CN109241261A - User's intension recognizing method, device, mobile terminal and storage medium - Google Patents
User's intension recognizing method, device, mobile terminal and storage medium Download PDFInfo
- Publication number
- CN109241261A CN109241261A CN201811004715.9A CN201811004715A CN109241261A CN 109241261 A CN109241261 A CN 109241261A CN 201811004715 A CN201811004715 A CN 201811004715A CN 109241261 A CN109241261 A CN 109241261A
- Authority
- CN
- China
- Prior art keywords
- user
- keyword
- text
- vectorization
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Abstract
The invention discloses a kind of user's intension recognizing methods, this method comprises: obtaining the dialog text of user's input, dialog text is pre-processed, obtain the keyword in dialog text, keyword is inputted to sample training model, each keyword of vectorization, keyword after input vector is to Naive Bayes Classifier, determine that user is intended to, the invention also discloses a kind of user's intention assessment device, mobile terminal and storage mediums, it can solve the problem of that robot can not provide the answer for meeting user's intention when human-computer dialogue.
Description
Technical field
The present invention relates to interactive field more particularly to a kind of user's intension recognizing methods, device, mobile terminal
And storage medium.
Background technique
With the fast development of interactive, user's intention assessment is particularly important for interactive.Mesh
Before, in live streaming field, session content substantially includes four plates, respectively game, face value, open air and Quadratic Finite Element.So, when
When user and robot chat, robot must determine that user thinks merely which plate first, raw from corresponding plate again afterwards
At reply, the answer for meeting user's intention can be just provided.For example, when user inputs: I likes the song of small elder sister, the sheet of user
Expect and chats face value, and when there is no to carry out user's intention assessment, robot not can determine that user thinks the content of merely which plate, from
And the content of user's game or other plates may be replied, the answer for meeting user's intention cannot be provided, user satisfaction is reduced.
Summary of the invention
The main purpose of the present invention is to provide a kind of user's intension recognizing method, device, mobile terminal and storage medium,
When aiming to solve the problem that human-computer dialogue in the prior art, robot can not provide the problem of answer for meeting user's intention.
To achieve the above object, first aspect of the embodiment of the present invention provides a kind of user's intension recognizing method, comprising:
Obtain the dialog text of user's input；
The dialog text is pre-processed, the keyword in the dialog text is obtained；
The keyword is inputted to sample training model, each keyword of vectorization；
Keyword after inputting the vectorization determines that user is intended to Naive Bayes Classifier.
Second aspect of the embodiment of the present invention provides a kind of user's intention assessment device, comprising:
Module is obtained, for obtaining the dialog text of user's input；
Preprocessing module pre-processes the dialog text, obtains the keyword in the dialog text；
First input module, for inputting the keyword to sample training model, each keyword of vectorization；
Second input module determines user for inputting the keyword after the vectorization to Naive Bayes Classifier
It is intended to.
The third aspect of the embodiment of the present invention provides a kind of mobile terminal, comprising:
Memory, processor and storage are on a memory and the computer program that can run on a processor, feature exist
In the processor realizes user's intension recognizing method that first aspect of the embodiment of the present invention provides when executing described program.
Fourth aspect of the embodiment of the present invention provides a kind of computer readable storage medium, is stored thereon with computer journey
Sequence realizes the user intention assessment side that first aspect of the embodiment of the present invention provides when the computer program is executed by processor
Method.
From the embodiments of the present invention it is found that user's intension recognizing method provided by the invention, device, mobile terminal and depositing
Storage media obtains the dialog text of user's input, dialog text is pre-processed, obtain keyword in dialog text it
Afterwards, input keyword is to sample training model, each keyword of vectorization, the keyword after then inputting the vectorization to simple shellfish
This classifier of leaf, to determine that user is intended to, when can solve human-computer dialogue, robot can not provide the answer for meeting user's intention
Problem.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical 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 skilled in the art without creative efforts, can also basis
These attached drawings obtain other attached drawings.
Fig. 1 is the flow diagram for user's intension recognizing method that first embodiment of the invention provides；
Fig. 2 is the flow diagram for user's intension recognizing method that second embodiment of the invention provides；
Fig. 3 is the structural schematic diagram for user's intention assessment device that third embodiment of the invention provides；
Fig. 4 shows a kind of hardware structure diagram of mobile terminal.
Specific embodiment
In order to make the invention's purpose, features and advantages of the invention more obvious and easy to understand, below in conjunction with the present invention
Attached drawing in embodiment, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described reality
Applying example is only a part of the embodiment of the present invention, and not all embodiments.Based on the embodiments of the present invention, those skilled in the art
Member's every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
Referring to Fig. 1, Fig. 1 is the flow diagram for user's intension recognizing method that first embodiment of the invention provides, it should
Method includes:
S101, the dialog text for obtaining user's input；
In human-computer dialogue field, if user is talked with by way of voice, the voice data of user's input is obtained, is used
Above-mentioned voice data is converted to dialog text by speech-to-text method, convenient for processing.If user is by way of text pair
Words directly acquire the dialog text of user's input, are handled.
It should be noted that speech-to-text method is the prior art, details are not described herein again.
S102, dialog text is pre-processed, obtains the keyword in dialog text；
Dialog text is subjected to word segmentation processing first, each participle in dialog text is obtained, then removes stop words.Its
In, removal stop words can be realized by preset deactivated vocabulary, and the stop words present in the dialog text preset is deactivated at this
When in vocabulary, the stop words is deleted, finally obtains the keyword in the dialog text.
Illustratively, dialog text: I likes the song of small elder sister, carries out word segmentation processing to it first, obtains after participle:
I likes the song of small elder sister, due to exist in default deactivated vocabulary " " word, obtain after removing stop words: I likes
Small elder sister's song.
S103, input keyword to sample training model, each keyword of vectorization；
Sample training model is intended to form by a plurality of corpus text and the corresponding user of corpus text, by improved
Feature Words in corpus text are carried out vectorization processing by TF-IDF algorithm.
Wherein, TF-IDF (term frequency-inverse document frequency) is a kind of for information
The common weighting technique of retrieval and data mining.TF means word frequency (Term Frequency) that IDF means inverse text frequency
Index (Inverse Document Frequency).
Specifically, corpus text is carried out word segmentation processing and removal stop words processing in above-mentioned steps S102, then lead to
Improved TF-IDF algorithm is crossed, the Feature Words in corpus text are subjected to vectorization processing.Enabling corpus text is Di, D=
(t1,t2,...,tm), wherein tjIt is j-th of Feature Words after participle, 1≤j≤m, wijIndicate Feature Words tjIn corpus text Di
In weight, tfijIndicate Feature Words tjIn corpus text DiIn word frequency, njThere is Feature Words t in expressionjCorpus textual data, N
Indicate corpus text sum.
Wherein, log (tfij+ 1.0) TF is indicated,Indicate IDF.
The improved TF-IDF algorithm enables algorithm normally to realize in extreme circumstances.For example, working as tfij、nj、
When N is 0, behind plus 1, it is smoothed, avoids algorithm meaningless.
S104, input vector keyword to Naive Bayes Classifier, determine that user is intended to.
It is Bayes' theorem as follows:
Where it is assumed that the value of Feature Words is independent of one another, i.e., there is no the relationship of interaction between Feature Words.Y={ y1,
y2,...,yiShare i type, X={ x1,x2,...,xnThere is n independent characteristic.
So, due toThe value of P (X) is fixed, therefore compares P (X | Y) * P (Y), i.e. By (yi)=max (P (X |
yi) * P (yi)), it is then, mutually indepedent between each Feature Words according to Bayes' theorem, therefore, P (X | yi)=P (x1, x2, x3...,
xn|yi)=P (x1yi)P(x2yi)...P(xnyi)。
So,
Wherein, P (yi) indicate the total degree of i-th of classification appearance divided by the sum of all categories frequency of occurrence.
P(xn|yi) indicate under i-th of classification, the probability that n-th of feature lexical item occurs.
In the above-mentioned TF-IDF matrix found out, Sum (xn∈yi) need to count under i-th of classification, all the specific words
The sum of weight, Sum (yi) the sum of the weight of this all Feature Words i.e. under i-th of classification.
It is so available to appear in the probability highest of Y classification under that is, each feature set of words determines at P (Y | X), it may be assumed that
Y=max P (x1 | yi) * P (x2 | yi) * ... * P (xn | yi) * P (yi) }
The keyword of input vector can determine that user is intended to above-mentioned Naive Bayes Classifier.
