CN110413875A - A kind of method and relevant apparatus of text information push - Google Patents
A kind of method and relevant apparatus of text information push Download PDFInfo
- Publication number
- CN110413875A CN110413875A CN201910564424.3A CN201910564424A CN110413875A CN 110413875 A CN110413875 A CN 110413875A CN 201910564424 A CN201910564424 A CN 201910564424A CN 110413875 A CN110413875 A CN 110413875A
- Authority
- CN
- China
- Prior art keywords
- topic
- keyword
- user
- text
- text information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
Abstract
The embodiment of the present application discloses the method and relevant apparatus of a kind of text information push, after text information and suitable topic are established corresponding relationship, the practical content expressed of article can be covered by the topic keyword of topic, it has filled up and has contacted keyword that is close but not occurring in article with article, user's matching is carried out by topic keyword, matched keyword can reach preferably push effect more close to the practical content expressed of article.The embodiment of the present application method includes: acquisition text information；According to default topic matching relationship, target topic corresponding to target text keyword described in the text information is determined with matching for topic keyword by the target text keyword；At least one topic keyword according in the target topic determines user to be pushed by the matching of at least one described topic keyword and user's keyword；The text information is pushed to the user to be pushed.
Description
Technical field
A kind of method and relevant apparatus pushed this application involves Internet technical field more particularly to text information.
Background technique
With the development of modern society, people require to browse a large amount of information daily, growing to meet people
Information requirement.User can search for oneself desired information on the net, can also passively receive the information of push, for example, with
When family point opens news website homepage, the news of website push will be shown in news website homepage, in another example, user refreshes information
Homepage will acquire new hot spot information, in another example, the classification corresponding article is then shown when user opens certain classification interface.
If user can push user's interested article when receiving PUSH message, the reading experience of user will mention significantly
It is high.Therefore, usually the one or more keywords for including in article are matched with the label of user at present, when keyword with
When user tag is identical, the corresponding article of the keyword is matched with user, then the corresponding article of the keyword is pushed to user.
However, this method for pushing compares the extractability for relying on article keyword, moreover, the key extracted from article
Word is the word occurred in article, and keyword that is close but not occurring in article can be contacted with article by missing, and is caused
The practical content expressed of article can not be completely covered in the corresponding keyword of article, and push effect is poor.
Summary of the invention
The embodiment of the present application provides the method and relevant apparatus of a kind of text information push, by determining text information
Corresponding target topic, so that the topic keyword using target topic determines user to be pushed, with more close to article reality
The topic keyword of expression content determines user to be pushed, and solves the technical problem of push effect difference at present.
In view of this, the embodiment of the present application first aspect provides a kind of method of text information push, comprising:
Obtain text information, wherein the text information includes target text keyword；
According to default topic matching relationship, described in the matching determination by the target text keyword and topic keyword
Target topic corresponding to target text keyword described in text information, wherein the default topic matching relationship includes words
The incidence relation between keyword and topic is inscribed, the target topic includes at least one topic keyword；
At least one topic keyword according in the target topic passes through at least one described topic keyword
Matching with user's keyword determines user to be pushed, wherein the user to be pushed corresponds to user's keyword, the user
Keyword is for the determining topic keyword with the user-association to be pushed；
The text information is pushed to the user to be pushed.
The embodiment of the present application second aspect provides a kind of text information driving means, comprising:
Module is obtained, for obtaining text information, wherein the text information includes target text keyword；
Processing module, for passing through the target text keyword and topic keyword according to topic matching relationship is preset
Matching determine target topic corresponding to target text keyword described in the text information, wherein the default topic
Matching relationship includes the incidence relation between topic keyword and topic, and the target topic includes at least one topic key
Word；
Processing module is also used at least one topic keyword according in the target topic, by it is described at least
The matching of one topic keyword and user's keyword determines user to be pushed, wherein the user to be pushed corresponds to user
Keyword, user's keyword is for the determining topic keyword with the user-association to be pushed；
Pushing module, for pushing the text information to the user to be pushed.
In a kind of possible design, in the first implementation of the second aspect of the embodiment of the present application, mould is handled
Block is also used to:
The term vector of the target text keyword is generated by term vector model；
Accumulation process and normalized are carried out to the term vector of the target text keyword, obtain the text envelope
The cumulative term vector of breath；
The term vector of the topic keyword is generated by term vector model；
Accumulation process and normalized are carried out to the term vector of the topic keyword, obtain the cumulative of the topic
Term vector；
The first cosine calculated between the cumulative term vector of the text information and the cumulative term vector of the topic is similar
Spend score；
The target topic is chosen from the M topics according to the first cosine similarity score, wherein the mesh
Topic described in the top n that topic is the descending sequence of the first cosine similarity score is marked, the M is more than or equal to 1
Integer, the N be more than or equal to 1, and be less than or equal to the M integer.
In a kind of possible design, in second of implementation of the second aspect of the embodiment of the present application, mould is handled
Block is also used to: if user's keyword is identical as at least one described topic keyword in the target topic, basis
User's keyword determines the user to be pushed.
In a kind of possible design, in the third implementation of the second aspect of the embodiment of the present application, text envelope
Ceasing driving means further includes text information processing module, if text information processing module is for user's keyword and the mesh
At least one the described topic keyword marked in topic is identical, then according to the topic keyword from the X text informations
Y text informations are chosen to be pushed, at least one described topic keyword of the Y text informations with it is described
User's keyword is identical, and the X is integer more than or equal to 1, and the Y is and to be less than or equal to the X more than or equal to 1
Integer.
In a kind of possible design, in the 4th kind of implementation of the second aspect of the embodiment of the present application, text envelope
Breath driving means further includes default topic matching relationship module, and default topic matching relationship module is for obtaining sample text letter
Breath, wherein the sample text information includes sample text keyword；
The sample text keyword is clustered by clustering algorithm；
The topic is chosen from the sample text keyword in the cluster；
The topic keyword is chosen from default dictionary according to the topic, wherein the topic keyword with it is described
Topic has incidence relation；
The default topic matching relationship is established according to the topic and the topic keyword.
In a kind of possible design, in the 5th kind of implementation of the second aspect of the embodiment of the present application, words are preset
Topic matching relationship module is also used to: generating candidate words in the term vector and the default dictionary of the topic by term vector model
Inscribe the term vector of keyword；
Calculate the second cosine similarity between the term vector of the topic and the term vector of the candidate topics keyword
Score；
The topic keyword is chosen from the default dictionary, wherein the topic keyword is described second similar
The preceding L candidate topics keywords of the descending sequence of score are spent, the L is the integer more than or equal to 1.
