CN107066541A - The processing method and system of customer service question and answer data - Google Patents
The processing method and system of customer service question and answer data Download PDFInfo
- Publication number
- CN107066541A CN107066541A CN201710147702.6A CN201710147702A CN107066541A CN 107066541 A CN107066541 A CN 107066541A CN 201710147702 A CN201710147702 A CN 201710147702A CN 107066541 A CN107066541 A CN 107066541A
- Authority
- CN
- China
- Prior art keywords
- answer
- question
- sentence
- customer service
- log
- 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
- 238000003672 processing method Methods 0.000 title claims abstract description 17
- 238000012545 processing Methods 0.000 claims abstract description 30
- 238000004458 analytical method Methods 0.000 claims description 52
- 230000011218 segmentation Effects 0.000 claims description 6
- 238000000034 method Methods 0.000 description 11
- 238000012163 sequencing technique Methods 0.000 description 4
- 238000010276 construction Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000000638 solvent extraction Methods 0.000 description 2
- 206010044565 Tremor Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
Abstract
The present invention relates to a kind of processing method and system of customer service question and answer data, the processing method of the customer service question and answer data includes：The processing system of customer service question and answer data obtains the Question and Answer log prestored from customer service Q ＆ A database, and the question and answer corresponding relation based on customer service in preset time and client is grouped to the Question and Answer log；To the Question and Answer log in each packet according to it is predetermined the problem of matched with the matched rule of answer, with draw it is to be added into service knowledge base the problem of answer pair.The Question and Answer log Auto-matching based on customer service Q ＆ A database of the invention draws problem answers pair, it is not necessary to which contact staff is write, the broad covered areas of problem answers pair, structure efficiency high and renewal speed is fast.
Description
Technical field
The present invention relates to computer software technical field, more particularly to a kind of customer service question and answer data processing method and be
System.
Background technology
At present, customer service automatically request-answering system is generally based on problem answers to being designed as knowledge base, according to
The problem of the problem of family, retrieves most like in knowledge base, further according to existing problem answers in knowledge base to (i.e. problem is with answering
The mapping of case to), directly give answer.Under this system design framework, there are three essential cores：Problem understands, retrieval is calculated
Method and construction of knowledge base.For knowledge base, traditional construction method is usually by possessing the contact staff of professional experiences for many years
Manually write, the weak point that this mode for building knowledge base is present is：The coverage rates of problem answers pair is narrow, build effect
Rate is low and renewal speed is slow.
The content of the invention
It is an object of the invention to provide a kind of processing method and system of customer service question and answer data, it is intended to based on customer service question and answer
The Question and Answer log Auto-matching of database draws problem answers pair, and the broad covered area of problem answers pair, builds efficiency high and more
New speed is fast.
To achieve the above object, the present invention provides a kind of processing method of customer service question and answer data, the customer service question and answer data
Processing method include：
S1, the processing system of customer service question and answer data obtains the Question and Answer log prestored from customer service Q ＆ A database, and based on pre-
If the question and answer corresponding relation of customer service and client are grouped to the Question and Answer log in the time；
S2, to the Question and Answer log in each packet according to it is predetermined the problem of matched with the matched rule of answer, to draw
It is to be added into service knowledge base the problem of answer pair.
Preferably, the step S1 includes：
The Question and Answer log of same client in preset time and same customer service is divided into same packet；Or
The Question and Answer log of same client in preset time and each customer service is divided into same packet；Or
The Question and Answer log of each client in preset time and same customer service is divided into same packet.
Preferably, the step S2 includes：
S21, data prediction is carried out to the Question and Answer log in each packet；
S22, sentence and the answer statement of customer service are according to pre- the problem of to client in the Question and Answer log after data prediction
If analysis rule analyzed, sentence and corresponding answer statement the problem of to analyze to be matched；
S23, the problem of extracting to be matched sentence and corresponding answer statement, by sentence the problem of extracting and corresponding time
Answer sentence and be associated as problem answers pair and by problem answers to adding into service knowledge base.
Preferably, the step S21 includes：
Font in Question and Answer log in each packet is uniformly converted into the complex form of Chinese characters or simplified Chinese character, by each packet
Question and Answer log in the unified upper of letter or lowercase, the particular statement that removes in Question and Answer log, with
And word segmentation processing is carried out to the sentence in the Question and Answer log in each packet.
Preferably, the default analysis rule includes the analysis rule of problem sentence and the analysis rule of answer statement,
The analysis rule of described problem sentence is：
The key index value S of computational problem sentence：
S=Sa+Sb+Sc,Wherein, l is the length of problem sentence, m
For the number of the business keyword contained by problem sentence, n is the number of the stop words of problem sentence, and a is the first of problem sentence
Length preset value, b is the second length preset value of problem sentence；
If key index value S be more than pre-set level threshold value, retain described problem sentence as it is to be matched the problem of sentence,
Or, if key index value S is less than or equal to pre-set level threshold value, remove described problem sentence；
The analysis rule of the answer statement is：If answer statement is not declarative sentence, the answer statement is removed
And its sentence the problem of correspondence.
To achieve the above object, the present invention also provides a kind of processing system of customer service question and answer data, the customer service question and answer number
According to processing system include：
Grouping module, for obtaining the Question and Answer log prestored from customer service Q ＆ A database, and based on customer service in preset time
The Question and Answer log is grouped with the question and answer corresponding relation of client；
Matching module, for the Question and Answer log in each packet according to it is predetermined the problem of and answer matched rule carry out
Match somebody with somebody, with draw it is to be added into service knowledge base the problem of answer pair.
Preferably, the grouping module is specifically for by the Question and Answer log of the same client in preset time and same customer service
It is divided into same packet；Or the Question and Answer log of the same client in preset time and each customer service is divided into same packet；Or will be default
Each client and the Question and Answer log of same customer service in time are divided into same packet.
Preferably, the matching module includes：
Pretreatment unit, for carrying out data prediction to the Question and Answer log in each packet；
Analytic unit, sentence and the answer language of customer service the problem of for client in the Question and Answer log after data prediction
Sentence is analyzed according to default analysis rule, sentence and corresponding answer statement the problem of to analyze to be matched；
Associative cell, sentence and corresponding answer statement the problem of for extracting to be matched, by sentence the problem of extracting
Problem answers pair are associated as with corresponding answer statement and by problem answers to adding into service knowledge base.
Preferably, the pretreatment unit is specifically for the font in the Question and Answer log in each packet is uniformly converted to
The complex form of Chinese characters or simplified Chinese character, by the unified upper of letter in the Question and Answer log in each packet or small letter
Particular statement that is female, removing in Question and Answer log, and word segmentation processing is carried out to the sentence in the Question and Answer log in each packet.
Preferably, the default analysis rule includes the analysis rule of problem sentence and the analysis rule of answer statement,
The analysis rule of described problem sentence is：
The key index value S of computational problem sentence：
S=Sa+Sb+Sc,Wherein, l is the length of problem sentence, m
For the number of the business keyword contained by problem sentence, n is the number of the stop words of problem sentence, and a is the first of problem sentence
Length preset value, b is the second length preset value of problem sentence；
If key index value S be more than pre-set level threshold value, retain described problem sentence as it is to be matched the problem of sentence,
Or, if key index value S is less than or equal to pre-set level threshold value, remove described problem sentence；
The analysis rule of the answer statement is：If answer statement is not declarative sentence, the answer statement is removed
And its sentence the problem of correspondence.
The beneficial effects of the invention are as follows：The present invention is when building service knowledge base, by the processing system of customer service question and answer data
Automatically Question and Answer log is obtained from customer service Q ＆ A database, Question and Answer log is grouped so that the question and answer day in same packet
Will has some identical attributes, in each packet with some same alike results Question and Answer log according to it is predetermined the problem of with
The matched rule of answer is matched, and draws to be added to service knowledge base problem answers pair to match, the present invention is based on customer service
The Question and Answer log Auto-matching of Q ＆ A database draws problem answers pair, it is not necessary to which contact staff is write, problem answers pair
Broad covered area, build efficiency high and renewal speed is fast.
Brief description of the drawings
Fig. 1 is the schematic flow sheet of the embodiment of processing method one of customer service question and answer data of the present invention；
Fig. 2 is the refinement schematic flow sheet of step S2 shown in Fig. 1；
Fig. 3 unifies the structural representation of embodiment for the processing system of customer service question and answer data of the present invention；
Fig. 4 is the structural representation of matching module shown in Fig. 3.
Embodiment
The principle and feature of the present invention are described below in conjunction with accompanying drawing, the given examples are served only to explain the present invention, and
It is non-to be used to limit the scope of the present invention.
As shown in figure 1, schematic flow sheets of the Fig. 1 for the embodiment of processing method one of customer service question and answer data of the present invention, the visitor
The processing method for taking question and answer data comprises the following steps：
Step S1, the processing system of customer service question and answer data obtains the Question and Answer log prestored, and base from customer service Q ＆ A database
The Question and Answer log is grouped in the question and answer corresponding relation of customer service in preset time and client；
In the present embodiment, it is stored with what is produced in the customer service of all people's work and client's dialog procedure in customer service Q ＆ A database
Question and Answer log, processing system and the customer service Q ＆ A database of customer service question and answer data are communicated to connect, the processing system of customer service question and answer data
System can obtain Question and Answer log from customer service Q ＆ A database automatically, to be further processed.
In the present embodiment, the processing system of customer service question and answer data can obtain question and answer from customer service Q ＆ A database in real time
Daily record, for example, obtain newly-increased customer service and the Question and Answer log of client in real time；Can also be periodically from customer service Q ＆ A database
Question and Answer log is obtained, for example, obtained Question and Answer log from customer service Q ＆ A database every 6 hours.By obtaining in real time or periodically
Question and Answer log in customer service Q ＆ A database, to obtain newest Question and Answer log.
