CN108170859B - Voice query method, device, storage medium and terminal equipment - Google Patents
Voice query method, device, storage medium and terminal equipment Download PDFInfo
- Publication number
- CN108170859B CN108170859B CN201810059882.7A CN201810059882A CN108170859B CN 108170859 B CN108170859 B CN 108170859B CN 201810059882 A CN201810059882 A CN 201810059882A CN 108170859 B CN108170859 B CN 108170859B
- Authority
- CN
- China
- Prior art keywords
- query
- intention
- statement
- voice
- query statement
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
- G06F16/63—Querying
- G06F16/632—Query formulation
Abstract
The invention provides a voice query method, a voice query device, a storage medium and terminal equipment, wherein the method comprises the steps of receiving a voice query request of a user; performing voice recognition on the voice query request to obtain a query statement; performing intention identification and object identification on the query statement to obtain a query intention and a query object word of the query statement; wherein the query object terms correspond to at least one query object; determining a query object of the query statement based on the query object terms; and querying a retrieval database according to the query intention and the query object of the query statement to obtain a query result. By adopting the invention, the query accuracy can be improved.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a method, an apparatus, a storage medium, and a terminal device for voice query.
Background
The information inquiry system is a system for solving information inquiry, and the inquiry system based on the search engine is the most popular information inquiry system at present, so that people can acquire information more conveniently and conveniently. However, as information is explosively increased, types and modes of user information demands are more and more abundant, and the traditional search engine query system cannot effectively solve the increasingly abundant information demands of people.
In the past PC internet era, the conventional search engine query system can meet the general information requirements of people, and a PC computer which can be networked is provided as information acquisition equipment based on an information demand party.
In the current popular mobile internet era, the mobile internet can meet the personalized information requirements of people, such as query within the range of personalized information according to geographic positions, and the requirements on information acquisition equipment are greatly reduced, only one mobile terminal is needed, but the mobile terminal corresponding to the information acquisition equipment has the information request mode corresponding to the requirements only including a text typing mode and an information sorting mode which are intentions for matching and understanding texts based on the typed texts, so that the newly added information request mode of the information acquisition equipment, such as a voice input mode, cannot be followed. With the arrival of the AI era, speech recognition, speech synthesis and man-machine conversation techniques are becoming more and more mature, and the frequency of use of speech input to request query information is also qualitatively changed. However, the query process based on the voice input is not improved in a matching way, and the following defects exist:
1. the traditional search engine query system only carries out a text query interaction mode of one question and one answer, and cannot realize voice query and carry out multiple rounds of voice interaction under the condition of fuzzy voice query input.
2. Ai system, but this system needs to manually label the label data of each intention for identifying the intention, but because the data of the query is too rich and the types of the intentions are too large, the intention identification system cannot be used to train each intention one by one, and the identification efficiency is low.
Disclosure of Invention
Embodiments of the present invention provide a method, an apparatus, a storage medium, and a terminal device for voice query, so as to solve or alleviate the above technical problems in the prior art.
In a first aspect, an embodiment of the present invention provides a method for voice query, including:
receiving a voice query request of a user;
performing voice recognition on the voice query request to obtain a query statement;
performing intention identification and object identification on the query statement to obtain a query intention and a query object word of the query statement; wherein the query object terms correspond to at least one query object;
determining a query object of the query statement based on the query object terms; and
and querying a retrieval database according to the query intention and the query object of the query statement to obtain a query result.
With reference to the first aspect, in a first implementation manner of the first aspect, the determining a query object of the query statement based on the query object term includes:
judging whether the number of the query objects corresponding to the query object terms is greater than 1;
if yes, returning all query objects corresponding to the query object terms to the user for the user to select and confirm;
receiving a query object returned by the user; and
and taking the received query object as the query object of the query statement.
With reference to the first aspect, in a second implementation manner of the first aspect, the performing intent recognition and object recognition on the query statement to obtain a query intent and a query object term of the query statement includes:
identifying the query statement according to an intention classification model to obtain a query intention of the query statement; and
and identifying the query statement according to the query intention and the object identification model to obtain a query object word of the query statement.
With reference to the second embodiment of the first aspect, in a third embodiment of the first aspect, the method further includes:
obtaining a historical query statement and an access address selected by a user based on a query result of the historical query statement from a retrieval log of the retrieval database;
identifying the intention of the historical query statement according to a regular expression intention identification algorithm;
associating the historical query statement with the identified intention with the intention word corresponding to the identified intention;
for the historical query sentences without the recognized intentions, judging whether the user-selected access addresses corresponding to the historical query sentences without the recognized intentions are the same as the user-selected access addresses corresponding to the historical query sentences with the recognized intentions; if so, associating the historical query statement without the identified intention with the intention word associated with the historical query statement with the same access address;
forming data pairs by the correlated historical query sentences and the intention words, and updating the data pairs in a training database; and
and training and updating the intention classification model according to the updated training database.
