CN107633093A - A kind of structure and its querying method of DECISION KNOWLEDGE collection of illustrative plates of powering - Google Patents
A kind of structure and its querying method of DECISION KNOWLEDGE collection of illustrative plates of powering Download PDFInfo
- Publication number
- CN107633093A CN107633093A CN201710934931.2A CN201710934931A CN107633093A CN 107633093 A CN107633093 A CN 107633093A CN 201710934931 A CN201710934931 A CN 201710934931A CN 107633093 A CN107633093 A CN 107633093A
- Authority
- CN
- China
- Prior art keywords
- power supply
- knowledge graph
- query
- association rule
- entities
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of construction method for DECISION KNOWLEDGE collection of illustrative plates of powering, specific steps include：1) item data storehouse DBT is built；2) Strong association rule is excavated；3) structure power supply DECISION KNOWLEDGE collection of illustrative plates.Through the above way, the present invention a kind of structure and its querying method of DECISION KNOWLEDGE collection of illustrative plates of powering, this method provides more intuitive decision references for power supply decision-maker, so as to improve constantly Power supply market occupation rate, to meet the needs to client's fast accurate service of power supply enterprise.
Description
Technical Field
The invention relates to the field of big data technology and knowledge engineering, in particular to a power supply decision knowledge graph construction and query method.
Background
The knowledge map is a knowledge base with a map structure, is different from a common knowledge base, and associates knowledge with different sources, different types and different structures into a map through semantic analysis, so that a wider and deeper knowledge system is provided for a user and is continuously expanded, and the knowledge map essentially systematizes, relates and visualizes domain knowledge. Briefly, the knowledge graph systematically displays complex domain knowledge through technologies such as data mining, knowledge conversion, semantic association and visualization, and the like, and reveals dynamic development rules of the domain knowledge.
The construction of knowledge graphs is generally divided into two categories: one is the knowledge graph of the special domain, and the other is the construction of the knowledge graph itself.
At present, the application of knowledge maps in specific fields is still limited to the aspects of search engines, question-answering systems, epidemic disease diagnosis and the like, the application of other aspects is less, along with the further deepening of the reform of the power system, the staged achievement of the construction of a three-set five-large system is obvious, and the twelve-five marketing planning of a national grid indicates that the large marketing reform needs to be changed from a service-oriented type to a client-oriented type, so that the requirements of clients need to be comprehensively mastered, and the change of the client requirements is known to guide the development of marketing work.
By means of the power supply decision knowledge graph, a power supply enterprise can realize accurate service, deep understanding and cognition of the power supply enterprise on all-around information of power utilization customers are behind the accurate service, and the power supply enterprise can find potential customers with high power failure sensitivity in the process; which customers are difficult to accept the new service mode only have a feeling of traditional business; which customers have low credit, high risk or the potential for malicious arrears.
The method comprises the steps of accurately outlining power consumption customers, acquiring potential decision-making knowledge after deep analysis and mining by combining multidimensional data such as power supply service operation data and power consumption customer behavior data, performing semantic association on the customers according to service targets, describing the customers in a graph mode, and improving service modes and customer satisfaction by constructing a power supply decision-making knowledge graph; the working means is expanded, and the working efficiency is improved; customer segmentation is strengthened, and active service is provided; and (4) increasing the prediction strength and providing decision support.
Disclosure of Invention
The invention mainly solves the technical problem of providing a power supply decision knowledge map construction and query method, which can facilitate multi-dimensional knowledge mining, knowledge conversion, semantic association and knowledge reasoning and provide more intuitive decision reference for power supply decision-making personnel, thereby continuously improving the power supply market share and meeting the requirements of power supply enterprises on quick and accurate service of customers.
