CN108170780A - A kind of the problem of self-service question and answer matching process and device - Google Patents
A kind of the problem of self-service question and answer matching process and device Download PDFInfo
- Publication number
- CN108170780A CN108170780A CN201711433681.0A CN201711433681A CN108170780A CN 108170780 A CN108170780 A CN 108170780A CN 201711433681 A CN201711433681 A CN 201711433681A CN 108170780 A CN108170780 A CN 108170780A
- Authority
- CN
- China
- Prior art keywords
- face
- topic face
- default
- synonym
- topic
- 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
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
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- 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
- G06F40/237—Lexical tools
- G06F40/247—Thesauruses; Synonyms
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Mathematical Physics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Machine Translation (AREA)
Abstract
The invention discloses matching process the problem of a kind of self-service question and answer and device, the method includes:Acquisition asks questions, and determines that this asks questions the synonym union with presetting topic face in self-service question and answer exam pool, the synonym union in a default topic face is asked questions including this presets the word that semanteme is identical in topic face with this;The first semantic effect degree in the synonym and set pair in topic the face default topic face is preset according to this, topic face set is obtained from self-service question and answer library, topic face set includes the default topic face that at least one first semantic effect degree is more than first threshold;The synonym and the set pair second semantic effect degree asked questions in topic face are preset according to this, target topic face is obtained in gathering from the topic face, target topic face includes the default topic face that at least one second semantic effect degree is more than second threshold.So that the target topic face is increased with the semantic matching degree asked questions, further so that the accuracy rate of the self-service answer of output increases.
Description
Technical field
The present invention relates to field of computer technology, in particular to the problem of a kind of self-service question and answer matching process and device.
Background technology
At present, in order to improve service quality, the quick FAQs for solving user and proposing, most of commercial city that services is in website
On provide self-service question and answer service.According to the FAQs that user proposes, multiple default topic faces are pre-set, and according to practical feelings
Condition sets each default corresponding self-service answer in topic face.Create a self-service question and answer library, store multiple default topic face and from
Help the correspondence of answer.
When website provides the user with the service of self-service question and answer, obtain it is input by user asks questions, analysis ask questions in advance
If the synonym union in the face of topic, which includes asking questions the words with representing same semanteme in default topic face, root
According to the semantic effect degree in the synonym and set pair default topic face, determine to ask questions the semantic matching degree with default topic face.
It is understood that if the semantic effect degree in synonym and the set pair default topic face is high, then it represents that asks questions and default topic
The semantic matching degree in face is high;If the semantic effect degree in synonym and the set pair default topic face is low, then it represents that ask questions in advance
If the semantic matching degree in the face of topic is low.The corresponding self-service answer in the default topic face high with asking questions semantic matching degree is exported, is realized
To the self-service response input by user asked questions.
But since in self-service library of responses, default topic face is to refine to summarize from the FAQs that user proposes to obtain,
The word negligible amounts in the default topic face are formed, when there are many word included in input by user ask questions, the default topic
Most word all occurs in this is asked questions in face, i.e. the semantic effect degree in synonym and the set pair default topic face
Height, still, the semantic effect degree that the synonym and set pair ask questions are really very low.At this point, using above problem matching process
Obtained default topic face, it is actually very low with the semantic matching degree asked questions, lead to the self-service answer mistake of output.
Invention content
Present invention solves the technical problem that a kind of the problem of being to provide self-service question and answer matching process and device, so as to
To ask questions matching and asking questions the high default topic face of semantic matching degree, so as to export the higher self-service answer of accuracy rate.
For this purpose, the technical solution that the present invention solves technical problem is:
A kind of the problem of self-service question and answer matching process, the method includes:
Acquisition asks questions;
Determine the synonym union asked questions with presetting topic face in self-service question and answer library, default topic face
The synonym union includes described ask questions and presets the words that semanteme is identical in topic face with this;
The first semantic effect degree in topic face is preset according to the synonym in the default topic face and set pair, from it is described from
Acquisition topic face set in question and answer library is helped, the topic face set includes at least one first semantic effect degree more than the first threshold
The default topic face of value;
The the second semantic effect degree asked questions according to the synonym in the default topic face and set pair, from the topic
Target topic face is obtained in the set of face, the target topic face includes at least one second semantic effect degree and is more than second threshold
The default topic face.
Optionally, the first semantic effect degree packet in default topic face described in the synonym in the default topic face and set pair is determined
It includes:
Count the synonym and concentration in the default topic face, the sum of the corresponding default score value of each words, as the first value;
It counts in the default topic face, the sum of the corresponding default score value of each words, as second value;
Determine the ratio of first value and the second value as the first semantic effect degree.
Optionally, the second semantic effect degree packet asked questions described in the synonym in the default topic face and set pair is determined
It includes:
In being asked questions described in statistics, the sum of the corresponding default score value of each words, as third value;
Determine the ratio of first value and the third value as the second semantic effect degree.
Optionally, it is described to determine that described ask questions includes with presetting the synonym union in topic face in self-service question and answer library:
Described ask questions is segmented with the default topic face in the self-service question and answer library;
The words identical with semanteme in the default topic face is asked questions described in acquisition, as the synonymous of the default topic face
Word union.
Optionally, the method further includes:
The corresponding self-service answer in the target topic face is exported, as to the self-service response asked questions.