Then, question and answer corresponding relation of the Question and Answer log based on customer service in preset time and client of acquisition is grouped,
For example, in one day on the day of a certain customer service in carry out question answer dialog with k client, then will on the day of the customer service in k client's progress
The corresponding Question and Answer log of question answer dialog is divided into one group.The main purpose being grouped to Question and Answer log is will there is common category
Property Question and Answer log be classified as one group, follow-up carry out Treatment Analysis to facilitate.
Step S2, to the Question and Answer log in each packet according to it is predetermined the problem of matched with the matched rule of answer, with
Draw it is to be added into service knowledge base the problem of answer pair.
In the present embodiment, to the Question and Answer log in each packet according to it is predetermined the problem of and answer matched rule carry out
Match somebody with somebody, the problem of for example first being extracted according to the time of enquirement and the sequencing of time of answer in Question and Answer log sentence and answer
Length in sentence, problem analysis sentence and included business keyword etc., the length and business for analyzing answer statement are crucial
Whether word is corresponding with the business keyword in problem sentence etc., according to above-mentioned analysis result by problem sentence and answer statement
Matched；And for example, the problem of first being extracted according to the time of enquirement and the sequencing of time of answer in Question and Answer log language
Sentence and answer statement, business keyword in problem analysis sentence and are extended to business keyword, and analysis answer statement is
No corresponding with the business keyword or after being extended with the business keyword word is corresponding, according to above-mentioned analysis result
Problem sentence and answer statement progress are matched, etc..Certainly, the present embodiment according to it is predetermined the problem of and answer matched rule
Matched and do not limit both above-mentioned modes, other can be matched problem sentence and answer statement to complete pairing
Mode also in the protection domain of the present embodiment.
In the present embodiment, answer is gone wrong to after to match problem sentence is matched with answer statement, will be asked
Answer is inscribed to adding into service knowledge base, to be constantly increasing or enrich the knowledge in service knowledge base.
Compared with prior art, the present embodiment build service knowledge base when, by customer service question and answer data processing system from
It is dynamic to obtain Question and Answer log from customer service Q ＆ A database, Question and Answer log is grouped so that the Question and Answer log in same packet
With some identical attributes, for the Question and Answer log with some same alike results in each packet according to it is predetermined the problem of with answering
The matched rule of case is matched, and draws to be added to service knowledge base problem answers pair to match, the present embodiment is based on customer service
The Question and Answer log Auto-matching of Q ＆ A database draws problem answers pair, it is not necessary to which contact staff is write, problem answers pair
Broad covered area, build efficiency high and renewal speed is fast.
In a preferred embodiment, on the basis of above-mentioned Fig. 1 embodiment, above-mentioned steps S1 includes：Will be same pre-
If same client in the time and the Question and Answer log of same customer service are divided into same packet, for example with preset time (such as one day or
Person one week) it is partitioning standards, the Question and Answer log of same client interior on the same day and same customer service is divided into same packet, for example, asked
Answering daily record includes：Client A and customer service a Question and Answer log 1, client A and customer service a Question and Answer log 2, client A and customer service b question and answer
Daily record 3 ,/customer service B and client a Question and Answer log 4, customer service B and client b Question and Answer log 5, then be one by Question and Answer log 1 and 2 points
Group, other Question and Answer logs 3,4 and 5 are respectively one group；Or
The Question and Answer log of same client in same preset time and each customer service is divided into same packet, for example will be above-mentioned
Question and Answer log 1,2 and 3 points be one group, be one group by Question and Answer log 4 and 5 points；Or
The Question and Answer log of each client in preset time and same customer service is divided into same packet, such as by above-mentioned question and answer
Daily record 1,2 and 4 points be one group, be one group by Question and Answer log 3 and 5 points.Certain the present embodiment is not limited to above-mentioned packet mode.
Question and Answer log of the present embodiment after packet has some identical attributes, such as same client may put question to phase
As problem, it puts question to sentence the characteristics of have similar, or same customer service is the problem of may answer similar, its answer statement tool
The characteristics of having similar, the Question and Answer log with identical attribute can provide facility for follow-up further analysis.
In a preferred embodiment, as shown in Fig. 2 on the basis of above-mentioned Fig. 1 embodiment, above-mentioned steps S2 bags
Include：
S21, data prediction is carried out to the Question and Answer log in each packet；
S22, sentence and the answer statement of customer service are according to pre- the problem of to client in the Question and Answer log after data prediction
If analysis rule analyzed, sentence and corresponding answer statement the problem of to analyze to be matched；
S23, the problem of extracting to be matched sentence and corresponding answer statement, by sentence the problem of extracting and corresponding time
Answer sentence and be associated as problem answers pair and by problem answers to adding into service knowledge base.
In the present embodiment, data prediction is carried out to the Question and Answer log in each packet, it is preferable that carry out data prediction
Including：Font in Question and Answer log in each packet is uniformly converted into the complex form of Chinese characters or simplified Chinese character, by each packet
The particular statement in the unified upper of letter or lowercase, removal Question and Answer log in Question and Answer log, and
Word segmentation processing etc. is carried out to the sentence in the Question and Answer log in each packet.Particular statement includes the sentence of repetition, commonly uses and tremble with fear
Warm and sunny language (such as commonly using some term of courtesy " hello ", " thanks ")；Sentence in Question and Answer log is carried out at participle
Reason can use the segmenting method of following three major types to perform：Segmenting method based on string matching, the participle side based on understanding
Method and the segmenting method based on statistics.
Then, to the client in the Question and Answer log after data prediction the problem of sentence and customer service answer statement according to pre-
If analysis rule analyzed, it is preferable that the analysis rule of default analysis rule including problem sentence and answer statement
Analysis rule, the analysis rule of described problem sentence is：
The key index value S of computational problem sentence：
S=Sa+Sb+Sc,Wherein, l is the length of problem sentence,
Length l can be measured with the word number after participle, i.e. word number more at most length l is bigger, and m is contained by problem sentence
Business keyword number, m is more big, and then key index value S is bigger, and n is the number of the stop words of problem sentence, and n is more big, is closed
Key desired value S is smaller, and problem quality score is lower, and a is the first length preset value of problem sentence, and b is the second of problem sentence
Length preset value, when length l is within [a, b], the moderate length of the problem sentence, problem quality score is high；Length l [a,
B] outside when, differ it is more, problem quality score is lower.
If key index value S is more than pre-set level threshold value, such as S>0, then retain described problem sentence as to be matched
Problem sentence, or, if key index value S is less than or equal to pre-set level threshold value, such as S<0, then remove described problem language
Sentence.
Preferably, the analysis rule of answer statement is：If answer statement is not declarative sentence, the answer statement is removed
And its sentence the problem of correspondence.
In addition, the problem of analyzing to be matched sentence with after corresponding answer statement, extract it is to be matched the problem of language
When sentence is with corresponding answer statement, sentence and corresponding answer statement the problem of if repeating, then only extraction repeatedly the problem of
Sentence and the problem sentence and corresponding answer statement in corresponding answer statement；And, the problem of will be to be matched
Sentence is arranged from large to small with the key business word weight that corresponding answer statement is included according to it, is extracted in queue and is closed
Key business word weight be more than a certain predetermined threshold value it is to be matched the problem of sentence and corresponding answer statement, by what is extracted
Problem sentence is associated as problem answers pair with corresponding answer statement, by problem answers to adding into service knowledge base.
As shown in figure 3, Fig. 3 unifies the structural representation of embodiment, the visitor for the processing system of customer service question and answer data of the present invention
Taking the processing system of question and answer data includes：
Grouping module 101, for obtaining the Question and Answer log prestored from customer service Q ＆ A database, and based on visitor in preset time
The question and answer corresponding relation taken with client is grouped to the Question and Answer log；
In the present embodiment, it is stored with what is produced in the customer service of all people's work and client's dialog procedure in customer service Q ＆ A database
Question and Answer log, processing system and the customer service Q ＆ A database of customer service question and answer data are communicated to connect, the processing system of customer service question and answer data
System can obtain Question and Answer log from customer service Q ＆ A database automatically, to be further processed.
In the present embodiment, the processing system of customer service question and answer data can obtain question and answer from customer service Q ＆ A database in real time
Daily record, for example, obtain newly-increased customer service and the Question and Answer log of client in real time；Can also be periodically from customer service Q ＆ A database
Question and Answer log is obtained, for example, obtained Question and Answer log from customer service Q ＆ A database every 6 hours.By obtaining in real time or periodically
Question and Answer log in customer service Q ＆ A database, to obtain newest Question and Answer log.
Then, question and answer corresponding relation of the Question and Answer log based on customer service in preset time and client of acquisition is grouped,
For example, in one day on the day of a certain customer service in carry out question answer dialog with k client, then will on the day of the customer service in k client's progress
The corresponding Question and Answer log of question answer dialog is divided into one group.The main purpose being grouped to Question and Answer log is will there is common category
Property Question and Answer log be classified as one group, follow-up carry out Treatment Analysis to facilitate.
Matching module 102, for the Question and Answer log in each packet according to it is predetermined the problem of and the matched rule of answer enter
Row matching, with draw it is to be added into service knowledge base the problem of answer pair.
In the present embodiment, to the Question and Answer log in each packet according to it is predetermined the problem of and answer matched rule carry out
Match somebody with somebody, the problem of for example first being extracted according to the time of enquirement and the sequencing of time of answer in Question and Answer log sentence and answer
Length in sentence, problem analysis sentence and included business keyword etc., the length and business for analyzing answer statement are crucial
Whether word is corresponding with the business keyword in problem sentence etc., according to above-mentioned analysis result by problem sentence and answer statement
Matched；And for example, the problem of first being extracted according to the time of enquirement and the sequencing of time of answer in Question and Answer log language
Sentence and answer statement, business keyword in problem analysis sentence and are extended to business keyword, and analysis answer statement is
No corresponding with the business keyword or after being extended with the business keyword word is corresponding, according to above-mentioned analysis result
Problem sentence and answer statement progress are matched, etc..Certainly, the present embodiment according to it is predetermined the problem of and answer matched rule
Matched and do not limit both above-mentioned modes, other can be matched problem sentence and answer statement to complete pairing
Mode also in the protection domain of the present embodiment.
In the present embodiment, answer is gone wrong to after to match problem sentence is matched with answer statement, will be asked
Answer is inscribed to adding into service knowledge base, to be constantly increasing or enrich the knowledge in service knowledge base.
In a preferred embodiment, on the basis of above-mentioned Fig. 1 embodiment, the grouping module 101 specifically for
The Question and Answer log of same client in same preset time and same customer service is divided into same packet, such as with preset time (example
Such as one day or one week) it is partitioning standards, the Question and Answer log of same client interior on the same day and same customer service is divided into same point
Group, such as Question and Answer log include：Client A and customer service a Question and Answer log 1, client A and customer service a Question and Answer log 2, client A and visitor
B Question and Answer log 3 ,/customer service B and client a Question and Answer log 4, customer service B and client b Question and Answer log 5 are taken, then by Question and Answer log 1
And 2 points are one group, other Question and Answer logs 3,4 and 5 are respectively one group；Or
The Question and Answer log of same client in same preset time and each customer service is divided into same packet, for example will be above-mentioned
Question and Answer log 1,2 and 3 points be one group, be one group by Question and Answer log 4 and 5 points；Or
The Question and Answer log of each client in preset time and same customer service is divided into same packet, such as by above-mentioned question and answer
Daily record 1,2 and 4 points be one group, be one group by Question and Answer log 3 and 5 points.Certain the present embodiment is not limited to above-mentioned packet mode.
Question and Answer log of the present embodiment after packet has some identical attributes, such as same client may put question to phase
As problem, it puts question to sentence the characteristics of have similar, or same customer service is the problem of may answer similar, its answer statement tool
The characteristics of having similar, the Question and Answer log with identical attribute can provide facility for follow-up further analysis.
In a preferred embodiment, as shown in figure 4, on the basis of above-mentioned Fig. 3 embodiment, matching module 102 is wrapped
Include：
Pretreatment unit 1021, for carrying out data prediction to the Question and Answer log in each packet；
Analytic unit 1022, sentence and customer service returns the problem of for client in the Question and Answer log after data prediction
Answer sentence to be analyzed according to default analysis rule, sentence and corresponding answer statement the problem of to analyze to be matched；
Associative cell 1023, sentence and corresponding answer statement the problem of for extracting to be matched, the problem of by extracting
Sentence is associated as problem answers pair and by problem answers to adding into service knowledge base with corresponding answer statement.
In the present embodiment, data prediction is carried out to the Question and Answer log in each packet, it is preferable that pretreatment unit 1021
Specifically for the font in the Question and Answer log in each packet is uniformly converted into the complex form of Chinese characters or simplified Chinese character, by each packet
Question and Answer log in the unified upper of letter or lowercase, the particular statement that removes in Question and Answer log, with
And word segmentation processing etc. is carried out to the sentence in the Question and Answer log in each packet.Particular statement includes the sentence of repetition, commonly used
Greeting language (such as commonly using some term of courtesy " hello ", " thanks ")；Participle is carried out to the sentence in Question and Answer log
Processing can use the segmenting method of following three major types to perform：Segmenting method based on string matching, the participle based on understanding
Method and the segmenting method based on statistics.
Then, to the client in the Question and Answer log after data prediction the problem of sentence and customer service answer statement according to pre-
If analysis rule analyzed, it is preferable that the analysis rule of default analysis rule including problem sentence and answer statement
Analysis rule, the analysis rule of described problem sentence is：
The key index value S of computational problem sentence：
S=Sa+Sb+Sc,Wherein, l is the length of problem sentence,
Length l can be measured with the word number after participle, i.e. word number more at most length l is bigger, and m is contained by problem sentence
Business keyword number, m is more big, and then key index value S is bigger, and n is the number of the stop words of problem sentence, and n is more big, is closed
Key desired value S is smaller, and problem quality score is lower, and a is the first length preset value of problem sentence, and b is the second of problem sentence
Length preset value, when length l is within [a, b], the moderate length of the problem sentence, problem quality score is high；Length l [a,
B] outside when, differ it is more, problem quality score is lower.
If key index value S is more than pre-set level threshold value, such as S>0, then retain described problem sentence as to be matched
Problem sentence, or, if key index value S is less than or equal to pre-set level threshold value, such as S<0, then remove described problem language
Sentence.
Preferably, the analysis rule of answer statement is：If answer statement is not declarative sentence, the answer statement is removed
And its sentence the problem of correspondence.
In addition, the problem of analyzing to be matched sentence with after corresponding answer statement, extract it is to be matched the problem of language
When sentence is with corresponding answer statement, sentence and corresponding answer statement the problem of if repeating, then only extraction repeatedly the problem of
Sentence and the problem sentence and corresponding answer statement in corresponding answer statement；And, the problem of will be to be matched
Sentence is arranged from large to small with the key business word weight that corresponding answer statement is included according to it, is extracted in queue and is closed
Key business word weight be more than a certain predetermined threshold value it is to be matched the problem of sentence and corresponding answer statement, by what is extracted
Problem sentence is associated as problem answers pair with corresponding answer statement, by problem answers to adding into service knowledge base.
The foregoing is only presently preferred embodiments of the present invention, be not intended to limit the invention, it is all the present invention spirit and
Within principle, any modification, equivalent substitution and improvements made etc. should be included in the scope of the protection.
Claims (10)
1. a kind of processing method of customer service question and answer data, it is characterised in that the processing method of the customer service question and answer data includes：
S1, the processing system of customer service question and answer data obtains the Question and Answer log prestored from customer service Q ＆ A database, and during based on presetting
The question and answer corresponding relation of interior customer service and client are grouped to the Question and Answer log；
S2, to the Question and Answer log in each packet according to it is predetermined the problem of matched with the matched rule of answer, it is to be added to draw
The problem of enter'sing into service knowledge base answer pair.
2. the processing method of customer service question and answer data according to claim 1, it is characterised in that the step S1 includes：
The Question and Answer log of same client in preset time and same customer service is divided into same packet；Or
The Question and Answer log of same client in preset time and each customer service is divided into same packet；Or
The Question and Answer log of each client in preset time and same customer service is divided into same packet.
3. the processing method of customer service question and answer data according to claim 1 or 2, it is characterised in that the step S2 includes：
S21, data prediction is carried out to the Question and Answer log in each packet；
S22, sentence and the answer statement of customer service are according to default the problem of to client in the Question and Answer log after data prediction
Analysis rule is analyzed, sentence and corresponding answer statement the problem of to analyze to be matched；
S23, the problem of extracting to be matched sentence and corresponding answer statement, by sentence the problem of extracting and corresponding answer language
Sentence is associated as problem answers pair and by problem answers to adding into service knowledge base.
4. the processing method of customer service question and answer data according to claim 3, it is characterised in that the step S21 includes：
Font in Question and Answer log in each packet is uniformly converted into the complex form of Chinese characters or simplified Chinese character, by asking in each packet
The unified upper of the letter in daily record or lowercase, the particular statement removed in Question and Answer log are answered, and it is right
Sentence in Question and Answer log in each packet carries out word segmentation processing.
5. the processing method of customer service question and answer data according to claim 3, it is characterised in that the default analysis rule
The analysis rule of analysis rule and answer statement including problem sentence, the analysis rule of described problem sentence is：
The key index value S of computational problem sentence：
S=Sa+Sb+Sc,Wherein, l is the length of problem sentence, and m is to ask
The number of the business keyword contained by sentence is inscribed, n is the number of the stop words of problem sentence, and a is the first length of problem sentence
Preset value, b is the second length preset value of problem sentence；
If key index value S be more than pre-set level threshold value, retain described problem sentence as it is to be matched the problem of sentence, or
Person, if key index value S is less than or equal to pre-set level threshold value, removes described problem sentence；
The analysis rule of the answer statement is：If answer statement is not declarative sentence, remove the answer statement and its
The problem of corresponding to sentence.
6. a kind of processing system of customer service question and answer data, it is characterised in that the processing system of the customer service question and answer data includes：
Grouping module, for obtaining the Question and Answer log prestored from customer service Q ＆ A database, and based on customer service in preset time and visitor
The question and answer corresponding relation at family is grouped to the Question and Answer log；
Matching module, for the Question and Answer log in each packet according to it is predetermined the problem of matched with the matched rule of answer,
With draw it is to be added into service knowledge base the problem of answer pair.
7. the processing system of customer service question and answer data according to claim 6, it is characterised in that the grouping module is specifically used
Same client and the Question and Answer log of same customer service in by preset time are divided into same packet；Or will be same in preset time
Client and the Question and Answer log of each customer service are divided into same packet；Or by the Question and Answer log of each client in preset time and same customer service
It is divided into same packet.
8. the processing system of the customer service question and answer data according to claim 6 or 7, it is characterised in that the matching module bag
Include：
Pretreatment unit, for carrying out data prediction to the Question and Answer log in each packet；
Analytic unit, sentence and the answer statement of customer service are pressed the problem of for client in the Question and Answer log after data prediction
Analyzed according to default analysis rule, sentence and corresponding answer statement the problem of to analyze to be matched；
Associative cell, sentence and corresponding answer statement the problem of for extracting to be matched, by sentence the problem of extracting with it is right
The answer statement answered is associated as problem answers pair and by problem answers to adding into service knowledge base.
9. the processing system of customer service question and answer data according to claim 8, it is characterised in that the pretreatment unit is specific
For the font in the Question and Answer log in each packet to be uniformly converted into the complex form of Chinese characters or simplified Chinese character, by asking in each packet
The unified upper of the letter in daily record or lowercase, the particular statement removed in Question and Answer log are answered, and it is right
Sentence in Question and Answer log in each packet carries out word segmentation processing.
10. the processing system of customer service question and answer data according to claim 8, it is characterised in that the default analysis rule
Then include the analysis rule of problem sentence and the analysis rule of answer statement, the analysis rule of described problem sentence is：
The key index value S of computational problem sentence：
S=Sa+Sb+Sc,Wherein, l is the length of problem sentence, and m is to ask
The number of the business keyword contained by sentence is inscribed, n is the number of the stop words of problem sentence, and a is the first length of problem sentence
Preset value, b is the second length preset value of problem sentence；
If key index value S be more than pre-set level threshold value, retain described problem sentence as it is to be matched the problem of sentence, or
Person, if key index value S is less than or equal to pre-set level threshold value, removes described problem sentence；
The analysis rule of the answer statement is：If answer statement is not declarative sentence, remove the answer statement and its
The problem of corresponding to sentence.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710147702.6A CN107066541A (en) | 2017-03-13 | 2017-03-13 | The processing method and system of customer service question and answer data |
| PCT/CN2017/091372 WO2018166115A1 (en) | 2017-03-13 | 2017-06-30 | Method for processing customer service question-answer data, electronic device, and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710147702.6A CN107066541A (en) | 2017-03-13 | 2017-03-13 | The processing method and system of customer service question and answer data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107066541A true CN107066541A (en) | 2017-08-18 |
Family
ID=59622075
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710147702.6A Pending CN107066541A (en) | 2017-03-13 | 2017-03-13 | The processing method and system of customer service question and answer data |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN107066541A (en) |
| WO (1) | WO2018166115A1 (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108415980A (en) * | 2018-02-09 | 2018-08-17 | 平安科技（深圳）有限公司 | Question and answer data processing method, electronic device and storage medium |
| CN109189803A (en) * | 2018-09-03 | 2019-01-11 | 深圳前海微众银行股份有限公司 | Question and answer are to construction method, device and computer readable storage medium |
| CN109272129A (en) * | 2018-09-20 | 2019-01-25 | 重庆先特服务外包产业有限公司 | Call center's business management system |
| CN109670021A (en) * | 2018-12-12 | 2019-04-23 | 安徽讯呼信息科技有限公司 | It is a kind of freely to circulate and the customer service system of custom field |
| CN110321416A (en) * | 2019-05-23 | 2019-10-11 | 深圳壹账通智能科技有限公司 | Intelligent answer method, apparatus, computer equipment and storage medium based on AIML |
| CN111949772A (en) * | 2019-05-17 | 2020-11-17 | 上海戈吉网络科技有限公司 | Intelligent customer service and knowledge base system and management method |
| CN112148743A (en) * | 2020-09-18 | 2020-12-29 | 北京达佳互联信息技术有限公司 | Method, device, equipment and storage medium for updating intelligent customer service knowledge base |
| WO2021051558A1 (en) * | 2019-09-18 | 2021-03-25 | 平安科技（深圳）有限公司 | Knowledge graph-based question and answer method and apparatus, and storage medium |
| CN112800219A (en) * | 2021-01-19 | 2021-05-14 | 苏宁金融科技(南京)有限公司 | Method and system for feeding back customer service log to return database |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110213344B (en) * | 2019-05-14 | 2022-05-17 | 株洲手之声信息科技有限公司 | Multi-center remote sign language online translation system and method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101520802A (en) * | 2009-04-13 | 2009-09-02 | 腾讯科技（深圳）有限公司 | Question-answer pair quality evaluation method and system |
| US8204751B1 (en) * | 2006-03-03 | 2012-06-19 | At&T Intellectual Property Ii, L.P. | Relevance recognition for a human machine dialog system contextual question answering based on a normalization of the length of the user input |
| CN103455535A (en) * | 2013-05-08 | 2013-12-18 | 深圳市明唐通信有限公司 | Method for establishing knowledge base based on historical consultation data |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2220576A2 (en) * | 2007-10-23 | 2010-08-25 | Grape Technology Group, Inc. | Method and computer system for automatically answering natural language questions |
| US20140122619A1 (en) * | 2012-10-26 | 2014-05-01 | Xiaojiang Duan | Chatbot system and method with interactive chat log |
| CN104679815B (en) * | 2014-12-08 | 2018-02-23 | 北京云知声信息技术有限公司 | It is a kind of to screen question and answer pair and the method and system in real-time update question and answer storehouse |
| CN105786851A (en) * | 2014-12-23 | 2016-07-20 | 北京奇虎科技有限公司 | Question and answer knowledge base construction method as well as search provision method and apparatus |
| CN105608221B (en) * | 2016-01-11 | 2018-08-21 | 北京光年无限科技有限公司 | A kind of self-learning method and device towards question answering system |
-
2017
- 2017-03-13 CN CN201710147702.6A patent/CN107066541A/en active Pending
- 2017-06-30 WO PCT/CN2017/091372 patent/WO2018166115A1/en active Application Filing
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8204751B1 (en) * | 2006-03-03 | 2012-06-19 | At&T Intellectual Property Ii, L.P. | Relevance recognition for a human machine dialog system contextual question answering based on a normalization of the length of the user input |
| CN101520802A (en) * | 2009-04-13 | 2009-09-02 | 腾讯科技（深圳）有限公司 | Question-answer pair quality evaluation method and system |
| CN103455535A (en) * | 2013-05-08 | 2013-12-18 | 深圳市明唐通信有限公司 | Method for establishing knowledge base based on historical consultation data |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108415980A (en) * | 2018-02-09 | 2018-08-17 | 平安科技（深圳）有限公司 | Question and answer data processing method, electronic device and storage medium |
| CN109189803A (en) * | 2018-09-03 | 2019-01-11 | 深圳前海微众银行股份有限公司 | Question and answer are to construction method, device and computer readable storage medium |
| CN109272129A (en) * | 2018-09-20 | 2019-01-25 | 重庆先特服务外包产业有限公司 | Call center's business management system |
| CN109670021A (en) * | 2018-12-12 | 2019-04-23 | 安徽讯呼信息科技有限公司 | It is a kind of freely to circulate and the customer service system of custom field |
| CN111949772A (en) * | 2019-05-17 | 2020-11-17 | 上海戈吉网络科技有限公司 | Intelligent customer service and knowledge base system and management method |
| CN110321416A (en) * | 2019-05-23 | 2019-10-11 | 深圳壹账通智能科技有限公司 | Intelligent answer method, apparatus, computer equipment and storage medium based on AIML |
| WO2021051558A1 (en) * | 2019-09-18 | 2021-03-25 | 平安科技（深圳）有限公司 | Knowledge graph-based question and answer method and apparatus, and storage medium |
| CN112148743A (en) * | 2020-09-18 | 2020-12-29 | 北京达佳互联信息技术有限公司 | Method, device, equipment and storage medium for updating intelligent customer service knowledge base |
| CN112800219A (en) * | 2021-01-19 | 2021-05-14 | 苏宁金融科技(南京)有限公司 | Method and system for feeding back customer service log to return database |
| CN112800219B (en) * | 2021-01-19 | 2022-12-30 | 苏宁金融科技(南京)有限公司 | Method and system for feeding back customer service log to return database |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2018166115A1 (en) | 2018-09-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107066541A (en) | The processing method and system of customer service question and answer data | |
| CN106776574B (en) | User comment text mining method and device | |
| CN107818164A (en) | A kind of intelligent answer method and its system | |
| CN108595696A (en) | A kind of human-computer interaction intelligent answering method and system based on cloud platform | |
| CN107506389B (en) | Method and device for extracting job skill requirements | |
| CN108364199B (en) | Data analysis method and system based on Internet user comments | |
| CN107766371A (en) | A kind of text message sorting technique and its device | |
| CN103617290B (en) | Chinese machine-reading system | |
| CN110717018A (en) | Industrial equipment fault maintenance question-answering system based on knowledge graph | |
| CN104281653A (en) | Viewpoint mining method for ten million microblog texts | |
| CN108052505A (en) | Text emotion analysis method and device, storage medium, terminal | |
| CN107145573A (en) | The problem of artificial intelligence customer service robot, answers method and system | |
| CN113505586A (en) | Seat-assisted question-answering method and system integrating semantic classification and knowledge graph | |
| CN109033166B (en) | Character attribute extraction training data set construction method | |
| CN106909573A (en) | A kind of method and apparatus for evaluating question and answer to quality | |
| CN111191463A (en) | Emotion analysis method and device, electronic equipment and storage medium | |
| CN108536673B (en) | News event extraction method and device | |
| CN111177401A (en) | Power grid free text knowledge extraction method | |
| CN111062211A (en) | Information extraction method and device, electronic equipment and storage medium | |
| CN106227720A (en) | A kind of APP software users comment mode identification method | |
| CN109214445A (en) | A kind of multi-tag classification method based on artificial intelligence | |
| CN110309513B (en) | Text dependency analysis method and device | |
| CN103092838B (en) | A kind of method and device for obtaining English words | |
| CN110489514B (en) | System and method for improving event extraction labeling efficiency, event extraction method and system | |
| CN111199208A (en) | Head portrait gender identification method and system based on deep learning framework |
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: 20170818 |
|
| RJ01 | Rejection of invention patent application after publication |