CN110765338A - Data processing method and device and data processing device - Google Patents
Data processing method and device and data processing device Download PDFInfo
- Publication number
- CN110765338A CN110765338A CN201810835658.2A CN201810835658A CN110765338A CN 110765338 A CN110765338 A CN 110765338A CN 201810835658 A CN201810835658 A CN 201810835658A CN 110765338 A CN110765338 A CN 110765338A
- Authority
- CN
- China
- Prior art keywords
- replied
- information
- reply
- determining
- message
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
Abstract
The embodiment of the invention provides a data processing method and device and a device for data processing. The method specifically comprises the following steps: determining information to be replied and the corresponding text of the information to be replied; determining a reply candidate corresponding to the information to be replied according to the information to be replied and the above; and outputting the reply candidate. The embodiment of the invention can improve the correlation degree between the reply candidate and the reply intention corresponding to the message to be replied, thereby improving the accuracy of the reply candidate.
Description
Technical Field
The present invention relates to the field of communications technologies, and in particular, to a data processing method and apparatus, and an apparatus for data processing.
Background
With the development of communication technology, communication applications such as short message applications and instant messaging applications have been accepted by more and more users. For example, the user can transmit text and multimedia (including video and audio) information with his family, friends or colleagues through the instant messaging application, thereby realizing the communication function.
At present, in order to improve the efficiency of information reply, the existing scheme may provide an intelligent reply function, which may generate reply candidates for the information to be replied from the sender, so that the user directly selects a desired reply candidate for reply; the operation of manually inputting the reply content by the user can be saved, so that the reply efficiency of the user can be improved.
The reply candidates of the existing schemes are usually content generated on the basis of analyzing information from the reply-to-be-replied. For example, the information to be replied is "how much is the weather today? The content of the information to be replied can be analyzed by using a natural language processing technology, and reply candidates such as "sunny days", "cloudy days", and "cloudy" are generated according to the analysis result.
However, in some scenarios, existing solutions may have a lower accuracy in analyzing reply candidates generated based on information from the message to be replied.
For example, in a conversation scenario, the content of the communication conversation between the user a and the user B is as follows:
the user A: how do the weather today?
And a user B: in sunny days.
The user A: that sky wool?
Then for the terminal used by user B, the existing scheme running on it can be to transmit "that tomorrow? "as the information to be replied, further obtain reply candidates such as" tomorrow can "," tomorrow should not be done "and so on; however, these reply candidates are not related to the reply intention corresponding to the message to be replied, and thus have lower accuracy.
Disclosure of Invention
Embodiments of the present invention provide a data processing method, an apparatus, and an apparatus for data processing, which can improve a correlation between a reply candidate and a reply intention corresponding to a message to be replied, and further can improve accuracy of the reply candidate.
In order to solve the above problem, an embodiment of the present invention discloses a data processing method, including:
determining information to be replied and the corresponding text of the information to be replied;
determining a reply candidate corresponding to the information to be replied according to the information to be replied and the above;
and outputting the reply candidate.
On the other hand, the embodiment of the invention discloses a data processing device, which comprises:
the information upper text determining module is used for determining the information to be replied and the upper text corresponding to the information to be replied;
a reply candidate determining module, configured to determine, according to the information to be replied and the above, a reply candidate corresponding to the information to be replied; and
and the reply candidate output module is used for outputting the reply candidate.
In yet another aspect, an embodiment of the present invention discloses an apparatus for data processing, including a memory, and one or more programs, where the one or more programs are stored in the memory, and configured to be executed by the one or more processors includes instructions for:
determining information to be replied and the corresponding text of the information to be replied;
determining a reply candidate corresponding to the information to be replied according to the information to be replied and the above;
and outputting the reply candidate.
In yet another aspect, an embodiment of the invention discloses a machine-readable medium having stored thereon instructions, which, when executed by one or more processors, cause an apparatus to perform a data processing method as described in one or more of the preceding.
The embodiment of the invention has the following advantages:
in the embodiment of the invention, the text corresponding to the information to be replied is used in the process of determining the reply candidate corresponding to the information to be replied, and the text can contain the relevant information corresponding to the information to be replied, and particularly, the text can reflect the reply intention corresponding to the information to be replied to a certain extent; therefore, the correlation between the reply candidate and the reply intention corresponding to the message to be replied can be improved, and the accuracy of the reply candidate can be further improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments of the present invention will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without inventive labor.
FIG. 1 is a schematic representation of an application environment for a data processing method of an embodiment of the present invention;
FIG. 2 is a flow chart of steps of a first embodiment of a data processing method of the present invention;
FIG. 3 is an example of communication session content according to an embodiment of the present invention;
FIG. 4 is a flowchart illustrating steps of a second embodiment of a data processing method according to the present invention;
FIG. 5 is a block diagram of an embodiment of a data processing apparatus of the present invention;
FIG. 6 is a block diagram of an apparatus 800 for data processing of the present invention; and
fig. 7 is a schematic diagram of a server in some embodiments of the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The embodiment of the invention provides a data processing scheme, which can determine information to be replied and the corresponding text of the information to be replied; determining a reply candidate corresponding to the information to be replied according to the information to be replied and the above; and outputs the reply candidate.
The embodiment of the invention can automatically determine the reply candidate corresponding to the information to be replied so that the user can directly select the required reply candidate to reply; the operation of manually inputting the reply content by the user can be saved, so that the reply efficiency of the user can be improved.
In the embodiment of the invention, the text corresponding to the information to be replied is used in the process of determining the reply candidate corresponding to the information to be replied, and the text can contain the relevant information corresponding to the information to be replied, and particularly, the text can reflect the reply intention corresponding to the information to be replied to a certain extent; therefore, the correlation between the reply candidate and the reply intention corresponding to the message to be replied can be improved, and the accuracy of the reply candidate can be further improved.
In an application example 1 of the present invention, the content of the communication session between the user a and the user B is as follows:
the user A: how do the weather today?
And a user B: in sunny days.
The user A: that sky wool?