In order to solve the technical problems, the invention adopts a technical scheme that: the method for constructing the power supply decision knowledge graph is provided, and comprises the following specific steps:
1) Constructing a thing database DBT: the method comprises the steps of utilizing 95598 worksheet data S as a data source for constructing a power supply decision knowledge graph, and converting the data source into a things database DBT by taking a client as a core concept;
2) Mining strong association rules: mining all strong association rules in the business database DBT by using an association rule mining algorithm;
3) Constructing a power supply decision knowledge graph: extracting concepts, entities, attributes of the entities and relations among the entities from the object database DBT in the step 1), constructing a data body according to the extracted concepts and entities, performing semantic analysis on the data body, extracting entity features to construct a feature body, performing association analysis on the feature body, extracting relations among the entities to construct a client body, and further completing construction of a power supply decision knowledge graph, wherein the data body comprises an object T and an item set I, and the feature body is an item I k And the client ontology is a strong association rule.
In a preferred embodiment of the present invention, the specific step of step 1) is to convert the numerical variables into factor variables by discretization, take the factor levels as terms, create a pivot table by using the client ID as an index for the data source, obtain some grouping statistics, and connect the relevant pivot table by the client ID to obtain a table, i.e. the object database DBT.
In a preferred embodiment of the present invention, the specific mining step of the strong association rule in step 2) includes:
1) Let DBT = { T l ,T 2 ,…T n Where T is i Is called a transaction T, where i takes on the value 1 k Called term, k takes the value 1.. Eta., m, the set of all terms in DBT is I = { I = { I } l ,i 2 ,…i m }，
2) Let A = { i = } A1 ,i A2 ,…i At 1 < t < m), wherein A is called a t-item set in DBT;
3) Calculating item set support by formula (1)
4) Calculating the confidence coefficient of the association rule by formula (2), and setting the association rule
Wherein A and B are both sets of terms in DBT, anda is the condition of the association rule and B is called the conclusion of the association rule.
5) The strong association rule is calculated by equation (3):
when in useAnd isThen the rule is associatedIs a strong association rule, otherwise is not a strong association rule, wherein A c The right part B of the strong association rule is a complement of A and is a dependent variable.
In a preferred embodiment of the present invention, the concepts, entities, attributes of entities and relationships between the entities in the step 3) are things T and item sets I, T-item set A, item i, respectively k (1. Ltoreq. K. Ltoreq.m) and strong association rules
The invention adopts another technical scheme that: the method for inquiring the power supply decision knowledge graph comprises the following specific steps:
1) Receiving a query: performing word segmentation on a query sentence W input by a user by using a Chinese academy Chinese word segmentation system NLPIR to obtain W = { W1, W2,..., wn }, wherein wi (i =1,..., n) is a word or a character;
2) Query and analysis: mapping each vocabulary in the word segmentation result in the step 1) to four types of query elements to obtain an element pair of the query W, and recording the element pair as:<P，W>＝{<p1，w1>，<p2，w2>,..,<pn，wn>}，pi∈{NULL，A ik wherein i =1,.. N, NULL does not correspond to any element, and the four types of query elements are thing T, item set a, item i k And others;
3) Forming a query pattern: according to the analyzed item set and item set supportThe support degree and the associated rule confidence degree form a query mode, and the query mode of W is recorded as X = f ({ A) ik })；
4) Displaying a knowledge graph: and executing a query mode through a knowledge graph reasoning engine based on the Cypher language of the Neo4j, and displaying the queried knowledge graph data to a decision maker.
The invention has the beneficial effects that: the invention relates to a power supply decision knowledge graph constructing and inquiring method, which comprises the steps of converting a 95598 data source into an object database, extracting concepts, entities and attribute information of the entities for constructing the power supply decision knowledge graph and relationship information among the entities by using an association rule mining algorithm, then storing the power supply decision knowledge graph by using a graph mode, inputting and analyzing a received decision maker, establishing an inquiring mode, displaying a result returned by the inquiring mode to the decision maker according to an inference engine, and providing a more intuitive decision reference for the power supply decision maker, thereby continuously improving the power supply market share so as to meet the requirement of a power supply enterprise on quick and accurate service of a client.
Drawings
Fig. 1 is a flow chart of a power supply decision knowledge graph construction and query method thereof.
Fig. 2 is a schematic diagram of a power supply decision knowledge graph of a method for constructing and querying the power supply decision knowledge graph.
Detailed Description
The following detailed description of the preferred embodiments of the present invention, taken in conjunction with the accompanying drawings, will make the advantages and features of the invention easier to understand by those skilled in the art, and thus will clearly and clearly define the scope of the invention.
Referring to fig. 1 and fig. 2, an embodiment of the present invention includes a method for constructing a power supply decision knowledge graph, which includes the following specific steps:
1) Constructing a thing database DBT: the method comprises the steps of utilizing 95598 worksheet data S as a data source for constructing a power supply decision knowledge graph, converting the data source into an object database DBT by taking a client as a core concept, converting numerical variables into factor variables through discretization, taking factor levels as items, establishing a pivot table by taking a client ID as an index for the data source, obtaining some grouping statistics, and connecting the client ID with a relevant pivot table to obtain a table, namely the object database DBT.
And taking the client ID as a grouping variable, acquiring statistical characteristics about the client, such as: the number of workers, the number of consultation times, the number of complaints and the like, and the acquired statistical characteristics and the clients form a pivot table T1.
And taking the client ID as a grouping variable, acquiring physical characteristics of the client, such as: urban and rural categories (town, countryside, suburban area, and the like) and electricity utilization categories (civil, commercial, hospital, and the like), and the acquired physical characteristics and clients form a perspective table T2.
And with the client ID as a grouping variable, acquiring semantic features related to the client through machine learning, such as: and power failure sensitivity, user experience degree and the like, and the acquired semantic features and clients form a pivot table T3.
And combining the pivot table T1, the pivot table T2 and the pivot table T3 to obtain a feature FILE FILE of the data source.
Discretizing the numerical features in the feature FILE.
And performing correlation analysis on the variables, and removing the variables with strong correlation [ if the variables A and B are too correlated (vectors are parallel), the information for identifying the client characteristics A and B is redundant, for example, the total number of the work order receptions and the average number of the work order receptions are correlated, and only one variable can be taken as the identified variable. Obtaining a things database DBT.
The 95598 work order data S is used to construct a data source of a power supply decision knowledge graph, and includes: work order acceptance, telephone consultation, work order supervision, power failure event, fault handling and other information.
TABLE 1 work order acceptance information
TABLE 2 telephone advisory information
Taking the customer call information table in table 2 as an example, the total number of records 1220000 and the total number of customers 7000, the pivot table T1 is obtained by data analysis, wherein the pivot table T1 is too simplified for convenience of description of the embodiment, and the actual pivot table is much more complicated than table 3, and generally exceeds 100 columns.
TABLE 3 perspective table T1
| USERID | Power supply unit |
| 1000000 | Development area |
| 1000003 | Development area |
| 1000012 | Harbor gate area |
| 2000010 | Chongchuan district |
| 2000014 | Chongchuan district |
TABLE 4 perspective table T2
| USERID | Susceptibility to blackout |
| 1000000 | 1 |
| 1000003 | 1 |
| 1000012 | 0 |
| 2000010 | 1 |
| 2000014 | 1 |
TABLE 5 perspective table T3
TABLE 6 Profile FILE
TABLE7 transaction database DBT
The term "more" in table7 means that the number of the work orders is 3 times larger than the number of the average user, the number of the work orders is less than the number of the average user, and the other cases are normal, and since the number of the work orders is 4 pieces per the average user, the number of the work orders is more than 12, and the number of the work orders is less than 4 pieces.
2) Mining strong association rules: and mining all strong association rules in the service database DBT by using an association rule mining algorithm, wherein the specific mining steps of the strong association rules comprise:
1) Let DBT = { T l ,T 2 ,…T n In which T is i Is called a transaction T, where i takes on the value 1 k Called term, k takes the value 1.. Eta., m, the set of all terms in DBT is I = { I = { I } l ,i 2 ,…i m }，
2) Let A = { i = } A1 ,i A2 ,…i At 1 ≦ t ≦ m), where A is called a t-term set in the DBT;
3) Calculating item set support by formula (1)
4) Calculating the confidence coefficient of the association rule by formula (2), and setting the association rule
WhereinA and B are both sets of items in the DBT, anda is a condition of the association rule, and B is called a conclusion of the association rule.
5) The strong association rule is calculated by equation (3):
when in useAnd isThen the rule is associatedIs a strong association rule, otherwise is not a strong association rule, wherein A c The right part B of the strong association rule is a complement of A and is a dependent variable.
Taking the example of the transaction database DBT of table7, the total number of transactions | DBT | =5,1-item set = { { many }0.6, { few }0.2, { normal }0.2, { sensitive }0.8, { insensitive }0.2, { development area }0.4, { harbor gate area }0.2, { Chongchuan area }0.4}, the number following an item indicates support.
This yields 1-item set confidence, P (sensitive | more) =3/3=1, P (sensitive | less) =1/1=1, P (sensitive | more) =3/4=0.75, P (less sensitive) =1/4=0.25, P (sensitive | normal) =0/1=0, P (insensitive | normal) =1/1=1.
P (sensitive | multi) -P (multi) =1-0.6=0.4>, 0,P (sensitive | Dou |) c ) -P (poly) c )＝1/2-3/5＝-0.1<, 0, so a strong association rule { many } is obtained{ sensitive }.
P (sensitive | normal) -P (normal) =0-0.2= -0.2, so rule { normal }Sensitive is not a strongly associated rule.
P (insensitive | normal) -P (normal) =1-0.2=0.8, P (insensitive | multi) -P (multi) =0-0.6= -0.6, so a strongly associated rule { normal }results{ insensitive }.
Finally obtaining strong association rule { multiple }Sensitive and normal{ insensitive }, which shows that the power failure sensitivity is a dependent variable, and because the confidence coefficient and the support degree of the 2-item set which takes 'sensitive' and 'insensitive' as conclusions are both 1, the probability of the obtained strong association rule is less than or equal to 0, and no 2-item set strong association rule exists.
3) Constructing a power supply decision knowledge graph: extracting concepts, entities, attributes of entities and relations between the entities from the things database DBT in the step 1), wherein the concepts, the entities, the attributes of the entities and the relations between the entities are the things T and the item set I, T-the item set A and the item i k (1. Ltoreq. K. Ltoreq.m) and strong association rulesConstructing a data body according to the extracted concepts and entities, performing semantic analysis on the data body, extracting entity features to construct a feature body, performing association analysis on the feature body, extracting relationships among the entities to construct a client body, and further completing construction of a power supply decision knowledge graph, wherein the feature body is an item I k And the client ontology is a strong association rule.
And converting the relationships among the entities into a power supply decision knowledge graph by using a Cypher language based on Neo4 j.
When strong association rules existIf the nodes A and B exist, the execution is executed
If node A or B does not exist, assuming A does not exist and knowing A { i1, i2, …, ik }, CREATE (A: table { Varname1: i1, varname2: i2, …, varname k: ik }) is executed first and then
Taking an object database DBT in a table7 as an example, the construction process of the power supply decision knowledge graph is explained;
discovery of strongly associated rules (multiple){ sensitive }, thus performing: CREATE [:0.4 [ ]]->, (sensitive).
Attribute of discovery entity "10000000": "number of accepted work orders" and "power outage sensitivity", so table7 (number of accepted work orders:, power outage sensitivity: sensitivity) is executed.
Similar operations are performed for entities "10000003", "10000012", "20000010", "20000014" as well.
The knowledge graph corresponding to table7 is shown in fig. 2, and is stored in a graph form, and Neo4j is used in a concrete implementation manner, so that data for constructing the power supply decision knowledge graph is simplified, and a knowledge graph with a client as a center is embodied, so that more intuitive decision reference content can be provided for decision-making personnel applying the power supply decision knowledge graph.
Through the knowledge graph, the following can be identified: information such as service satisfaction, power failure sensitivity, user experience, customer value, customer classification and the like;
the power supply decision knowledge graph is constructed in real time because the power supply decision knowledge graph is based on the object database DBT which adopts an incremental generation technology;
suppose that a client 200032 adds 1 new acceptance sheet, the value of the attribute 'accepted work singular number' may be changed from 'normal' to 'much', and at this time, a query is made as to whether the client 200032 is sensitive to power failure, and the 'much-to-many' rule is strongly associated according to fig. 2Sensitive ", return to customer 200032 is sensitive to power outage.
Identifying which customers are difficult to accept new service patterns, only in the traditional business; what customers have low credit, high risk or malicious arrears may be "method as" blackout sensitivity analysis ", the key is to modify table 5.
See table 8 for a perspective to identify "which customers are difficult to accept the new service patterns, only in the traditional business":
TABLE 8 new perspective table T3
See table 9 for a perspective identifying "which customers have low credit, high risk, or the potential for malicious arrears":
TABLE 9 new perspective table T3
A power supply decision knowledge graph query method specifically comprises the following steps:
1) Receiving a query: performing word segmentation on a query sentence W input by a user by using a Chinese word segmentation system NLPIR of a Chinese academy to obtain W = { W1, W2,. And wn }, wherein wi (i =1,. And n) is a word or a character;
2) Query and analysis: mapping each vocabulary in the word segmentation result in the step 1) to four types of query elements to obtain an element pair of the query W, and recording the element pair as:<P，W>＝{<p1，w1>，<p2，w2>,..,<pn，wn>}，pi∈{NULL，A ik wherein i =1,.. N, NULL does not correspond to any element, and the four types of query elements are thing T, item set a, item i k And other parts, because the units of the divided words are small and are generally two-word words, and concepts, entities, attributes and relations can be compound words, elements are combined properly, the sequence is adjusted, and the semantics are normalized to form the compound words.
3) Forming a query pattern: forming a query pattern according to the analyzed item set, item set support degree and association rule confidence degree, wherein the query pattern of W is marked as X = f ({ A) ik })；
4) Displaying a knowledge graph: and executing a query mode through a knowledge graph reasoning engine based on the Neo4j Cypher language, and displaying the queried knowledge graph data to a decision maker.
Take "whether a customer sensitive to blackout is price sensitive" as an example
The segmentation result is W = { W1= pair, W2= power failure, W3= sensitive, W4= client, W5= client, W6= yes, W7= no, W8= price, W9= sensitive }, the segmentation result obtained after merging, adjusting the order and semantic normalization is W = { W1= pair, W2= power failure sensitive, W3= client, W4= client, W5= yes, W6= no, W7= price sensitive }.
And mapping each vocabulary in the word segmentation result into the three types of query elements, namely obtaining an element pair of query W, < P, W > = { < MULL, pair >, < A1, power failure sensitive >, < MULL, yes >, < MULL, client >, < MULL, yes >, < MULL, no >, < B1, price sensitive > }.
Forming a query pattern according to the analyzed item set, item set support and association rule confidence, wherein the query pattern of W is marked as X = f ({ A) ik })。
The query mode is executed through a knowledge graph reasoning engine of the Cypher language based on the Neo4j, and the Cypher query language can return results to decision-making personnel as long as the query mode is formed.