With reference to the third implementation manner of the first aspect, in a fourth implementation manner of the first aspect, the forming the associated query statement and intention word into a data pair includes:
carrying out structuralization processing on the history query sentences in the history query sentences and the intention words which are mutually associated to obtain the history query sentences with syntactic characteristics and semantic characteristics;
and forming a data pair by the structured historical query statement and the intention word.
With reference to the fourth implementation manner of the first aspect, in a fifth implementation manner of the first aspect, before identifying the query statement according to the intention classification model, the method further includes:
and carrying out structural processing on the query statement to obtain the query statement with syntactic characteristics and semantic characteristics.
With reference to the first aspect, in a sixth implementation manner of the first aspect, all query objects corresponding to the query object terms are returned to the user in the form of a voice signal for the user to select and confirm, and the method further includes: and returning the query result to the user in a voice signal form. .
In a second aspect, an embodiment of the present invention further provides an apparatus for voice query, including:
the voice query receiving module is used for receiving a voice query request of a user;
the query statement acquisition module is used for carrying out voice recognition on the voice query request to acquire a query statement;
the intention and object identification module is used for carrying out intention identification and object identification on the query statement to obtain a query intention and a query object word of the query statement; wherein the query object terms correspond to at least one query object;
the query object confirming module is used for determining a query object of the query statement based on the query object words; and
and the query database module is used for querying and retrieving a database according to the query intention and the query object of the query statement to obtain a query result.
With reference to the second aspect, in a first implementation manner of the second aspect, the query object confirmation module includes:
the quantity judging unit is used for judging whether the quantity of the query objects corresponding to the query object terms is greater than 1 or not;
the object returning unit is used for returning all query objects corresponding to the query object terms to the user for the user to select and confirm if the query objects are the same as the query objects;
the object receiving unit is used for receiving the query object returned by the user; and
and the object confirmation unit is used for taking the received query object as the query object of the query statement.
With reference to the second aspect, in a second implementation manner of the second aspect, the intention and object recognition module further includes:
the intention identification unit is used for identifying the query statement according to an intention classification model to obtain the query intention of the query statement; and
and the object identification unit is used for identifying the query statement according to the query intention and the object identification model to obtain a query object term of the query statement.
With reference to the second embodiment of the second aspect, in a third embodiment of the second aspect, the apparatus further comprises:
the log acquisition module is used for acquiring a historical query statement and an access address selected by a user based on a query result of the historical query statement from a retrieval log of the retrieval database;
the preliminary intention identification module is used for identifying the intention of the historical query statement according to a regular expression intention identification algorithm;
the first association module is used for associating the historical query statement with the identified intention with the intention word corresponding to the identified intention;
the second correlation module is used for judging whether the user-selected access address corresponding to the historical query statement without the identified intention and the user-selected access address corresponding to the historical query statement with the identified intention have the same access address or not; if so, associating the historical query statement without the identified intention with the intention word associated with the historical query statement with the same access address;
the training data updating module is used for forming data pairs by the correlated historical query sentences and the intention words and updating the data pairs in a training database; and
and the model updating module is used for training and updating the intention classification model according to the updated training database.
The functions of the device can be realized by hardware, and can also be realized by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above-described functions.
In one possible design, the structure of the voice query includes a processor and a memory, the memory is used for storing a program of the apparatus supporting the voice query to execute the method of the voice query in the first aspect, and the processor is configured to execute the program stored in the memory. The apparatus for voice querying may further comprise a communication interface for communicating the apparatus for voice querying with other devices or a communication network.
In a third aspect, an embodiment of the present invention provides a computer-readable storage medium, computer software instructions for an apparatus for voice query, which includes a program for performing the method for voice query in the first aspect to the apparatus for voice query.
Any one of the above technical solutions has the following advantages or beneficial effects:
the embodiment of the invention converts the voice query request of a user into a query statement when receiving the voice query request of the user, then identifies the query intent and query object terms of the query statement, and directly queries to obtain a query result if the query object terms only correspond to one query object; when the query object term includes a plurality of query objects, the query request is ambiguous, and it is necessary to confirm the query object of the query sentence based on the query object term to improve the query accuracy.
The foregoing summary is provided for the purpose of description only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features of the present invention will be readily apparent by reference to the drawings and following detailed description.
Drawings
In the drawings, like reference numerals refer to the same or similar parts or elements throughout the several views unless otherwise specified. The figures are not necessarily to scale. It is appreciated that these drawings depict only some embodiments in accordance with the disclosure and are therefore not to be considered limiting of its scope.