For the terminal used by the user B, it may perform the scheme of the embodiment of the present invention, and specifically, may transmit "the sky? "as the information to be replied, since the above corresponding to the information to be replied can reflect the reply intention corresponding to the information to be replied: "weather", therefore, the embodiment of the present invention may obtain recovery candidates such as "sunny day", "unknown", "cloudy day", and the like. The reply candidates have a certain correlation with the reply intentions corresponding to the information to be replied, so that the accuracy of the reply candidates can be improved.
The data processing method provided by the embodiment of the invention can be applied to Application environments such as websites and/or APPs (Application programs), can improve the correlation between the reply candidates and the reply intents corresponding to the information to be replied, and further can improve the accuracy of the reply candidates. The application environment of the website may include: webpage version instant messaging environment, etc., the APP application environment may include: and the application type of instant messaging environment, such as a WeChat environment, a short message environment, an email environment and the like.
The data processing method provided by the embodiment of the invention can be applied to the application environment shown in FIG. 1, such as
As shown in fig. 1, the client 100 and the server 200 are located in a wired or wireless network, through which the client 100 and the server 200 perform data interaction.
Optionally, the client 100 may be run on a communication terminal, which specifically includes but is not limited to: smart phones, tablet computers, electronic book readers, MP3 (Moving picture Experts Group Audio Layer III) players, MP4 (Moving picture Experts Group Audio Layer IV) players, laptop portable computers, car-mounted computers, desktop computers, set-top boxes, smart televisions, wearable devices, and the like.
The communication opposite end is other participants of the communication session, which the communication terminal participates in, and may only have one other participant or may be one of a plurality of participants. The communication Number may be an identifier that can uniquely identify each participating communication terminal in the communication session, and may be, for example, a fixed phone Number, an MDN (Mobile Directory Number), an MSISDN (Mobile station international Subscriber Directory Number), or an MIN (Mobile identification Number), and may also be an instant messaging user account or a social network user account.
In an embodiment of the present invention, the client 100 running on the communication terminal may establish a communication session with the communication peer, where the communication session content specifically includes: the information received from the correspondent node and/or the information sent to the correspondent node. The client 100 may determine the message to be replied and the text thereof from the communication session content, determine the reply candidate corresponding to the message to be replied according to the message to be replied and the text, and display the reply candidate to the user for the user to select.
Of course, the method of the embodiment of the present invention executed by the client 100 is only an example, and actually, the server 200 may also determine the reply candidate corresponding to the information to be replied by executing the method of the embodiment of the present invention.
Method embodiment one
Referring to fig. 2, a flowchart illustrating steps of a first embodiment of a data processing method according to the present invention is shown, where the method may specifically include the following steps:
and step 203, outputting the reply candidate.
At least one step of the embodiment shown in fig. 2 may be performed by a server and/or a client, and of course, the embodiment of the present invention does not limit the specific execution subject of each step.
In an alternative embodiment of the invention, at least one step of the embodiment shown in fig. 2 may be performed by the input method APP.
The Input Method (IME) mentioned here means: a user of the device inputs characters and symbols into the device using an input tool such as a keyboard, a tablet, a touch panel, or a voice capturing device. The input method uses a certain coding rule to convert the input information (pinyin string, pencils string, handwriting board input track and the like) of a user into character codes which can be processed by a machine. The input method is an entry program for a user to perform computer information processing. The characters supported by the input method may include: words, letters, numbers, symbols, etc., the words may further include: english, Chinese, Japanese, Korean, etc.
As a hosted program, the input method APP can be applied to various Application scenarios, for example, a user can input characters in an instant messaging APP (Application program) to communicate with other users, or can input characters in a document APP to edit a document, or can input characters in a search APP to search, and so on.
In step 201, the message to be replied and its text can be determined from the communication session content.
Optionally, the step 201 determines the information to be replied and the above process corresponding to the information to be replied, which may specifically include:
taking the information received from the communication opposite terminal for the last time or a plurality of times as the information to be replied;
and acquiring information except the information to be replied from the communication conversation content corresponding to the communication opposite terminal as the information corresponding to the information to be replied.
Referring to fig. 3, an example of a communication session content is shown, which may display information through a communication session window 300 in order of time from far to near; the displayed communication session content may specifically include: information 1 sent by the correspondent node and information 2 … sent by the communication terminal are information n sent by the correspondent node, where n is a natural number greater than 1, and information n is the information received from the correspondent node most recently.
According to an embodiment, the information n may be used as the information to be replied 301, and the previous (n-1) pieces of information may be used as the above 302 corresponding to the information to be replied.
According to another embodiment, when the value of N is larger, N pieces of information can be selected from the N pieces of information according to the sequence from near to far according to time; further, the information N may be used as the information to be replied, and the first (N-1) information of the N information may be used as the above corresponding to the information to be replied; n is a natural number less than N.
According to still another embodiment, when the value of n is large, M pieces of information generated within a preset time period may be selected from the n pieces of information according to the generation time of the information; further, the information n may be regarded as the information to be replied, and the first (M-1) information of the M information may be regarded as the above corresponding to the information to be replied. The end time of the preset time period may be the current time, and the length of the preset time period may be a time threshold, for example, the time threshold may be 1 hour, 2 hours, and the like.
In an optional embodiment of the present application, considering a situation that some users are used to send short messages, that is, used to divide a sentence or an expression into multiple messages for sending, the messages received from the opposite communication terminal for the last time can be used as the messages to be replied, so that the habit that the users divide a sentence or an expression into multiple messages for sending can be satisfied.
In an optional embodiment of the present application, the above text corresponding to the information to be replied may also be filtered, and the corresponding filtering process may include: determining the correlation degree between the information to be replied and the text; and screening the corresponding upper text of the information to be replied according to the correlation. For example, the above that the correlation is less than the correlation threshold may be filtered out, and the above that the correlation exceeds the correlation threshold may be retained. The filtered text above may be used as input to step 202.