A kind of the problem of self-service question and answer coalignment, the method includes:
Acquiring unit is asked questions, is asked questions for obtaining;
Synonym union determination unit, for determining the synonym asked questions with presetting topic face in self-service question and answer library
Union, the synonym union in a default topic face include described asking questions that preset semanteme in topic face identical with this
Words;
Acquiring unit is gathered in topic face, for presetting the first of topic face described in the synonym according to the default topic face and set pair
Semantic effect degree, obtains topic face set from the self-service question and answer library, and the topic face set includes at least one described first
Semantic effect degree is more than the default topic face of first threshold;
Target inscribes face acquiring unit, for asked questions described in the synonym according to the default topic face and set pair second
Semantic effect degree, obtains target topic face in gathering from the topic face, the target topic face includes at least one second language
Adopted influence degree is more than the default topic face of second threshold.
Optionally, the topic face set acquiring unit includes:
First Data-Statistics subelement, for counting the synonym and concentration in the default topic face, each words is corresponding pre-
If the sum of score value, as the first value;
Second Data-Statistics subelement, for counting in the default topic face, the sum of the corresponding default score value of each words is made
For second value;
First semantic effect degree determination subelement, for determining the ratio of first value and the second value as institute
State the first semantic effect degree.
Optionally, the target topic face acquiring unit, including:
Third Data-Statistics subelement, in asking questions for counting described, sum of the corresponding default score value of each words, work
For third value;
Second semantic effect degree determination subelement, for determining the ratio of first value and the third value as institute
State the second semantic effect degree.
Optionally, the synonym union determination unit includes:
Subelement is segmented, for dividing with the default topic face in the self-service question and answer library described ask questions
Word;
Synonym union obtains subelement, for obtaining described ask questions and identical word semantic in the default topic face
Word, the synonym union as the default topic face.
Optionally, described device further includes:
Self-service answer output unit for exporting the corresponding self-service answer in the target topic face, is asked as to the consulting
The self-service response of topic.
According to the above-mentioned technical solution, the method have the advantages that:
The problem of self-service question and answer provided by the invention matching process and device, acquisition ask questions, determine that this is asked questions
Synonym union with presetting topic face in self-service question and answer exam pool, the synonym union in a default topic face include this ask questions and
Semantic identical word in the default topic face;First semantic effect in the synonym and set pair in topic the face default topic face is preset according to this
Degree, obtains topic face set from self-service question and answer library, and topic face set includes at least one first semantic effect degree more than the
The default topic face of one threshold value;The synonym and the set pair second semantic effect degree asked questions in topic face are preset according to this, from
Target topic face is obtained in the set of the topic face, target topic face includes at least one second semantic effect degree and is more than second threshold
Default topic face.It follows that not only analyzing the semantic effect degree in synonym and the set pair default topic face, characterize from default
From the point of view of topic face, the semantic dependency with default topic face is asked questions;Also analyzing synonym and set pair, this is asked questions
Semantic effect degree, characterizes from the point of view of asking questions, and asks questions the semantic dependency with default topic face;So as to make
It obtains using the obtained target topic face of problem matching process provided by the invention, is increased with the semantic matching degree asked questions,
Further so that the accuracy rate of the self-service answer of output increases.
Description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention, for those of ordinary skill in the art, without creative efforts, can be with
Other attached drawings are obtained according to these attached drawings.
The flow chart of the problem of Fig. 1 is self-service question and answer provided in an embodiment of the present invention matching process;
Fig. 2 determines to ask questions the synonym union with presetting topic face in self-service question and answer library to be provided in an embodiment of the present invention
Method flow diagram;
Fig. 3 is the first semantic shadow that the synonym provided in an embodiment of the present invention for determining default topic face and set pair preset topic face
The method flow diagram of the degree of sound;
Fig. 4 is the second semantic shadow that the synonym provided in an embodiment of the present invention for determining default topic face and set pair ask questions
The method flow diagram of the degree of sound;
Another flow chart of the problem of Fig. 5 is self-service question and answer provided in an embodiment of the present invention matching process;
The schematic diagram of the problem of Fig. 6 is self-service question and answer provided in an embodiment of the present invention coalignment.
Specific embodiment
For the implementation in the higher default topic face of the semantic matching degree that provides to obtain and ask questions, the present invention is implemented
A kind of the problem of example provides self-service question and answer matching process and device, below in conjunction with preferred implementation of the Figure of description to the present invention
Example illustrates, it should be understood that preferred embodiment described herein is merely to illustrate and explain the present invention, and is not used to limit
The present invention.And in the absence of conflict, the feature in the embodiment and embodiment in the application can be combined with each other.
On the one hand, inventor has found under study for action, due in self-service library of responses, default topic face be proposed from user it is common
It is refined in problem and summarizes what is obtained, formed the word negligible amounts in the default topic face, included when in input by user ask questions
Word it is many when, most word all occurs in this is asked questions in the default topic face, i.e., this is pre- for synonym and set pair
If the semantic effect degree in the face of topic is high, still, the semantic effect degree that the synonym and set pair ask questions is really very low.For example,
It is input by user ask questions for:" I has registered user and then how to operate in Bang Bang communities APP", in self-service library of responses
It is " how registered user " there are one default topic face, is asked since the word preset in topic face is both present in consulting input by user
In topic, i.e., the synonym union for asking questions and presetting topic face be this it is default carry face, so, existing self-service question answering system
It will be considered that the semantic effect degree in the synonym and the set pair default topic face is very high, inputted so this is preset topic face with user
Ask questions and match, and export and preset the corresponding self-service answer in topic face with this.But due to synonym and set pair presets topic
The influence degree height in face can only illustrate that synonym union can characterize the core semanteme in default topic face, and cannot illustrate the synonym simultaneously
It is semantic that collection can characterize the core that asks questions, and therefore, which asks questions with input by user semantically not
Therefore matching, causes the self-service answer of output to be inaccurate.
On the other hand, inventor has found under study for action, can further analyze synonym and set pair consulting input by user is asked
The semantic effect degree of topic, analyzes whether the synonym union can characterize the core semanteme asked questions, with reference to synonym simultaneously
The set pair semantic effect degree input by user asked questions and synonym and set pair preset the semantic effect degree in topic face, obtain
The high default topic face of semantic matching degree is asked questions with input by user.
In consideration of it, a kind of the problem of self-service question and answer of offer of embodiment of the present invention matching process and device, acquisition ask questions,
Determine that this asks questions the synonym union with presetting topic face in self-service question and answer exam pool, the synonym union packet in a default topic face
It includes this and asks questions and preset the word that semanteme is identical in topic face with this;Synonym and the set pair default topic face in topic face are preset according to this
The first semantic effect degree, obtained from self-service question and answer library topic face set, the topic face set includes it is at least one first semanteme
Influence degree is more than the default topic face of first threshold;The synonym and set pair second language asked questions in topic face are preset according to this
Adopted influence degree, obtains target topic face in gathering from the topic face, target topic face includes at least one second semantic effect degree
More than the default topic face of second threshold.It follows that not only analyze the semantic effect journey in synonym and the set pair default topic face
Degree, characterizes from the point of view of default topic face, asks questions the semantic dependency with default topic face;Also synonym union is analyzed
It to the semantic effect degree asked questions, characterizes from the point of view of asking questions, asks questions the semanteme with default topic face
Correlation;So as to so that using the obtained target topic face of problem matching process provided by the invention, with the language asked questions
Adopted matching degree increases, further so that the accuracy rate of the self-service answer of output increases.
Illustrative methods
Referring to Fig. 1, the problem of which is self-service question and answer provided in this embodiment matching process flow chart.
The problem of self-service question and answer provided in this embodiment matching process, include the following steps:
S101:Acquisition asks questions.
What is referred in the present embodiment asks questions the consulting for referring to that user inputs in the website for providing self-service question and answer service
Problem.User can be inputted in the website for providing self-service question and answer service by terminal and be asked questions.User can also be at other
The platform input for providing self-service question and answer service asks questions, and the present embodiment does not limit specifically.
In the present embodiment, after user's input asks questions, the server for providing the website of self-service question and answer service can be with
Obtain it is input by user ask questions, the present embodiment, which does not limit specifically, to be provided the server acquisition of self-service question and answer service and asks questions
Specific implementation.
S102:Determine that this asks questions the synonym union with presetting topic face in self-service question and answer library, default topic face
Synonym union asks questions including this and presets the words that semanteme is identical in topic face with this.
It is quick to solve asking questions for user's proposition in order to improve service quality, a self-service question and answer library can be created, it should
It is stored in self-service question and answer library and multiple ask questions and each ask questions corresponding self-service answer.It is referred in the present embodiment
Default topic face refer to that is stored in self-service question and answer library asks questions.Multiple default topic faces are saved in self-service question and answer library, so as to
After input by user ask questions is obtained, this is asked questions and is matched with the default topic face in self-service question and answer library.
In the present embodiment, it matches with the default topic face in self-service question and answer library to asking questions, can determine first
Ask questions the synonym union with presetting topic face in self-service question and answer library.
It should be noted that referred in the present embodiment it is determining ask questions it is same with the default topic face in self-service question and answer library
Adopted word union, in order to it which of determines to ask questions words and appears in the default topic face in self-service question and answer library, from
And analyze the matching degree asked questions with default topic face.That is, for a certain default topic face, this presets the same of topic face
Adopted word union only can include asking questions presetting words identical in topic face with this.
It is understood that for same semanteme, the description of different user may be different, therefore, is determining consulting
When the synonym union in topic face is preset in problem and self-service question and answer library, can be combined with the semanteme of words determine to ask questions with
The synonym union in topic face is preset in self-service question and answer library, the words identical with semanteme in default topic face is also received in will asking questions
Enter the synonym and concentration in the default topic face.
Specifically, it is determined that ask questions the synonym union with presetting topic face in self-service question and answer library, can by S201 and
S202 is realized.
S201:It is segmented to asking questions with the default topic face in self-service question and answer library.
S202:It obtains and asks questions and preset semantic identical words, the synonym union as default topic face in topic face.