The embodiment of the present application third aspect provides a kind of server, comprising: memory, transceiver, processor and bus
System；
Wherein, the memory is for storing program；
The processor is used to execute the program in the memory, includes the following steps:
Obtain text information, wherein the text information includes target text keyword；
According to default topic matching relationship, described in the matching determination by the target text keyword and topic keyword
Target topic corresponding to target text keyword described in text information, wherein the default topic matching relationship includes words
The incidence relation between keyword and topic is inscribed, the target topic includes at least one topic keyword；
At least one topic keyword according in the target topic passes through at least one described topic keyword
Matching with user's keyword determines user to be pushed, wherein the user to be pushed corresponds to user's keyword, the user
Keyword is for the determining topic keyword with the user-association to be pushed；
The text information is pushed to the user to be pushed；
The bus system is for connecting the memory and the processor, so that the memory and the place
Reason device is communicated.
The embodiment of the present application fourth aspect provides a kind of computer readable storage medium, including instruction, when it is in computer
When upper operation, so that the method that computer executes above-mentioned first aspect.
The 5th aspect of the embodiment of the present application provides a kind of computer program product comprising instruction, when it is at computer or place
When being run on reason device, so that the method that computer or processor execute above-mentioned first aspect.
As can be seen from the above technical solutions, the embodiment of the present application has the advantage that
After text information and suitable topic are established corresponding relationship by the embodiment of the present application, pass through the topic keyword of topic
The practical content expressed of article can be covered, has filled up and has contacted keyword that is close but not occurring in article with article, passed through
Topic keyword carries out user's matching, and matched keyword close to the practical content expressed of article, can more reach better
Push effect.
Detailed description of the invention
Fig. 1 is a configuration diagram of the system of text information push in the embodiment of the present application；
Fig. 2 is the flow chart of one embodiment of the method for text information push in the embodiment of the present application；
Fig. 3 is that topic and text information carry out matched schematic diagram in the embodiment of the present application；
Fig. 4 is text information and the matched schematic diagram of user in the embodiment of the present application；
Fig. 5 is one embodiment schematic diagram of text information push-delivery apparatus in the embodiment of the present application；
Fig. 6 is an alternative embodiment schematic diagram of text information push-delivery apparatus in the embodiment of the present application；
Fig. 7 is an alternative embodiment schematic diagram of text information push-delivery apparatus in the embodiment of the present application；
Fig. 8 is a kind of server architecture schematic diagram provided by the embodiments of the present application.
Specific embodiment
The embodiment of the present application provides the method and relevant apparatus of a kind of text information push, by determining text information
Corresponding target topic, so that the topic keyword using target topic determines user to be pushed, with more close to article reality
The topic keyword of expression content determines user to be pushed, and solves the technical problem of push effect difference at present.
The description and claims of this application and term " first ", " second ", " third ", " in above-mentioned attached drawing
The (if present)s such as four " are to be used to distinguish similar objects, without being used to describe a particular order or precedence order.It should manage
The data that solution uses in this way are interchangeable under appropriate circumstances, so that embodiments herein described herein for example can be to remove
Sequence other than those of illustrating or describe herein is implemented.In addition, term " includes " and " corresponding to " and their times
What is deformed, it is intended that cover it is non-exclusive include, for example, contain the process, method of a series of steps or units, system,
Product or equipment those of are not necessarily limited to be clearly listed step or unit, but may include be not clearly listed or for
The intrinsic other step or units of these process, methods, product or equipment.
It should be understood that people enter big data era with the development of modern society, bulk information is all browsed daily,
In main information be text information.Text information is usually the set of text, for example, article, short essay, news, information, microblogging,
Short message and message, the language of use can be any languages such as Chinese, English, Japanese, the text stored and transmitted in a computer
Part format can be the arbitrary formats such as doc, txt.These text information storages in the server, when user prepare browsing text envelope
When breath, text information is pushed to the client of user by server, for example, user opens client homepage, refreshes client head
It when page, the page under a certain classification of opening and opening a certain application program (Application, APP), or is user
When client is standby, short message, notice or message that server is sent are received, for example, user has subscribed the message that starts broadcasting of A main broadcaster,
When A main broadcaster starts broadcasting, even if user client is not turned on live streaming client, but server still pushes background message and gives the use
Family client.As can be seen that text message is usually that the user of push is selected by Keywords matching, " A master as escribed above
Broadcast " keyword.However the different range that text message is surely completely covered of these keywords, for example, certain text message describes A
Financial swindling case in electric business website can then be pushed text message by keyword " A electric business website " and " financial swindling case "
To user's keyword it is the user of " A electric business website " or " financial swindling case ", but not being pushed to user's keyword is " B electric business
The user of website ".But actually user's keyword is that the user of " B electric business website " is likely to feel emerging to " A electric business website "
Interest, i.e., the text message that the financial swindling case in A electric business website is described to this is interested, therefore, if server is by the text
Message is pushed to the user that user's keyword is " B electric business website ", actually can be improved the reading experience of user, realizes more preferable
Push effect.
The embodiment of the present application provides the method and relevant apparatus of a kind of text information push, can say and describe A electricity
The text message of financial swindling case on quotient website is pushed to the user that user's keyword is " B electric business website ", improves user's
Reading experience realizes preferably push effect.The embodiment of the present application will be described in detail below.
It should be understood that carrying out push by server is common technological means, it is in practical applications, negative to mitigate server
Text message can also all be downloaded to client, then carry out pushing away for text message by analysis of key word by client by load
It send, the embodiment of the present application is not specifically limited the executing subject.
Fig. 1 shows a configuration diagram of the system that text information pushes in the embodiment of the present application, as shown in Figure 1,
Server is established by network and terminal device and is communicated to connect, and after selecting text information by server, text information is pushed to
Terminal device shows text information by the client on terminal device.Wherein, terminal device shown in Fig. 1 is only one
A signal, in practical applications, terminal device include but be not limited only to mobile phone, desktop computer, tablet computer, laptop with
And palm PC.
It is understood that text information is pushed to terminal device by server in the embodiment of the present application, and pass through terminal
Client in equipment shows text information, and the client on terminal device can be small in application program or application program
Program can also be subscription number, service number, enterprise number, microblogging, circle of friends, video website and other portal websites etc., herein
It is not especially limited.