In this embodiment of the present invention, the type of the information to be replied may include, but is not limited to: short messages, QQ information, WeChat information, WhatsApp (Watson Purpt) information, Google Talk messages, or MSN Messenger (Microsoft Network Messenger) messages, etc.
In an application example 2 of the present invention, the contents of the communication session between the user C and the user D are arranged in the order of time from far to near:
and a user C: is there available at night?
And a user D: and is empty.
And a user C: that sky wool?
Then, the embodiment of the present invention may use "the sky wool" as the information to be replied, and "will be available at night? "and" none "as above for the information to be replied.
For application example 2, the above corresponding to the information to be replied may reflect the reply intention corresponding to the information to be replied: the present invention can obtain the reply candidates of "available", "not available in tomorrow", "available again in tomorrow", and "not necessary". The reply candidates have a certain correlation with the reply intentions corresponding to the information to be replied, so that the accuracy of the reply candidates can be improved.
It should be noted that the information to be replied corresponding to the application examples 1 and 2 is the same, and both are "namingtze", and since only the information to be replied is considered in the process of determining the reply candidate, the existing scheme will give the same reply candidate for the application examples 1 and 2. In the process of determining the reply candidate corresponding to the information to be replied according to the embodiment of the present invention, the information to be replied and the corresponding text thereof may be used, and since the text corresponding to application example 1 and application example 2 is different, different reply candidates may be given for application example 1 and application example 2 according to the embodiment of the present invention, so that the accuracy of the reply candidate may be improved.
The embodiment of the invention can provide the following technical scheme for determining the reply candidate corresponding to the information to be replied:
determination of scheme 1,
In the scheme 1, the step 202 of determining the reply candidate corresponding to the message to be replied according to the message to be replied and the above, specifically may include:
respectively determining the information to be replied and the first representation and the second representation corresponding to the information to be replied;
inputting the first representation and the second representation into a first reply model to obtain a reply candidate output by the first reply model; the first recovery model may be obtained by learning first training data, and the first training data specifically may include: a plurality of training samples, one training sample may specifically include: a to-be-replied information sample, a above sample, and a reply sample.
The first representation and the second representation are used for representing the information to be replied and the above, respectively. Optionally, the form of the first representation and the second representation may include: vectors or text.
The first recovery model may be a supervised machine learning model, and the machine learning method may include: bayesian method, or neural network method, etc. The machine learning method can design and analyze algorithms which can enable a computer to automatically learn, the algorithms can automatically analyze and obtain rules from training data, and the rules are used for predicting unknown data, so that the method has better robustness and can obtain higher precision.
Specifically, in the embodiments of the present invention, parameters may be defined as follows: c. m and r, m represents information to be replied, c represents the above, and r represents a reply candidate, embodiments of the present invention may learn P (r | c, m) from training data by a machine learning method, where P (r | c, m) represents a probability that r appears on the premise that c and m appear.
In an alternative embodiment of the present invention, the training data may be learned by a neural network method to obtain the parameters of the first recovery model.
Alternatively, the parameters of the first regression model may be parameters of a fully-connected matrix, which is relative to a fully-connected layer of the neural network, and each node of the fully-connected layer is connected to all nodes of a previous layer for integrating the extracted features.
Learning the training data by adopting a neural network method, wherein the learning target is as follows: the parameters of the full-connection matrix are iteratively learned to enable the conditional probability P (r) corresponding to the training samplei︱ci，mi) More closely to 1, wherein ci、mi、riRespectively representing the above sample, the information sample to be replied and the reply sample included in one training sample.
The iterative method employed for iterative learning may include, but is not limited to: gradient descent, least squares, newtons, etc., it being understood that embodiments of the present invention are not limited to particular iterative methods.
The training data in the embodiment of the present invention may be derived from a session database, and the session data in the session database may be derived from client data collected by a communication application.
The embodiment of the invention can obtain the first representation or the second representation in the form of the vector through the following technical scheme:
technical solution 1
Technical solution 1 may obtain the first representation or the second representation in the form of a vector through a classification model.
In an alternative embodiment of the invention, the second representation may be of the object class to which it belongs. A person skilled in the art or a user may determine the plurality of categories according to actual application requirements, and thus, which of the plurality of categories the above belongs to may be determined through the classification model, that is, the target category corresponding to the first representation may be determined. Examples of preset categories may include: "weather", "free", "movie", "television", "music", "soccer", etc. In this way, it can be determined by the classification model that the above in application example 1 belongs to the "weather" category, and it can be determined by the classification model that the above in application example 2 belongs to the "whether or not to be idle" category, and the like.
The corresponding category can reflect the reply intention corresponding to the information to be replied to a certain extent; therefore, the correlation between the reply candidate and the reply intention corresponding to the message to be replied can be improved, and the accuracy of the reply candidate can be further improved.
Alternatively, the classification model may be a fastText model. The fastText model may output, for an input sequence of words (a piece of text or a sentence), the probability that the sequence of words belongs to different categories. The fastText model can form a feature vector by words and phrases in a word sequence, the feature vector is mapped to the middle layer through linear transformation, and the middle layer is mapped to the corresponding preset category. Optionally, the fastText may use a non-linear activation function in the process of mapping to the corresponding preset category. The fastText has the advantages of high speed and high accuracy, and the specific classification model is not limited in the embodiment of the invention.
In an alternative embodiment of the present invention, the first representation may be a target category to which the information to be replied belongs. For the determination process of the target category to which the to-be-replied message belongs, since it is similar to the determination process of the target category to which the to-be-replied message belongs, it is not described herein again, and it is sufficient to refer to each other. Of course, the first representation of the embodiment of the present invention may also be the message to be replied itself.
The classification result output by the fastText model can be in a vector form, so that the embodiment of the invention can obtain the first representation or the second representation in the vector form.
Technical scheme 2,
Technical solution 1 may obtain a first representation or a second representation in a vector form through a word vector generation model; the word vector generation model can convert words in natural language into dense vectors that can be understood by a computer. Optionally, word vectors of the word sequence may be spliced to obtain word vectors corresponding to the word sequence.