About S201 and S202, it is to be understood that asking questions may wrap in the default topic face in self-service question and answer library
When including multiple words, and determining to ask questions the synonym union with presetting topic face in self-service question and answer library, if directly to entirely consulting
Inquiry inscribes and presets topic face and handled, then can increase the complexity of actual algorithm.Therefore, in the present embodiment, first to consulting
Inquiry is inscribed to be segmented with the default topic face in self-service question and answer library, and the multiple words obtained later to participle are analyzed, so as to
Obtain the synonym union asked questions with presetting topic face in self-service question and answer library.
Since the synonym union in a default topic face includes asking questions presetting the words that semanteme is identical in topic face with this,
Therefore, after being segmented to default topic face and asking questions, can to ask questions segmented obtain later it is multiple
Words is compared one by one with being segmented the multiple words obtained later to default topic face, carries out segmenting it when to asking questions
When the words obtained afterwards is identical with being segmented the words semanteme obtained later to default topic face, then it is pre- the words to be added to this
If the synonym and concentration in the face of topic.
S103:The first semantic effect degree in topic face is preset according to the synonym in default topic face and set pair, from self-service question and answer
Acquisition topic face set, topic face set include the default topic that at least one first semantic effect degree is more than first threshold in library
Face.
The the first semantic effect degree referred in the present embodiment refers to the semantic effect in synonym and the set pair default topic face
Degree.The first semantic effect degree in the present embodiment can represent with the numerical value of a value range between zero and one, and
And the numerical value is bigger, represents that the first semantic effect degree is higher, the numerical value is smaller, represents that the first semantic effect degree is lower.
It is understood that for one asks questions, there may be multiple default topic faces in self-service question and answer library so that
The semantic effect degree in the synonym and set pair in multiple default topic faces default topic face is all higher.Multiple default topic face institute group
Into set be the present embodiment in refer to topic face set.
The first threshold referred in the present embodiment is a pre-set value, when the synonym and set pair in default topic face are pre-
If the first semantic effect degree in the face of topic is more than first threshold, it is believed that this is default for the synonym and set pair in the default topic face
The semantic effect degree in topic face is higher.As a kind of example, the value of first threshold can be 0.4 or 0.5.
In the present embodiment, determine the synonym in default topic face and set pair preset the first semantic effect degree in topic face can be with
It is realized by S301-S303.
S301:The synonym and concentration in the default topic face of statistics, the sum of the corresponding default score value of each words, as first
Value.
As it was noted above, the first semantic effect degree, refer to the semantic effect degree in synonym and the set pair default topic face.
That is, the first semantic effect degree, it can be by judging synonym and concentrating the words included shared in default topic face
Semantic proportion determine.
It is understood that multiple words can be obtained after being segmented to default topic face, and can in multiple words
Can be comprising vocabulary of different nature, and vocabulary of different nature is different in the significance level in presetting topic face.Generally
For, the significance level of noun is higher than verb, and the significance level of verb is higher than modal particle, and the significance level of modal particle is higher than Jie
Word, etc..
Therefore, in the present embodiment, the multiple words obtained later to topic face is preset to be segmented are set first corresponding
Default score value.By synonym and the default score value of each words is concentrated to be added, obtain the total score of synonym union, this is synonymous
The total score of word union is the first value.
S302:In the default topic face of statistics, the sum of the corresponding default score value of each words, as second value.
The second value referred in the present embodiment refers to the default score value of words each in default topic face being added, acquisition
The total score in default topic face.
S303:Determine the ratio of the first value and second value as the first semantic effect degree.
The ratio of first value and second value may be considered synonym and concentrate the words included shared in default topic face
Semantic proportion.
It is understood that since the first value is the synonym and concentration in default topic face, each words is default point corresponding
Sum of value, second value are preset in topic face, the sum of the corresponding default score value of each words.And the synonym and concentration in default topic face
Comprising words it is at most identical with the words included in default topic face.Therefore, the first value should be less than or equal to second value, that is,
It says, the maximum value of the ratio of the first value and second value is equal to 1, when the ratio of the first value and second value is equal to 1, represents default topic
The words included in face, which is both present in, to be asked questions, and illustrates the synonym in the default topic face and set pair presets the influence journey in topic face
Degree is high.
S104:According to the second semantic effect degree that the synonym in default topic face and set pair ask questions, from topic face set
Middle acquisition target topic face, target topic face include the default topic face that at least one second semantic effect degree is more than second threshold.
The the second semantic effect degree referred in the present embodiment refers to synonym and the set pair semantic effect asked questions
Degree.The second semantic effect degree in the present embodiment can represent with the numerical value of a value range between zero and one, and
And the numerical value is bigger, represents that the second semantic effect degree is higher, the numerical value is smaller, represents that the second semantic effect degree is lower.
The second threshold referred in the present embodiment is a pre-set value, when the synonym and set pair in default topic face are consulted
When second semantic effect degree of inquiry topic is more than second threshold, it is believed that the simultaneously set pair consulting of the synonym in the default topic face
The semantic effect degree of problem is higher.As a kind of example, the value of second threshold can be 0.4 or 0.5.
In the multiple default topic faces included in gathering for topic face, it is understood that there may be the default topic face of one or more so that should
The semantic effect degree that the synonym and set pair in multiple default topic faces ask questions is all higher.Multiple default topic face is this
The target topic face referred in embodiment.It is understood that target topic face can include a default topic face, can also include more
A default topic face.
In the present embodiment, the first semantic effect degree in the synonym in target topic face and set pair target topic face is more than first
Threshold value, and when the synonym in target topic face and the second semantic effect degree for asking questions of set pair are more than second threshold, can recognize
It is higher with the semantic matching degree that asks questions that face is inscribed for target.
In the present embodiment, determine that the second semantic effect degree that the synonym in default topic face and set pair ask questions can be with
It is realized by S401-S402.
S401:In statistical consultation problem, the sum of the corresponding default score value of each words, as third value.
As it was noted above, the second semantic effect degree, refers to the semantic effect degree that synonym and set pair ask questions.
It, can be by judging synonym and concentrating the words included shared in asking questions that is the second semantic effect degree
Semantic proportion determines.
Therefore, in the present embodiment, first to being set accordingly to asking questions the multiple words for being segmented and being obtained later
Default score value.The default score value that each words obtained later is segmented to asking questions is added, is asked questions
Total score, which is third value.
It should be noted that in the present embodiment, to being set to asking questions the multiple words for being segmented and being obtained later
The rule of corresponding default score value, default score value corresponding with being segmented the multiple words obtained later setting to default topic face
Rule should be identical.As a kind of example, the rule for setting corresponding default score value can be rule as shown in table 1 below.
Table 1
The property of words | Default score value |
Noun | 4 |
Verb | 2 |
Modal particle | 1 |
Preposition | 1 |
S402:Determine the ratio of the first value and third value as the second semantic effect degree.
The ratio of first value and third value may be considered synonym and concentrate the words included shared in asking questions
Semantic proportion.
It is understood that since the first value is the synonym and concentration in default topic face, each words is default point corresponding
The sum of value, second value be in asking questions, the sum of the corresponding default score value of each words.And the synonym and concentration in default topic face
Comprising words it is at most identical with the words included in asking questions.Therefore, the first value should be less than or equal to third value, that is,
It says, the maximum value of the ratio of the first value and third value is equal to 1, when the ratio of the first value and third value is equal to 1, represents that consulting is asked
The words included in topic is both present in default topic face, illustrates the influence journey that the synonym in the default topic face and set pair ask questions
Degree is high.
It should be noted that after S104 obtains target topic face, the present embodiment can also export target topic face and correspond to
Self-service answer, as the self-service response to asking questions.
It should be noted that due to saving self-service answer corresponding with each default topic face respectively in self-service question and answer library,
Therefore, after obtaining and asking questions the high target topic face of matching degree, it is possible to which directly output is corresponding with target topic face
Self-service answer.
Multiple target topic faces may be corresponded to by being asked questions in view of one, in the present embodiment, can be exported respectively each
The corresponding self-service answer in target topic face can also combine the first semantic effect degree and the second semantic effect in multiple targets topic face
Degree therefrom selects a target topic face as final target and inscribes face, and exports corresponding with the final target topic face
Self-service answer as to the self-service response asked questions, multiple targets can also be inscribed according to the second semantic effect degree face into
Row sequence under normal circumstances, according to the sequence of the second semantic effect degree from high to low, is sequentially output each target topic face and corresponds to
Self-service answer as to the self-service response asked questions, the present embodiment is not specifically limited this.Based on above example
The problem of self-service question and answer provided matching process, matching process carries out the problem of below with reference to specific example to the self-service question and answer
It introduces.
Ask questions 1:I is bundled with bank card on APP, and how cell-phone number is changed
Default topic face 1:Bind cell-phone number.
Default topic face 2:How modification binding bank card cell-phone number
Wherein, first threshold and the value of second threshold are 0.5.
Referring to Fig. 5, the problem of which is self-service question and answer provided in this embodiment matching process again
One flow chart.
Method provided in this embodiment, includes the following steps:
S501:To asking questions carry out word segmentation processing, and it is corresponding default for each participle setting that participle obtains later
Score value, specific word segmentation result and default score value setting result are as follows:
My (1 point)/in (1 point)/APP (4 points)/upper (1 point)/binding (2 points)/(1 point)/bank (4 points)/card (4
Point)/, mobile phone (4 points)/number (1 point)/how (1 point)/modification (2 points).
S502:Word segmentation processing is carried out, and set for each participle that participle obtains later to default topic face 1 and default topic face 2
Corresponding default score value is put, specific word segmentation result and default score value setting result are as follows:
Default topic face 1:Bind (2 points)/mobile phone (4 points)/number (1 point);
Default topic face 2:How (1 point)/modification (2 points)/binding (2 points)/bank (4 points)/card (4 points)/(1 point)/hand
Machine (4 points)/number (1 point).
S503:Obtain the synonym union in default topic face 1 and the synonym union in default topic face 2.
Presetting 1 corresponding synonym union of topic face is:Binding, mobile phone, number;
Presetting 2 corresponding synonym union of topic face is:How, modification, binding, bank, card, mobile phone, number.
S504:The total score of the synonym union in the default topic face 1 of statistics, i.e. the first value;The total score in the default topic face 1 of statistics
Value, i.e. second value;And calculate the first semantic effect degree of the synonym union in default topic face 1.
First value is:2+4+1=7;
Second value is:2+4+1=7;
First semantic effect degree:First value/second value=7/7=1.0;
S505:The total score of the synonym union in the default topic face 2 of statistics, i.e. the first value;The total score in the default topic face 2 of statistics
Value, i.e. second value;And calculate the first semantic effect degree of the synonym union in default topic face 2.
First value is:1+2+2+4+4+4+1=18
Second value is:1+2+2+4+4+1+4+1=19;
First semantic effect degree:First value/second value=18/19=0.95;
S506:Determine that topic face is gathered.
Since the first semantic effect degree for presetting topic face 1 and default topic face 2 is all higher than first threshold 0.5, it presets
Topic face 1 and default topic face 2 are all contained in the set of topic face.
S507:The total score of statistical consultation problem, i.e. third value.
Third value is:1+1+4+1+2+1+4+4+4+1+1+2=26
S508:Calculate the second semantic effect degree of the synonym union in default topic face 1;Calculate the synonymous of default topic face 2
Second semantic effect degree of word union.
Second semantic effect degree of the synonym union in default topic face 1:First value/third value=7/26=0.27;
Second semantic effect degree of the synonym union in default topic face 2:First value/third value=18/26=0.69.
S509:Obtain target topic face.
Since the second semantic effect degree for presetting topic face 1 is less than second threshold, and preset second semantic effect in topic face 2
Degree is more than second threshold, and therefore, default topic face 2 is to inscribe face with asking questions matched target.
S510:The corresponding self-service answer in target topic face is exported, as the self-service response to asking questions.
The present embodiment is not specifically limited the setting of the corresponding self-service answer in target topic face, and the setting of the self-service answer can
It is specifically set with the content that face is inscribed according to target.
By being calculated above it is found that the first semantic effect degree in default topic face 1 is very high, but its semantically with consulting
Problem simultaneously mismatches.Although the and first semantic effect degree in the first semantic effect degree Non-precondition topic face 1 in default topic face 2
Height still, is more matched semantically with asking questions, and can be by the second language in matching semantically with asking questions
Adopted influence degree embodies.Can be that consulting is asked that is, with reference to the first semantic effect degree and the second semantic effect degree
Topic matches the higher target topic face of semantic matching degree, so as to be user's output more accurately autonomous answer.
The problem of self-service question and answer provided in this embodiment matching process, not only analyze synonym and the set pair default topic face
Semantic effect degree, synonym and the set pair semantic effect degree asked questions are also analyzed, so that utilizing this hair
The obtained target topic face of the problem of bright offer matching process, increases with the semantic matching degree asked questions, further, makes
The accuracy of self-service answer that must be exported increases.
Exemplary means
In view of the problem of self-service question and answer of above example offer matching process, the present embodiment provides a kind of self-service question and answer
Problem coalignment is introduced the device below with reference to attached drawing.
Referring to Fig. 6, the problem of which is self-service question and answer provided in this embodiment coalignment schematic diagram.
The problem of self-service question and answer provided in this embodiment coalignment, including:Ask questions acquiring unit 610, synonym
Union determination unit 620, topic face set acquiring unit 630 and target topic face acquiring unit 640.
Wherein:
Acquiring unit 610 is asked questions, is asked questions for obtaining;
Synonym union determination unit 620, for determining to ask questions the synonym with presetting topic face in self-service question and answer library
Union, the synonym union in a default topic face include asking questions presetting the words that semanteme is identical in topic face with this;
Acquiring unit 630 is gathered in topic face, semantic for presetting the first of topic face according to the synonym and set pair of presetting topic face
Influence degree obtains topic face set from self-service question and answer library, and it is big that topic face set includes at least one first semantic effect degree
In the default topic face of first threshold;
Target inscribes face acquiring unit 640, the second semanteme asked questions for the synonym according to default topic face and set pair
Influence degree, obtains target topic face in gathering from topic face, target topic face is more than including at least one second semantic effect degree
The default topic face of second threshold.
In one embodiment, set acquiring unit 630 in topic face includes:First Data-Statistics subelement, the second Data-Statistics
Subelement and the first semantic effect degree determination subelement.
First Data-Statistics subelement, for counting the synonym and concentration in default topic face, each words is default point corresponding
The sum of value, as the first value;
Second Data-Statistics subelement, for counting in default topic face, the sum of the corresponding default score value of each words, as
Two-value;
First semantic effect degree determination subelement, for determining the ratio of the first value and the second value as described
One semantic effect degree.
In one embodiment, target topic face acquiring unit 640, including:
Third Data-Statistics subelement, in statistical consultation problem, the sum of the corresponding default score value of each words, as
Three values;
Second semantic effect degree determination subelement, for determining the ratio of the first value and the third value as described
Two semantic effect degree.
In one embodiment, synonym union determination unit 620 includes:
Subelement is segmented, for being segmented to asking questions with the default topic face in the self-service question and answer library;
Synonym union obtains subelement, and the words identical with semanteme in the default topic face is asked questions for obtaining,
Synonym union as default topic face.
In one embodiment, described device further includes:
Self-service answer output unit for exporting the target corresponding self-service answer in topic face, is asked questions as to described
Self-service response.
The problem of self-service question and answer provided in this embodiment coalignment, not only analyze synonym and the set pair default topic face
Semantic effect degree, synonym and the set pair semantic effect degree asked questions are also analyzed, so that utilizing this hair
The obtained target topic face of the problem of bright offer matching process, increases with the semantic matching degree asked questions, further, makes
The accuracy of self-service answer that must be exported increases.
The above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, without departing from the principle of the present invention, several improvements and modifications can also be made, these improvements and modifications also should
It is considered as protection scope of the present invention.
Claims (10)
1. a kind of the problem of self-service question and answer matching process, which is characterized in that the method includes:
Acquisition asks questions;
Determine the synonym union asked questions with presetting topic face in self-service question and answer library, default topic face it is described
Synonym union includes described ask questions and presets the words that semanteme is identical in topic face with this;
The first semantic effect degree in topic face is preset according to the synonym in the default topic face and set pair, self-service is asked from described
Acquisition topic face set in library is answered, the topic face set includes at least one first semantic effect degree more than first threshold
The default topic face;
The the second semantic effect degree asked questions according to the synonym in the default topic face and set pair, from topic face collection
Target topic face is obtained in conjunction, the target topic face includes the institute that at least one second semantic effect degree is more than second threshold
State default topic face.
2. it according to the method described in claim 1, it is characterized in that, determines pre- described in the synonym in the default topic face and set pair
If the first semantic effect degree in topic face includes:
Count the synonym and concentration in the default topic face, the sum of the corresponding default score value of each words, as the first value;
It counts in the default topic face, the sum of the corresponding default score value of each words, as second value;
Determine the ratio of first value and the second value as the first semantic effect degree.
3. according to the method described in claim 2, it is characterized in that, determine official communication described in the synonym in the default topic face and set pair
Second semantic effect degree of inquiry topic includes:
In being asked questions described in statistics, the sum of the corresponding default score value of each words, as third value;
Determine the ratio of first value and the third value as the second semantic effect degree.
4. according to the method described in claim 1-3 any one, which is characterized in that it is described determine it is described ask questions with it is self-service
The synonym union that topic face is preset in question and answer library includes:
Described ask questions is segmented with the default topic face in the self-service question and answer library;
The words identical with semanteme in the default topic face is asked questions described in acquisition, the synonym as the default topic face is simultaneously
Collection.
5. according to the method described in any of claim 1 to 4, which is characterized in that the method further includes:
The corresponding self-service answer in the target topic face is exported, as to the self-service response asked questions.
6. a kind of the problem of self-service question and answer coalignment, which is characterized in that the method includes:
Acquiring unit is asked questions, is asked questions for obtaining;
Synonym union determination unit, for determining described ask questions with the synonym in topic face default in self-service question and answer library simultaneously
Collection, the synonym union in a default topic face include described ask questions and preset the word that semanteme is identical in topic face with this
Word;
Acquiring unit is gathered in topic face, semantic for presetting the first of topic face described in the synonym according to the default topic face and set pair
Influence degree obtains topic face set from the self-service question and answer library, and it is semantic that the topic face set includes at least one described first
Influence degree is more than the default topic face of first threshold;
Target inscribes face acquiring unit, for the second semanteme asked questions described in the synonym according to the default topic face and set pair
Influence degree, obtains target topic face in gathering from the topic face, the target topic face includes at least one described second semantic shadow
The degree of sound is more than the default topic face of second threshold.
7. device according to claim 6, which is characterized in that the topic face set acquiring unit includes:
First Data-Statistics subelement, for counting the synonym and concentration in the default topic face, each words is default point corresponding
The sum of value, as the first value;
Second Data-Statistics subelement, for counting in the default topic face, the sum of the corresponding default score value of each words, as
Two-value;
First semantic effect degree determination subelement, for determining the ratio of first value and the second value as described
One semantic effect degree.
8. device according to claim 7, which is characterized in that the target inscribes face acquiring unit, including:
Third Data-Statistics subelement, in asking questions for counting described, the sum of the corresponding default score value of each words, as the
Three values;
Second semantic effect degree determination subelement, for determining the ratio of first value and the third value as described
Two semantic effect degree.
9. according to the device described in claim 6-8 any one, which is characterized in that the synonym union determination unit packet
It includes:
Subelement is segmented, for being segmented to described ask questions with the default topic face in the self-service question and answer library;
Synonym union obtains subelement, for obtain it is described ask questions with identical words semantic in the default topic face,
Synonym union as the default topic face.
10. according to the device described in claim 6-9 any one, which is characterized in that described device further includes:
Self-service answer output unit for exporting the corresponding self-service answer in target topic face, is asked questions as to described
Self-service response.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711433681.0A CN108170780A (en) | 2017-12-26 | 2017-12-26 | A kind of the problem of self-service question and answer matching process and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711433681.0A CN108170780A (en) | 2017-12-26 | 2017-12-26 | A kind of the problem of self-service question and answer matching process and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108170780A true CN108170780A (en) | 2018-06-15 |
Family
ID=62521097
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711433681.0A Pending CN108170780A (en) | 2017-12-26 | 2017-12-26 | A kind of the problem of self-service question and answer matching process and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108170780A (en) |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20010008121A (en) * | 2000-11-09 | 2001-02-05 | 김진문 | System for measuring an annual salary through communications and a method thereof and a method for seeking jobs using these |
CN101097573A (en) * | 2006-06-28 | 2008-01-02 | 腾讯科技(深圳)有限公司 | Automatically request-answering system and method |
CN101286161A (en) * | 2008-05-28 | 2008-10-15 | 华中科技大学 | Intelligent Chinese request-answering system based on concept |
CN101465749A (en) * | 2008-12-29 | 2009-06-24 | 武汉大学 | Method for building interlocution service based on Web Service combination |
CN103902652A (en) * | 2014-02-27 | 2014-07-02 | 深圳市智搜信息技术有限公司 | Automatic question-answering system |
CN104050256A (en) * | 2014-06-13 | 2014-09-17 | 西安蒜泥电子科技有限责任公司 | Initiative study-based questioning and answering method and questioning and answering system adopting initiative study-based questioning and answering method |
CN105608218A (en) * | 2015-12-31 | 2016-05-25 | 上海智臻智能网络科技股份有限公司 | Intelligent question answering knowledge base establishment method, establishment device and establishment system |
CN105824933A (en) * | 2016-03-18 | 2016-08-03 | 苏州大学 | Automatic question-answering system based on theme-rheme positions and realization method of automatic question answering system |
KR101662450B1 (en) * | 2015-05-29 | 2016-10-05 | 포항공과대학교 산학협력단 | Multi-source hybrid question answering method and system thereof |
CN106874441A (en) * | 2017-02-07 | 2017-06-20 | 腾讯科技(上海)有限公司 | Intelligent answer method and apparatus |
-
2017
- 2017-12-26 CN CN201711433681.0A patent/CN108170780A/en active Pending
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20010008121A (en) * | 2000-11-09 | 2001-02-05 | 김진문 | System for measuring an annual salary through communications and a method thereof and a method for seeking jobs using these |
CN101097573A (en) * | 2006-06-28 | 2008-01-02 | 腾讯科技(深圳)有限公司 | Automatically request-answering system and method |
CN101286161A (en) * | 2008-05-28 | 2008-10-15 | 华中科技大学 | Intelligent Chinese request-answering system based on concept |
CN101465749A (en) * | 2008-12-29 | 2009-06-24 | 武汉大学 | Method for building interlocution service based on Web Service combination |
CN103902652A (en) * | 2014-02-27 | 2014-07-02 | 深圳市智搜信息技术有限公司 | Automatic question-answering system |
CN104050256A (en) * | 2014-06-13 | 2014-09-17 | 西安蒜泥电子科技有限责任公司 | Initiative study-based questioning and answering method and questioning and answering system adopting initiative study-based questioning and answering method |
KR101662450B1 (en) * | 2015-05-29 | 2016-10-05 | 포항공과대학교 산학협력단 | Multi-source hybrid question answering method and system thereof |
CN105608218A (en) * | 2015-12-31 | 2016-05-25 | 上海智臻智能网络科技股份有限公司 | Intelligent question answering knowledge base establishment method, establishment device and establishment system |
CN105824933A (en) * | 2016-03-18 | 2016-08-03 | 苏州大学 | Automatic question-answering system based on theme-rheme positions and realization method of automatic question answering system |
CN106874441A (en) * | 2017-02-07 | 2017-06-20 | 腾讯科技(上海)有限公司 | Intelligent answer method and apparatus |
Non-Patent Citations (1)
Title |
---|
唐素勤等: "基于句型模板的智能问答系统", 《广西师范大学学报:自然科学版》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108897867B (en) | Data processing method, device, server and medium for knowledge question answering | |
CN109981910B (en) | Service recommendation method and device | |
CN108682420B (en) | Audio and video call dialect recognition method and terminal equipment | |
CN110347863B (en) | Speaking recommendation method and device and storage medium | |
CN112365894B (en) | AI-based composite voice interaction method and device and computer equipment | |
WO2019062001A1 (en) | Intelligent robotic customer service method, electronic device and computer readable storage medium | |
CN108962283A (en) | A kind of question terminates the determination method, apparatus and electronic equipment of mute time | |
CN109597874B (en) | Information recommendation method, device and server | |
CN110782962A (en) | Hearing language rehabilitation device, method, electronic equipment and storage medium | |
CN107580155B (en) | Network telephone quality determination method, network telephone quality determination device, computer equipment and storage medium | |
CN108833595B (en) | Computer readable storage medium for online customer service | |
CN111192170B (en) | Question pushing method, device, equipment and computer readable storage medium | |
CN114547293A (en) | Cross-platform false news detection method and system | |
CN110348539B (en) | Short text relevance judging method | |
CN116821290A (en) | Multitasking dialogue-oriented large language model training method and interaction method | |
CN111221945A (en) | Method and device for generating standard question based on user question | |
CN115017289A (en) | Method and device for serving customers based on knowledge base | |
Ono et al. | Lexical acquisition through implicit confirmations over multiple dialogues | |
CN105701208A (en) | Questions and answers evaluation method and device for questions and answers system | |
CN108170780A (en) | A kind of the problem of self-service question and answer matching process and device | |
CN110795630A (en) | Learning scheme recommendation method and device | |
CN111818290B (en) | Online interviewing method and system | |
CN111970311B (en) | Session segmentation method, electronic device and computer readable medium | |
CN111565254B (en) | Call data quality inspection method and device, computer equipment and storage medium | |
CN110399462B (en) | Information query method and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180615 |
|
WD01 | Invention patent application deemed withdrawn after publication |