The method pushed from server side to text information in the embodiment of the present application is described in detail below, is asked
Referring to Fig.2, the one embodiment for the method that text information pushes in the embodiment of the present application includes:
201, text information is obtained, wherein text information includes target text keyword；
In the embodiment of the present application, text information be can store in the database, also can store in a network, or storage
In specific server.It can be server and obtain text information, treat after push user selects and push away text information
It is sent to subscriber terminal equipment.
In the embodiment of the present application, the target text keyword in text information is the key extracted from text information
Word can be extracted by keyword extraction algorithm.The quantity of text information can be one or more.
202, according to topic matching relationship is preset, text is determined by the matching of target text keyword and topic keyword
Target topic corresponding to target text keyword in information, wherein default topic matching relationship includes topic keyword and words
Incidence relation between topic, target topic include at least one topic keyword；
It is understood that target text keyword can be closed with the topic in default topic matching relationship in text information
Keyword is matched, and the matching process between target text keyword and topic keyword (between two groups of words) can be searching
The mode or neural network of cosine similarity are calculated between the word term vector occurred jointly in two groups of words
The mode of Similarity matching, is not specifically limited herein.
In the embodiment of the present application, has incidence relation between topic and topic keyword, which can be preparatory
Setting.Topic keyword specifically can be topic label, and the incidence relation of topic and topic keyword is realized by label form,
After determining the corresponding relationship of text information and target topic, topic label and user's Keywords matching can be passed through.
It, can be true after topic Keywords matching in text information in target text keyword and default topic matching relationship
Determine target topic corresponding to target text keyword in text information, wherein the corresponding target topic numbers of text information are not
Fixed, as shown in table 1, the target text keyword of text information 1 is that (financial swindling case is in table 1 for A electric business website and financial swindling case
Do not write out), the topic keyword of electric business website topic is A electric business website, B electric business website and C electric business website.The application is implemented
In example, the corresponding electric business of text information 1 corresponding A electric business website (target text keyword) corresponding A electric business website (topic keyword)
A electric business website keyword in website (topic), i.e. A electric business website keyword in text information 1 and electric business website topic
Match, can determine that target topic corresponding to target text keyword is electric business website topic in text information 1.Table 1 is please referred to,
Table 1 is a signal of the topic certain situation based on default topic matching relationship.
Table 1
203, it according at least one topic keyword in target topic, is closed by least one topic keyword and user
The matching of keyword determines user to be pushed, wherein user to be pushed corresponds to user's keyword, user's keyword for determine with
The topic keyword of user-association to be pushed；
It is understood that after the corresponding target topic of text information has been determined, it can be crucial by the topic of target topic
Word and user's Keywords matching of user to be pushed determine user to be pushed.Wherein, between topic keyword and user's keyword
The matching process of (between two groups of words) can be to be counted between the word or term vector found and occurred jointly in two groups of words
The mode of cosine similarity or the mode of neural network Similarity matching are calculated, is not specifically limited herein.
In the embodiment of the present application, user's keyword and user have incidence relation, which can preset.With
Family keyword can be user's portrait, i.e. user draws a portrait label (portrait tag), and text information also has its corresponding text information to draw
Picture, i.e. article are drawn a portrait label (article portrait tag), and text information portrait includes under target topic corresponding with text information
Keyword is inscribed, therefore text information portrait draws a portrait to match and can be the topic keyword of feeling the pulse with the finger-tip mark topic and the use of user with user
Family keyword is matched, so that it is determined that the incidence relation of text information and user, that is, determine which text information should be pushed to
User to be pushed a bit.
The topic keyword of target topic can determine use to be pushed with after the user's Keywords matching for pushing user
Family, as shown in table 2, the corresponding target topic of text information 1 are electric business website and fraud case, and wherein electric business website topic has B
The topic keyword of electric business website matches with user's keyword B electric business website, then can determine that text information 1 can be pushed to
User 1 to be pushed.It can be seen that the target text keyword " A even if user 1 to be pushed without text information 1 from above-mentioned example
Electric business website " can also obtain the push of text information 1 by the embodiment of the present application, and text information page 1 is user 1 to be pushed
It is more interested, therefore can be improved the reading experience of user 1 to be pushed, realize preferably push effect.Table 2 is please referred to,
Table 2 is the signal that the user based on default topic matching relationship pushes situation.
Table 2
204, text information is pushed to user to be pushed.
It should be understood that server can push text information to the terminal device of user to be pushed by network, so that wait push
The terminal device of user shows text information, can also be stored in the memory module of terminal device, when user to be pushed needs
It is shown again when checking relevant textual information, or can be using text information as the generation of data code insertion terminal device
In code library.Text information can be used as backstage notice display, also may be displayed in application program.
In the embodiment of the present application, it by determining the target topic of text information, is then closed according to the topic of target topic
Keyword is matched, and can be covered the practical content expressed of article by the topic keyword of topic, filled up and contacted with article
Keyword that is close but not occurring in article, by topic keyword progress user's matching, matched keyword more close to
The practical content expressed of article can reach preferably push effect.
Optionally, on the basis of above-mentioned Fig. 2 corresponding each embodiment, text information provided by the embodiments of the present application is pushed away
In one alternative embodiment of the method sent, according to default topic matching relationship, pass through the target text keyword and topic
The matching of keyword determines that target topic corresponding to target text keyword includes: in text information
The term vector of target text keyword is generated by term vector model；
Accumulation process and normalized are carried out to the term vector of target text keyword, obtain the cumulative of text information
Term vector；
The term vector of topic keyword is generated by term vector model；
Accumulation process and normalized are carried out to the term vector of topic keyword, obtain the cumulative term vector of topic；
Calculate the first cosine similarity score between the cumulative term vector of text information and the cumulative term vector of topic；
Target topic is chosen from M topic according to the first cosine similarity score, wherein target topic is the first cosine
The top n topic of the descending sequence of similarity score, M are integer more than or equal to 1, and N is and to be less than more than or equal to 1
Or the integer equal to M.
In the embodiment of the present application, term vector model can be each model in Word2vec, such as bag of words,
Skip-gram model or continuous bag of words (Continuous Bag-Of-Words Model, CBOW Model).It can be with
By gensim tool training term vector model, the term vector of target text keyword is can be generated in the term vector model after training
Or generate the term vector of topic keyword.The side that term vector is one of vectorization processing is generated by term vector model
Formula can also generate term vector in such a way that other vectorizations are handled, be not specifically limited herein.
Multiple vectors can be accumulated as one vector, the application by carrying out accumulation process and normalized to term vector
In embodiment, text information includes one or more target text keywords, therefore has also corresponded to one or more
The term vector of these target text keywords is carried out accumulation process and normalized by the term vector of target text keyword
An available cumulative term vector, the i.e. corresponding cumulative term vector of text information.And topic includes one or more words
Keyword is inscribed, each topic keyword generates a term vector, carries out accumulation process and normalizing to the term vector of these topics
Change and handles an available cumulative term vector, the i.e. cumulative term vector of topic.
It is understood that calculating the first cosine between the cumulative term vector of text information and the cumulative term vector of topic
Similarity score can be calculated by formula.Their similarity is assessed by calculating the included angle cosine value of two vectors, this
Place repeats no more.
In the embodiment of the present application, a text information can be matched with multiple topics, therefore can be calculated more
These cosine similarities are arranged in descending order by a cosine similarity, take the wherein corresponding topic of top n cosine similarity
As target topic, that is, have selected with the highest top n topic of a text information similarity as target topic.This top n
Topic can also be by artificial further screening, herein without limitation.
In the embodiment of the present application, it is also possible to multiple text informations and is matched with a topic, as shown in figure 3, can be with
Multiple cosine similarities are calculated, these cosine similarities are arranged in descending order, several cosine phases before taking wherein
The text information recycled from Text Information Data library like the corresponding text information of degree as server.Server is from text information
After obtaining these text informations in database, processing push can be carried out for these text informations, without being directed to text
All text informations in information database are handled and are pushed, and the burden of server, also, these texts are significantly reduced
Information is all the text information arranged in pairs or groups with the topic, can be directly displayed in the corresponding interface of the topic.For example, electric business website
There are 30 text informations under topic, then can show these text informations in the interface of electric business website topic.
Fig. 3 shows topic and text information carries out matched schematic diagram, and the embodiment of the present application is corresponding to each topic
Topic keyword set does vectorization processing, generates the term vector of topic keyword, then to the word of each topic keyword to
Amount is cumulative and does normalized, obtains the cumulative term vector (topic vector) of topic.Meanwhile passing through keyword extraction techniques
The keyword for extracting text information does vectorization processing to the keyword of text information, generates text information keyword (target
Text key word) term vector, accumulation process then is done to these term vectors and does normalized, obtains text information
Vector finally calculates the cosine similarity (cos similarity) between topic vector and text information vector, before finally taking
The text information that the corresponding text information of several cosine similarities is recycled from Text Information Data library as server.Recycling
It afterwards, include topic keyword in the keyword of these text informations.
In the embodiment of the present application, it is also possible to multiple text informations and is matched with multiple topics, can equally be calculated
To multiple cosine similarities, these cosine similarities are arranged in descending order, each cosine similarity is corresponding with text
Information and topic combination, for example, first cosine similarity corresponds to text information 1 and topic 1, second cosine similarity is corresponding
Text information 1 and topic 2, several cosine similarities before taking wherein, several available preceding most like combinations.Pass through system again
Meter can also be selected with some highest topic of text information similarity as target topic.
Optionally, on the basis of above-mentioned Fig. 2 corresponding each embodiment, text information provided by the embodiments of the present application is pushed away
In one alternative embodiment of the method sent, according at least one topic keyword in target topic, pass through described at least one
The matching of a topic keyword and user's keyword determines that user to be pushed includes:
If user's keyword is identical as at least one topic keyword in target topic, determined according to user's keyword
User to be pushed.
In the embodiment of the present application, the matching process between topic keyword and user's keyword (between two groups of words) can
To be the word occurred jointly in two groups of words of searching, i.e. user's keyword and at least one topic keyword in target topic
It is identical, for example, the topic keyword of text information 1 has A electric business website, B electric business website, C electric business website, financial swindling in table 2
Six case, telephone fraud case and network fraud case words, and user's keyword of user to be pushed 1 has stock, B electric business website and base
Golden three words, topic keyword " B electric business website " is identical as user's keyword " B electric business website ", then can be according to user's key
Word " B electric business website " determines user 1 to be pushed, and can push text information 1 to user 1 to be pushed in next step.
It is understood that server can first retrieve user's keyword in the embodiment of the present application, i.e., first search
Whether there is user's keyword identical with first topic keyword in user's keyword, then has searched whether and second topic
The identical user's keyword of keyword, until having searched whether user's keyword identical with the last one topic keyword, example
As can first be searched in user's keyword in table 2 whether have with " A electric business website " identical user's keyword, then searched whether
Identical user's keyword with " B electric business website ", until having searched whether and " network fraud case " identical user's keyword, In
During this, " B electric business website " identical user's keyword has been found, then by the corresponding use to be pushed of this user's keyword
Family 1 is found out, then can determine user 1 to be pushed according to user's keyword " B electric business website ", in next step can be to wait push
User 1 pushes text information 1.If also having found out user 2 to be pushed also has user's keyword " B electric business website ", it is determined that
User 2 to be pushed is also the target of push text information 1.
Fig. 4 shows text information and the matched schematic diagram of user in the embodiment of the present application, it can be seen that the text in Fig. 4
The corresponding target topic of this information has: TV play, lift face and variety show, and the corresponding topic keyword of TV play has idol
Play, the user's keyword for finding Fig. 4 upper user also have idol acute, it is determined that Fig. 4 upper user is user to be pushed, immediately
Text information is pushed to Fig. 4 upper user or the corresponding topic keyword of lift face cuts double-edged eyelid, and above Fig. 4
User's keyword of user also cuts double-edged eyelid, it is determined that Fig. 4 upper user is user to be pushed, with pushing away text information
Give Fig. 4 upper user.And the corresponding topic keyword of variety show has XX satellite TV and talk show, finds user below Fig. 4
User's keyword Ye You XX satellite TV and talk show, it is determined that user is user to be pushed below Fig. 4, with i.e. by text information
It is pushed to user below Fig. 4.
Optionally, on the basis of above-mentioned Fig. 2 corresponding each embodiment, text information provided by the embodiments of the present application is pushed away
In one alternative embodiment of the method sent, text envelope is determined by the matching of the target text keyword and topic keyword
In breath after target topic corresponding to target text keyword, method further include:
If user's keyword is identical as at least one topic keyword in target topic, according to topic keyword from X
It chooses Y text information in a text information to be pushed, at least one topic keyword of Y text information and user are crucial
Word is identical, and X is the integer more than or equal to 1, and Y is the integer more than or equal to 1, and less than or equal to X.
In the embodiment of the present application, X text information can be handled simultaneously.At the same time to X text information into
When row processing, if user's keyword is identical as at least one topic keyword in target topic, according to the identical topic
Keyword can determine Y text information, can be and carry out in X text information to the topic keyword of X text information
Retrieval is searched, if finding in text information has the identical topic keyword, text information can be extracted, most
It obtains Y text information eventually, there is the identical topic keyword in this Y text information, i.e., at least the one of Y text information
A topic keyword is identical as user's keyword.
For example, identical topic keyword is B electric business website in table 2, then it can search the corresponding topic of text information and close
There is the text information of B electric business website in keyword, these text informations are extracted to obtain Y text information, this Y text envelope
There is B electric business website in breath.
It is understood that choosing to after Y text information, this Y text information can be pushed on demand has
The user to be pushed of same subscriber keyword.
It, can also be by the text envelope with other same topic keyword according to the method for Y text information of above-mentioned selection
Breath, which selects, to be come, and is obtained all text informations that can match user, is formed Text Information Data library.
Optionally, on the basis of above-mentioned Fig. 2 corresponding each embodiment, text information provided by the embodiments of the present application is pushed away
In one alternative embodiment of the method sent, according to default topic matching relationship, pass through the target text keyword and topic
In the determining text information of the matching of keyword before target topic corresponding to target text keyword, method further include:
Obtain sample text information, wherein sample text information includes sample text keyword；
Sample text keyword is clustered by clustering algorithm；
Topic is chosen from the sample text keyword in cluster；
Topic keyword is chosen from default dictionary according to topic, wherein topic keyword and topic have incidence relation；
Default topic matching relationship is established according to topic and topic keyword.
In the embodiment of the present application, the sample text keyword in sample text information can pass through keyword extraction techniques
It extracts, is not repeating herein.Clustering algorithm in the embodiment of the present application can be kmeans algorithm.Pass through clustering algorithm pair
Sample text keyword is clustered, at least one cluster is formed, and words are then chosen from the sample text keyword in cluster
Topic, for example, the sample text keyword in some cluster has electric business website, A electric business website, B electric business website and C electric business website,
Electric business website can then be chosen as topic, the process of selection, which can be, manually to be selected, and is also possible to cluster by calculating
The algorithm at center calculates cluster centre, selects the sample text keyword of cluster centre as topic, or passes through other algorithms
A sample text keyword is chosen from sample text keyword as topic, herein without limitation.After selection obtains topic,
Topic keyword relevant to the topic can be chosen from default dictionary, default topic matching relationship is then established, such as 3 institute of table
Show, default topic matching relationship includes topic and topic keyword, has incidence relation between topic and topic keyword.It please join
Read table 3, a signal of the corresponding relationship between topic and topic keyword of table 3.
Table 3
It is understood that the word in topic can repeat in topic keyword, can not also occur.Table 3
In topic and topic keyword be only used as example, other default topics can be established according to method provided by the embodiments of the present application
With relationship, details are not described herein again.
Optionally, on the basis of above-mentioned Fig. 2 corresponding each embodiment, text information provided by the embodiments of the present application is pushed away
In one alternative embodiment of the method sent, choosing topic keyword from default dictionary according to topic includes:
Pass through the term vector of candidate topics keyword in the raw newsy term vector of term vector model and default dictionary；
Calculate the second cosine similarity score between the term vector of topic and the term vector of candidate topics keyword；
Topic keyword is chosen from default dictionary, wherein topic keyword is the descending row of the second similarity score
The preceding L candidate topics keyword of sequence, L are the integer more than or equal to 1.
Topic keyword is chosen from default dictionary according to topic, can be and manually chosen, algorithm can also be passed through
It extracts, in the embodiment of the present application, is extracted by the algorithm of term vector cosine similarity.Pass through term vector mould first
The term vector of candidate topics keyword, term vector model can be Word2vec in the raw newsy term vector of type and default dictionary
In arbitrary model, preset dictionary in candidate topics keyword quantity be one or more.It, can after generating term vector
The second cosine similarity score is calculated by cosine similarity algorithm, the second cosine similarity score can describe topic
With the similarity degree between candidate topics keyword, the second cosine similarity score is higher, then similarity degree is higher.It can be by
For two cosine similarity scores by small sequence is arrived greatly, the second cosine similarity score of candidate topics keyword is high, then comes front,
Then L candidate topics keyword is as the corresponding topic keyword of the topic before taking.
It should be noted that can also carry out artificial screening in preceding L candidate topics keyword, then will manually sieve
Candidate topics keyword after choosing is as the corresponding topic keyword of the topic.It is of course also possible to without artificial screening, or
It is screened by other algorithms, is not specifically limited herein.
Text information driving means provided by the embodiments of the present application is described in detail below, referring to Fig. 5, Fig. 5
For one embodiment schematic diagram of text information push-delivery apparatus in the embodiment of the present application, text information driving means 500 includes:
Module 501 is obtained, for obtaining text information, wherein text information includes target text keyword；
Processing module 502, for passing through the target text keyword and topic key according to topic matching relationship is preset
The matching of word determines target topic corresponding to target text keyword in text information, wherein default topic matching relationship packet
The incidence relation between topic keyword and topic is included, target topic includes at least one topic keyword；
Processing module 502 is also used to according at least one topic keyword in target topic, by it is described at least one
The matching of topic keyword and user's keyword determines user to be pushed, wherein user to be pushed corresponds to user's keyword, uses
Family keyword is for the determining topic keyword with user-association to be pushed；
Pushing module 503, for pushing text information to user to be pushed.
Optionally, on the basis of above-mentioned Fig. 5 corresponding each embodiment, text information provided by the embodiments of the present application is pushed away
It send in an alternative embodiment of device 500, processing module 502 is also used to:
The term vector of target text keyword is generated by term vector model；
Accumulation process and normalized are carried out to the term vector of target text keyword, obtain the cumulative of text information
Term vector；
The term vector of topic keyword is generated by term vector model；
Accumulation process and normalized are carried out to the term vector of topic keyword, obtain the cumulative term vector of topic；
Calculate the first cosine similarity score between the cumulative term vector of text information and the cumulative term vector of topic；
Target topic is chosen from M topic according to the first cosine similarity score, wherein target topic is the first cosine
The top n topic of the descending sequence of similarity score, M are integer more than or equal to 1, and N is and to be less than more than or equal to 1
Or the integer equal to M.
Optionally, on the basis of above-mentioned Fig. 5 corresponding each embodiment, text information provided by the embodiments of the present application is pushed away
It send in an alternative embodiment of device 500, processing module 502 is also used to: if in user's keyword and target topic at least
One topic keyword is identical, then determines user to be pushed according to user's keyword.
Optionally, on the basis of above-mentioned Fig. 5 corresponding each embodiment, text information provided by the embodiments of the present application is pushed away
It send in an alternative embodiment of device 500, referring to Fig. 6, text information driving means 500 further includes text information processing mould
Block 504, if text information processing module 504 is used for user's keyword and at least one topic keyword phase in target topic
Together, then Y text information is chosen from X text information according to topic keyword to be pushed, at least the one of Y text information
A topic keyword is identical as user's keyword, and X is integer more than or equal to 1, and Y is and to be less than or equal to more than or equal to 1
The integer of X.
Optionally, on the basis of above-mentioned Fig. 5 corresponding each embodiment, text information provided by the embodiments of the present application is pushed away
It send in an alternative embodiment of device 500, referring to Fig. 7, text information driving means 500 further includes that default topic matching is closed
It is module 505, default topic matching relationship module is for obtaining sample text information, wherein sample text information includes sample
Text key word；
Sample text keyword is clustered by clustering algorithm；
Topic is chosen from the sample text keyword in cluster；
Topic keyword is chosen from default dictionary according to topic, wherein topic keyword and topic have incidence relation；
Default topic matching relationship is established according to topic and topic keyword.
Optionally, on the basis of above-mentioned Fig. 5 corresponding each embodiment, text information provided by the embodiments of the present application is pushed away
It send in an alternative embodiment of device 500, default topic matching relationship module 505 is also used to: being generated by term vector model
The term vector of candidate topics keyword in the term vector of topic and default dictionary；
Calculate the second cosine similarity score between the term vector of topic and the term vector of candidate topics keyword；
Topic keyword is chosen from default dictionary, wherein topic keyword is the descending row of the second similarity score
The preceding L candidate topics keyword of sequence, L are the integer more than or equal to 1.
Fig. 8 is a kind of server architecture schematic diagram provided in an embodiment of the present invention, which can be because of configuration or performance
It is different and generate bigger difference, it may include one or more central processing units (central processing
Units, CPU) 822 (for example, one or more processors) and memory 832, one or more storages apply journey
The storage medium 830 (such as one or more mass memory units) of sequence 842 or data 844.Wherein, 832 He of memory
Storage medium 830 can be of short duration storage or persistent storage.The program for being stored in storage medium 830 may include one or one
With upper module (diagram does not mark), each module may include to the series of instructions operation in server.Further, in
Central processor 822 can be set to communicate with storage medium 830, execute on server 800 a series of in storage medium 830
Instruction operation.
Server 800 can also include one or more power supplys 826, one or more wired or wireless networks
Interface 850, one or more input/output interfaces 858, and/or, one or more operating systems 841, such as
Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM etc..
The step as performed by server 800 can be based on the server architecture shown in Fig. 8 in above-described embodiment.
In the embodiment of the present application, central processing unit 822 included by the server 800 is also with the following functions:
Obtain text information, wherein text information includes target text keyword；
According to default topic matching relationship, text is determined by the matching of the target text keyword and topic keyword
Target topic corresponding to target text keyword in information, wherein default topic matching relationship includes topic keyword and words
Incidence relation between topic, target topic include at least one topic keyword；
According at least one topic keyword in target topic, closed by least one described topic keyword and user
The matching of keyword determines user to be pushed, wherein user to be pushed corresponds to user's keyword, user's keyword for determine with
The topic keyword of user-association to be pushed；
Text information is pushed to user to be pushed.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
In several embodiments provided herein, it should be understood that disclosed system, device and method can be with
It realizes by another way.For example, the apparatus embodiments described above are merely exemplary, for example, the unit
It divides, only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components
It can be combined or can be integrated into another system, or some features can be ignored or not executed.Another point, it is shown or
The mutual coupling, direct-coupling or communication connection discussed can be through some interfaces, the indirect coupling of device or unit
It closes or communicates to connect, can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, each functional unit in each embodiment of the application can integrate in one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, the technical solution of the application is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies, which is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server or the network equipment etc.) executes the complete of each embodiment the method for the application
Portion or part steps.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only
Memory), random access memory (RAM, Random Access Memory), magnetic or disk etc. are various can store journey
The medium of sequence code.
Claims (10)
1. a kind of method of text information push characterized by comprising
Obtain text information, wherein the text information includes target text keyword；
According to default topic matching relationship, the text is determined by the matching of the target text keyword and topic keyword
Target topic corresponding to target text keyword described in information, wherein the default topic matching relationship includes the words
The incidence relation between keyword and topic is inscribed, the target topic includes at least one topic keyword；
At least one topic keyword according in the target topic by least one described topic keyword and is used
The matching of family keyword determines user to be pushed, wherein the user to be pushed corresponds to user's keyword, the user
Keyword is for the determining topic keyword with the user-association to be pushed；
The text information is pushed to the user to be pushed.
2. passing through the mesh the method according to claim 1, wherein the basis presets topic matching relationship
Mark text key word determines mesh corresponding to target text keyword described in the text information with matching for topic keyword
Marking topic includes:
The term vector of the target text keyword is generated by term vector model；
Accumulation process and normalized are carried out to the term vector of the target text keyword, obtain the text information
Cumulative term vector；
The term vector of the topic keyword is generated by term vector model；
Accumulation process and normalized are carried out to the term vector of the topic keyword, obtain the cumulative word of the topic to
Amount；
Calculate the first cosine similarity point between the cumulative term vector of the text information and the cumulative term vector of the topic
Number；
The target topic is chosen from the M topics according to the first cosine similarity score, wherein the target words
Topic described in the top n of the entitled descending sequence of first cosine similarity score, the M are whole more than or equal to 1
Number, the N are the integer more than or equal to 1, and less than or equal to the M.
3. the method according to claim 1, wherein it is described according in the target topic at least one
Topic keyword determines that user to be pushed includes: by the matching of at least one described topic keyword and user's keyword
If user's keyword is identical as at least one described topic keyword in the target topic, according to the use
Family keyword determines the user to be pushed.
4. passing through the mesh the method according to claim 1, wherein the basis presets topic matching relationship
Mark text key word determines mesh corresponding to target text keyword described in the text information with matching for topic keyword
After marking topic, the method also includes:
If user's keyword is identical as at least one described topic keyword in the target topic, according to the words
Topic keyword is chosen the Y text informations from the X text informations and is pushed, a text information of the Y
At least one described topic keyword is identical as user's keyword, and the X is the integer more than or equal to 1, and the Y is big
In or be equal to 1, and be less than or equal to the X integer.
5. passing through the mesh the method according to claim 1, wherein the basis presets topic matching relationship
Mark text key word determines mesh corresponding to target text keyword described in the text information with matching for topic keyword
Before marking topic, the method also includes:
Obtain sample text information, wherein the sample text information includes sample text keyword；
The sample text keyword is clustered by clustering algorithm；
The topic is chosen from the sample text keyword in the cluster；
The topic keyword is chosen from default dictionary according to the topic, wherein the topic keyword and the topic
With incidence relation；
The default topic matching relationship is established according to the topic and the topic keyword.
6. according to the method described in claim 5, it is characterized in that, described choose topic from default dictionary according to the topic
Keyword includes:
The term vector of candidate topics keyword in the term vector and the default dictionary of the topic is generated by term vector model；
Calculate the second cosine similarity score between the term vector of the topic and the term vector of the candidate topics keyword；
The topic keyword is chosen from the default dictionary, wherein the topic keyword is second similarity point
The preceding L candidate topics keywords of the descending sequence of number, the L are the integer more than or equal to 1.
7. a kind of text information driving means characterized by comprising
Module is obtained, for obtaining text information, wherein the text information includes target text keyword；
Processing module, for passing through of the target text keyword and topic keyword according to topic matching relationship is preset
With target topic corresponding to target text keyword described in the determination text information, wherein the default topic matching
Relationship includes the incidence relation between topic keyword and topic, and the target topic includes at least one topic keyword；
Processing module is also used at least one topic keyword according in the target topic, by it is described at least one
The matching of topic keyword and user's keyword determines user to be pushed, wherein it is crucial that the user to be pushed corresponds to user
Word, user's keyword is for the determining topic keyword with the user-association to be pushed；
Pushing module, for pushing the text information to the user to be pushed.
8. a kind of server characterized by comprising memory, transceiver, processor and bus system；
Wherein, the memory is for storing program；
The processor is used to execute the program in the memory, includes the following steps:
Obtain text information, wherein the text information includes target text keyword；
According to default topic matching relationship, the text is determined by the matching of the target text keyword and topic keyword
Target topic corresponding to target text keyword described in information, wherein the default topic matching relationship includes that topic closes
Incidence relation between keyword and topic, the target topic include at least one topic keyword；
At least one topic keyword according in the target topic by least one described topic keyword and is used
The matching of family keyword determines user to be pushed, wherein the user to be pushed corresponds to user's keyword, and the user is crucial
Word is for the determining topic keyword with the user-association to be pushed；
The text information is pushed to the user to be pushed；
The bus system is for connecting the memory and the processor, so that the memory and the processor
It is communicated.
9. a kind of computer readable storage medium, which is characterized in that including instruction, when run on a computer, make to succeed in one's scheme
Method described in any one of calculation machine perform claim requirement 1 to 6.
10. a kind of computer program product comprising instruction, which is characterized in that when it runs on a computer or a processor,
So that method described in any one of computer or processor perform claim requirement 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910564424.3A CN110413875A (en) | 2019-06-26 | 2019-06-26 | A kind of method and relevant apparatus of text information push |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910564424.3A CN110413875A (en) | 2019-06-26 | 2019-06-26 | A kind of method and relevant apparatus of text information push |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110413875A true CN110413875A (en) | 2019-11-05 |
Family
ID=68358382
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910564424.3A Pending CN110413875A (en) | 2019-06-26 | 2019-06-26 | A kind of method and relevant apparatus of text information push |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110413875A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110888990A (en) * | 2019-11-22 | 2020-03-17 | 深圳前海微众银行股份有限公司 | Text recommendation method, device, equipment and medium |
| CN111143506A (en) * | 2019-12-27 | 2020-05-12 | 汉海信息技术(上海)有限公司 | Topic content ordering method, device, server and storage medium |
| CN111159566A (en) * | 2019-12-31 | 2020-05-15 | 中国银行股份有限公司 | Information pushing method and device for financial market products |
| CN111368063A (en) * | 2020-03-06 | 2020-07-03 | 腾讯科技（深圳）有限公司 | Information pushing method based on machine learning and related device |
| CN111914079A (en) * | 2020-08-07 | 2020-11-10 | 上海梅斯医药科技有限公司 | Topic recommendation method and system based on user tags |
| WO2021092803A1 (en) * | 2019-11-13 | 2021-05-20 | 深圳市欢太科技有限公司 | Push user determination method and apparatus, server, and storage medium |
| CN113379481A (en) * | 2021-05-25 | 2021-09-10 | 北京大米科技有限公司 | Data processing method and device |
| CN113505293A (en) * | 2021-06-15 | 2021-10-15 | 深圳追一科技有限公司 | Information pushing method and device, electronic equipment and storage medium |
| CN114357278A (en) * | 2020-09-28 | 2022-04-15 | 腾讯科技（深圳）有限公司 | Topic recommendation method, device and equipment |
| CN115150462A (en) * | 2022-05-25 | 2022-10-04 | 东风柳州汽车有限公司 | Driving topic pushing method, device, equipment and storage medium |
| CN116992111A (en) * | 2023-09-28 | 2023-11-03 | 中国科学技术信息研究所 | Data processing method, device, electronic equipment and computer storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2014110005A (en) * | 2012-12-04 | 2014-06-12 | Nec Software Tohoku Ltd | Information search device and information search method |
| CN107220386A (en) * | 2017-06-29 | 2017-09-29 | 北京百度网讯科技有限公司 | Information-pushing method and device |
| CN107767172A (en) * | 2017-10-12 | 2018-03-06 | 百度在线网络技术（北京）有限公司 | Information-pushing method, device, server and medium |
| CN107943895A (en) * | 2017-11-16 | 2018-04-20 | 百度在线网络技术（北京）有限公司 | Information-pushing method and device |
| CN107944033A (en) * | 2017-12-13 | 2018-04-20 | 北京百度网讯科技有限公司 | Associate topic and recommend method and apparatus |
| CN108153824A (en) * | 2017-12-06 | 2018-06-12 | 阿里巴巴集团控股有限公司 | The determining method and device of targeted user population |
| CN108920675A (en) * | 2018-07-09 | 2018-11-30 | 北京百悟科技有限公司 | A kind of method, apparatus of information processing, computer storage medium and terminal |
-
2019
- 2019-06-26 CN CN201910564424.3A patent/CN110413875A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2014110005A (en) * | 2012-12-04 | 2014-06-12 | Nec Software Tohoku Ltd | Information search device and information search method |
| CN107220386A (en) * | 2017-06-29 | 2017-09-29 | 北京百度网讯科技有限公司 | Information-pushing method and device |
| CN107767172A (en) * | 2017-10-12 | 2018-03-06 | 百度在线网络技术（北京）有限公司 | Information-pushing method, device, server and medium |
| CN107943895A (en) * | 2017-11-16 | 2018-04-20 | 百度在线网络技术（北京）有限公司 | Information-pushing method and device |
| CN108153824A (en) * | 2017-12-06 | 2018-06-12 | 阿里巴巴集团控股有限公司 | The determining method and device of targeted user population |
| CN107944033A (en) * | 2017-12-13 | 2018-04-20 | 北京百度网讯科技有限公司 | Associate topic and recommend method and apparatus |
| CN108920675A (en) * | 2018-07-09 | 2018-11-30 | 北京百悟科技有限公司 | A kind of method, apparatus of information processing, computer storage medium and terminal |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021092803A1 (en) * | 2019-11-13 | 2021-05-20 | 深圳市欢太科技有限公司 | Push user determination method and apparatus, server, and storage medium |
| CN110888990B (en) * | 2019-11-22 | 2024-04-12 | 深圳前海微众银行股份有限公司 | Text recommendation method, device, equipment and medium |
| CN110888990A (en) * | 2019-11-22 | 2020-03-17 | 深圳前海微众银行股份有限公司 | Text recommendation method, device, equipment and medium |
| CN111143506A (en) * | 2019-12-27 | 2020-05-12 | 汉海信息技术(上海)有限公司 | Topic content ordering method, device, server and storage medium |
| CN111143506B (en) * | 2019-12-27 | 2023-11-03 | 汉海信息技术(上海)有限公司 | Topic content ordering method, topic content ordering device, server and storage medium |
| CN111159566A (en) * | 2019-12-31 | 2020-05-15 | 中国银行股份有限公司 | Information pushing method and device for financial market products |
| CN111368063B (en) * | 2020-03-06 | 2023-03-17 | 腾讯科技（深圳）有限公司 | Information pushing method based on machine learning and related device |
| CN111368063A (en) * | 2020-03-06 | 2020-07-03 | 腾讯科技（深圳）有限公司 | Information pushing method based on machine learning and related device |
| CN111914079A (en) * | 2020-08-07 | 2020-11-10 | 上海梅斯医药科技有限公司 | Topic recommendation method and system based on user tags |
| CN114357278A (en) * | 2020-09-28 | 2022-04-15 | 腾讯科技（深圳）有限公司 | Topic recommendation method, device and equipment |
| CN114357278B (en) * | 2020-09-28 | 2024-03-19 | 腾讯科技（深圳）有限公司 | Topic recommendation method, device and equipment |
| CN113379481A (en) * | 2021-05-25 | 2021-09-10 | 北京大米科技有限公司 | Data processing method and device |
| CN113505293A (en) * | 2021-06-15 | 2021-10-15 | 深圳追一科技有限公司 | Information pushing method and device, electronic equipment and storage medium |
| CN113505293B (en) * | 2021-06-15 | 2024-03-19 | 深圳追一科技有限公司 | Information pushing method and device, electronic equipment and storage medium |
| CN115150462A (en) * | 2022-05-25 | 2022-10-04 | 东风柳州汽车有限公司 | Driving topic pushing method, device, equipment and storage medium |
| CN115150462B (en) * | 2022-05-25 | 2023-10-31 | 东风柳州汽车有限公司 | Driving topic pushing method, device, equipment and storage medium |
| CN116992111A (en) * | 2023-09-28 | 2023-11-03 | 中国科学技术信息研究所 | Data processing method, device, electronic equipment and computer storage medium |
| CN116992111B (en) * | 2023-09-28 | 2023-12-26 | 中国科学技术信息研究所 | Data processing method, device, electronic equipment and computer storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110413875A (en) | A kind of method and relevant apparatus of text information push | |
| US11194965B2 (en) | Keyword extraction method and apparatus, storage medium, and electronic apparatus | |
| CN109885773B (en) | Personalized article recommendation method, system, medium and equipment | |
| CN111324713B (en) | Automatic replying method and device for conversation, storage medium and computer equipment | |
| CN107943792B (en) | Statement analysis method and device, terminal device and storage medium | |
| WO2014127673A1 (en) | Method and apparatus for acquiring hot topics | |
| CN112749280A (en) | Internet public opinion classification method, device, electronic device and storage medium | |
| CN111061837A (en) | Topic identification method, device, equipment and medium | |
| CN111125457A (en) | Deep cross-modal Hash retrieval method and device | |
| CN110717038B (en) | Object classification method and device | |
| CN116109732A (en) | Image labeling method, device, processing equipment and storage medium | |
| CN109634436B (en) | Method, device, equipment and readable storage medium for associating input method | |
| CN114490923A (en) | Training method, device and equipment for similar text matching model and storage medium | |
| CN112784141A (en) | Search result quality determination method and device, storage medium and computer equipment | |
| CN114428842A (en) | Method and device for expanding question-answer library, electronic equipment and readable storage medium | |
| CN111429200B (en) | Content association method and device, storage medium and computer equipment | |
| CN112417210A (en) | Body-building video query method, device, terminal and storage medium | |
| CN107992501B (en) | Social network information identification method, processing method and device | |
| CN111737288B (en) | Search control method, device, terminal equipment, server and storage medium | |
| CN111753194B (en) | Information pushing method and device, electronic equipment and storage medium | |
| CN110110202A (en) | A kind of information flow method for pushing and device | |
| CN110351183B (en) | Resource collection method and device in instant messaging | |
| CN113704462A (en) | Text processing method and device, computer equipment and storage medium | |
| CN111125387A (en) | Multimedia list generation and naming method and device, electronic equipment and storage medium | |
| CN113486184B (en) | Keyword determination method, device, equipment and storage medium |
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 |