Word vectors have good semantic properties and are a common way to represent word features. The value of each dimension of the word vector represents a feature with a certain semantic and grammatical interpretation. Therefore, each dimension of a word vector may be referred to as a word feature.
Examples of the word vector generation model may include: word2 vec. Word2vec may project words into a K-dimensional vector space, and each Word may be represented by a K-dimensional vector. It is to be understood that embodiments of the present invention are not limited to generating models of word vectors based thereon.
Determination of scheme 2,
In the determining scheme 2, the step 202 determines, according to the information to be replied and the above, a reply candidate corresponding to the information to be replied, and specifically may include:
determining the corresponding theme;
and determining reply candidates corresponding to the information to be replied according to the theme and the information to be replied.
A theme may refer to the central idea presented above. The embodiment of the invention can represent the theme through the theme key words, and the theme key words can refer to the key words capable of embodying the theme of the text content.
The embodiment of the present invention may provide the following determination manner for determining the above corresponding theme:
determining mode 1, and determining the corresponding topic keyword by adopting a TF-IDF (term frequency-inverse document frequency algorithm) method.
The main idea of TF-IDF is: if a word or phrase appears frequently in a document or a piece of text with high frequency TF and rarely appears in other documents or texts, the word or phrase is considered to have good classification capability and is suitable for classification.
Determining mode 2, determining the corresponding topic keyword by using an LDA (Latent Dirichlet Allocation) model.
The LDA model is a document generation model and is an unsupervised machine learning technology. It considers a document with multiple topics, and each topic corresponds to a different topic keyword. The construction process of a document includes firstly selecting a theme with a certain probability, and then selecting a theme keyword with a certain probability under the theme, so that a first theme keyword of the document is generated. This process is repeated over and over again to produce the entire article. The LDA is used as the inverse process of the document generation process, namely, the topics of the document and topic keywords corresponding to the topics are found according to the obtained document.
And determining a mode 3, determining the corresponding category of the text by adopting a classification model, and obtaining the topic keyword according to the information of the category. The classification model may include: the aforementioned fasttext model.
It can be understood that, according to the practical application requirement, a person skilled in the art may adopt any one or a combination of the above determination modes 1 to 3, and the embodiment of the present invention does not impose a limitation on the specific process for determining the above corresponding subject.
In an alternative embodiment of the present application, the above may correspond to a plurality of topics, in which case, the target topic may be determined from the plurality of topics.
In the embodiment of the present invention, the target theme may satisfy at least one of the following conditions:
the condition 1 is that the reply content corresponding to the target theme is not included in the text; and
and 2, the target subject is related to the information to be replied.
Condition 1 may state that the target topic has not been answered yet to rule out the questions that have been answered; condition 2 may exclude topics that are not related to the message to be replied to.
Optionally, the process of determining the target topic from the plurality of topics may include: and judging whether the reply content corresponding to one theme is contained in the text, and if so, filtering the theme.
In one example of the present invention, it is assumed that the above includes: "will not rain in tomorrow? "," do you have an umbrella in your home "," do you sit in a subway and get away? "," rainy tomorrow, I have an umbrella ", suppose that the information to be replied includes: "can sit on a subway"; in this example, the above corresponding subject matter may include: whether it rains in tomorrow, whether there is an umbrella, and whether it sits on a subway; the reply contents of 'whether it rains in tomorrow' and 'whether there is an umbrella' are included in the above text, so that the theme of 'whether to sit on the subway' can be kept, that is, the target theme of 'whether to sit on the subway' can be determined, and therefore the accuracy of the corresponding theme can be improved.
According to an embodiment, the determining, according to the topic and the to-be-replied message, a reply candidate corresponding to the to-be-replied message may specifically include: and screening reply contents corresponding to the information to be replied according to the theme so as to obtain reply candidates corresponding to the information to be replied.
The embodiment of the invention can determine the reply content corresponding to the information to be replied by utilizing a text analysis technology, and the reply content can be a plurality of. Furthermore, a plurality of reply contents can be screened according to the theme, and the reply contents matched with the theme are used as reply candidates corresponding to the information to be replied. The filtering can filter out the reply content which is not related to the subject, and can keep the reply content which is related to the subject.
Taking application example 1 as an example, the multiple reply contents corresponding to the information to be replied include: the "sunny day" and "sunny day" are not known, the "sunny day" and "sunny day" are filtered, and the "sunny day" and "unknown" and "cloudy day" are reserved according to the theme "weather".
Taking application example 2 as an example, the multiple reply contents corresponding to the information to be replied include: "tomorrow can", "tomorrow should have nothing, better", "drive do", "weather", etc., then "can be better", "drive do", "weather" is filtered out according to the theme "whether idle", and keep "tomorrow can", "tomorrow should have nothing", etc.
According to another embodiment, the determining the reply candidate corresponding to the message to be replied according to the topic and the message to be replied may specifically include: inputting the information to be replied into a second reply model corresponding to the theme to obtain reply candidates output by the second reply model; the second reply model is obtained by learning second training data, where the second training data may include: and the information sample to be replied and the reply sample correspond to the theme.
The second reply model may be a reply model corresponding to a topic, and the second training data may include: the training sample corresponds to a topic, so that the reply candidate output by the second reply model can be related to the topic, and the accuracy of the reply candidate can be improved.
According to another embodiment, the determining, according to the theme and the to-be-replied message, the reply candidate corresponding to the to-be-replied message may specifically include: and searching in the data source corresponding to the theme according to the information to be replied and the environmental characteristics of the user to obtain a reply candidate corresponding to the information to be replied.
The data source corresponding to the subject can be used to provide data corresponding to the subject. The data source corresponding to the theme may include: databases of APPs, databases of websites, etc. Taking the theme as "weather", for example, the data source corresponding to the theme may include: a database of weather APPs, a database of weather websites, etc. Taking the theme as "idle or not" as an example, the data source corresponding to the theme may include: memorandum APP or reminding data recorded by reminding APP, and the like, wherein the reminding data can include the user's travel, so that the memorandum APP or the reminding data can be used for judging whether the user is idle. It is understood that the embodiments of the present invention do not limit the specific data sources corresponding to the specific subjects.