FIG. 1 is a flow chart diagram illustrating one embodiment of a method for voice querying provided by the present invention;
FIG. 2 is a flow chart illustrating one embodiment of a query object validation process of the method for voice querying provided by the present invention;
FIG. 3 is a flow diagram illustrating one embodiment of an intent recognition and object recognition process of the method for voice querying provided by the present invention;
FIG. 4 is a schematic diagram of the structure of another embodiment of intent recognition and object recognition provided by the present invention;
FIG. 5 is a flowchart illustrating an embodiment of a training process of an intention classification model in the method of voice query provided by the present invention;
FIG. 6 is a schematic diagram of one implementation of training data mining in the method of voice querying provided by the present invention;
FIG. 7 is a structural diagram of an embodiment of an intention classification model in the method of voice query provided by the present invention;
FIG. 8 is a schematic structural diagram of an apparatus for voice query according to an embodiment of the present invention;
fig. 9 is a schematic structural diagram of an embodiment of a terminal device provided by the present invention.
Detailed Description
In the following, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments may be modified in various different ways, all without departing from the spirit or scope of the present invention. Accordingly, the drawings and description are to be regarded as illustrative in nature, and not as restrictive.
Example one
Referring to fig. 1, an embodiment of the present invention provides a method for voice query, including steps S10 through S50, as follows:
and S10, receiving the voice query request of the user.
In the embodiment of the invention, the method can be executed by the user terminal, and the user terminal directly receives the voice of the user and performs voice interaction with the user; the method can also be executed by the server, the user can be referred to as a user terminal, and the user terminal receives the voice of the user, transmits the voice signal to the server for voice recognition and executes subsequent query operation. The user terminal includes but is not limited to a mobile phone, a wearable device, a vehicle-mounted device, and the like.
S20, carrying out voice recognition to the voice inquiry request to obtain the inquiry statement.
In the embodiment of the present invention, the voice query request is a voice signal, and the voice signal may be converted by a voice recognition technology of asr (auto spoke recognition) to obtain a corresponding text signal, so as to recognize that the text signal is a query statement.
S30, performing intention identification and object identification on the query statement to obtain a query intention and a query object word of the query statement; wherein the query object terms correspond to at least one query object.
In particular, a query statement may include a plurality of query object terms, one query object term for at least one query object.
In one particular example, a query statement may be structured such that the query statement has syntactic characteristics and semantic characteristics, the syntactic characteristics referring to syntactic structure information of the sentence, and the semantic characteristics referring to hypernym information of the sentence. Furthermore, the structured text is identified to obtain an inquiry intention and slot position information of the text, the inquiry intention is used for representing the type or purpose to be inquired by the user, the slot position information is used for representing the object to be inquired by the user, for example, for the inquiry sentence which is "li na years", the inquiry intention is "age", the slot position (namely the inquiry object word) is "li na", and the slot position identification can use a proper name identification technology for identifying the name of a person, the place name, the time, various types of works and the like in the text. However, there may be multiple meanings for the query object term, for example, lina may be singer, lina may also be lina of playing tennis, if query is directly performed with the query object term, information about multiple linas may be detected, so that the user is required to clarify the query object that the user wants to query before retrieval, specifically referring to step S40.
S40, determining the query object of the query statement based on the query object terms.
In one specific example, this step S40 may be as shown in fig. 2, such that voice interaction with the user confirms the query object:
s41, judging whether the number of the query objects corresponding to the query object terms is more than 1;
s42, if yes, returning all query objects corresponding to the query object terms to the user for the user to select and confirm;
in a specific example, all query objects corresponding to the query object terms may be returned to the user in a form of a voice signal for the user to select and confirm, or all query objects corresponding to the query object terms may be directly returned to the user, so that returned information is displayed in a display interface of the user, and the user performs selection and confirmation in the display interface.
And S43, receiving the query object returned by the user.
In a specific example, a voice confirmation signal returned by the user can be received in the form of a voice signal, the voice confirmation signal includes a query object confirmed by the user, and then the query object is obtained through a voice conversion mode, so that the query request is confirmed in a voice interaction mode. And if the returned voice confirmation signal is fuzzy, continuously returning all the query objects corresponding to the query object words to the user for the user to select and confirm until a clear voice confirmation signal capable of identifying the query objects is received.
And S44, taking the received query object as the query object of the query statement.
In the embodiment of the present invention, generally, the user selects one confirmed query object as the query object of the query statement, but in a special case, the user may also select to confirm a plurality of query objects, and then the selected plurality of query objects may also be used as the query object of the query statement.
And S50, querying a retrieval database according to the query intention and the query object of the query statement to obtain a query result. Preferably, the query result is returned to the user in the form of a voice signal.
In the embodiment of the invention, when the query object words comprise the meanings of a plurality of query objects, the query object words can be searched after being confirmed by the user in advance, the query request of the user can be accurately met, and compared with the mode that the search is carried out when the query request of the user is not confirmed, the search is carried out on the basis that the detection results comprise information corresponding to a plurality of query objects, the search is confirmed to the user, and the detection results corresponding to the information confirmed by the user are screened from the detection results, the embodiment of the invention is obviously beneficial to improving the detection efficiency and reducing the working pressure of the system.
Example two
Referring to fig. 3, in the present embodiment, on the basis of the first embodiment, a process of performing intent recognition and object recognition by an intent classification model and an object recognition model is provided, that is, the specific implementation process of step S30 in the first embodiment includes:
and S31, identifying the query statement according to the intention classification model, and obtaining the query intention of the query statement.
In this embodiment, the intention classification model may be trained by machine learning, that is, a query sentence and an intention word are associated to form a training data pair, and then a large amount of training data pairs are input into the intention classification model for training.
And S32, identifying the query statement according to the query intention and the object identification model, and obtaining a query object term of the query statement.
Because the query statement contains a plurality of terms, when the slot position identification is carried out on the query statement by using the object identification model, the query intent is also required to further assist in confirming the query statement. For example, as shown in fig. 4, for the query sentence queryA "how the temperature change of tomorrow beijing is, the intention classification model identifies the query sentence as intent" weather ", and further, the object identification model identifies the query target words" tomorrow "," beijing "," temperature ", and excludes" temperature "with the assistance of the query intention, so that the query target words are slot 1" tomorrow "and slot 12" beijing ", and then the whole intention query structure for subsequent database query is" beijing tomorrow ".
Before step S31, the intention classification model may be generated by training, or after step S31, the training process of the intention classification model includes:
s311, obtaining a historical query statement and an access address selected by a user based on a query result of the historical query statement from a retrieval log of the retrieval database.
In this embodiment, the search database may be a search database of a search engine, and the search engine may be a hundredth, 360 or google search engine.
S312, identifying the intention of the historical query statement according to a regular expression intention identification algorithm.
In this embodiment, a heuristic intent mining method may be adopted, and first, intent words, such as "weather", "movie", "age", and other concepts, attributes, or categories, are defined; then, because the query information types of the user are very rich and the intention word set is difficult to define manually, the embodiment mines candidate concepts and attribute words from the search log of the search engine, the search engine can be a hundred-degree search engine, a 360-degree search engine or a google search engine, the intention words are filtered out through manual review, and 36000 intention words are mined in the specific test process of the embodiment.
Further, setting a recognition rule, i.e., a regular expression intention recognition algorithm, based on the mined intention words, and recognizing the intention of a part of the historical query sentences query of the search log, i.e., including the historical query sentences in which the intention has been recognized and the historical query sentences in which the intention has not been recognized, by this simple recognition algorithm, the following operations of step S313 and step S314 are performed, respectively.
S313, the historical query statement for which the intention has been identified is associated with the intention word corresponding to the intention identified.
S314, judging whether the user-selected access address corresponding to the historical query sentence without the identified intention and the user-selected access address corresponding to the historical query sentence with the identified intention have the same access address or not for the historical query sentence without the identified intention; and if so, associating the historical query statement without the identified intention with the intention word associated with the historical query statement with the same access address.
Since the search engine can retrieve a plurality of query link addresses, and excavate parallel linguistic data according to the clicking behavior of the user, namely synonymous query, if two different query statements query, and the user clicks the same link address url, the two query statements are considered to be synonymous, for example, two query statements of "how high liudeluhua" and "how high liudeluhua" are clicked on the same webpage, namely, the same access address is stored, the two query statements are considered to be synonymous, and then the simple intention identification system identifies how high "liudeluhua" is "height", and the intention of "how high liudeluhua" without identifying the intention is also "height". Thus, massive training data can be mined.
And S315, forming data pairs by the correlated historical query sentences and the intention words, and updating the data pairs in the training database.
Taking fig. 6 as an example, determining an intention word through intention word definition and intention word mining, then obtaining a query statement query from a search log, performing preliminary intention recognition on the query, directly forming a data pair with the corresponding intention word for the query statement query with the intention word recognized, and mining parallel corpora for the query statement query without the intention word recognized, that is, step S314, then combining data pairs with the same intention word in the formed data pairs, forming a data pair in a manner that a plurality of queries correspond to one intention word, and updating a training database.
And S316, training and updating the intention classification model according to the updated training database.
In this embodiment, the query sentence of the training data may be subjected to structural processing before the input of the intention classification model, specifically, the historical query sentence in the associated historical query sentence and the historical query sentence in the intention word are subjected to structural processing to obtain a historical query sentence with syntactic and semantic features; and forming a data pair by the structured historical query statement and the intention word. Preferably, the intention classification model is a neural network model, as shown in fig. 7, an input query sentence may have word features, syntactic features, and semantic features, the extracted features are input into an input layer, the input layer emb layer vectorizes corresponding feature pairs, the vectorized vector features are input into a loop network RNN layer, recursion is performed on all feature vectors into a vector form, and finally the vector is output into a full connection layer softmat layer.
EXAMPLE III
Referring to fig. 8, an embodiment of the present invention further provides a device for voice query, including:
a voice query receiving module 10, configured to receive a voice query request of a user;
a query sentence acquisition module 20, configured to perform voice recognition on the voice query request to acquire a query sentence;
an intention and object recognition module 30, configured to perform intention recognition and object recognition on the query statement, and obtain a query intention and a query object term of the query statement; wherein the query object terms correspond to at least one query object;
a query object confirmation module 40, configured to determine a query object of the query statement based on the query object term; and
and the query database module 50 is used for querying and retrieving the database according to the query intention and the query object of the query statement to obtain a query result.
With reference to the third embodiment, in a first implementation manner of the third embodiment, the query object confirming module includes:
the quantity judging unit is used for judging whether the quantity of the query objects corresponding to the query object terms is greater than 1 or not;
the object returning unit is used for returning all query objects corresponding to the query object terms to the user for the user to select and confirm if the query objects are the same as the query objects;
the object receiving unit is used for receiving the query object returned by the user; and
and the object confirmation unit is used for taking the received query object as the query object of the query statement.
With reference to the third embodiment, in a second implementation manner of the third embodiment, the intention and object identification module further includes:
the intention identification unit is used for identifying the query statement according to an intention classification model to obtain the query intention of the query statement; and
and the object identification unit is used for identifying the query statement according to the query intention and the object identification model to obtain a query object term of the query statement.
With reference to the second implementation manner of the third embodiment, in a third implementation manner of the third embodiment, the apparatus further includes:
the log acquisition module is used for acquiring a historical query statement and an access address selected by a user based on a query result of the historical query statement from a retrieval log of the retrieval database;
the preliminary intention identification module is used for identifying the intention of the historical query statement according to a regular expression intention identification algorithm;
the first association module is used for associating the historical query statement with the identified intention with the intention word corresponding to the identified intention;
the second correlation module is used for judging whether the user-selected access address corresponding to the historical query statement without the identified intention and the user-selected access address corresponding to the historical query statement with the identified intention have the same access address or not; if so, associating the historical query statement without the identified intention with the intention word associated with the historical query statement with the same access address;
the training data updating module is used for forming data pairs by the correlated historical query sentences and the intention words and updating the data pairs in a training database; and
and the model updating module is used for training and updating the intention classification model according to the updated training database.
The functions of the device can be realized by hardware, and can also be realized by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above-described functions.
In one possible design, the structure of the voice query includes a processor and a memory, the memory is used for storing a program of the apparatus supporting the voice query to execute the method of the voice query in the first aspect, and the processor is configured to execute the program stored in the memory. The apparatus for voice querying may further comprise a communication interface for communicating the apparatus for voice querying with other devices or a communication network.
Example four
An embodiment of the present invention further provides a terminal device, as shown in fig. 9, where the terminal device includes: a memory 21 and a processor 22, the memory 21 having stored therein a computer program operable on the processor 22. The processor 22, when executing the computer program, implements the method of voice query in the above embodiments. The number of the memory 21 and the processor 22 may be one or more.
The apparatus further comprises:
a communication interface 23 for communication between the processor 22 and an external device.
The memory 21 may comprise a high-speed RAM memory, and may further include a non-volatile memory (non-volatile memory), such as at least one disk memory.
If the memory 21, the processor 22 and the communication interface 23 are implemented independently, the memory 21, the processor 22 and the communication interface 23 may be connected to each other through a bus and perform communication with each other. The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown in FIG. 9, but this does not indicate only one bus or one type of bus.
Optionally, in a specific implementation, if the memory 21, the processor 22 and the communication interface 23 are integrated on a chip, the memory 21, the processor 22 and the communication interface 23 may complete mutual communication through an internal interface.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality" means two or more unless specifically defined otherwise.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and alternate implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present invention.
The logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
The computer readable medium described in embodiments of the present invention may be a computer readable signal medium or a computer readable storage medium or any combination of the two. More specific examples (a non-exhaustive list) of the computer-readable storage medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable read-only memory (CDROM). Additionally, the computer-readable storage medium may even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
In embodiments of the present invention, a computer readable signal medium may comprise a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, input method, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, Radio Frequency (RF), etc., or any suitable combination of the preceding.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.
In addition, functional units in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a separate product, may also be stored in a computer readable storage medium. The storage medium may be a read-only memory, a magnetic or optical disk, or the like.
The above description is only for the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive various changes or substitutions within the technical scope of the present invention, and these should be covered by the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.
Claims (11)
1. A method of voice querying, comprising:
receiving a voice query request of a user;
performing voice recognition on the voice query request to obtain a query statement;
performing intention identification and object identification on the query statement to obtain a query intention and a query object word of the query statement; wherein the query object terms correspond to at least one query object, the query statement being identified according to an intent classification model;
determining a query object of the query statement based on the query object terms; and
inquiring a retrieval database according to the inquiry intention and the inquiry object of the inquiry statement to obtain an inquiry result; wherein the method further comprises:
obtaining a historical query statement and an access address selected by a user based on a query result of the historical query statement from a retrieval log of the retrieval database;
identifying the intention of the historical query statement according to a regular expression intention identification algorithm;
associating the historical query statement with the identified intention with the intention word corresponding to the identified intention;
for the historical query sentences without the recognized intentions, judging whether the user-selected access addresses corresponding to the historical query sentences without the recognized intentions are the same as the user-selected access addresses corresponding to the historical query sentences with the recognized intentions; and
if yes, associating the historical query statement without the identified intention with the intention word associated with the historical query statement with the same access address;
forming data pairs by the correlated historical query sentences and the intention words, and updating the data pairs in a training database; and
and training and updating the intention classification model according to the updated training database.
2. The method of voice query of claim 1, wherein said determining a query object for the query statement based on the query object terms comprises:
judging whether the number of the query objects corresponding to the query object terms is greater than 1;
if yes, returning all query objects corresponding to the query object terms to the user for the user to select and confirm;
receiving a query object returned by the user; and
and taking the received query object as the query object of the query statement.
3. The method of voice query according to claim 1, wherein the performing intent recognition and object recognition on the query statement to obtain query intent and query object terms of the query statement comprises:
identifying the query statement according to an intention classification model to obtain a query intention of the query statement; and
and identifying the query statement according to the query intention and the object identification model to obtain a query object word of the query statement.
4. The method of voice querying according to claim 3, wherein said composing the associated query statement and intent word into a data pair comprises:
carrying out structuralization processing on the history query sentences in the history query sentences and the intention words which are mutually associated to obtain the history query sentences with syntactic characteristics and semantic characteristics; and
and forming a data pair by the structured historical query statement and the intention word.
5. The method of voice querying of claim 4, prior to identifying the query statement according to an intent classification model, further comprising:
and carrying out structural processing on the query statement to obtain the query statement with syntactic characteristics and semantic characteristics.
6. The method of voice query according to claim 2, wherein all query objects corresponding to the query object terms are returned to the user in a form of voice signals for selective confirmation by the user, and the method further comprises: and returning the query result to the user in a voice signal form.
7. An apparatus for voice querying, comprising:
the voice query receiving module is used for receiving a voice query request of a user;
the query statement acquisition module is used for carrying out voice recognition on the voice query request to acquire a query statement;
the intention and object identification module is used for carrying out intention identification and object identification on the query statement to obtain a query intention and a query object word of the query statement; wherein the query object terms correspond to at least one query object, the query statement being identified according to an intent classification model;
the query object confirming module is used for determining a query object of the query statement based on the query object words; and
the query database module is used for querying a retrieval database according to the query intention and the query object of the query statement to obtain a query result;
the device further comprises:
the log acquisition module is used for acquiring a historical query statement and an access address selected by a user based on a query result of the historical query statement from a retrieval log of the retrieval database;
the preliminary intention identification module is used for identifying the intention of the historical query statement according to a regular expression intention identification algorithm;
the first association module is used for associating the historical query statement with the identified intention with the intention word corresponding to the identified intention;
the second correlation module is used for judging whether the user-selected access address corresponding to the historical query statement without the identified intention and the user-selected access address corresponding to the historical query statement with the identified intention have the same access address or not; if so, associating the historical query statement without the identified intention with the intention word associated with the historical query statement with the same access address;
the training data updating module is used for forming data pairs by the correlated historical query sentences and the intention words and updating the data pairs in a training database; and
and the model updating module is used for training and updating the intention classification model according to the updated training database.
8. The apparatus for voice query according to claim 7, wherein the query object confirmation module comprises:
the quantity judging unit is used for judging whether the quantity of the query objects corresponding to the query object terms is greater than 1 or not;
the object returning unit is used for returning all query objects corresponding to the query object terms to the user for the user to select and confirm if the query objects are the same as the query objects;
the object receiving unit is used for receiving the query object returned by the user; and
and the object confirmation unit is used for taking the received query object as the query object of the query statement.
9. The apparatus of voice query of claim 7, wherein the intent and object recognition module further comprises:
the intention identification unit is used for identifying the query statement according to an intention classification model to obtain the query intention of the query statement; and
and the object identification unit is used for identifying the query statement according to the query intention and the object identification model to obtain a query object term of the query statement.
10. A terminal device for implementing voice query, the terminal device comprising:
one or more processors;
storage means for storing one or more programs;
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of voice querying as recited in any of claims 1-6.
11. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the method of voice query according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810059882.7A CN108170859B (en) | 2018-01-22 | 2018-01-22 | Voice query method, device, storage medium and terminal equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810059882.7A CN108170859B (en) | 2018-01-22 | 2018-01-22 | Voice query method, device, storage medium and terminal equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108170859A CN108170859A (en) | 2018-06-15 |
| CN108170859B true CN108170859B (en) | 2020-07-28 |
Family
ID=62515162
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810059882.7A Active CN108170859B (en) | 2018-01-22 | 2018-01-22 | Voice query method, device, storage medium and terminal equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108170859B (en) |
Families Citing this family (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110619050B (en) * | 2018-06-20 | 2023-05-09 | 华为技术有限公司 | Intention recognition method and device |
| CN109033075B (en) * | 2018-06-29 | 2022-04-15 | 北京百度网讯科技有限公司 | Intention matching method and device, storage medium and terminal equipment |
| CN108897867B (en) * | 2018-06-29 | 2021-02-09 | 北京百度网讯科技有限公司 | Data processing method, device, server and medium for knowledge question answering |
| CN109299129A (en) * | 2018-09-05 | 2019-02-01 | 深圳壹账通智能科技有限公司 | Data query method, apparatus, computer equipment and the storage medium of natural language |
| CN110968686A (en) * | 2018-09-28 | 2020-04-07 | 百度在线网络技术（北京）有限公司 | Intention recognition method, device, equipment and computer readable medium |
| CN109522393A (en) * | 2018-10-11 | 2019-03-26 | 平安科技（深圳）有限公司 | Intelligent answer method, apparatus, computer equipment and storage medium |
| CN109582690A (en) * | 2018-10-23 | 2019-04-05 | 上海航空工业(集团)有限公司 | A kind of quality audit information query method based on artificial intelligent voice interaction |
| CN109859752A (en) * | 2019-01-02 | 2019-06-07 | 珠海格力电器股份有限公司 | A kind of sound control method, device, storage medium and voice joint control system |
| CN109753436A (en) * | 2019-01-09 | 2019-05-14 | 中国联合网络通信集团有限公司 | Test method, device and the storage medium of intelligent customer service system |
| CN110377812A (en) * | 2019-06-14 | 2019-10-25 | 平安科技（深圳）有限公司 | Self-help shopping method, apparatus, equipment and computer readable storage medium |
| CN110427611A (en) * | 2019-06-26 | 2019-11-08 | 深圳追一科技有限公司 | Text handling method, device, equipment and storage medium |
| CN112307154A (en) * | 2019-07-17 | 2021-02-02 | 百度时代网络技术（北京）有限公司 | Advertisement promotion result display method and device, electronic equipment and storage medium |
| CN110378747A (en) * | 2019-07-25 | 2019-10-25 | 苏州思必驰信息科技有限公司 | Speech advertising distribution method and device |
| CN112416984A (en) * | 2019-08-21 | 2021-02-26 | 华为技术有限公司 | Data processing method and device |
| CN110737687A (en) * | 2019-09-06 | 2020-01-31 | 平安普惠企业管理有限公司 | Data query method, device, equipment and storage medium |
| CN110765342A (en) * | 2019-09-12 | 2020-02-07 | 竹间智能科技（上海）有限公司 | Information query method and device, storage medium and intelligent terminal |
| CN110825949A (en) * | 2019-09-19 | 2020-02-21 | 平安科技（深圳）有限公司 | Information retrieval method based on convolutional neural network and related equipment thereof |
| CN110799973A (en) * | 2019-09-27 | 2020-02-14 | 京东方科技集团股份有限公司 | Information query method and display device |
| CN111324626B (en) * | 2020-01-21 | 2022-07-12 | 思必驰科技股份有限公司 | Search method and device based on voice recognition, computer equipment and storage medium |
| CN111429903B (en) * | 2020-03-19 | 2021-02-05 | 百度在线网络技术（北京）有限公司 | Audio signal identification method, device, system, equipment and readable medium |
| CN112069206B (en) * | 2020-03-31 | 2022-02-22 | 北京来也网络科技有限公司 | Data query method, device, medium and computing device based on RPA and AI |
| CN111782763A (en) * | 2020-05-22 | 2020-10-16 | 平安科技（深圳）有限公司 | Information retrieval method based on voice semantics and related equipment thereof |
| CN111539208B (en) * | 2020-06-22 | 2023-11-14 | 北京百度网讯科技有限公司 | Sentence processing method and device, electronic device and readable storage medium |
| CN112399019A (en) * | 2020-09-16 | 2021-02-23 | 中国农业银行股份有限公司河北省分行 | Intelligent outbound call method, terminal equipment and readable storage medium |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103064936A (en) * | 2012-12-24 | 2013-04-24 | 北京百度网讯科技有限公司 | Voice-input-based image information extraction analysis method and device |
| CN104424216A (en) * | 2013-08-23 | 2015-03-18 | 佳能株式会社 | Method and device for intention digging |
| CN104598445A (en) * | 2013-11-01 | 2015-05-06 | 腾讯科技（深圳）有限公司 | Automatic question-answering system and method |
| CN104951458A (en) * | 2014-03-26 | 2015-09-30 | 华为技术有限公司 | Method and equipment for helping processing based on semantic recognition |
| CN105912630A (en) * | 2016-04-07 | 2016-08-31 | 北京搜狗科技发展有限公司 | Information expansion method and device |
| US9460209B1 (en) * | 2010-02-09 | 2016-10-04 | Google Inc. | Identifying non-search actions based on a search query |
| CN106328166A (en) * | 2016-08-31 | 2017-01-11 | 上海交通大学 | Man-machine dialogue anomaly detection system and method |
| CN106372132A (en) * | 2016-08-25 | 2017-02-01 | 北京百度网讯科技有限公司 | Artificial intelligence-based query intention prediction method and apparatus |
| CN106407333A (en) * | 2016-09-05 | 2017-02-15 | 北京百度网讯科技有限公司 | Artificial intelligence-based spoken language query identification method and apparatus |
| CN106682192A (en) * | 2016-12-29 | 2017-05-17 | 北京奇虎科技有限公司 | Method and device for training answer intention classification model based on search keywords |
| CN107045496A (en) * | 2017-04-19 | 2017-08-15 | 畅捷通信息技术股份有限公司 | The error correction method and error correction device of text after speech recognition |
| CN107357875A (en) * | 2017-07-04 | 2017-11-17 | 北京奇艺世纪科技有限公司 | A kind of voice search method, device and electronic equipment |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8190627B2 (en) * | 2007-06-28 | 2012-05-29 | Microsoft Corporation | Machine assisted query formulation |
-
2018
- 2018-01-22 CN CN201810059882.7A patent/CN108170859B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9460209B1 (en) * | 2010-02-09 | 2016-10-04 | Google Inc. | Identifying non-search actions based on a search query |
| CN103064936A (en) * | 2012-12-24 | 2013-04-24 | 北京百度网讯科技有限公司 | Voice-input-based image information extraction analysis method and device |
| CN104424216A (en) * | 2013-08-23 | 2015-03-18 | 佳能株式会社 | Method and device for intention digging |
| CN104598445A (en) * | 2013-11-01 | 2015-05-06 | 腾讯科技（深圳）有限公司 | Automatic question-answering system and method |
| CN104951458A (en) * | 2014-03-26 | 2015-09-30 | 华为技术有限公司 | Method and equipment for helping processing based on semantic recognition |
| CN105912630A (en) * | 2016-04-07 | 2016-08-31 | 北京搜狗科技发展有限公司 | Information expansion method and device |
| CN106372132A (en) * | 2016-08-25 | 2017-02-01 | 北京百度网讯科技有限公司 | Artificial intelligence-based query intention prediction method and apparatus |
| CN106328166A (en) * | 2016-08-31 | 2017-01-11 | 上海交通大学 | Man-machine dialogue anomaly detection system and method |
| CN106407333A (en) * | 2016-09-05 | 2017-02-15 | 北京百度网讯科技有限公司 | Artificial intelligence-based spoken language query identification method and apparatus |
| CN106682192A (en) * | 2016-12-29 | 2017-05-17 | 北京奇虎科技有限公司 | Method and device for training answer intention classification model based on search keywords |
| CN107045496A (en) * | 2017-04-19 | 2017-08-15 | 畅捷通信息技术股份有限公司 | The error correction method and error correction device of text after speech recognition |
| CN107357875A (en) * | 2017-07-04 | 2017-11-17 | 北京奇艺世纪科技有限公司 | A kind of voice search method, device and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108170859A (en) | 2018-06-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108170859B (en) | Voice query method, device, storage medium and terminal equipment | |
| US11322153B2 (en) | Conversation interaction method, apparatus and computer readable storage medium | |
| CN109918673B (en) | Semantic arbitration method and device, electronic equipment and computer-readable storage medium | |
| CN111625635A (en) | Question-answer processing method, language model training method, device, equipment and storage medium | |
| CN109299320B (en) | Information interaction method and device, computer equipment and storage medium | |
| CN112100354B (en) | Man-machine conversation method, device, equipment and storage medium | |
| CN114549874B (en) | Training method of multi-target image-text matching model, image-text retrieval method and device | |
| CN112035599B (en) | Query method and device based on vertical search, computer equipment and storage medium | |
| US11651014B2 (en) | Source code retrieval | |
| CN109492081B (en) | Text information searching and information interaction method, device, equipment and storage medium | |
| CN113836925B (en) | Training method and device for pre-training language model, electronic equipment and storage medium | |
| CN107958059B (en) | Intelligent question answering method, device, terminal and computer readable storage medium | |
| CN111177358B (en) | Intention recognition method, server and storage medium | |
| CN111488468A (en) | Geographic information knowledge point extraction method and device, storage medium and computer equipment | |
| CN113282754A (en) | Public opinion detection method, device, equipment and storage medium for news events | |
| CN111274822A (en) | Semantic matching method, device, equipment and storage medium | |
| CN111737990A (en) | Word slot filling method, device, equipment and storage medium | |
| CN110795942A (en) | Keyword determination method and device based on semantic recognition and storage medium | |
| CN111639160A (en) | Domain identification method, interaction method, electronic device and storage medium | |
| CN107368525B (en) | Method and device for searching related words, storage medium and terminal equipment | |
| CN112541091A (en) | Image searching method, device, server and storage medium | |
| CN111930919B (en) | Enterprise online education APP voice interaction implementation method | |
| CN112447173A (en) | Voice interaction method and device and computer storage medium | |
| CN104731918A (en) | Voice search method and device | |
| CN114970541A (en) | Text semantic understanding method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |