CN105630917A - Intelligent answering method and intelligent answering device - Google Patents
Intelligent answering method and intelligent answering device Download PDFInfo
- Publication number
- CN105630917A CN105630917A CN201510970008.5A CN201510970008A CN105630917A CN 105630917 A CN105630917 A CN 105630917A CN 201510970008 A CN201510970008 A CN 201510970008A CN 105630917 A CN105630917 A CN 105630917A
- Authority
- CN
- China
- Prior art keywords
- answer
- keyword
- customer
- matching degree
- disaggregated model
- 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
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention provides an intelligent answering method and an intelligent answering device. The method comprises the following steps of receiving a customer problem sent by a client; extracting a keyword in the customer problem; obtaining at least one answer preset by aiming at the keyword; by aiming at each answer, building feature vectors according to the customer problem, the keyword and the answer; predicting matching degrees of the answer corresponding to the feature vector and the customer problem according to each built feature vector by using a preset classification model; and when at least one matching degree in the calculated matching degrees is greater than a first preset threshold value, returning the answer corresponding to the matching degree with the greatest value in the matching degrees greater than the first preset threshold value as the answer for answering the customer problem through the client to the client. The method and the device have the advantages that the accuracy rate of matching the answer with the problem can be improved when the client problem is automatically answered.
Description
Technical field
The present invention relates to computer internet field, in particular to a kind of intelligent response method and device.
Background technology
At present, in electricity business's transaction system, a large amount of businessmans adopt robot chat tool automatically to answer the various problems that buyer proposes. The key word matching method of the concrete character string aspect that the is based on coupling adopted, for instance, businessman presets keyword " logistics+information " and the answer that sets for this keyword for " side parent follows up here, urges in time! ". Therefore, in the process using robot chat tool and buyer to talk with, as long as the language of buyer occurs " logistics " and " information " two words simultaneously, for instance " how but without logistics information ", robot chat tool will automatically reply " here side parent follow-up, urge in time! ".
But, such answer provided under many circumstances based on the Keywords matching scheme of character string aspect can not be well matched with the problem that buyer proposes, and even there is the situation given an irrelevant answer. It is to say, the accuracy rate that the Keywords matching scheme that in prior art, robot chat tool adopts when automatic-answering back device has answer matches problem is relatively low, it is impossible to the problem better meeting dialogue demand.
Summary of the invention
Given this, it is an object of the invention to provide a kind of intelligent response method and device, there is the problem that the accuracy rate of answer matches problem is relatively low, can not better meet dialogue demand improving the Keywords matching scheme that in prior art, robot chat tool adopts when automatic-answering back device.
To achieve these goals, the technical scheme that the embodiment of the present invention adopts is as follows:
First aspect, embodiments provides a kind of intelligent response method, including: receive the customer issue that client sends, extract the keyword in described customer issue; Obtain at least one answer set in advance of described keyword; For answer each described, according to described customer issue, described keyword and this answer construction feature vector; Utilize the disaggregated model preset, predict the matching degree of the answer corresponding with this characteristic vector and described customer issue according to each constructed characteristic vector; When having at least one matching degree in calculated matching degree more than the first predetermined threshold value, will be greater than the answer that matching degree that in the matching degree of the first predetermined threshold value, numerical value is maximum is corresponding, as being used for returning to described client by the answer of customer issue described in described client response.
Second aspect, the embodiment of the present invention additionally provides a kind of intelligent response device, including: extraction module, for receiving the customer issue that client sends, extracts the keyword in described customer issue; Acquisition module, for obtaining at least one answer set in advance of described keyword; Build module, for for answer each described, according to described customer issue, described keyword and this answer construction feature vector; Prediction module, for utilizing default disaggregated model, predicting the matching degree of the answer corresponding with this characteristic vector and described customer issue according to each constructed characteristic vector; Choose module, during for having at least one matching degree in calculated matching degree more than the first predetermined threshold value, will be greater than the answer that matching degree that in the matching degree of the first predetermined threshold value, numerical value is maximum is corresponding, as being used for returning to described client by the answer of customer issue described in described client response.
The intelligent response method of embodiment of the present invention offer and device, by extracting the keyword in the customer issue received, according to described customer issue, described keyword and build corresponding characteristic vector for each answer set in advance of this keyword, utilize default disaggregated model afterwards, the matching degree of the corresponding answer of this characteristic vector and described customer issue is predicted according to each constructed characteristic vector, only by answer corresponding for matching degree maximum more than numerical value in the matching degree of the first predetermined threshold value in calculated matching degree, as the answer of customer issue described in response. thus, adopt the robot chat tool of such scheme can improve the accuracy rate of answer matches problem when automatic-answering back device, meet dialogue demand preferably.
For making the above and other purpose of the present invention, feature and advantage to become apparent, preferred embodiment cited below particularly, and coordinate institute's accompanying drawings, it is described in detail below.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, the accompanying drawing used required in embodiment will be briefly described below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the premise not paying creative work, it is also possible to obtain other accompanying drawing according to these accompanying drawings. Shown in accompanying drawing, above-mentioned and other purpose, feature and the advantage of the present invention will become apparent from. The part that accompanying drawing labelling instruction identical in whole accompanying drawings is identical. Deliberately do not draw accompanying drawing by actual size equal proportion convergent-divergent, it is preferred that emphasis is the purport of the present invention is shown.
Fig. 1 is the applied environment schematic diagram of the embodiment of the present invention;
Fig. 2 illustrates the structured flowchart of a kind of server that can be applicable to the embodiment of the present invention;
Fig. 3 illustrates the flow chart of the intelligent response method that first embodiment of the invention provides;
Fig. 4 illustrates about customer issue, the keyword extracted and several the example data for the answer set in advance of extracted keyword;
Fig. 5 illustrates several exemplary sample data;
Fig. 6 illustrates the schematic diagram of the intelligent response device that second embodiment of the invention provides.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete description, it is clear that described embodiment is only a part of embodiment of the present invention, rather than whole embodiments. Based on the embodiment in the present invention, the every other embodiment that those of ordinary skill in the art obtain under not making creative work premise, broadly fall into the scope of protection of the invention.
It should also be noted that similar label and letter below figure represent similar terms, therefore, once a certain Xiang Yi accompanying drawing is defined, then it need not be carried out definition further and explain in accompanying drawing subsequently. Meanwhile, in describing the invention, term " first ", " second " etc. are only used for distinguishing description, and it is not intended that indicate or hint relative importance.
The following each embodiment of the present invention all can be applicable in environment as shown in Figure 1. As it is shown in figure 1, server 100 is communicatively coupled by network 300 and one or more user terminals 200, to carry out data communication or mutual. Described server 100 can be multiple servers such as instant communication server, the webserver, database server, authentication server, it is also possible to is a server. Described user terminal 200 can be PC (personalcomputer, PC), panel computer, smart mobile phone, personal digital assistant (personaldigitalassistant, PDA) etc. User terminal 200 and server 100 can set up communication connection in Wi-Fi (Wireless Fidelity) network, 2G/3G/4G network or LAN.
Fig. 2 illustrates the structured flowchart of a kind of server that can be applicable in the embodiment of the present invention. As in figure 2 it is shown, described server 100 can include intelligent response device, memorizer 102, storage control 103, processor 104 and the mixed-media network modules mixed-media 105 that the embodiment of the present invention provides.
Electrically connect directly or indirectly between memorizer 102, storage control 103, processor 104, each element of mixed-media network modules mixed-media 105, to realize the transmission of data or mutual. Such as, one or more communication bus can be passed through between these elements or signal bus realizes electrical connection. Described intelligent response device includes at least one can be stored in the software function module in memorizer 102 with the form of software or firmware (firmware), for instance software function module that described intelligent response device includes or computer program.
Memorizer 102 can store various software program and module, the intelligent response method provided such as the embodiment of the present invention and programmed instruction/module corresponding to device, processor 104 is by running the software program and module stored in the memory 102, thus performing the application of various function and data process, namely realize the intelligent response method in the embodiment of the present invention. Memorizer 102 can include but not limited to random access memory (RandomAccessMemory, RAM), read only memory (ReadOnlyMemory, ROM), programmable read only memory (ProgrammableRead-OnlyMemory, PROM), erasable read-only memory (ErasableProgrammableRead-OnlyMemory, EPROM), electricallyerasable ROM (EEROM) (ElectricErasableProgrammableRead-OnlyMemory, EEPROM) etc.
Processor 104 can be a kind of IC chip, has signal handling capacity. Above-mentioned processor can be general processor, including central processing unit (CentralProcessingUnit is called for short CPU), network processing unit (NetworkProcessor is called for short NP) etc.; Can also is that digital signal processor (DSP), special IC (ASIC), ready-made programmable gate array (FPGA) or other PLDs, discrete gate or transistor logic, discrete hardware components. It can realize or perform the disclosed each method in the embodiment of the present invention, step and logic diagram. The processor etc. that general processor can be microprocessor or this processor can also be any routine.
Mixed-media network modules mixed-media 105 is used for receiving and sending network signal. Above-mentioned network signal can include wireless signal or wire signal.
Being appreciated that the structure shown in Fig. 2 is only signal, server 100 can also include the assembly more or more less than shown in Fig. 2, or has the configuration different from shown in Fig. 2. Each assembly shown in Fig. 2 can adopt hardware, software or its combination to realize. It addition, the server in the embodiment of the present invention can also include the server of multiple concrete difference in functionality.
In embodiments of the present invention, being provided with client in user terminal 200, this client can be third-party application software (such as Ali Wang Wang), corresponding with server end, thus providing the user service, for instance instant messaging, chat. The embodiment of the present invention is mainly based upon the scene using buyer/customer issue involved by robot chat tool response electricity business's transaction system at server end, provides satisfied answer for buyer/client.
First embodiment
Fig. 3 illustrates the flow chart of the intelligent response method that first embodiment of the invention provides. Referring to Fig. 3, the intelligent response method that first embodiment provides may include that
Step S11, receives the customer issue that client sends, extracts the keyword in described customer issue.
When buyer such as wants about certain part commodity consultation businessman, the client (Ali Wang Wang as installed on mobile phone) that buyer can pass through to install on its user terminal held sends relevant issues. After receiving the customer issue that buyer is sent by client, server (can be specifically the server being mounted with the robot chat tool for automatic-answering back device that businessman uses) can extract the keyword in described customer issue after text-processing (such as word segmentation processing etc.). Described keyword can be that businessman is pre-set, and being set in the memorizer being stored in afterwards included by server. Therefore, server after receiving customer issue, can identify whether contain businessman's keyword set in advance in described customer issue, and when recognizing such keyword by this keyword extraction out.
Step S12, obtains at least one answer set in advance of described keyword.
After extracting described keyword, server can obtain be stored in data base at least one answer set in advance of described keyword. Described answer is also set in advance by businessman. It should be noted that when extracting multiple keyword from customer issue, each keyword is obtained for its at least one answer set in advance. Refer to Fig. 4, it is shown that about customer issue, the keyword extracted, several example data for the answer set in advance of extracted keyword, the example shown in Fig. 4 has preset only one answer for each keyword.
Step S13, for answer each described, according to described customer issue, described keyword and this answer construction feature vector.
In a kind of detailed description of the invention, the characteristic vector built for answer each described can include at least one in following component: the linked character of the feature of described keyword, the feature of described customer issue, the feature of described answer, described customer issue and the linked character of described keyword, described customer issue and the linked character of described answer, described keyword and described answer. Preferably, the characteristic vector built for answer each described includes each in above-mentioned component.
Specifically, the feature of described keyword can include the length of described keyword (keyword is more long, and its linguistic information comprised is more many; When extracting longer keyword, customer issue more likely sets the problem of anticipation when keyword is answered and is consistent with businessman), (word frequency of keyword is more little, and its linguistic information comprised is more many for the word frequency of described keyword; When extracting longer keyword, customer issue more likely sets the problem of anticipation when keyword is answered and is consistent with businessman), the number of word that described keyword obtains after participle, numerical character number in described keyword account for the ratio that the chinese character number in the length of chinese character in the ratio of total number of characters of this keyword, described keyword and described keyword accounts for total number of characters of this keyword.
The number of the word that the feature of described customer issue can include the number of the chinese character in described customer issue and described customer issue obtains after participle.
Whether number and the described answer of the word that the feature of described answer can include the number of the comprised character of described answer, described answer obtains after participle include described keyword.
The number of same words that the linked character of described customer issue and described keyword can include described customer issue and described keyword obtains after participle, described customer issue and described keyword are based on the editing distance based on word of character edit distance with described keyword of the cosine similarity of reverse document frequency (InverseDocumentFrequency, IDF) value of word, described customer issue and described customer issue and described keyword.
The linked character of described customer issue and described answer includes the editing distance based on word of the number of same words that described customer issue obtains after participle, described customer issue and the described answer character edit distance based on the cosine similarity of the IDF value of word, described customer issue and described answer and described customer issue and described answer with described answer.
The editing distance based on word of character edit distance based on the cosine similarity of the IDF value of word, described keyword and described answer of the number of same words that the linked character of described keyword and described answer can include described keyword and described answer obtains after participle, described keyword and described answer and described keyword and described answer.
Specifically, such as described customer issue and described keyword may include that described customer issue and described keyword participle based on the calculating of the cosine similarity of the IDF value of word, add up the IDF value of each word that described customer issue obtains after participle respectively and the IDF value of each word that described keyword obtains after participle, for described customer issue, the IDF vector of described customer issue is built using its all words as vector space, and for described keyword, the IDF vector of described keyword is built using its all words as vector space, then described customer issue and the described keyword cosine similarity based on the IDF value of word is calculated according to following formula:
Wherein, similarity represents described customer issue and the described keyword cosine similarity based on the IDF value of word, AiRepresent the IDF value of each word that described customer issue obtains after participle, BiRepresent the IDF value of each word that described keyword obtains after participle. Described customer issue and described answer can adopt similar approach to calculate based on the cosine similarity of the IDF value of word, described keyword and described answer based on the cosine similarity of the IDF value of word.
In embodiments of the present invention, described character edit distance refers to that the Levenstein of character compares distance.
It should be noted that if extracting multiple keyword from customer issue, then should about each keyword with for its answer preset corresponding characteristic vector of structure.
Step S14, utilizes the disaggregated model preset, predicts the matching degree of the answer corresponding with this characteristic vector and described customer issue according to each constructed characteristic vector.
In a kind of detailed description of the invention, described default disaggregated model is pre-build by following step: the multiple sample data of automatic-answering back device session establishment according to historical record, and each described sample data includes in the historic customer problem extracted from described dialogue, the keyword extracted from this historic customer problem, described dialogue for this keyword identification matching degree answering answer and described answer answer and this historic customer problem for answering this historic customer problem set in advance; The plurality of sample data is divided into one group of training sample data and one group of test specimens notebook data; The described one group of training sample data training grader for identifying answer matches degree is utilized to obtain housebroken disaggregated model, utilize described one group of test specimens notebook data to check described housebroken disaggregated model, and when assay meets pre-conditioned, described housebroken disaggregated model is set as described default disaggregated model.
Specifically, the automatic-answering back device dialogue of described historical record can be the real dialog extracted from businessman's daily record. The identification matching degree answering answer and historic customer problem that described sample data includes is to be demarcated according to the standard intraocular of normal chat by businessman. When answer answer matches with historic customer problem, businessman can set that the identification matching degree both it is 1; When answer answer is not mated with historic customer problem, businessman can set that the identification matching degree both it is 0. Such as, when buyer inquiry historic customer problem for " being dealt into how many days Shenzhen needs? " time, the answer of answering of " needing to two days " can be the answer mated; When buyer inquiry historic customer problem for " how also not delivering? " time, the answer of answering of " order arranges " can be the answer mated. Refer to Fig. 5, it is shown that several exemplary sample data.
In a kind of detailed description of the invention, described described one group of training sample data training grader for identifying answer matches degree is utilized to obtain housebroken disaggregated model, may include that for each the training sample data in described one group of training sample data, build training feature vector according to historic customer problem, keyword and the answer answer that these training sample data include; Identification matching degree in training feature vector each described and the training sample data corresponding to this training feature vector is performed classifier algorithm, obtains housebroken disaggregated model.
In a kind of detailed description of the invention, described described one group of test specimens notebook data is utilized to check described housebroken disaggregated model, and when assay meets pre-conditioned, described housebroken disaggregated model is set as described default disaggregated model, may include that for each the test specimens notebook data in described one group of test specimens notebook data, build testing feature vector according to historic customer problem, keyword and the answer answer that this test specimens notebook data includes; Utilize described housebroken disaggregated model that testing feature vector each described is predicted, obtain corresponding to the historic customer problem in the test specimens notebook data of this testing feature vector and the prediction and matching degree answering answer; Calculate the difference between the identification matching degree relevant to test specimens notebook data each described and test matching degree, add up the first number of the obtained difference less than the second predetermined threshold value, and set described housebroken disaggregated model as described default disaggregated model when the percentage ratio of total number that described first number accounts for described test specimens notebook data exceedes preset percentage. Described second predetermined threshold value and described preset percentage all can be set by the user, and the percentage ratio of the total number that described first number accounts for described test specimens notebook data is more high, represent that the prediction accuracy of housebroken disaggregated model is more high.
Specifically, for instance the most of sample datas in multiple sample datas can be used as training sample data, remaining fraction sample data is used as test specimens notebook data. each described training feature vector can include at least one in following component: the feature of the keyword in training sample data corresponding with this training feature vector, the feature of the historic customer problem in these training sample data, the feature answering answer in these training sample data, the linked character of the historic customer problem in these training sample data and keyword, historic customer problem in these training sample data and the linked character answering answer, keyword in these training sample data and the linked character answering answer.
Each described testing feature vector can include at least one in following component: the feature of the keyword in test specimens notebook data corresponding with this testing feature vector, the feature of the historic customer problem in this test specimens notebook data, the feature answering answer in this test specimens notebook data, the linked character of the historic customer problem in this test specimens notebook data and keyword, historic customer problem in this test specimens notebook data and the linked character answering answer, keyword in this test specimens notebook data and the linked character answering answer. the characteristic vector described above that is configured similarly to of training feature vector and testing feature vector builds, and no longer repeats at this.
The classifier algorithm adopted can be logistic regression algorithm, boosting sorting algorithm or other suitable algorithms, and the housebroken disaggregated model that correspondence obtains can be logistic regression disaggregated model, boosting disaggregated model etc. Preferably select boosting sorting algorithm and corresponding boosting disaggregated model. Boosting disaggregated model belongs to integrated study model, and its basic thought is that tree-model relatively low for hundreds and thousands of classification accuracies is combined, and becomes a significantly high model of accuracy rate. This model can constantly iteration, each iteration is generated as a new tree. The embodiment of the present invention is preferably by GradientBoosting disaggregated model. This disaggregated model adopts the thought of gradient decline based on all trees being previously created when generating every one tree, makes a move towards the direction making given the minimization of object function more. Can adopting the open source projects xgboost (https: //github.com/dmlc/xgboost) of github, for above-mentioned sample data, arameter optimization is for using 400 decision trees, and the depth capacity of each tree is 3. Utilizing test specimens notebook data that the boosting disaggregated model trained is tested, it is known that the predictablity rate of boosting disaggregated model is 87%, recall rate is 96% (performance is better than Logic Regression Models).
Once establish default disaggregated model, it is possible to predict the matching degree of the answer corresponding with this characteristic vector and described customer issue according to each constructed characteristic vector with it.
Step S15, when having at least one matching degree in calculated matching degree more than the first predetermined threshold value, will be greater than the answer that matching degree that in the matching degree of the first predetermined threshold value, numerical value is maximum is corresponding, as being used for returning to described client by the answer of customer issue described in described client response.
Described first predetermined threshold value can be set by the user, for instance is set to 0.5. Specifically, if calculated matching degree having a matching degree more than the first predetermined threshold value, then show that the corresponding answer of this matching degree matches with customer issue, it is possible to using answer corresponding for this matching degree as the answer being used for customer issue described in response. If calculated matching degree having multiple matching degree more than the first predetermined threshold value, then can will be greater than the corresponding answer of matching degree that in the matching degree of the first predetermined threshold value, numerical value is maximum as the answer of customer issue described in response, it is to say, select the answer that in the answer of Satisfying Matching Conditions, an answer mating most is the most final. If calculated matching degree is respectively less than or is equal to the first predetermined threshold value, then show answer set in advance does not have answer and customer issue match, now do not select for keyword answer acknowledged client problem set in advance, and general answer acknowledged client problem can be adopted, such as " it is sorry; the problem that still can not answer you at present, we can actively handle through consultation, please bear with! " etc.
In the intelligent response method that first embodiment of the invention provides, by extracting the keyword in the customer issue received, according to described customer issue, described keyword and build corresponding characteristic vector for each answer set in advance of this keyword, utilize default disaggregated model afterwards, the matching degree of the corresponding answer of this characteristic vector and described customer issue is predicted according to each constructed characteristic vector, only by answer corresponding for matching degree maximum more than numerical value in the matching degree of the first predetermined threshold value in calculated matching degree, as the answer of customer issue described in response. thus, adopt the robot chat tool of such scheme can improve the accuracy rate of answer matches problem when automatic-answering back device, meet dialogue demand preferably.
Second embodiment
Second embodiment of the invention provides a kind of intelligent response device. Fig. 6 has illustrated the schematic diagram of the intelligent response device that second embodiment of the invention provides. Referring to Fig. 6, the intelligent response device 400 that the second embodiment provides can include extraction module 410, acquisition module 420, builds module 430, prediction module 440 and choose module 450.
Extraction module 410, for receiving the customer issue that client sends, extracts the keyword in described customer issue.
Acquisition module 420 is for obtaining at least one answer set in advance of described keyword.
Build module 430 for for answer each described, according to described customer issue, described keyword and this answer construction feature vector.
Prediction module 440 is for utilizing default disaggregated model, predicting the matching degree of the answer corresponding with this characteristic vector and described customer issue according to each constructed characteristic vector.
When choosing module 450 for having at least one matching degree in calculated matching degree more than the first predetermined threshold value, will be greater than the answer that matching degree that in the matching degree of the first predetermined threshold value, numerical value is maximum is corresponding, as being used for returning to described client by the answer of customer issue described in described client response.
In a kind of detailed description of the invention, as shown in Figure 6, intelligent response device 400 can also include disaggregated model presetting module 460, and it is for building described default disaggregated model. Specifically, disaggregated model presetting module 460 may be used for: the multiple sample data of automatic-answering back device session establishment according to historical record, and each described sample data includes in the historic customer problem extracted from described dialogue, the keyword extracted from this historic customer problem, described dialogue for this keyword identification matching degree answering answer and described answer answer and this historic customer problem for answering this historic customer problem set in advance; The plurality of sample data is divided into one group of training sample data and one group of test specimens notebook data; The described one group of training sample data training grader for identifying answer matches degree is utilized to obtain housebroken disaggregated model, utilize described one group of test specimens notebook data to check described housebroken disaggregated model, and when assay meets pre-conditioned, described housebroken disaggregated model is set as described default disaggregated model.
Specifically, housebroken disaggregated model is obtained about utilizing the described one group of training sample data training grader for identifying answer matches degree, described disaggregated model presetting module 460 can: for each the training sample data in described one group of training sample data, build training feature vector according to historic customer problem, keyword and the answer answer that these training sample data include; Identification matching degree in training feature vector each described and the training sample data corresponding to this training feature vector is performed classifier algorithm, obtains housebroken disaggregated model.
Specifically, described housebroken disaggregated model is checked about utilizing described one group of test specimens notebook data, and when assay meets pre-conditioned, described housebroken disaggregated model is set as described default disaggregated model, described disaggregated model presetting module 460 can: for each the test specimens notebook data in described one group of test specimens notebook data, build testing feature vector according to historic customer problem, keyword and the answer answer that this test specimens notebook data includes; Utilize described housebroken disaggregated model that testing feature vector each described is predicted, obtain corresponding to the historic customer problem in the test specimens notebook data of this testing feature vector and the prediction and matching degree answering answer; Calculate the difference between the identification matching degree relevant to test specimens notebook data each described and test matching degree, add up the first number of the obtained difference less than the second predetermined threshold value, and set described housebroken disaggregated model as described default disaggregated model when the percentage ratio of total number that described first number accounts for described test specimens notebook data exceedes preset percentage.
The present embodiment detailed process to each Implement of Function Module each function of intelligent response device 400, refers to above-mentioned Fig. 1 to the particular content described in embodiment illustrated in fig. 5, repeats no more herein.
It should be noted that each embodiment in this specification all adopts the mode gone forward one by one to describe, what each embodiment stressed is the difference with other embodiments, between each embodiment identical similar part mutually referring to. For device class embodiment, due to itself and embodiment of the method basic simlarity, so what describe is fairly simple, relevant part illustrates referring to the part of embodiment of the method.
In several embodiments provided herein, it should be understood that disclosed apparatus and method, it is also possible to realize by another way. Device embodiment described above is merely schematic, for instance, flow chart and block diagram in accompanying drawing show according to the device of multiple embodiments of the present invention, the architectural framework in the cards of method and computer program product, function and operation. In this, flow chart or each square frame in block diagram can represent a part for a module, program segment or code, and a part for described module, program segment or code comprises the executable instruction of one or more logic function for realizing regulation. It should also be noted that at some as in the implementation replaced, the function marked in square frame can also to be different from the order generation marked in accompanying drawing. Such as, two continuous print square frames can essentially perform substantially in parallel, and they can also perform sometimes in the opposite order, and this determines according to involved function. It will also be noted that, the combination of the square frame in each square frame in block diagram and/or flow chart and block diagram and/or flow chart, can realize by the special hardware based system of the function or action that perform regulation, or can realize with the combination of specialized hardware Yu computer instruction.
It addition, each functional module in each embodiment of the present invention can integrate one independent part of formation, it is also possible to be modules individualism, it is also possible to the integrally formed independent part of two or more modules.
If described function is using the form realization of software function module and as independent production marketing or use, it is possible to be stored in a computer read/write memory medium. Based on such understanding, part or the part of this technical scheme that prior art is contributed by technical scheme substantially in other words can embody with the form of software product, this computer software product is stored in a storage medium, including some instructions with so that a computer equipment (can be personal computer, server, or the network equipment etc.) perform all or part of step of method described in each embodiment of the present invention. And aforesaid storage medium includes: USB flash disk, portable hard drive, read only memory (ROM, Read-OnlyMemory), the various media that can store program code such as random access memory (RAM, RandomAccessMemory), magnetic disc or CD. It should be noted that, in this article, the relational terms of such as first and second or the like is used merely to separate an entity or operation with another entity or operating space, and not necessarily requires or imply the relation that there is any this reality between these entities or operation or sequentially. And, term " includes ", " comprising " or its any other variant are intended to comprising of nonexcludability, so that include the process of a series of key element, method, article or equipment not only include those key elements, but also include other key elements being not expressly set out, or also include the key element intrinsic for this process, method, article or equipment. When there is no more restriction, statement " including ... " key element limited, it is not excluded that there is also other identical element in including the process of described key element, method, article or equipment.
The foregoing is only the preferred embodiments of the present invention, be not limited to the present invention, for a person skilled in the art, the present invention can have various modifications and variations. All within the spirit and principles in the present invention, any amendment of making, equivalent replacement, improvement etc., should be included within protection scope of the present invention. It should also be noted that similar label and letter below figure represent similar terms, therefore, once a certain Xiang Yi accompanying drawing is defined, then it need not be carried out definition further and explain in accompanying drawing subsequently.
The above; being only the specific embodiment of the present invention, but protection scope of the present invention is not limited thereto, any those familiar with the art is in the technical scope that the invention discloses; change can be readily occurred in or replace, all should be encompassed within protection scope of the present invention. Therefore, protection scope of the present invention should described be as the criterion with scope of the claims.
Claims (10)
1. an intelligent response method, it is characterised in that including:
Receive the customer issue that client sends, extract the keyword in described customer issue;
Obtain at least one answer set in advance of described keyword;
For answer each described, according to described customer issue, described keyword and this answer construction feature vector;
Utilize the disaggregated model preset, predict the matching degree of the answer corresponding with this characteristic vector and described customer issue according to each constructed characteristic vector;
When having at least one matching degree in calculated matching degree more than the first predetermined threshold value, will be greater than the answer that matching degree that in the matching degree of the first predetermined threshold value, numerical value is maximum is corresponding, as being used for returning to described client by the answer of customer issue described in described client response.
2. method according to claim 1, it is characterised in that described default disaggregated model is pre-build by following step:
The multiple sample data of automatic-answering back device session establishment according to historical record, each described sample data includes in the historic customer problem extracted from described dialogue, the keyword extracted from this historic customer problem, described dialogue for this keyword identification matching degree answering answer and described answer answer and this historic customer problem for answering this historic customer problem set in advance;
The plurality of sample data is divided into one group of training sample data and one group of test specimens notebook data;
The described one group of training sample data training grader for identifying answer matches degree is utilized to obtain housebroken disaggregated model, utilize described one group of test specimens notebook data to check described housebroken disaggregated model, and when assay meets pre-conditioned, described housebroken disaggregated model is set as described default disaggregated model.
3. method according to claim 2, it is characterised in that described utilize described one group of training sample data training grader for identifying answer matches degree to obtain housebroken disaggregated model, including:
For each the training sample data in described one group of training sample data, build training feature vector according to historic customer problem, keyword and the answer answer that these training sample data include;
Identification matching degree in training feature vector each described and the training sample data corresponding to this training feature vector is performed classifier algorithm, obtains housebroken disaggregated model.
4. method according to claim 2, it is characterized in that, described utilize described one group of test specimens notebook data to check described housebroken disaggregated model, and when assay meets pre-conditioned, described housebroken disaggregated model is set as described default disaggregated model, including:
For each the test specimens notebook data in described one group of test specimens notebook data, build testing feature vector according to historic customer problem, keyword and the answer answer that this test specimens notebook data includes;
Utilize described housebroken disaggregated model that testing feature vector each described is predicted, obtain corresponding to the historic customer problem in the test specimens notebook data of this testing feature vector and the prediction and matching degree answering answer;
Calculate the difference between the identification matching degree relevant to test specimens notebook data each described and test matching degree, add up the first number of the obtained difference less than the second predetermined threshold value, and set described housebroken disaggregated model as described default disaggregated model when the percentage ratio of total number that described first number accounts for described test specimens notebook data exceedes preset percentage.
5. method according to claim 3, it is characterized in that, each described training feature vector includes at least one in following component: the feature of the keyword in training sample data corresponding with this training feature vector, the feature of the historic customer problem in these training sample data, the feature answering answer in these training sample data, the linked character of the historic customer problem in these training sample data and keyword, historic customer problem in these training sample data and the linked character answering answer, keyword in these training sample data and the linked character answering answer.
6. method according to claim 4, it is characterized in that, each described testing feature vector includes at least one in following component: the feature of the keyword in test specimens notebook data corresponding with this testing feature vector, the feature of the historic customer problem in this test specimens notebook data, the feature answering answer in this test specimens notebook data, the linked character of the historic customer problem in this test specimens notebook data and keyword, historic customer problem in this test specimens notebook data and the linked character answering answer, keyword in this test specimens notebook data and the linked character answering answer.
7. method according to claim 1, it is characterized in that, the characteristic vector built for answer each described includes at least one in following component: the linked character of the feature of described keyword, the feature of described customer issue, the feature of described answer, described customer issue and the linked character of described keyword, described customer issue and the linked character of described answer, described keyword and described answer.
8. method according to claim 7, it is characterised in that
The feature of described keyword includes the length of described keyword, numerical character number in the number of word that the word frequency of described keyword, described keyword obtain after participle, described keyword accounts for the length of chinese character in the ratio of total number of characters of this keyword, described keyword and chinese character number in described keyword accounts for the ratio of total number of characters of this keyword
The number of the word that the feature of described customer issue includes the number of the chinese character in described customer issue and described customer issue obtains after participle,
Whether number and the described answer of the word that the feature of described answer includes the number of the comprised character of described answer, described answer obtains after participle include described keyword,
The number of same words that the linked character of described customer issue and described keyword includes described customer issue and described keyword obtains after participle, described customer issue and described keyword are based on the editing distance based on word of character edit distance with described keyword of the cosine similarity of reverse document frequency value of word, described customer issue and described customer issue and described keyword
The linked character of described customer issue and described answer includes the number of same words that described customer issue obtains after participle, described customer issue and described answer with described answer based on the editing distance based on word of character edit distance with described answer of the cosine similarity of reverse document frequency value of word, described customer issue and described customer issue and described answer
The number of same words that the linked character of described keyword and described answer includes described keyword and described answer obtains after participle, described keyword and described answer are based on the editing distance based on word of character edit distance with described answer of the cosine similarity of reverse document frequency value of word, described keyword and described keyword and described answer.
9. method according to claim 1, it is characterised in that described default disaggregated model is boosting disaggregated model.
10. an intelligent response device, it is characterised in that including:
Extraction module, for receiving the customer issue that client sends, extracts the keyword in described customer issue;
Acquisition module, for obtaining at least one answer set in advance of described keyword;
Build module, for for answer each described, according to described customer issue, described keyword and this answer construction feature vector;
Prediction module, for utilizing default disaggregated model, predicting the matching degree of the answer corresponding with this characteristic vector and described customer issue according to each constructed characteristic vector;
Choose module, during for having at least one matching degree in calculated matching degree more than the first predetermined threshold value, will be greater than the answer that matching degree that in the matching degree of the first predetermined threshold value, numerical value is maximum is corresponding, as being used for returning to described client by the answer of customer issue described in described client response.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510970008.5A CN105630917A (en) | 2015-12-22 | 2015-12-22 | Intelligent answering method and intelligent answering device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510970008.5A CN105630917A (en) | 2015-12-22 | 2015-12-22 | Intelligent answering method and intelligent answering device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105630917A true CN105630917A (en) | 2016-06-01 |
Family
ID=56045850
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510970008.5A Pending CN105630917A (en) | 2015-12-22 | 2015-12-22 | Intelligent answering method and intelligent answering device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105630917A (en) |
Cited By (33)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106202301A (en) * | 2016-07-01 | 2016-12-07 | 武汉泰迪智慧科技有限公司 | A kind of intelligent response system based on degree of depth study |
| CN106789595A (en) * | 2017-01-17 | 2017-05-31 | 北京诸葛找房信息技术有限公司 | Information-pushing method and device |
| CN106777257A (en) * | 2016-12-28 | 2017-05-31 | 厦门快商通科技股份有限公司 | The constructing system and method for the Intelligent dialogue model based on words art |
| CN106878819A (en) * | 2017-01-20 | 2017-06-20 | 合网络技术（北京）有限公司 | The method, system and device of information exchange in a kind of network direct broadcasting |
| CN106951473A (en) * | 2017-03-06 | 2017-07-14 | 浙江大学 | Towards the construction method of the deep vision question answering system of dysopia personage |
| CN107038196A (en) * | 2016-11-28 | 2017-08-11 | 阿里巴巴集团控股有限公司 | A kind of customer service question answering processing method and processing device |
| WO2018006469A1 (en) * | 2016-07-07 | 2018-01-11 | 深圳狗尾草智能科技有限公司 | Knowledge graph-based human-robot interaction method and system |
| WO2018006472A1 (en) * | 2016-07-07 | 2018-01-11 | 深圳狗尾草智能科技有限公司 | Knowledge graph-based human-robot interaction method and system |
| CN107609092A (en) * | 2017-09-08 | 2018-01-19 | 北京百度网讯科技有限公司 | Intelligent response method and apparatus |
| CN108052577A (en) * | 2017-12-08 | 2018-05-18 | 北京百度网讯科技有限公司 | A kind of generic text content mining method, apparatus, server and storage medium |
| CN108153800A (en) * | 2016-12-06 | 2018-06-12 | 松下知识产权经营株式会社 | Information processing method, information processing unit and program |
| WO2018121380A1 (en) * | 2016-12-30 | 2018-07-05 | 华为技术有限公司 | Community question and answer-based article recommendation method, system, and user equipment |
| CN108388563A (en) * | 2017-02-03 | 2018-08-10 | 北京京东尚科信息技术有限公司 | Information output method and device |
| CN108427685A (en) * | 2017-02-15 | 2018-08-21 | 北京京东尚科信息技术有限公司 | A kind of intelligent response system automatic-answering back device acquisition methods |
| CN108491433A (en) * | 2018-02-09 | 2018-09-04 | 平安科技（深圳）有限公司 | Chat answer method, electronic device and storage medium |
| WO2018165932A1 (en) * | 2017-03-16 | 2018-09-20 | Microsoft Technology Licensing, Llc | Generating responses in automated chatting |
| CN108920291A (en) * | 2018-06-06 | 2018-11-30 | 阿里巴巴集团控股有限公司 | A kind of collection method of fault message, device and equipment |
| TWI643080B (en) * | 2017-05-09 | 2018-12-01 | 遊戲橘子數位科技股份有限公司 | A method to parse network data and simulate specific objects accordingly |
| CN109086400A (en) * | 2018-07-30 | 2018-12-25 | 深圳追科技有限公司 | The self checking method and device of customer service robot |
| CN109359182A (en) * | 2018-10-08 | 2019-02-19 | 网宿科技股份有限公司 | A kind of answer method and device |
| CN109493186A (en) * | 2018-11-20 | 2019-03-19 | 北京京东尚科信息技术有限公司 | The method and apparatus for determining pushed information |
| CN109800292A (en) * | 2018-12-17 | 2019-05-24 | 北京百度网讯科技有限公司 | The determination method, device and equipment of question and answer matching degree |
| CN109903134A (en) * | 2019-02-26 | 2019-06-18 | 北京多点在线科技有限公司 | A kind of order automatic generation method and system, storage medium |
| CN109947909A (en) * | 2018-06-19 | 2019-06-28 | 平安科技（深圳）有限公司 | Intelligent customer service answer method, equipment, storage medium and device |
| CN110020181A (en) * | 2018-01-02 | 2019-07-16 | 中国移动通信有限公司研究院 | A kind of processing method of recommendation information, device and computer readable storage medium |
| CN110059231A (en) * | 2019-04-19 | 2019-07-26 | 话媒(广州)科技有限公司 | A kind of generation method and device of reply content |
| CN110263135A (en) * | 2019-05-20 | 2019-09-20 | 北京字节跳动网络技术有限公司 | A kind of data exchange matching process, device, medium and electronic equipment |
| CN111191018A (en) * | 2019-12-30 | 2020-05-22 | 华为技术有限公司 | Response method and device of dialog system, electronic equipment and intelligent equipment |
| CN111309875A (en) * | 2018-12-10 | 2020-06-19 | 百度在线网络技术（北京）有限公司 | Question answering method, device, equipment and storage medium |
| CN111752437A (en) * | 2020-06-29 | 2020-10-09 | 上海寻梦信息技术有限公司 | Comment method and device, electronic equipment and storage medium |
| CN111985238A (en) * | 2020-06-30 | 2020-11-24 | 联想(北京)有限公司 | Answer generation method and equipment |
| CN112966085A (en) * | 2021-03-11 | 2021-06-15 | 平安国际智慧城市科技股份有限公司 | Intelligent control method and device for human-computer conversation, electronic equipment and storage medium |
| CN116860951A (en) * | 2023-09-04 | 2023-10-10 | 贵州中昂科技有限公司 | Information consultation service management method and management system based on artificial intelligence |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1741013A (en) * | 2004-08-27 | 2006-03-01 | 英业达股份有限公司 | Automatic adaptive system for customer service and its method |
| CN101097573A (en) * | 2006-06-28 | 2008-01-02 | 腾讯科技（深圳）有限公司 | Automatically request-answering system and method |
| CN101178711A (en) * | 2006-11-30 | 2008-05-14 | 腾讯科技（深圳）有限公司 | Chinese auto-answer method and system |
| US20100332500A1 (en) * | 2009-06-26 | 2010-12-30 | Iac Search & Media, Inc. | Method and system for determining a relevant content identifier for a search |
| CN102184225A (en) * | 2011-05-09 | 2011-09-14 | 北京奥米时代生物技术有限公司 | Method for searching preferred expert information in question-answering system |
| 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 |
| CN104137102A (en) * | 2012-02-23 | 2014-11-05 | 独立行政法人情报通信研究机构 | Non-factoid question answering system and computer program |
-
2015
- 2015-12-22 CN CN201510970008.5A patent/CN105630917A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1741013A (en) * | 2004-08-27 | 2006-03-01 | 英业达股份有限公司 | Automatic adaptive system for customer service and its method |
| CN101097573A (en) * | 2006-06-28 | 2008-01-02 | 腾讯科技（深圳）有限公司 | Automatically request-answering system and method |
| CN101178711A (en) * | 2006-11-30 | 2008-05-14 | 腾讯科技（深圳）有限公司 | Chinese auto-answer method and system |
| US20100332500A1 (en) * | 2009-06-26 | 2010-12-30 | Iac Search & Media, Inc. | Method and system for determining a relevant content identifier for a search |
| CN102184225A (en) * | 2011-05-09 | 2011-09-14 | 北京奥米时代生物技术有限公司 | Method for searching preferred expert information in question-answering system |
| CN104137102A (en) * | 2012-02-23 | 2014-11-05 | 独立行政法人情报通信研究机构 | Non-factoid question answering system and computer program |
| 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 |
Cited By (53)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106202301A (en) * | 2016-07-01 | 2016-12-07 | 武汉泰迪智慧科技有限公司 | A kind of intelligent response system based on degree of depth study |
| CN106202301B (en) * | 2016-07-01 | 2019-10-08 | 武汉泰迪智慧科技有限公司 | A kind of intelligent response system based on deep learning |
| WO2018006472A1 (en) * | 2016-07-07 | 2018-01-11 | 深圳狗尾草智能科技有限公司 | Knowledge graph-based human-robot interaction method and system |
| WO2018006469A1 (en) * | 2016-07-07 | 2018-01-11 | 深圳狗尾草智能科技有限公司 | Knowledge graph-based human-robot interaction method and system |
| CN107038196A (en) * | 2016-11-28 | 2017-08-11 | 阿里巴巴集团控股有限公司 | A kind of customer service question answering processing method and processing device |
| CN108153800B (en) * | 2016-12-06 | 2023-05-23 | 松下知识产权经营株式会社 | Information processing method, information processing apparatus, and recording medium |
| CN108153800A (en) * | 2016-12-06 | 2018-06-12 | 松下知识产权经营株式会社 | Information processing method, information processing unit and program |
| CN106777257B (en) * | 2016-12-28 | 2020-06-05 | 厦门快商通科技股份有限公司 | Intelligent dialogue model construction system and method based on dialect |
| CN106777257A (en) * | 2016-12-28 | 2017-05-31 | 厦门快商通科技股份有限公司 | The constructing system and method for the Intelligent dialogue model based on words art |
| CN108269110B (en) * | 2016-12-30 | 2021-10-26 | 华为技术有限公司 | Community question and answer based item recommendation method and system and user equipment |
| WO2018121380A1 (en) * | 2016-12-30 | 2018-07-05 | 华为技术有限公司 | Community question and answer-based article recommendation method, system, and user equipment |
| CN108269110A (en) * | 2016-12-30 | 2018-07-10 | 华为技术有限公司 | Item recommendation method, system and user equipment based on community's question and answer |
| CN106789595A (en) * | 2017-01-17 | 2017-05-31 | 北京诸葛找房信息技术有限公司 | Information-pushing method and device |
| WO2018133518A1 (en) * | 2017-01-20 | 2018-07-26 | 优酷网络技术（北京）有限公司 | Method, system, and apparatus for information interaction in live webcast |
| CN106878819A (en) * | 2017-01-20 | 2017-06-20 | 合网络技术（北京）有限公司 | The method, system and device of information exchange in a kind of network direct broadcasting |
| CN108388563A (en) * | 2017-02-03 | 2018-08-10 | 北京京东尚科信息技术有限公司 | Information output method and device |
| CN108427685A (en) * | 2017-02-15 | 2018-08-21 | 北京京东尚科信息技术有限公司 | A kind of intelligent response system automatic-answering back device acquisition methods |
| CN106951473B (en) * | 2017-03-06 | 2019-11-26 | 浙江大学 | The construction method of deep vision question answering system towards dysopia personage |
| CN106951473A (en) * | 2017-03-06 | 2017-07-14 | 浙江大学 | Towards the construction method of the deep vision question answering system of dysopia personage |
| US11729120B2 (en) | 2017-03-16 | 2023-08-15 | Microsoft Technology Licensing, Llc | Generating responses in automated chatting |
| WO2018165932A1 (en) * | 2017-03-16 | 2018-09-20 | Microsoft Technology Licensing, Llc | Generating responses in automated chatting |
| TWI643080B (en) * | 2017-05-09 | 2018-12-01 | 遊戲橘子數位科技股份有限公司 | A method to parse network data and simulate specific objects accordingly |
| CN107609092A (en) * | 2017-09-08 | 2018-01-19 | 北京百度网讯科技有限公司 | Intelligent response method and apparatus |
| CN107609092B (en) * | 2017-09-08 | 2021-03-09 | 北京百度网讯科技有限公司 | Intelligent response method and device |
| US11062090B2 (en) | 2017-12-08 | 2021-07-13 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and apparatus for mining general text content, server, and storage medium |
| CN108052577B (en) * | 2017-12-08 | 2022-06-14 | 北京百度网讯科技有限公司 | Universal text content mining method, device, server and storage medium |
| CN108052577A (en) * | 2017-12-08 | 2018-05-18 | 北京百度网讯科技有限公司 | A kind of generic text content mining method, apparatus, server and storage medium |
| CN110020181B (en) * | 2018-01-02 | 2021-05-25 | 中国移动通信有限公司研究院 | Processing method and device of recommendation information and computer readable storage medium |
| CN110020181A (en) * | 2018-01-02 | 2019-07-16 | 中国移动通信有限公司研究院 | A kind of processing method of recommendation information, device and computer readable storage medium |
| CN108491433A (en) * | 2018-02-09 | 2018-09-04 | 平安科技（深圳）有限公司 | Chat answer method, electronic device and storage medium |
| WO2019153613A1 (en) * | 2018-02-09 | 2019-08-15 | 平安科技（深圳）有限公司 | Chat response method, electronic device and storage medium |
| CN108920291A (en) * | 2018-06-06 | 2018-11-30 | 阿里巴巴集团控股有限公司 | A kind of collection method of fault message, device and equipment |
| CN109947909B (en) * | 2018-06-19 | 2024-03-12 | 平安科技（深圳）有限公司 | Intelligent customer service response method, equipment, storage medium and device |
| CN109947909A (en) * | 2018-06-19 | 2019-06-28 | 平安科技（深圳）有限公司 | Intelligent customer service answer method, equipment, storage medium and device |
| CN109086400A (en) * | 2018-07-30 | 2018-12-25 | 深圳追科技有限公司 | The self checking method and device of customer service robot |
| CN109359182A (en) * | 2018-10-08 | 2019-02-19 | 网宿科技股份有限公司 | A kind of answer method and device |
| CN109493186A (en) * | 2018-11-20 | 2019-03-19 | 北京京东尚科信息技术有限公司 | The method and apparatus for determining pushed information |
| CN111309875A (en) * | 2018-12-10 | 2020-06-19 | 百度在线网络技术（北京）有限公司 | Question answering method, device, equipment and storage medium |
| CN111309875B (en) * | 2018-12-10 | 2023-08-04 | 百度在线网络技术（北京）有限公司 | Method, device, equipment and storage medium for answering questions |
| CN109800292A (en) * | 2018-12-17 | 2019-05-24 | 北京百度网讯科技有限公司 | The determination method, device and equipment of question and answer matching degree |
| CN109903134A (en) * | 2019-02-26 | 2019-06-18 | 北京多点在线科技有限公司 | A kind of order automatic generation method and system, storage medium |
| CN110059231B (en) * | 2019-04-19 | 2021-10-19 | 康忠华 | Reply content generation method and device |
| CN110059231A (en) * | 2019-04-19 | 2019-07-26 | 话媒(广州)科技有限公司 | A kind of generation method and device of reply content |
| CN110263135A (en) * | 2019-05-20 | 2019-09-20 | 北京字节跳动网络技术有限公司 | A kind of data exchange matching process, device, medium and electronic equipment |
| CN111191018B (en) * | 2019-12-30 | 2023-10-20 | 华为技术有限公司 | Response method and device of dialogue system, electronic equipment and intelligent equipment |
| CN111191018A (en) * | 2019-12-30 | 2020-05-22 | 华为技术有限公司 | Response method and device of dialog system, electronic equipment and intelligent equipment |
| CN111752437B (en) * | 2020-06-29 | 2021-07-16 | 上海寻梦信息技术有限公司 | Comment method and device, electronic equipment and storage medium |
| CN111752437A (en) * | 2020-06-29 | 2020-10-09 | 上海寻梦信息技术有限公司 | Comment method and device, electronic equipment and storage medium |
| CN111985238A (en) * | 2020-06-30 | 2020-11-24 | 联想(北京)有限公司 | Answer generation method and equipment |
| CN112966085B (en) * | 2021-03-11 | 2023-10-13 | 平安国际智慧城市科技股份有限公司 | Man-machine conversation intelligent control method and device, electronic equipment and storage medium |
| CN112966085A (en) * | 2021-03-11 | 2021-06-15 | 平安国际智慧城市科技股份有限公司 | Intelligent control method and device for human-computer conversation, electronic equipment and storage medium |
| CN116860951A (en) * | 2023-09-04 | 2023-10-10 | 贵州中昂科技有限公司 | Information consultation service management method and management system based on artificial intelligence |
| CN116860951B (en) * | 2023-09-04 | 2023-11-14 | 贵州中昂科技有限公司 | Information consultation service management method and management system based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105630917A (en) | Intelligent answering method and intelligent answering device | |
| CN110502608B (en) | Man-machine conversation method and man-machine conversation device based on knowledge graph | |
| CN111368043A (en) | Event question-answering method, device, equipment and storage medium based on artificial intelligence | |
| CN113326991B (en) | Automatic authorization method, device, computer equipment and storage medium | |
| CN110866115B (en) | Sequence labeling method, system, computer equipment and computer readable storage medium | |
| CN109886719B (en) | Data mining processing method and device based on grid and computer equipment | |
| CN110781381B (en) | Data verification method, device, equipment and storage medium based on neural network | |
| CN107644106B (en) | Method, terminal device and storage medium for automatically mining service middleman | |
| CN109543005A (en) | The dialogue state recognition methods of customer service robot and device, equipment, storage medium | |
| CN104142990A (en) | Search method and device | |
| CN112995414A (en) | Behavior quality inspection method, device, equipment and storage medium based on voice call | |
| CN114090601B (en) | Data screening method, device, equipment and storage medium | |
| CN112235470A (en) | Incoming call client follow-up method, device and equipment based on voice recognition | |
| CN111415196A (en) | Advertisement recall method, device, server and storage medium | |
| CN113569118B (en) | Self-media pushing method, device, computer equipment and storage medium | |
| CN110096593A (en) | A method of the outer paging system of building intelligence | |
| EP4198777A1 (en) | Sample classification method and apparatus, electronic device and storage medium | |
| CN113191137A (en) | Operation risk obtaining method and system, electronic equipment and storage medium | |
| CN104572820A (en) | Method and device for generating model and method and device for acquiring importance degree | |
| CN113407579A (en) | Group query method and device, electronic equipment and readable storage medium | |
| CN112417886A (en) | Intention entity information extraction method and device, computer equipment and storage medium | |
| CN111027296A (en) | Report generation method and system based on knowledge base | |
| CN113536788B (en) | Information processing method, device, storage medium and equipment | |
| CN114238574B (en) | Intention recognition method based on artificial intelligence and related equipment thereof | |
| CN114238583B (en) | Natural language processing method, device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: 610000 Huayang Street, Tianfu New District, Chengdu City, Sichuan Province, No. 1, No. 2, No. 19 Building, Unit 2, 1903 Applicant after: Chengdu Xiaoduo Technology Co., Ltd. Address before: 610000 846, southern section of Tianfu Avenue, Huayang street, Tianfu New District, Chengdu, Sichuan Applicant before: CHENGDU XIAODUO TECH CO., LTD. |
|
| CB02 | Change of applicant information | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160601 |
|
| RJ01 | Rejection of invention patent application after publication |