Compared with the prior art, the method comprises the steps of converting 95598 data sources into an object database, extracting concept, entity and entity attribute information for constructing the power supply decision knowledge map and relationship information among the entities by using an association rule mining algorithm, storing the power supply decision knowledge map by using a graph mode, inputting and analyzing received decision makers, establishing a query mode, displaying results returned by the query mode to the decision makers according to an inference engine, and providing more intuitive decision references for power supply decision makers, so that the power supply market occupancy is continuously improved, and the requirement of power supply enterprises for quick and accurate service of customers is met.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes performed by the present specification and drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (5)
1. A power supply decision knowledge graph construction method is characterized by comprising the following specific steps:
1) Constructing a transaction database DBT: the method comprises the steps that 95598 worksheet data S are used as a data source for constructing a power supply decision knowledge graph, and the data source is converted into an object database DBT by taking a client as a core concept;
2) Mining strong association rules: mining all strong association rules in the object database DBT by using an association rule mining algorithm;
3) Constructing a power supply decision knowledge graph: extracting concepts, entities, attributes of the entities and relations among the entities from the object database DBT in the step 1), constructing a data body according to the extracted concepts and entities, and carrying out processing on the data bodySemantic analysis, namely extracting entity features to construct a feature ontology, performing association analysis on the feature ontology, extracting relationships among entities to construct a client ontology, and further completing construction of a power supply decision knowledge graph, wherein the data ontology comprises an object T and an item set I, and the feature ontology is an item I k And the client ontology is a strong association rule.
2. The method for constructing a power supply decision knowledge graph according to claim 1, wherein the specific steps of step 1) are converting numerical variables into factor-type variables through discretization, taking factor levels as terms, creating pivot tables by using a client ID as an index of a data source to obtain a plurality of grouping statistics, and connecting tables obtained by using the client ID and related pivot tables to obtain a transaction database DBT.
3. The method for constructing a power supply decision knowledge graph according to claim 1, wherein the specific mining step of the strong association rule in the step 2) comprises:
1) Let DBT = { T l ,T 2 ,…T n Where T is i Called transaction T, where i takes on the value 1.., n, the element i that constitutes T k Called term, k takes the value 1.. Eta., m, the set of all terms in DBT is I = { I = { I } l ,i 2 ,…i m }，
2) Let A = { i = } A1 ,i A2 ,…i At 1 < t < m), wherein A is called a t-item set in DBT;
3) Calculating item set support by formula (1)
4) Calculating the confidence coefficient of the association rule by formula (2), and setting the association rule
Wherein A and B are both sets of terms in DBT, anda is the condition of the association rule and B is called the conclusion of the association rule.
5) The strong association rule is calculated by equation (3):
when the temperature is higher than the set temperatureAnd isThen the rule is associatedIs a strong association rule, otherwise is not a strong association rule, wherein A c The right part B of the strong association rule is a complement of A and is a dependent variable.
4. The method for constructing power supply decision knowledge graph according to claim 1, wherein the concepts, entities, attributes of the entities and the relationships among the entities in the step 3) are things T and item sets I, T-item set A and item i, respectively k (1. Ltoreq. K. Ltoreq.m) and strong association rules
5. A power supply decision knowledge graph query method is characterized by comprising the following specific steps:
1) Receiving a query: performing word segmentation on a query sentence W input by a user by using a Chinese word segmentation system NLPIR of a Chinese academy to obtain W = { W1, W2,. And wn }, wherein wi (i =1,. And n) is a word or a character;
2) Query and analysis: mapping each vocabulary in the word segmentation result in the step 1) to four types of query elements to obtain an element pair of the query W, and recording the element pair as:<P，W>＝{<p1，w1>，<p2，w2>,..,<pn，wn>}，pi∈{NULL，A ik wherein i =1,.. N, NULL does not correspond to any element, and the four types of query elements are thing T, item set a, item i k And others;
3) Forming a query pattern: forming a query pattern according to the analyzed item set, item set support degree and association rule confidence degree, wherein the query pattern of W is marked as X = f ({ A) ik })；
4) Displaying a knowledge graph: and executing a query mode through a knowledge graph reasoning engine based on the Cypher language of the Neo4j, and displaying the queried knowledge graph data to a decision maker.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710934931.2A CN107633093A (en) | 2017-10-10 | 2017-10-10 | A kind of structure and its querying method of DECISION KNOWLEDGE collection of illustrative plates of powering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710934931.2A CN107633093A (en) | 2017-10-10 | 2017-10-10 | A kind of structure and its querying method of DECISION KNOWLEDGE collection of illustrative plates of powering |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107633093A true CN107633093A (en) | 2018-01-26 |
Family
ID=61104524
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710934931.2A Pending CN107633093A (en) | 2017-10-10 | 2017-10-10 | A kind of structure and its querying method of DECISION KNOWLEDGE collection of illustrative plates of powering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107633093A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108182295A (en) * | 2018-02-09 | 2018-06-19 | 重庆誉存大数据科技有限公司 | A kind of Company Knowledge collection of illustrative plates attribute extraction method and system |
| CN108460136A (en) * | 2018-03-08 | 2018-08-28 | 国网福建省电力有限公司 | Electric power O＆M information knowledge map construction method |
| CN110096598A (en) * | 2019-04-25 | 2019-08-06 | 广州供电局有限公司 | Power distribution network knowledge mapping method for building up, device, computer equipment and storage medium |
| CN110222127A (en) * | 2019-06-06 | 2019-09-10 | 中国电子科技集团公司第二十八研究所 | The converging information method, apparatus and equipment of knowledge based map |
| CN110457403A (en) * | 2019-08-12 | 2019-11-15 | 南京星火技术有限公司 | The construction method of figure network decision system, method and knowledge mapping |
| CN110543951A (en) * | 2018-05-28 | 2019-12-06 | 中国铁道科学研究院铁道建筑研究所 | Virtual assistant system for maintenance of railway bridge |
| CN111026883A (en) * | 2019-12-11 | 2020-04-17 | 南方电网数字电网研究院有限公司 | Knowledge graph construction method, device, equipment and medium of power business data |
| CN116680445A (en) * | 2023-05-05 | 2023-09-01 | 国网吉林省电力有限公司 | Knowledge-graph-based multi-source heterogeneous data fusion method and system for electric power optical communication system |
| CN117313855A (en) * | 2023-11-28 | 2023-12-29 | 支付宝(杭州)信息技术有限公司 | Rule decision method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103871003A (en) * | 2014-03-31 | 2014-06-18 | 国家电网公司 | Power distribution network fault diagnosis method utilizing historical fault data |
| CN104866593A (en) * | 2015-05-29 | 2015-08-26 | 中国电子科技集团公司第二十八研究所 | Database searching method based on knowledge graph |
| CN105868313A (en) * | 2016-03-25 | 2016-08-17 | 浙江大学 | Mapping knowledge domain questioning and answering system and method based on template matching technique |
| CN106447346A (en) * | 2016-08-29 | 2017-02-22 | 北京中电普华信息技术有限公司 | Method and system for construction of intelligent electric power customer service system |
-
2017
- 2017-10-10 CN CN201710934931.2A patent/CN107633093A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103871003A (en) * | 2014-03-31 | 2014-06-18 | 国家电网公司 | Power distribution network fault diagnosis method utilizing historical fault data |
| CN104866593A (en) * | 2015-05-29 | 2015-08-26 | 中国电子科技集团公司第二十八研究所 | Database searching method based on knowledge graph |
| CN105868313A (en) * | 2016-03-25 | 2016-08-17 | 浙江大学 | Mapping knowledge domain questioning and answering system and method based on template matching technique |
| CN106447346A (en) * | 2016-08-29 | 2017-02-22 | 北京中电普华信息技术有限公司 | Method and system for construction of intelligent electric power customer service system |
Non-Patent Citations (1)
| Title |
|---|
| 田晓 等: ""电网公司客户服务知识图谱构建的应用价值"", 《山东电力技术》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108182295B (en) * | 2018-02-09 | 2021-09-10 | 重庆电信系统集成有限公司 | Enterprise knowledge graph attribute extraction method and system |
| CN108182295A (en) * | 2018-02-09 | 2018-06-19 | 重庆誉存大数据科技有限公司 | A kind of Company Knowledge collection of illustrative plates attribute extraction method and system |
| CN108460136A (en) * | 2018-03-08 | 2018-08-28 | 国网福建省电力有限公司 | Electric power O＆M information knowledge map construction method |
| CN110543951B (en) * | 2018-05-28 | 2022-05-17 | 中国铁道科学研究院铁道建筑研究所 | Virtual assistant system for maintenance of railway bridge |
| CN110543951A (en) * | 2018-05-28 | 2019-12-06 | 中国铁道科学研究院铁道建筑研究所 | Virtual assistant system for maintenance of railway bridge |
| CN110096598A (en) * | 2019-04-25 | 2019-08-06 | 广州供电局有限公司 | Power distribution network knowledge mapping method for building up, device, computer equipment and storage medium |
| CN110222127A (en) * | 2019-06-06 | 2019-09-10 | 中国电子科技集团公司第二十八研究所 | The converging information method, apparatus and equipment of knowledge based map |
| CN110457403B (en) * | 2019-08-12 | 2022-04-22 | 南京星火技术有限公司 | Graph network decision system and method and knowledge graph construction method |
| CN110457403A (en) * | 2019-08-12 | 2019-11-15 | 南京星火技术有限公司 | The construction method of figure network decision system, method and knowledge mapping |
| CN111026883A (en) * | 2019-12-11 | 2020-04-17 | 南方电网数字电网研究院有限公司 | Knowledge graph construction method, device, equipment and medium of power business data |
| CN116680445A (en) * | 2023-05-05 | 2023-09-01 | 国网吉林省电力有限公司 | Knowledge-graph-based multi-source heterogeneous data fusion method and system for electric power optical communication system |
| CN117313855A (en) * | 2023-11-28 | 2023-12-29 | 支付宝(杭州)信息技术有限公司 | Rule decision method and device |
| CN117313855B (en) * | 2023-11-28 | 2024-03-15 | 支付宝(杭州)信息技术有限公司 | Rule decision method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107633093A (en) | A kind of structure and its querying method of DECISION KNOWLEDGE collection of illustrative plates of powering | |
| Carley et al. | ORA user’s guide 2013 | |
| Moges et al. | A multidimensional analysis of data quality for credit risk management: New insights and challenges | |
| El Morr et al. | Descriptive, predictive, and prescriptive analytics | |
| CN103226554A (en) | Automatic stock matching and classifying method and system based on news data | |
| Dwivedi et al. | Exploring artificial intelligence and big data scholarship in information systems: A citation, bibliographic coupling, and co-word analysis | |
| Wongthongtham et al. | Ontology and trust based data warehouse in new generation of business intelligence: State-of-the-art, challenges, and opportunities | |
| Wu et al. | Fuzzy multiattribute grey related analysis using DEA | |
| Aviad et al. | A decision support method, based on bounded rationality concepts, to reveal feature saliency in clustering problems | |
| Ridzuan et al. | Diagnostic analysis for outlier detection in big data analytics | |
| Ruiz et al. | Information fusion from multiple databases using meta-association rules | |
| Wang et al. | A knowledge graph–GCN–community detection integrated model for large-scale stock price prediction | |
| Fernando et al. | Empirical analysis of data mining techniques for social network websites | |
| Bakariya et al. | An efficient algorithm for extracting infrequent itemsets from weblog. | |
| Sumangali et al. | Determination of interesting rules in FCA using information gain | |
| Greenberg | Disclosure avoidance research at the census bureau | |
| Izadikhah | Modelling bank performance: a novel fuzzy two-stage DEA approach | |
| Khatun et al. | Machine Learning based Advanced Crime Prediction and Analysis | |
| Liang et al. | Distributed outlier detection in hierarchically structured datasets with mixed attributes | |
| Kasinadh et al. | Building fuzzy OLAP using multi-attribute summarization | |
| Wisnuwardhana et al. | Systematic Literature Review: Critical Success Factor in the Application of Data Mining | |
| Yasami | Anomaly Detection in Dynamic Complex Networks | |
| Balamane et al. | Descriptive group detection in two-mode data networks using biclustering | |
| Vanisri et al. | Fuzzy pattern cluster scheme for breast cancer datasets | |
| Signature | Signature. ca |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180126 |
|
| RJ01 | Rejection of invention patent application after publication |