The environmental characteristics of the user may refer to the environmental characteristics of the user within a preset time period, and the environmental characteristics may include, but are not limited to: location environment characteristics, network environment characteristics, time environment characteristics, and the like.
According to the embodiment of the invention, the reply candidate is searched according to the environmental characteristics of the user, so that the accuracy of the reply candidate can be improved. For example, a query of weather data at a specific time may be made according to the location environment feature and the time environment feature to improve the accuracy of the reply candidate. For another example, the reminding data at a specific time can be queried according to the time environment characteristics, so as to improve the accuracy of reply candidates. It is to be understood that embodiments of the present invention are not limited to the particular environmental features.
In step 203, if the execution subject is a server, the server may send a reply candidate to the client; or, if the execution subject is the client, the client may present the reply candidates to the user for selection by the user. Optionally, the reply candidates may be one or more, and the reply candidate selected by the user may be displayed in an input box of the communication APP in the embodiment of the present invention, or the reply candidate selected by the user may be sent to the correspondent node in the embodiment of the present invention, so that the reply candidate selected by the user is displayed in the communication dialog box.
To sum up, in the data processing method according to the embodiment of the present invention, the context corresponding to the to-be-replied information is used in the process of determining the reply candidate corresponding to the to-be-replied information, and the context may include the relevant information corresponding to the to-be-replied information, and particularly, the context may reflect the reply intention corresponding to the to-be-replied information to a certain extent; therefore, the correlation between the reply candidate and the reply intention corresponding to the message to be replied can be improved, and the accuracy of the reply candidate can be further improved.
Method embodiment two
Referring to fig. 4, a flowchart illustrating steps of a second embodiment of a data processing method according to the present invention is shown, where the method may specifically include the following steps:
and step 404, displaying the reply candidate.
The first category vector and the second category vector may both be category information in the form of vectors, and may optionally be determined by a classification model, such as a fasttext model.
The second category vector is used for representing the category information, so that the reply intention corresponding to the information to be replied can be reflected to a certain extent; therefore, the correlation between the reply candidate and the reply intention corresponding to the message to be replied can be improved, and the accuracy of the reply candidate can be further improved.
Moreover, the first category vector is used for representing category information of the information to be replied, so that the reply intention corresponding to the information to be replied can be reflected to a certain extent; therefore, the correlation between the reply candidate and the reply intention corresponding to the message to be replied can be improved, and the accuracy of the reply candidate can be further improved.
It should be noted that, for simplicity of description, the method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the embodiments are not limited by the order of acts described, as some steps may occur in other orders or concurrently depending on the embodiments. Further, those skilled in the art will also appreciate that the embodiments described in the specification are presently preferred and that no particular act is required of the embodiments of the application.
Device embodiment
Referring to fig. 5, a block diagram of a data processing apparatus according to an embodiment of the present invention is shown, which may specifically include: a message context determination module 501, a reply candidate determination module 502, and a reply candidate output module 503.
The information context determining module 501 is configured to determine information to be replied and contexts corresponding to the information to be replied;
a reply candidate determining module 502, configured to determine, according to the to-be-replied message and the above, a reply candidate corresponding to the to-be-replied message; and
a reply candidate output module 503, configured to output the reply candidate.
Optionally, the information context determining module 501 may include:
the information to be replied determining submodule is used for taking the information which is received from the communication opposite terminal for the last time or for the last time as the information to be replied;
the above-mentioned determining submodule is configured to obtain, from the communication session content corresponding to the communication peer, information other than the information to be replied, as the above-mentioned corresponding to the information to be replied.
Optionally, the reply candidate determination module 502 may include:
the representation determining submodule is used for respectively determining the information to be replied and the first representation and the second representation corresponding to the information to be replied;
a first model classification submodule, configured to input the first representation and the second representation into a first recovery model to obtain a recovery candidate output by the first recovery model; the first recovery model is obtained by learning first training data, where the first training data may include: a to-be-replied information sample, a above sample, and a reply sample.
Optionally, the reply candidate determination module 502 may include:
the theme determining submodule is used for determining the corresponding theme;
and the candidate determining submodule is used for determining a reply candidate corresponding to the information to be replied according to the theme and the information to be replied.
Optionally, the candidate determination sub-module may include:
and the screening unit is used for screening the reply content corresponding to the information to be replied according to the theme so as to obtain a reply candidate corresponding to the information to be replied.
Optionally, the candidate determination sub-module may include:
the second model classification unit is used for inputting the information to be replied into a second reply model corresponding to the theme so as to obtain reply candidates output by the second reply model; the second reply model is obtained by learning second training data, where the second training data may include: and the information sample to be replied and the reply sample correspond to the theme.
Optionally, the candidate determination sub-module may include:
and the searching unit is used for searching in the data source corresponding to the theme according to the information to be replied and the environment characteristics of the user so as to obtain a reply candidate corresponding to the information to be replied.
For the device embodiment, since it is basically similar to the method embodiment, the description is simple, and for the relevant points, refer to the partial description of the method embodiment.
The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
With regard to the apparatus in the above-described embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.
An embodiment of the present invention provides an apparatus for data processing, comprising a memory, and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs configured to be executed by the one or more processors include instructions for: determining information to be replied and the corresponding text of the information to be replied; determining a reply candidate corresponding to the information to be replied according to the information to be replied and the above; and outputting the reply candidate.
Fig. 6 is a block diagram illustrating an apparatus 800 for data processing in accordance with an example embodiment. For example, the apparatus 800 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, an exercise device, a personal digital assistant, and the like.
Referring to fig. 6, the apparatus 800 may include one or more of the following components: processing component 802, memory 804, power component 806, multimedia component 808, audio component 810, input/output (I/O) interface 812, sensor component 814, and communication component 816.
The processing component 802 generally controls overall operation of the device 800, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. The processing elements 802 may include one or more processors 820 to execute instructions to perform all or a portion of the steps of the methods described above. Further, the processing component 802 can include one or more modules that facilitate interaction between the processing component 802 and other components. For example, the processing component 802 can include a multimedia module to facilitate interaction between the multimedia component 808 and the processing component 802.
The memory 804 is configured to store various types of data to support operation at the device 800. Examples of such data include instructions for any application or method operating on device 800, contact data, phonebook data, messages, pictures, videos, and so forth. The memory 804 may be implemented by any type or combination of volatile or non-volatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.
Power components 806 provide power to the various components of device 800. The power components 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the apparatus 800.
The multimedia component 808 includes a screen that provides an output interface between the device 800 and a user. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 808 includes a front facing camera and/or a rear facing camera. The front-facing camera and/or the rear-facing camera may receive external multimedia data when the device 800 is in an operating mode, such as a shooting mode or a video mode. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.
The audio component 810 is configured to output and/or input audio signals. For example, the audio component 810 includes a Microphone (MIC) configured to receive external audio signals when the apparatus 800 is in an operational mode, such as a call mode, a recording mode, and a voice data processing mode. The received audio signals may further be stored in the memory 804 or transmitted via the communication component 816. In some embodiments, audio component 810 also includes a speaker for outputting audio signals.
The I/O interface 812 provides an interface between the processing component 802 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: a home button, a volume button, a start button, and a lock button.
The sensor assembly 814 includes one or more sensors for providing various aspects of state assessment for the device 800. For example, the sensor assembly 814 may detect the open/closed state of the device 800, the relative positioning of the components, such as a display and keypad of the apparatus 800, the sensor assembly 814 may also detect a change in position of the apparatus 800 or a component of the apparatus 800, the presence or absence of user contact with the apparatus 800, orientation or acceleration/deceleration of the apparatus 800, and a change in temperature of the apparatus 800. Sensor assembly 814 may include a proximity sensor configured to detect the presence of a nearby object without any physical contact. The sensor assembly 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
The communication component 816 is configured to facilitate communications between the apparatus 800 and other devices in a wired or wireless manner. The device 800 may access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 816 receives a broadcast signal or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 816 further includes a Near Field Communication (NFC) module to facilitate short-range communications. For example, the NFC module may be implemented based on radio frequency data processing (RFID) technology, infrared data association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.
In an exemplary embodiment, the apparatus 800 may be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, micro-controllers, microprocessors or other electronic components for performing the above-described methods.
In an exemplary embodiment, a non-transitory computer-readable storage medium comprising instructions, such as the memory 804 comprising instructions, executable by the processor 820 of the device 800 to perform the above-described method is also provided. For example, the non-transitory computer readable storage medium may be a ROM, a Random Access Memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.
Fig. 7 is a schematic diagram of a server in some embodiments of the invention. The server 1900, which may vary widely in configuration or performance, may include one or more Central Processing Units (CPUs) 1922 (e.g., one or more processors) and memory 1932, one or more storage media 1930 (e.g., one or more mass storage devices) storing applications 1942 or data 1944. Memory 1932 and storage medium 1930 can be, among other things, transient or persistent storage. The program stored in the storage medium 1930 may include one or more modules (not shown), each of which may include a series of instructions operating on a server. Still further, a central processor 1922 may be provided in communication with the storage medium 1930 to execute a series of instruction operations in the storage medium 1930 on the server 1900.
The server 1900 may also include one or more power supplies 1926, one or more wired or wireless network interfaces 1950, one or more input-output interfaces 1958, one or more keyboards 1956, and/or one or more operating systems 1941, such as Windows Server, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.
A non-transitory computer readable storage medium in which instructions, when executed by a processor of an apparatus (server or terminal), enable the apparatus to perform the data processing method shown in fig. 2 or fig. 3 or fig. 4.
A non-transitory computer readable storage medium in which instructions, when executed by a processor of an apparatus (server or terminal), enable the apparatus to perform a data processing method, the method comprising: determining information to be replied and the corresponding text of the information to be replied; determining a reply candidate corresponding to the information to be replied according to the information to be replied and the above; and outputting the reply candidate.
The embodiment of the invention discloses A1 and a data processing method, wherein the method comprises the following steps:
determining information to be replied and the corresponding text of the information to be replied;
determining a reply candidate corresponding to the information to be replied according to the information to be replied and the above;
and outputting the reply candidate.
A2, according to the method in A1, the determining the information to be replied and the corresponding text of the information to be replied includes:
taking the information received from the communication opposite terminal for the last time or a plurality of times as the information to be replied;
and acquiring information except the information to be replied from the communication conversation content corresponding to the communication opposite terminal as the information corresponding to the information to be replied.
A3, according to the method of A1 or A2, the determining reply candidates corresponding to the message to be replied according to the message to be replied and the above text includes:
respectively determining the information to be replied and the first representation and the second representation corresponding to the information to be replied;
inputting the first representation and the second representation into a first reply model to obtain a reply candidate output by the first reply model; the first recovery model is obtained by learning first training data, and the first training data includes: a to-be-replied information sample, a above sample, and a reply sample.
A4, according to the method of A1 or A2, the determining reply candidates corresponding to the message to be replied according to the message to be replied and the above text includes:
determining the corresponding theme;
and determining reply candidates corresponding to the information to be replied according to the theme and the information to be replied.
A5, according to the method of A4, determining reply candidates corresponding to the information to be replied according to the topic and the information to be replied includes:
and screening reply contents corresponding to the information to be replied according to the theme so as to obtain reply candidates corresponding to the information to be replied.
A6, according to the method of A4, determining reply candidates corresponding to the information to be replied according to the topic and the information to be replied includes:
inputting the information to be replied into a second reply model corresponding to the theme to obtain reply candidates output by the second reply model; the second reply model is obtained by learning second training data, and the second training data includes: and the information sample to be replied and the reply sample correspond to the theme.
A7, according to the method of A4, determining reply candidates corresponding to the information to be replied according to the topic and the information to be replied includes:
and searching in the data source corresponding to the theme according to the information to be replied and the environmental characteristics of the user to obtain a reply candidate corresponding to the information to be replied.
The embodiment of the invention discloses B8 and a data processing device, which comprises:
the information upper text determining module is used for determining the information to be replied and the upper text corresponding to the information to be replied;
a reply candidate determining module, configured to determine, according to the information to be replied and the above, a reply candidate corresponding to the information to be replied; and
and the reply candidate output module is used for outputting the reply candidate.
B9, the apparatus of B8, the information above determining module comprising:
the information to be replied determining submodule is used for taking the information which is received from the communication opposite terminal for the last time or for the last time as the information to be replied;
the above-mentioned determining submodule is configured to obtain, from the communication session content corresponding to the communication peer, information other than the information to be replied, as the above-mentioned corresponding to the information to be replied.
B10, the apparatus of B8 or B9, the reply candidate determination module comprising:
the representation determining submodule is used for respectively determining the information to be replied and the first representation and the second representation corresponding to the information to be replied;
a first model classification submodule, configured to input the first representation and the second representation into a first recovery model to obtain a recovery candidate output by the first recovery model; the first recovery model is obtained by learning first training data, and the first training data includes: a to-be-replied information sample, a above sample, and a reply sample.
B11, the apparatus of B8 or B9, the reply candidate determination module comprising:
the theme determining submodule is used for determining the corresponding theme;
and the candidate determining submodule is used for determining a reply candidate corresponding to the information to be replied according to the theme and the information to be replied.
B12, the apparatus of B11, the candidate determination submodule comprising:
and the screening unit is used for screening the reply content corresponding to the information to be replied according to the theme so as to obtain a reply candidate corresponding to the information to be replied.
B13, the apparatus of B11, the candidate determination submodule comprising:
the second model classification unit is used for inputting the information to be replied into a second reply model corresponding to the theme so as to obtain reply candidates output by the second reply model; the second reply model is obtained by learning second training data, and the second training data includes: and the information sample to be replied and the reply sample correspond to the theme.
B14, the apparatus of B11, the candidate determination submodule comprising:
and the searching unit is used for searching in the data source corresponding to the theme according to the information to be replied and the environment characteristics of the user so as to obtain a reply candidate corresponding to the information to be replied.
The embodiment of the invention discloses C15, an apparatus for data processing, comprising a memory, and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs configured to be executed by the one or more processors comprise instructions for:
determining information to be replied and the corresponding text of the information to be replied;
determining a reply candidate corresponding to the information to be replied according to the information to be replied and the above;
and outputting the reply candidate.
C16, the device according to C15, the determining the information to be replied and the corresponding text of the information to be replied includes:
taking the information received from the communication opposite terminal for the last time or a plurality of times as the information to be replied;
and acquiring information except the information to be replied from the communication conversation content corresponding to the communication opposite terminal as the information corresponding to the information to be replied.
C17, the apparatus according to C15 or C16, wherein the determining the reply candidate corresponding to the message to be replied according to the message to be replied and the above text includes:
respectively determining the information to be replied and the first representation and the second representation corresponding to the information to be replied;
inputting the first representation and the second representation into a first reply model to obtain a reply candidate output by the first reply model; the first recovery model is obtained by learning first training data, and the first training data includes: a to-be-replied information sample, a above sample, and a reply sample.
C18, the apparatus according to C15 or C16, wherein the determining the reply candidate corresponding to the message to be replied according to the message to be replied and the above text includes:
determining the corresponding theme;
and determining reply candidates corresponding to the information to be replied according to the theme and the information to be replied.
C19, according to the apparatus of C18, determining reply candidates corresponding to the information to be replied according to the topic and the information to be replied includes:
and screening reply contents corresponding to the information to be replied according to the theme so as to obtain reply candidates corresponding to the information to be replied.
C20, according to the apparatus of C18, determining reply candidates corresponding to the information to be replied according to the topic and the information to be replied includes:
inputting the information to be replied into a second reply model corresponding to the theme to obtain reply candidates output by the second reply model; the second reply model is obtained by learning second training data, and the second training data includes: and the information sample to be replied and the reply sample correspond to the theme.
C21, according to the apparatus of C18, determining reply candidates corresponding to the information to be replied according to the topic and the information to be replied includes:
and searching in the data source corresponding to the theme according to the information to be replied and the environmental characteristics of the user to obtain a reply candidate corresponding to the information to be replied.
Embodiments of the present invention disclose D22, a machine-readable medium having instructions stored thereon, which when executed by one or more processors, cause an apparatus to perform a data processing method as described in one or more of a 1-a 7.
Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.
It will be understood that the invention is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the invention is limited only by the appended claims.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
The data processing method, the data processing apparatus and the apparatus for data processing provided by the embodiments of the present invention are described in detail above, and specific examples are applied herein to explain the principles and embodiments of the present invention, and the above description of the embodiments is only used to help understanding the method and the core ideas of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (10)
1. A method of data processing, the method comprising:
determining information to be replied and the corresponding text of the information to be replied;
determining a reply candidate corresponding to the information to be replied according to the information to be replied and the above;
and outputting the reply candidate.
2. The method according to claim 1, wherein the determining the information to be replied and the context to which the information to be replied corresponds comprises:
taking the information received from the communication opposite terminal for the last time or a plurality of times as the information to be replied;
and acquiring information except the information to be replied from the communication conversation content corresponding to the communication opposite terminal as the information corresponding to the information to be replied.
3. The method according to claim 1 or 2, wherein the determining the reply candidate corresponding to the message to be replied according to the message to be replied and the above text comprises:
respectively determining the information to be replied and the first representation and the second representation corresponding to the information to be replied;
inputting the first representation and the second representation into a first reply model to obtain a reply candidate output by the first reply model; the first recovery model is obtained by learning first training data, and the first training data includes: a to-be-replied information sample, a above sample, and a reply sample.
4. The method according to claim 1 or 2, wherein the determining the reply candidate corresponding to the message to be replied according to the message to be replied and the above text comprises:
determining the corresponding theme;
and determining reply candidates corresponding to the information to be replied according to the theme and the information to be replied.
5. The method according to claim 4, wherein the determining the reply candidate corresponding to the message to be replied according to the topic and the message to be replied comprises:
and screening reply contents corresponding to the information to be replied according to the theme so as to obtain reply candidates corresponding to the information to be replied.
6. The method according to claim 4, wherein the determining the reply candidate corresponding to the message to be replied according to the topic and the message to be replied comprises:
inputting the information to be replied into a second reply model corresponding to the theme to obtain reply candidates output by the second reply model; the second reply model is obtained by learning second training data, and the second training data includes: and the information sample to be replied and the reply sample correspond to the theme.
7. The method according to claim 4, wherein the determining the reply candidate corresponding to the message to be replied according to the topic and the message to be replied comprises:
and searching in the data source corresponding to the theme according to the information to be replied and the environmental characteristics of the user to obtain a reply candidate corresponding to the information to be replied.
8. A data processing apparatus, comprising:
the information upper text determining module is used for determining the information to be replied and the upper text corresponding to the information to be replied;
a reply candidate determining module, configured to determine, according to the information to be replied and the above, a reply candidate corresponding to the information to be replied; and
and the reply candidate output module is used for outputting the reply candidate.
9. An apparatus for data processing, comprising a memory, and one or more programs, wherein the one or more programs are stored in the memory, and wherein execution of the one or more programs by one or more processors comprises instructions for:
determining information to be replied and the corresponding text of the information to be replied;
determining a reply candidate corresponding to the information to be replied according to the information to be replied and the above;
and outputting the reply candidate.
10. A machine-readable medium having stored thereon instructions which, when executed by one or more processors, cause an apparatus to perform a data processing method as claimed in one or more of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810835658.2A CN110765338A (en) | 2018-07-26 | 2018-07-26 | Data processing method and device and data processing device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810835658.2A CN110765338A (en) | 2018-07-26 | 2018-07-26 | Data processing method and device and data processing device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110765338A true CN110765338A (en) | 2020-02-07 |
Family
ID=69327638
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810835658.2A Pending CN110765338A (en) | 2018-07-26 | 2018-07-26 | Data processing method and device and data processing device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110765338A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106448670A (en) * | 2016-10-21 | 2017-02-22 | 竹间智能科技（上海）有限公司 | Dialogue automatic reply system based on deep learning and reinforcement learning |
| CN107193978A (en) * | 2017-05-26 | 2017-09-22 | 武汉泰迪智慧科技有限公司 | A kind of many wheel automatic chatting dialogue methods and system based on deep learning |
| WO2018036555A1 (en) * | 2016-08-25 | 2018-03-01 | 腾讯科技（深圳）有限公司 | Session processing method and apparatus |
| CN107885756A (en) * | 2016-09-30 | 2018-04-06 | 华为技术有限公司 | Dialogue method, device and equipment based on deep learning |
| CN108170749A (en) * | 2017-12-21 | 2018-06-15 | 北京百度网讯科技有限公司 | Dialogue method, device and computer-readable medium based on artificial intelligence |
-
2018
- 2018-07-26 CN CN201810835658.2A patent/CN110765338A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018036555A1 (en) * | 2016-08-25 | 2018-03-01 | 腾讯科技（深圳）有限公司 | Session processing method and apparatus |
| CN107885756A (en) * | 2016-09-30 | 2018-04-06 | 华为技术有限公司 | Dialogue method, device and equipment based on deep learning |
| CN106448670A (en) * | 2016-10-21 | 2017-02-22 | 竹间智能科技（上海）有限公司 | Dialogue automatic reply system based on deep learning and reinforcement learning |
| CN107193978A (en) * | 2017-05-26 | 2017-09-22 | 武汉泰迪智慧科技有限公司 | A kind of many wheel automatic chatting dialogue methods and system based on deep learning |
| CN108170749A (en) * | 2017-12-21 | 2018-06-15 | 北京百度网讯科技有限公司 | Dialogue method, device and computer-readable medium based on artificial intelligence |
Non-Patent Citations (1)
| Title |
|---|
| 宋皓宇;张伟男;刘挺;: "基于DQN的开放域多轮对话策略学习", 中文信息学报, no. 07, pages 99 - 108 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| 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 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20170154104A1 (en) | Real-time recommendation of reference documents | |
| CN108073606B (en) | News recommendation method and device for news recommendation | |
| CN107564526B (en) | Processing method, apparatus and machine-readable medium | |
| CN110391966B (en) | Message processing method and device and message processing device | |
| CN108304412B (en) | Cross-language search method and device for cross-language search | |
| CN108345625B (en) | Information mining method and device for information mining | |
| CN111160047A (en) | Data processing method and device and data processing device | |
| CN110232181B (en) | Comment analysis method and device | |
| CN109814730B (en) | Input method and device and input device | |
| CN113987128A (en) | Related article searching method and device, electronic equipment and storage medium | |
| CN109521888B (en) | Input method, device and medium | |
| CN108241614B (en) | Information processing method and device, and device for information processing | |
| CN112036195A (en) | Machine translation method, device and storage medium | |
| CN109144286B (en) | Input method and device | |
| CN110765338A (en) | Data processing method and device and data processing device | |
| CN111831132A (en) | Information recommendation method and device and electronic equipment | |
| CN110929122B (en) | Data processing method and device for data processing | |
| CN110362686B (en) | Word stock generation method and device, terminal equipment and server | |
| CN114610163A (en) | Recommendation method, apparatus and medium | |
| CN112181163A (en) | Input method, input device and input device | |
| CN108983992B (en) | Candidate item display method and device with punctuation marks | |
| CN113589949A (en) | Input method and device and electronic equipment | |
| CN112651221A (en) | Data processing method and device and data processing device | |
| CN111611030A (en) | Data processing method and device and data processing device | |
| CN113010768B (en) | Data processing method and device for data processing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |