is, the information after the Group by field; the index information participating in the calculation during the aggregation process is se_lpv_pc_1d_001, se_uv_pc_1d_001, that is, the information after the count (1) and count (distinct uid) fields; the time information of the index operation is '20160119', which is the score area after the where ds field.

Determine the polymer piece information of the current data as the dimensional characteristics of the current data, determine that the index information participating in the calculation during the current data aggregation is the index characteristic of the current data, and determine the time nuance of the current data based on the time information of the current data index calculation feature.

In addition, the time interval represented by the time information calculated by the above indicators can be used as the time subtle characteristics of the current data. For example, the time information calculated by the current data indicators, that is, the score area after the where field is "ds = '20160119" '", Then the time nuance characteristic of the current data is 1, for example, the time information of the current data index calculation, that is, the score area after the where field is" ds> =' 20160101'and ds <= '20160107' ", The temporal nuance characteristic of the current data is 7.

Step S13: The search engine database stores data in the data application and characteristic information of the data.

Finally, the data in the data application that the user is currently accessing Add dimensional features, index features, and temporal nuances, and store the added data in the search engine database.

In this embodiment, since the user accesses the data application once, the dimensional characteristics, index characteristics, and time nuance characteristics of the data in the data application currently accessed by the user can be obtained once, and it is The data in the accessed data application is added with the above-mentioned dimensional characteristics, index characteristics, and temporal nuance characteristics. Finally, the data after adding the above characteristics is stored in the search engine database. When the user accesses all the data in the data application, all the data in the data application can be stored in the search engine database. Each piece of data in the search engine database has dimensional characteristics, index characteristics, and Subtle characteristics of time.

Step S902: The search engine database stores data in the report, the knowledge base platform, and the cluster physical table.

Step S902 can be implemented through the following steps S21-S23:

Step S21: The search engine database obtains data in the report, the knowledge base platform, and the cluster physical table, respectively.

In this embodiment, each data in the report, the knowledge base platform, and the cluster physical table can be split by a TF-iDF algorithm.

Step S22: The search engine database determines the dimensional characteristics of each data in the report, the knowledge base platform, and the cluster physical table according to a preset algorithm.

LDA algorithm and TOPIC MODEL algorithm are used to extract the features of the split data, and the extracted features are used as the dimensional features of the corresponding data.

Step S23: The search engine database stores each report in the report, the knowledge base platform, and the cluster physical table, and the dimensional characteristics of the data.

Add dimensional features to each report in the report, the knowledge base platform, and the cluster physical table, and store the data after adding the dimensional features to the search engine database.

In this embodiment, the search engine database stores all the data in the data application, and each data stored from the data application to the search engine database is associated with dimensional characteristics, index characteristics, and temporal nuance characteristics; in addition, , The search engine database stores all data in the report, knowledge base platform, and cluster physical tables, and each data stored from the report, knowledge base platform, and cluster physical tables in the search engine database is associated with dimensional characteristics.

FIG. 11 is a flowchart of a data processing method provided in Embodiment 9 of the present invention. As shown in FIG. 11, the data processing method provided in this embodiment may include the following steps:

Step S1001: The search engine database acquires the first data in the data application, and the dimensional characteristics, index characteristics, and time nuance characteristics of the first data.

In this embodiment, the implementation of step S1001 may include the following two methods:

The first type: the search engine database receives the first data sent by a grammar parser, and the dimensional characteristics, index characteristics, and time nuance characteristics of the first data, and the grammar parser is used to collect the data The first data in the application, and the dimensional characteristics, index characteristics, and time nuance characteristics of the first data.

Specifically, the data in the data application program 15 can be stored in the search engine database 13 through the parser 19. Specifically, the data in the data application program 15 is acquired into the parser 19 through the SDK. Syntax parser 19 can parse a structured query language (Structured Query Language, SQL for short) of dimensional characteristics, index characteristics, time nuance characteristics and read table names. For example, a piece of SQL is as follows: SELECT stat_date AS date, user_type AS user type, se_lpv_pc_1d_001 AS Pv, se_uv_pc_1d_001 AS Uv

FROM tbbi.ads_tb_log_1d

WHERE ds = '20151026'

The parser 19 can parse the SQL dimension of this segment as "user type", the index feature is "Pv, Uv", the time nuance feature is "last day", and the table name read is "tbbi.ads_tb_log_1d" . Through the foregoing method, the grammar parser 19 can parse out the dimensional characteristics, index characteristics, and temporal nuance characteristics of each data in the data application program 15. The grammar parser 19 sends the parsed data to the search engine database 13. The search engine database 13 not only stores the data itself, but also stores the dimensional characteristics, index characteristics, and temporal nuance characteristics of the data.

The second type includes the following steps S31-S32:

Step S31: The search engine database obtains the data application The access logic of the data source is accessed in a manner that includes the first data in the data application, and the data source stores the output logic of the first data.

In this embodiment, the data application program may be specifically in the form of a Web page, which needs to frequently interact with background data sources; the background data source may specifically be a device that stores the data application operation data. Because the data application is developed according to the SDK, the SDK has the largest operation permission for the data application, so the SDK can capture the first access logic of the data application to the background data source through the SDK, and the first access logic includes data The time when the application accesses the background data source, the user ’s second access logic to the data application, and other fields. Therefore, through the existing analysis method, the second access logic of the user to the data application can be obtained. The second access logic of the user to the data application includes field information such as the time when the user accesses the data application, the data in the data application currently accessed by the user, and the like. Therefore, through the existing parsing method, the data in the data application currently accessed by the user can be obtained.

Suppose the user's second access logic to the data application is: SELECT stat_date AS date, user_type AS user type, se_lpv_pc_1d_001 AS Pv, se_uv_pc_1d_001 AS Uv

FROM tbbi.ads_tb_log_1d

WHERE ds = '20151026'

By analyzing the second access logic, the data in the data application currently accessed by the user can be obtained as "tbbi.ads_tb_log_1d", that is, the information behind the FROM field.

In addition, the output logic of each data is stored in the background data source. Therefore, in the background data source, the output logic of the data in the data application currently accessed by the user can be directly found.

The output logic includes the polymer piece information of the first data, the index information participating in the calculation during the aggregation process, and the time information of the index calculation. Specifically, the search engine database determines that the polymer piece information of the first data is a dimensional feature of the first data; the search engine database determines the index information of the first data participating in the calculation during the aggregation process Is the index characteristic of the first data; the search engine database determines the temporal nuance characteristic of the first data according to the time information calculated by the index.

Step S32: The search engine database determines feature information of the first data in the data application program according to the output logic, and the feature information includes dimensional features, index features, and temporal nuance features.

In this embodiment, it is assumed that the output logic of data in the data application currently accessed by the user is as follows: Select stat_date, user_type, count (1) se_kpv_pc_1d_001, count (distinct uid) se_uv_pc_1d_001

From tbcdm.dwd_tb_log_1d where ds = ’20160119’

Group by user_type, stat_date

By analyzing the above output logic, the user's current The polymer piece information of the data in the accessed data application is stat_date, user_type, that is, the information after the Group by field; the index information involved in the calculation during the aggregation process is se_lpv_pc_1d_001, se_uv_pc_1d_001, which is count (1) and count (distinct uid) Information after the field; the time information for the indicator calculation is '20160119', which is the score area after the where ds field.

Determine the polymer piece information of the current data as the dimensional characteristics of the current data, determine the index information participating in the calculation during the current data aggregation process as the index characteristics of the current data, and determine the time nuance of the current data based on the time information of the current data index calculation. feature.

In addition, the time interval represented by the time information calculated by the above indicators can be used as the time subtle characteristics of the current data. For example, the time information calculated by the current data indicators, that is, the score area after the where field is "ds = '20160119" '", Then the time nuance characteristic of the current data is 1, for example, the time information of the current data index calculation, that is, the score area after the where field is" ds> =' 20160101'and ds <= '20160107' ", The temporal nuance characteristic of the current data is 7.

Step S1002: The search engine database obtains the second data in the report, the knowledge base platform, and the cluster physical table, and the dimensional characteristics of the second data.

Specifically, the search engine database obtains the report, the knowledge base platform, and the second data in the cluster physical table respectively; the search engine database determines the report, The dimensional characteristics of the knowledge base platform and each second data in the cluster physical table are described.

In this embodiment, each data in the report, the knowledge base platform, and the cluster physical table can be split by a TF-iDF algorithm. LDA algorithm and TOPIC MODEL algorithm are used to extract the features of the split data, and the extracted features are used as the dimensional features of the corresponding data.

Step S1003: The search engine database stores the first data, and the dimensional characteristics, index characteristics, and time nuance characteristics of the first data.

Step S1004: The search engine database stores the second data and the dimensional characteristics of the second data.

Add dimensional features to each report in the report, the knowledge base platform, and the cluster physical table, and store the data after adding the dimensional features to the search engine database.

In this embodiment, data in a data application, a report, a knowledge base platform, and a cluster physical table are collected in advance into a search engine database, and dimensional features, index features, and time nuance features are added for each piece of collected data. At least one of: when a search engine receives a search keyword entered by a user, the search keyword is first split to obtain a dimensional keyword, an index keyword, and a time fineness keyword; then, in a pre-built search In the engine database, find the data that matches the dimensional keywords, index keywords, and time-specific keywords, and display the matched data to the user; the user does not need to go through each data exit to find the data, only Enter a search keyword once, and the search engine database can find all the data related to the search keyword in all data outlets, thereby improving the efficiency of finding data.

FIG. 12 is a schematic structural diagram of an inquiry terminal provided in Embodiment 1 of the present invention. As shown in FIG. 12, the inquiry terminal includes a receiving unit, a processing unit, and a sending unit.

The receiving unit is configured to receive a query request from a user, where the query request includes a search keyword.

The processing unit is coupled to the receiving unit and is configured to obtain a dimensional keyword, an index keyword, and a time nuance keyword in the search keywords.

The sending unit is coupled to the processing unit, and is configured to send the dimensional keyword, the index keyword, and the time fineness keyword to a search engine database, so that the search engine database Obtaining first data corresponding to a dimensional feature matching the dimensional keyword, second data corresponding to an index feature matching the index keyword, and time-corresponding features corresponding to the time-subtleness keyword match Third data, the search engine database pre-stores the data in the data export and the characteristic information of the data. The data export includes at least one of the following: data applications, reports, knowledge base platforms, and cluster physical tables , The feature information includes at least one of the following: a dimensional feature, an index feature, and a temporal nuance feature.

The receiving unit is further configured to receive the first data, the second data, and the third data sent by the search engine database.

The processing unit is further configured to determine target data to be returned to the user according to the first data, the second data, and the third data.

In this embodiment, through pre-collected data applications, reports, and knowledge The data in the knowledge base platform and the cluster physical table are added to the search engine database, and at least one of dimensional characteristics, index characteristics, and time nuance characteristics is added to each piece of collected data; when the search engine receives user input When searching for keywords, the search keywords are first split to obtain the dimensional keywords, index keywords, and time-specific keywords; then, in the search engine database that is established in advance, the keywords related to the dimensional keywords and indicators are found respectively. Words and time-specific keywords match the data, and display the matched data to the user; the user does not need to go through each data exit to find the data, just enter the search keyword once, and the search engine database can find out All the data related to the search keyword in the data export, thereby improving the efficiency of finding data.

Based on the embodiment shown in FIG. 12, the processing unit is specifically configured to perform word segmentation processing on the search keywords to obtain multiple target segmentations; query a preset mapping table according to each target segmentation, and the mapping table includes dimensions Segmentation, index segmentation, and time subtle segmentation; determining a target segmentation that matches the dimensional segmentation among the plurality of target segmentations as the dimension keyword; and matching among the plurality of target segmentations that matches the index segmentation A target participle is determined as the index keyword; and a target participle that matches the temporal subtlety part of the plurality of target participles is determined as the temporal subtleness key.

Further, the processing unit is specifically configured to determine whether the first data, the second data, and the third data are the same data; if the first data, the second data, and the third data are The data is the same data, the processing unit determines the same data as a feedback to the user If the first data, the second data, and the third data are not the same data, the processing unit performs the first data, the second data, and the third data Sorting, to determine the sorted data as the target data for feedback to the user.

In this embodiment, a plurality of target segmentations are obtained by performing segmentation processing on the search keywords, and the dimension keywords, index keywords, and time nuance keywords in the multiple target segmentations are queried according to a pre-established mapping table, thereby improving the determination Retrieve the efficiency of dimensional keywords, index keywords, and time nuance keywords in keywords.

FIG. 13 is a schematic structural diagram of an inquiry terminal provided in Embodiment 2 of the present invention. As shown in FIG. 13, the inquiry terminal further includes a display.

The receiving unit is further configured to receive a click operation on the target data by the user.

The processing unit is further configured to establish an association relationship between the user and the target data according to the click operation.

The display is coupled to the processing unit, and when the user does not enter the search keyword, the display displays the target data associated with the association relationship.

In this embodiment, by establishing the association between the user and the target data that they clicked on, when the user does not enter a search keyword, the user ’s clicked information can be displayed according to the association relationship between the user and the target data. The target data improves the convenience of user query data.

FIG. 14 is a schematic structural diagram of an inquiry terminal provided in Embodiment 3 of the present invention. Referring to FIG. 14, the inquiry terminal 1900 includes a processing element 1922, and its development One step includes one or more processors, and a memory resource represented by the memory 1932, for storing instructions executable by the processing element 1922, such as an application program. The application programs stored in the memory 1932 may include one or more modules each corresponding to a set of instructions. In addition, the processing element 1922 is configured to execute instructions to perform the methods of steps S201-S1004 described above.

The device 1900 may further include a power supply element 1926 configured to perform power management of the device 1900, a wired or wireless network interface 1950 configured to connect the device 1900 to a network, and an input / output (I / O) interface 1958. The device 1900 can operate based on an operating system stored in the memory 1932, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, or the like.

FIG. 15 is a schematic structural diagram of a search engine database provided by an embodiment of the present invention. As shown in FIG. 15, the search engine database includes a receiver, a memory, a processor, and a transmitter.

The receiver is configured to receive the dimensional keyword, the index keyword, and the time fineness keyword sent by the query terminal, and the dimensional keyword, the index keyword, and the time fineness keyword are the The query terminal receives a query request from a user, and obtains the query request from a search keyword included in the query request.

The memory is used to store data in a data export and characteristic information of the data. The data export includes at least one of the following: a data application, a report, a knowledge base platform, and a cluster physical table. The characteristic information Including at least one of the following: dimensional characteristics, index characteristics, and time detail Sexual characteristics.

The processor is coupled to the receiver and the memory, and is configured to obtain a first data corresponding to a dimensional feature matching the dimensional keyword, and a second data corresponding to an index feature matching the index keyword. The data and the third data corresponding to the time nuance feature matching the time nuance keyword.

The transmitter is coupled to the processor, and is configured to send the first data, the second data, and the third data to the query terminal, so that the query terminal is configured according to the first The data, the second data, and the third data determine target data to be returned to the user.

In this embodiment, data in a data application, a report, a knowledge base platform, and a cluster physical table are collected in advance into a search engine database, and dimensional features, index features, and time nuance features are added for each piece of collected data. At least one of: when a search engine receives a search keyword entered by a user, the search keyword is first split to obtain a dimensional keyword, an index keyword, and a time fineness keyword; then, in a pre-built search In the engine database, find the data that matches the dimensional keywords, index keywords, and time-specific keywords, and display the matched data to the user; the user does not need to go through each data exit to find the data, only Enter a search keyword once, and the search engine database can find all the data related to the search keyword in all data outlets, thereby improving the efficiency of finding data.

Based on the embodiment shown in FIG. 15, the processor is specifically configured to obtain access logic for the data application to access a data source, and the access The logic includes data in the data application, and the data source stores output logic of the data; according to the output logic, determining characteristic information of the data in the data application; applying the data The data in the program and characteristic information of the data are stored in the memory.

Alternatively, based on the embodiment shown in FIG. 15, the receiver is further configured to receive data sent by a parser, and the dimensional characteristics, index characteristics, and time nuance characteristics of the data, and the parser is used to collect The data in the data application, and the dimensional characteristics, index characteristics, and time nuance characteristics of the data; the processor is further configured to combine the data in the data application and the dimensional characteristics of the data Index characteristics, time nuance characteristics are stored in the memory.

Alternatively, based on the embodiment shown in FIG. 15, the processor is specifically configured to obtain data in the report, the knowledge base platform, and the cluster physical table respectively; and determine the report according to a preset algorithm , The dimensional characteristics of each material in the knowledge base platform and the cluster physical table; storing the report, each material in the knowledge base platform and the cluster physical table, and the dimensional characteristics of the data in The memory.

In this embodiment, the search engine database stores all the data in the data application, and each data stored from the data application to the search engine database is associated with dimensional characteristics, index characteristics, and temporal nuance characteristics; in addition, , The search engine database stores all data in the report, knowledge base platform, and cluster physical tables, and each data stored from the report, knowledge base platform, and cluster physical tables in the search engine database is associated with dimensional characteristics.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention, but not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: The technical solutions described in the foregoing embodiments can still be modified, or some or all of the technical features can be equivalently replaced; and these modifications or replacements do not depart from the essence of the corresponding technical solutions of the technical solutions of the embodiments of the present invention. range.

Claims (29)

A data processing system, comprising: a query terminal and a search engine database; the query terminal is configured to receive a user's query request, the query request includes a search keyword; and the query terminal obtains the search keyword Dimension keywords, index keywords, and time fineness keywords, and send the dimension keywords, the indicator keywords, and the time fineness keywords to the search engine database; the search engine database stores data in advance The data in the exit, and the characteristic information of the data, the characteristic information includes at least one of the following: dimensional characteristics, index characteristics, and temporal nuance characteristics; the search engine database is used to obtain dimensional characteristics that match the dimensional keywords The corresponding first data, the second data corresponding to the index characteristic matching the index keyword, and the third data corresponding to the temporal nuance characteristic matching the temporal nuance keyword, and the first data, the first data The second data and the third data are sent to the inquiry terminal; the inquiry terminal is further used for The second information and the third information feedback to the user to determine the target data, and displays the target information to the user. A data processing method, comprising: a query terminal receiving a query request from a user, the query request including a search keyword; The query terminal obtains the dimensional keywords, index keywords, and time fineness keywords in the search keywords; the query terminal sends the dimension keywords, the index keywords, and time fineness keywords to the search engine data Database, so that the search engine database obtains the first data corresponding to the dimensional feature matching the dimensional keyword, the second data corresponding to the index feature matching the metric keyword, and the time-specificity keyword matching The third data corresponding to the temporal nuance feature. The search engine database previously stores data in the data outlet and the characteristic information of the data. The characteristic information includes at least one of the following: dimensional features, index features, and temporal nuance features. ; The query terminal receives the first data, the second data, and the third data sent by the search engine database; the query terminal determines to return to the first data according to the first data, the second data, and the third data; User's goal data. The method according to item 2 of the scope of patent application, wherein the data export includes at least one of the following: a data application, a report, a knowledge base platform, and a cluster physical table. The method according to item 3 of the scope of patent application, wherein the query terminal obtains the dimensional keywords, index keywords, and time nuance keywords in the search keywords, including: the query terminal performs word segmentation processing on the search keywords to obtain Multiple target segmentation; The query terminal queries a preset mapping table according to each target segmentation, and the mapping table includes dimensional segmentation, index segmentation, and time subtle segmentation; the query terminal determines the target segmentation that matches the dimension segmentation among the plurality of target segmentations as the Dimension keyword; the query terminal determines the target segmentation that matches the index segmentation among the plurality of target segmentations as the index keyword; the query terminal determines the target segmentation that matches the time subtle segmentation among the plurality of target segmentations Keywords for that time nuance. The method according to item 4 of the scope of patent application, wherein the query terminal determines the target data to be returned to the user based on the first data, the second data, and the third data, including: the query terminal determines the first Whether the data, the second data, and the third data are the same data; if the first data, the second data, and the third data are the same data, the query terminal determines the same data as a feedback to the user The target data; if the first data, the second data, and the third data are not the same data, the query terminal sorts the first data, the second data, and the third data, and sorts the sorted data Determine the target data to give back to the user. The method according to item 5 of the scope of patent application, wherein the query terminal sorts the first data, the second data, and the third data, including Including: the query terminal calculates the weight value of each of the first data, the second data and the third data; the query terminal calculates each data of the first data, the second data and the third data Similarity to the search key; the query terminal calculates the ranking value of each material according to the weight value and similarity of the each material; the query terminal uses the ranking value of each material to the first material , Each of the second data and the third data is sorted. The method according to item 6 of the scope of patent application, wherein the query terminal sorts each of the first material, the second material, and the third material according to the ranking value of each material, including: the The query terminal determines, according to the ranking value of each material, the materials in which the ranking value of the first material, the second material, and the third material is greater than the first threshold; the query terminal determines the materials with the ranking value greater than the first threshold, Sort by the size of the sort value. The method according to any one of claims 2 to 7, wherein the query terminal determines the target data to be returned to the user according to the first data, the second data, and the third data, and further includes: : The query terminal receives the user's click operation on the target data; the query terminal establishes the user and the target data according to the click operation Data; when the user does not enter the search keyword, the query terminal displays the target data according to the relationship. The method according to item 8 of the scope of patent application, wherein the association relationship includes a degree of association that identifies the degree of association between the user and the target material; the query terminal displays the target material according to the relationship, including: the query The terminal displays the target data whose relevance is greater than a second threshold. A data processing method, comprising: a query terminal receiving a user's query request, the query request including a search keyword; the query terminal obtaining at least two types of keywords in the search keyword; the query terminal The class keywords are sent to the search engine database, so that the search engine database obtains source data corresponding to the at least two types of keywords respectively; the query terminal receives the source data sent by the search engine database; the query terminal is based on The source data determines the target data returned to the user. A data processing method, comprising: a search engine database receiving a dimensional keyword, a pointer Target keyword, and time fineness keyword, the dimension keyword, the index keyword, and the time fineness keyword are the query request received by the query terminal from the query terminal, and from the search keywords included in the query request Obtained; the search engine database pre-stores the data in the data outlet, and the characteristic information of the data, the characteristic information includes at least one of the following: dimensional characteristics, index characteristics and temporal nuance characteristics; the search engine data acquisition and The first data corresponding to the dimensional characteristics matched by the dimensional keyword, the second data corresponding to the index characteristics matched with the index keyword, and the third data corresponding to the time granularity characteristics matched with the time dimensionality keyword; the The search engine data sends the first data, the second data, and the third data to the query terminal, so that the query terminal determines to give back to the use according to the first data, the second data, and the third data. Target information. The method according to item 11 of the scope of patent application, wherein the data export includes at least one of the following: a data application, a report, a knowledge base platform, and a cluster physical table. The method according to item 12 of the patent application scope, wherein before the search engine database receives the dimensional keywords, index keywords, and time fineness keywords sent by the query terminal, the method further includes: the search engine database stores the data application Program, the report, the Knowledge base platform and materials in the physical table of the cluster. The method according to item 13 of the scope of patent application, wherein the search engine database stores data in the data application, including: the search engine database obtains access logic of the data application to access the data source, and the access logic includes the The data in the data application, the data source stores the output logic of the data; the search engine database determines the characteristic information of the data in the data application according to the output logic; the search engine database stores the data Data in the app, and information about the characteristics of that data. The method according to item 14 of the scope of patent application, wherein the output logic includes the polymer piece information of the data, the index information participating in the calculation during the aggregation process, and the time information of the index calculation; the search engine database is based on the output logic, Determine the characteristic information of the data in the data application, including: the search engine database determines that the polymer piece information of the data is the dimensional characteristics of the data; the search engine database determines the index information of the data participating in the calculation during the aggregation process Is the index characteristic of the data; the search engine database determines the time subtle characteristics of the data according to the time information calculated by the index. The method according to item 13 of the patent application scope, wherein the search engine database stores the report, the knowledge base platform, and the data in the cluster physical table, including: the search engine database obtains the report and the knowledge base platform, respectively And the data in the cluster physical table; the search engine database determines the dimensional characteristics of the report, the knowledge base platform, and each data in the cluster physical table according to a preset algorithm; the search engine database stores the report, Each material in the knowledge base platform and the cluster physical table, and the dimensional characteristics of the material. A data processing method, comprising: a search engine database acquiring first data in a data application, and dimensional characteristics, index characteristics, and time nuance characteristics of the first data; the search engine database obtaining reports separately , The knowledge base platform, the second data in the cluster physical table, and the dimensional characteristics of the second data; the search engine database stores the first data, and the dimensional characteristics, index characteristics, and time nuance characteristics of the first data ; The search engine database stores the second data and the dimensional characteristics of the second data. The method according to item 17 of the scope of patent application, wherein the search engine database obtains the first data in the data application, and the dimensional characteristics, index characteristics, and time nuance characteristics of the first data include the search engine data The library receives the first data sent by the parser Data, and the dimensional characteristics, index characteristics, and time nuances of the first data, the grammar parser is used to collect the first data in the data application, and parse the dimensional characteristics, index characteristics, and time details of the first data Sexual characteristics. The method according to item 17 of the scope of patent application, wherein the search engine database obtains the first data in the data application, and the dimensional characteristics, index characteristics, and time nuance characteristics of the first data include the search engine data The database obtains the access logic of the data application to access the data source. The access logic includes the first data in the data application. The data source stores the output logic of the first data. The search engine database is based on the output. Logic to determine characteristic information of the first data in the data application, the characteristic information including dimensional characteristics, index characteristics, and temporal nuance characteristics. The method according to item 19 of the scope of patent application, wherein the output logic includes the polymer piece information of the first data, the index information participating in the calculation during the aggregation process, and the time information of the index calculation; the search engine database is based on the output Logic to determine the characteristic information of the first data in the data application, including: the search engine database determines that the polymer piece information of the first data is a dimensional feature of the first data; the search engine database determines the first data The index information of the data participating in the calculation during the aggregation process is the index characteristic of the first data; The search engine database determines the temporal nuance characteristics of the first data according to the time information calculated by the index. The method according to any one of claims 17 to 20, wherein the search engine database obtains the second data in the report, the knowledge base platform, and the cluster physical table, and the dimensional characteristics of the second data, including : The search engine database obtains the report, the knowledge base platform, and the second data in the cluster physical table respectively; the search engine database determines the report, the knowledge base platform, and the cluster physical table according to a preset algorithm The dimensional characteristics of each second data in. A query terminal, comprising: a receiving unit, a processing unit, and a sending unit; the receiving unit is configured to receive a user's query request, the query request includes a search keyword; and the processing unit is coupled to the receiving unit To obtain the dimensional keyword, the index keyword, and the time subtlety keyword in the search keyword; the sending unit, coupled to the processing unit, is used to place the dimension keyword, the index keyword, and the time The subtle keywords are sent to the search engine database, so that the search engine database obtains the first data corresponding to the dimensional feature matching the dimensional keyword, the second data corresponding to the index feature matching the index keyword, and Matches the time nuanced keywords The third data corresponding to the time nuance characteristics is matched, and the search engine database stores the data in the data outlet and the characteristic information of the data in advance. The data outlet includes at least one of the following: data applications, reports, and knowledge bases Platform and cluster physical table, the feature information includes at least one of the following: a dimensional feature, an index feature, and a temporal detail feature; the receiving unit is further configured to receive the first data, the second data, and The third data; the processing unit is further configured to determine the target data to be returned to the user according to the first data, the second data, and the third data. The query terminal according to item 22 of the scope of patent application, wherein the processing unit is specifically configured to perform word segmentation processing on the search key to obtain multiple target segmentations; query a preset mapping table according to each target segmentation, and the mapping table includes dimension segmentation , Index segmentation, and time subtle segmentation; determine the target segmentation that matches the dimension segmentation among the multiple target segmentations as the dimension keyword; determine the target segmentation that matches the index segmentation among the multiple target segmentations as the index Keyword; the target participle that matches the time subtlety part of the plurality of target participle is determined as the time subtlety key. According to the query terminal of the scope of patent application No. 23, the processing unit is specifically configured to determine whether the first data, the second data, and the third data are the same data; if the first data, the second data, and If the third data is the same data, the processing unit determines the same data as the feedback to the user. Target data; if the first data, the second data, and the third data are not the same data, the processing unit sorts the first data, the second data, and the third data, and determines the sorted data Target data for giving back to the user. The query terminal according to item 24 of the patent application scope, which further includes: a display; the receiving unit is further configured to receive the user's click operation on the target data; the processing unit is further configured to establish the user according to the click operation Association with the target data; the display is coupled to the processing unit, and when the user does not enter the search keyword, the display displays the target data associated with the relationship. A search engine database, comprising: a receiver, a memory, a processor, and a transmitter; the receiver is used to receive a dimensional keyword, an index keyword, and a time nuance keyword sent by a query terminal The dimension keyword, the index keyword, and the time fineness keyword are obtained by the query terminal from a query request of a user and obtained from the search keywords included in the query request; the memory is used for storing The data in the data export, and the data The feature information includes at least one of the following: a data application, a report, a knowledge base platform, and a cluster physical table, and the feature information includes at least one of the following: a dimensional feature, an index feature, and a temporal nuance feature; the processor Is coupled to the receiver and the memory, and is used to obtain the first data corresponding to the dimensional feature matching the dimensional keyword, the second data corresponding to the index feature matching the index keyword, and the time nuance The third material corresponding to the time nuance characteristic of the keyword matching; the transmitter, coupled to the processor, for sending the first material, the second material and the third material to the query terminal, so that the The query terminal determines the target data to be returned to the user according to the first data, the second data, and the third data. The search engine database according to item 26 of the patent application scope, wherein the processor is specifically configured to obtain the access logic of the data application to access the data source, and the access logic includes the data in the data application, and the data source stores The output logic of the data; determining characteristic information of the data in the data application according to the output logic; storing the data in the data application and the characteristic information of the data into the memory. The search engine database according to item 26 of the patent application scope, wherein the receiver is also used to receive the data sent by the parser, as well as the dimensional characteristics, index characteristics, and time nuance characteristics of the data, and the parser is used to collect The data in the data application, and the dimensional characteristics, index characteristics, and time nuance characteristics of the data; The processor is also used to store the data in the data application, as well as the dimensional characteristics, index characteristics, and time nuance characteristics of the data into the memory. The search engine database according to item 26 of the patent application scope, wherein the processor is specifically configured to obtain data in the report, the knowledge base platform, and the cluster physical table respectively; according to a preset algorithm, determine the report, the The dimensional characteristics of each data in the knowledge base platform and the cluster physical table; the report, the each knowledge in the knowledge base platform and the cluster physical table, and the dimensional characteristics of the data are stored in the memory.