In embodiments of the present invention, the dialog text for obtaining user's input, dialog text is pre-processed, is talked with
After keyword in text, input keyword to sample training model, each keyword of vectorization, after then inputting the vectorization
Keyword to Naive Bayes Classifier, to determine that user is intended to, when can solve human-computer dialogue, robot, which can not provide, to be met
The problem of answer that user is intended to.
Referring to Fig. 2, Fig. 2 is the flow diagram for user's intension recognizing method that second embodiment of the invention provides, it should
Method includes:
S201, the dialog text for obtaining user's input；
S202, dialog text is pre-processed, obtains the keyword in dialog text；
Illustratively, dialog text: I likes the song of small elder sister, carries out word segmentation processing to it first, obtains after participle:
I likes the song of small elder sister, due to exist in default deactivated vocabulary " " word, obtain after removing stop words: I likes
Small elder sister's song.
S203, input keyword to sample training model, each keyword of vectorization；
In sample training model, illustratively, corpus text is denoted as D for two1And D2, then N=2, D1And D2
Corresponding user is intended to for game and face value, and the sample of the two composition is as shown in table 1 below:
Table 1
After carrying out word segmentation processing and removal stop words processing to above-mentioned two corpus texts, obtain:
D1: outstanding this of game direct broadcasting room is marched into the arena out greatly
D2: the small elder sister's song of direct broadcasting room is pleasing to the ear
Therefore all feature set of words are as follows: game direct broadcasting room, Jie Si, marches into the arena, opens large and small elder sister, is song, pleasing to the ear, amounts to 8
A Feature Words.Then, " game " is t1, " direct broadcasting room " is t2, " Jie Si " be t3, " marching into the arena " be t4, " open big " be t5, " small elder sister " be
t6, " song " be t7, it is " pleasing to the ear " be t8。
Then, by each Feature Words tjIt projects in feature set of words, obtains each Feature Words tjIn its corpus text DiIn word
Frequency tfij。
Specifically, in corpus text D1In, t1Occur 1 time, tf11=1；t2Occur 1 time, tf12=1；t3Occur 1 time, tf13
=1；t4Occur 1 time, tf14=1；t5Occur 1 time, tf15=1；t6Occur 0 time, tf16=0；t7Occur 0 time, tf17=0；t8Out
It is 0 time existing, tf18=0.Then, D1=[1,1,1,1,1,0,0,0].
In corpus text D2In, t1Occur 0 time, tf21=1；t2Occur 1 time, tf22=1；t3Occur 0 time, tf23=0；t4Out
It is 0 time existing, tf24=0；t5Occur 0 time, tf25=0；t6Occur 1 time, tf26=1；t7Occur 1 time, tf27=1；t8Occur 2 times, tf28
=2.Then, D2=[0,1,0,0,0,1,1,2].
Then, according to the part TF in improved TF-IDF formula, by tfijAll plus 1 and to counting mode normalization at
Reason, obtains word frequency TF:
D1=[log (2/13), log (2/13), log (2/13), log (2/13), log (2/13), log (1/13), log
(1/13),log(1/13)]
D2=[log (1/13), log (2/13), log (1/13), log (1/13), log (1/13), log (2/13), log
(2/13),log(3/13)]
Further, in feature set of words, " game " is in D1Middle appearance, therefore there is the corpus text of Feature Words " game "
Number is 1, i.e. n1=1." direct broadcasting room " is in D1And D2In occur, therefore occur Feature Words " direct broadcasting room " corpus textual data be 2, i.e.,
n2=2.It is identical, n3=1, n4=1, n5=1, n6=1, n7=1, " pleasing to the ear " is in D2Middle appearance (disregarding frequency of occurrence), therefore go out
The corpus textual data of existing Feature Words " pleasing to the ear " is 1, i.e. n8=1.
Further, by above-mentioned n1=1, n2=2, n3=1, n4=1, n5=1, n6=1, n7=1, n8After=1 brings improvement into
TF-IDF algorithm in the part IDF, find out inverse document frequency IDF are as follows: [log (2/3), log (3/3), log (2/3),
log(2/3),log(2/3),log(2/3),log(2/3),log(2/3)]。
Therefore, improved TF-IDF algorithm, obtained TF-IDF weight matrix, i.e., in corpus text after vectorization are based on
Each Feature Words are as follows:
D1=[log (2/13) * log (2/3), log (2/13) * log (3/3), log (2/13) * log (2/3), log (2/
13)*log(2/3),log(2/13)*log(2/3),log(1/13)*log(2/3),log(1/13)*log(2/3),log(1/
13)*log(2/3)]
D2=[log (1/13) * log (2/3), log (2/13) * log (3/3), log (1/13) * log (2/3), log (1/
13)*log(2/3),log(1/13)*log(2/3),log(2/13)*log(2/3),log(2/13)*log(2/3),log(3/
13)*log(2/3)]
Further, it by Feature Words each in TF-IDF matrix, is projected in feature set of words, even in corresponding corpus text
There is the specific word in this, then the Feature Words after retaining the vectorization.If there is not this in corresponding corpus text
Feature Words after the vectorization are then denoted as 0 by Feature Words, obtain TF-IDF weight matrix:
D1=[log (2/13) * log (2/3), log (2/13) * log (3/3), log (2/13) * log (2/3), log (2/
13)*log(2/3),log(2/13)*log(2/3),0,0,0]
D2=[0, log (2/13) * log (3/3), 0,0,0, log (2/13) * log (2/3), log (2/13) * log (2/
3),2*log(3/13)*log(2/3)]
It should be noted that if there is the specific word in corresponding corpus text, also need to calculate the specific word appearance
Number, for example, above-mentioned " pleasing to the ear " in D2Middle appearance twice, is denoted as 2*log (3/13) * log (2/3).
Above-mentioned steps are the overall process of sample training model realization.
Illustratively, user's input dialogue text is by taking " I likes the song of small elder sister " as an example.It is carried out above-mentioned identical
Word segmentation processing and removal stop words processing, obtain keyword: I likes small elder sister's song.
Further, above-mentioned keyword is input to TF-IDF weight matrix, each keyword of vectorization.Specifically, due to
Keyword " I ", " liking " are not present in above-mentioned matrix, therefore are denoted as 0.There are two kinds of users again to be intended to, therefore is generated after inputting
Two matrixes, conditional probability of the element representation keyword of each matrix in the case where user is intended to:
Game: [0,0, log (1/13) * log (2/3), log (1/13) * log (2/3)]
Face value: [0,0, log (2/13) * log (2/3), log (2/13) * log (2/3)]
Further, for it is above-mentioned there are conditional probability be 0 the case where, will after all conditions probability add 1, therefore
It arrives:
Game: [1,1, log (1/13) * log (2/3)+1, log (1/13) * log (2/3)+1]
Face value: [1,1, log (2/13) * log (2/3)+1, log (2/13) * log (2/3)+1]
Keyword after S204, input vector determines that user is intended to Naive Bayes Classifier.
In the case where mutually indepedent between each keyword, the Bayesian formula as shown in the embodiment of the present invention one, i.e.,
Naive Bayes Classifier calculates separately dialog text and corresponds to user's intention probability that each user is intended to, obtains dialog text pair
The user for answering each user to be intended to is intended to probability value.
Specifically, P (game | I likes small elder sister's song)=P (I | game) P (like | game) P (small elder sister | trip
Play) and P (song | game) * P (game)=1*1* (log (1/13) * log (2/3)+1) * (log (1/13) * log (2/3)+1) * 1/
2
P (face value | I likes small elder sister's song)=P (I | face value) P (like | face value) P (small elder sister | face value) P (song
Sound | face value) * P (face value)=1*1* (log (2/13) * log (2/3)+1) * (log (2/13) * log (2/3)+1) * 1/2
Further, more above-mentioned each user is intended to probability value, can obtain:
P (face value | I likes small elder sister's song) > P (game | I likes small elder sister's song)
Finally can be obtained from result above, user be intended to the maximum user of probability value be intended to face value, using face value as pair
The user for talking about text is intended to output.
In embodiments of the present invention, the dialog text for obtaining user's input, dialog text is pre-processed, is talked with
After keyword in text, input keyword to sample training model, each keyword of vectorization, after then inputting the vectorization
Keyword to Naive Bayes Classifier, to determine that user is intended to, when can solve human-computer dialogue, robot, which can not provide, to be met
The problem of answer that user is intended to.
Referring to Fig. 3, Fig. 3 is the structural schematic diagram for user's intention assessment device that third embodiment of the invention provides, it should
Device includes:
Obtain module 301, preprocessing module 302, the first input module 303 and the second input module 304.
Module 301 is obtained, for obtaining the dialog text of user's input.
Preprocessing module 302, dialog text is pre-processed, and obtains the keyword in the dialog text.
First input module 303, for inputting keyword to sample training model, each keyword of vectorization.
Second input module 304 determines that user anticipates for the keyword after input vector to Naive Bayes Classifier
Figure.
In embodiments of the present invention, the dialog text for obtaining user's input, dialog text is pre-processed, is talked with
After keyword in text, input keyword to sample training model, each keyword of vectorization, after then inputting the vectorization
Keyword to Naive Bayes Classifier, to determine that user is intended to, when can solve human-computer dialogue, robot, which can not provide, to be met
The problem of answer that user is intended to.
Fig. 4 is referred to, Fig. 4 shows a kind of hardware structure diagram of mobile terminal.
Mobile terminal as described in this embodiment, comprising:
Memory 41, processor 42 and it is stored in the computer program that can be run on memory 41 and on a processor, located
Reason device realizes user's intension recognizing method described in aforementioned embodiment illustrated in fig. 1 when executing the program.
Further, the mobile terminal further include:
At least one input equipment 43；At least one output equipment 44.
Above-mentioned memory 41,42 input equipment 43 of processor and output equipment 44 are connected by bus 45.
Wherein, input equipment 43 concretely camera, touch panel, physical button or mouse etc..Output equipment
44 concretely display screens.
Memory 41 can be high random access memory body (RAM, Random Access Memory) memory, can also
For non-labile memory (non-volatile memory), such as magnetic disk storage.Memory 41 can for storing one group
Program code is executed, processor 42 is coupled with memory 41.
Further, the embodiment of the invention also provides a kind of computer readable storage medium, the computer-readable storages
Medium can be in the terminal being set in the various embodiments described above, which can be shown in earlier figures 4
Memory in embodiment.It is stored with computer program on the computer readable storage medium, when which is executed by processor
Realize user's intension recognizing method described in aforementioned embodiment illustrated in fig. 1.Further, the computer can storage medium may be used also
To be USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random
Access Memory), the various media that can store program code such as magnetic or disk.
It should be noted that each functional module in each embodiment of the present invention can integrate in a processing module
In, it is also possible to modules and physically exists alone, can also be integrated in two or more modules in a module.It is above-mentioned
Integrated module both can take the form of hardware realization, can also be realized in the form of software function module.
If the integrated module is realized in the form of software function module and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies.
It should be noted that for the various method embodiments described above, describing for simplicity, therefore, it is stated as a series of
Combination of actions, but those skilled in the art should understand that, the present invention is not limited by the sequence of acts described because
According to the present invention, certain steps can use other sequences or carry out simultaneously.Secondly, those skilled in the art should also know
It knows, the embodiments described in the specification are all preferred embodiments, and related actions and modules might not all be this hair
Necessary to bright.
In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, there is no the portion being described in detail in some embodiment
Point, it may refer to the associated description of other embodiments.
The above are to a kind of user's intension recognizing method provided by the present invention, device, mobile terminal and storage medium
Description, for those skilled in the art, thought according to an embodiment of the present invention, in specific embodiments and applications
It will change, to sum up, the contents of this specification are not to be construed as limiting the invention.
Claims (10)
1. a kind of user's intension recognizing method, which is characterized in that the described method includes:
Obtain the dialog text of user's input；
The dialog text is pre-processed, the keyword in the dialog text is obtained；
The keyword is inputted to sample training model, each keyword of vectorization；
Keyword after inputting the vectorization determines that user is intended to Naive Bayes Classifier.
2. obtaining institute the method according to claim 1, wherein described pre-process the dialog text
The keyword stated in dialog text includes:
Word segmentation processing is carried out to the dialog text, obtains each participle in the dialog text；
According to preset deactivated vocabulary, the existing stop words in each participle is removed, the pass in the dialog text is obtained
Keyword.
3. method according to claim 1 or 2, which is characterized in that the keyword of the input vectorization to simplicity
Bayes classifier determines that user is intended to include:
Using Naive Bayes Classifier, calculates separately the dialog text and correspond to user's intention probability that each user is intended to, obtain
The user that each user is intended to, which is corresponded to, to dialog text is intended to probability value；
Compare each user and be intended to probability value, the user is intended to the maximum user of probability value and is intended to as the dialogue text
This user is intended to output.
4. according to the method described in claim 3, it is characterized in that, the sample training model uses a plurality of corpus text and institute
The sample that the corresponding user of predicate material text is intended to composition is trained, and the Feature Words in corpus text described in vectorization obtain
The sample training model.
5. according to the method described in claim 4, it is characterized in that, Feature Words in corpus text described in the vectorization, obtain
Include: to sample training model
The corpus text is pre-processed, the Feature Words in the corpus text are obtained；
By improved TF-IDF algorithm, each Feature Words of vectorization obtain the TF-IDF weight matrix of sample training；
DiFor the corpus text, D=(t1,t2,...,tm), wherein tjIt is j-th of Feature Words after participle, 1≤j≤m, wij
Indicate Feature Words tjIn corpus text DiIn weight, tfijIndicate Feature Words tjIn corpus text DiIn word frequency, njIt represents
Existing Feature Words tjCorpus textual data, N indicate corpus text sum.
6. according to the method described in claim 5, it is characterized in that, the input keyword to sample training model, to
Quantifying each keyword includes:
Input the TF-IDF weight matrix of the keyword to sample training, each keyword of vectorization.
7. according to the method described in claim 6, it is characterized in that, the value of each keyword after there are the vectorization is 0
When, the value of the keyword after each vectorization is added one.
8. a kind of user's intention assessment device, which is characterized in that described device includes:
Module is obtained, for obtaining the dialog text of user's input；
Preprocessing module pre-processes the dialog text, obtains the keyword in the dialog text；
First input module, for inputting the keyword to sample training model, each keyword of vectorization；
Second input module determines that user is intended to for inputting the keyword after the vectorization to Naive Bayes Classifier.
9. a kind of mobile terminal, comprising: memory, processor and storage are on a memory and the calculating that can run on a processor
Machine program, which is characterized in that when the processor executes the computer program, realize any one of claim 1 to 7 institute
The each step in user's intension recognizing method stated.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
When being executed by processor, each step in user's intension recognizing method described in any one of claim 1 to 7 is realized.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811004715.9A CN109241261A (en) | 2018-08-30 | 2018-08-30 | User's intension recognizing method, device, mobile terminal and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811004715.9A CN109241261A (en) | 2018-08-30 | 2018-08-30 | User's intension recognizing method, device, mobile terminal and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109241261A true CN109241261A (en) | 2019-01-18 |
Family
ID=65068985
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811004715.9A Pending CN109241261A (en) | 2018-08-30 | 2018-08-30 | User's intension recognizing method, device, mobile terminal and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109241261A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109840276A (en) * | 2019-02-12 | 2019-06-04 | 北京健康有益科技有限公司 | Intelligent dialogue method, apparatus and storage medium based on text intention assessment |
| CN110113422A (en) * | 2019-05-10 | 2019-08-09 | 南京硅基智能科技有限公司 | A kind of intension recognizing method and system of the virtual mobile phone based on cloud |
| CN110909543A (en) * | 2019-11-15 | 2020-03-24 | 广州洪荒智能科技有限公司 | Intention recognition method, device, equipment and medium |
| CN111324727A (en) * | 2020-02-19 | 2020-06-23 | 百度在线网络技术（北京）有限公司 | User intention recognition method, device, equipment and readable storage medium |
| CN111798853A (en) * | 2020-03-27 | 2020-10-20 | 北京京东尚科信息技术有限公司 | Method, device, equipment and computer readable medium for speech recognition |
| CN111985206A (en) * | 2020-07-17 | 2020-11-24 | 联想(北京)有限公司 | Corpus understanding method and equipment |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140172899A1 (en) * | 2012-12-14 | 2014-06-19 | Microsoft Corporation | Probability-based state modification for query dialogues |
| CN105389307A (en) * | 2015-12-02 | 2016-03-09 | 上海智臻智能网络科技股份有限公司 | Statement intention category identification method and apparatus |
| CN106599278A (en) * | 2016-12-23 | 2017-04-26 | 北京奇虎科技有限公司 | Identification method and method of application search intention |
| CN107346340A (en) * | 2017-07-04 | 2017-11-14 | 北京奇艺世纪科技有限公司 | A kind of user view recognition methods and system |
| CN107608999A (en) * | 2017-07-17 | 2018-01-19 | 南京邮电大学 | A kind of Question Classification method suitable for automatically request-answering system |
| CN107807968A (en) * | 2017-10-13 | 2018-03-16 | 上海壹账通金融科技有限公司 | Question and answer system, method and storage medium based on Bayesian network |
-
2018
- 2018-08-30 CN CN201811004715.9A patent/CN109241261A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140172899A1 (en) * | 2012-12-14 | 2014-06-19 | Microsoft Corporation | Probability-based state modification for query dialogues |
| CN105389307A (en) * | 2015-12-02 | 2016-03-09 | 上海智臻智能网络科技股份有限公司 | Statement intention category identification method and apparatus |
| CN106599278A (en) * | 2016-12-23 | 2017-04-26 | 北京奇虎科技有限公司 | Identification method and method of application search intention |
| CN107346340A (en) * | 2017-07-04 | 2017-11-14 | 北京奇艺世纪科技有限公司 | A kind of user view recognition methods and system |
| CN107608999A (en) * | 2017-07-17 | 2018-01-19 | 南京邮电大学 | A kind of Question Classification method suitable for automatically request-answering system |
| CN107807968A (en) * | 2017-10-13 | 2018-03-16 | 上海壹账通金融科技有限公司 | Question and answer system, method and storage medium based on Bayesian network |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109840276A (en) * | 2019-02-12 | 2019-06-04 | 北京健康有益科技有限公司 | Intelligent dialogue method, apparatus and storage medium based on text intention assessment |
| CN110113422A (en) * | 2019-05-10 | 2019-08-09 | 南京硅基智能科技有限公司 | A kind of intension recognizing method and system of the virtual mobile phone based on cloud |
| CN110909543A (en) * | 2019-11-15 | 2020-03-24 | 广州洪荒智能科技有限公司 | Intention recognition method, device, equipment and medium |
| CN111324727A (en) * | 2020-02-19 | 2020-06-23 | 百度在线网络技术（北京）有限公司 | User intention recognition method, device, equipment and readable storage medium |
| US11646016B2 (en) | 2020-02-19 | 2023-05-09 | Baidu Online Network Technology (Beijing) Co., Ltd. | Method and apparatus for recognizing user intention, device, and readable storage medium |
| CN111798853A (en) * | 2020-03-27 | 2020-10-20 | 北京京东尚科信息技术有限公司 | Method, device, equipment and computer readable medium for speech recognition |
| CN111985206A (en) * | 2020-07-17 | 2020-11-24 | 联想(北京)有限公司 | Corpus understanding method and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109241261A (en) | User's intension recognizing method, device, mobile terminal and storage medium | |
| AU2016256764B2 (en) | Semantic natural language vector space for image captioning | |
| GB2547068B (en) | Semantic natural language vector space | |
| US10049103B2 (en) | Author personality trait recognition from short texts with a deep compositional learning approach | |
| CN109923556B (en) | Pointer Sentinel Hybrid Architecture | |
| US9811765B2 (en) | Image captioning with weak supervision | |
| CN107273438B (en) | Recommendation method, device, equipment and storage medium | |
| WO2020177282A1 (en) | Machine dialogue method and apparatus, computer device, and storage medium | |
| RU2678716C1 (en) | Use of autoencoders for learning text classifiers in natural language | |
| US10354182B2 (en) | Identifying relevant content items using a deep-structured neural network | |
| US10635858B2 (en) | Electronic message classification and delivery using a neural network architecture | |
| US9875294B2 (en) | Method and apparatus for classifying object based on social networking service, and storage medium | |
| Zhao et al. | Shap values for explaining cnn-based text classification models | |
| CN110741364A (en) | Determining a state of an automated assistant dialog | |
| CN107112005A (en) | Depth nerve SVMs | |
| US10685012B2 (en) | Generating feature embeddings from a co-occurrence matrix | |
| CN110795542A (en) | Dialogue method and related device and equipment | |
| WO2019220113A1 (en) | Device and method for natural language processing | |
| WO2014073206A1 (en) | Information-processing device and information-processing method | |
| CN111967599B (en) | Method, apparatus, electronic device and readable storage medium for training model | |
| CN110390052A (en) | Search for recommended method, the training method of CTR prediction model, device and equipment | |
| CN113505198A (en) | Keyword-driven generating type dialogue reply method and device and electronic equipment | |
| CN113392179A (en) | Text labeling method and device, electronic equipment and storage medium | |
| US11561964B2 (en) | Intelligent reading support | |
| CN110781666B (en) | Natural language processing text modeling based on generative antagonism network |
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: 20190118 |
|
| RJ01 | Rejection of invention patent application after publication |