CN106528588A - Method and apparatus for matching resources for text information - Google Patents
Method and apparatus for matching resources for text information Download PDFInfo
- Publication number
- CN106528588A CN106528588A CN201610824080.1A CN201610824080A CN106528588A CN 106528588 A CN106528588 A CN 106528588A CN 201610824080 A CN201610824080 A CN 201610824080A CN 106528588 A CN106528588 A CN 106528588A
- Authority
- CN
- China
- Prior art keywords
- resource
- text
- items
- tag
- keyword
- 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
- G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
- G06F16/48—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Library & Information Science (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention discloses a method for matching resources for text information. The method comprises the steps of acquiring a text keyword according to text information; and determining resources matched with the text information according to similarity of the text keyword and a resource tag. Further, the method comprises the steps of calculating a matching degree of the text keyword and a resource item according to the similarity of the text keyword and the resource tag and a weight of the corresponding resource tag in the resource item to which the resource tag belongs; and determining the resource matched with the text information according to the matching degrees of the text keyword and all resource items. Further, the method comprises the step of adjusting the matching degree of the text keyword and the resource item according to the use information of a user. The present invention further discloses an apparatus for matching resources for text information. The apparatus comprises an acquisition module and a matching module. The matching module further comprises a calculation module and a resource determining module. Through adoption of the method and apparatus, the resource matched with the text information can be determined more accurately and comprehensively according to the feature of the text information, and demands of the user on diversified representation forms are met.
Description
Technical field
The application is related to field of computer technology, more particularly to a kind of method and device for matching resource for text message.
Background technology
User can record various information by the various network platforms, for example, record the information of daily life, equally,
User also can produce information when which is communicated with other users by communications platform.In this method, by obtaining above-mentioned use
Family information, and conversion is identified to information, one section of static word is changed into into the animation of vivid and interesting automatically.
With the development of Internet technology, various multimedia resources become increasingly abundant, and the form of expression of various information is also more
Flexibly, it is lively, interesting.In the practical application of user, it is sometimes desirable to which text message is shown using more diversified form
Come, this is accomplished by finding the multimedia resource matched with these text messages.For example, send the blessing of " happy birthday to you "
When, can be using the picture of the birthday cake for plugging candle, animation and/or musical, accordingly, it would be desirable to foundation " wishes your birthday
It is happy " text message search out matched picture, animation and/or music.
In order to find the resource for matching for text message, method of the existing way typically by Keywords matching,
For the certain keyword of each resource settings, after a series of words of user input, the word of user input is carried out synonymous
Word extends, and obtains inquiry word list (Query word lists), is inquired about further according to inquiry word list, obtains corresponding inquiry knot
Really.The shortcoming of prior art is:
(1) extend synonym the script of word may be caused to look like and deviation occur, and bad process negative, and negative
This relation of negative.
(2) when being inquired about according to inquiry word list, the word in inquiry word list is carried out with the keyword of resource
Matching, only on all four word are just regarded as matching, consequently, it is possible to because synonym extends inadequate or implication deviation
And the resource for matching cannot be found.
The content of the invention
The embodiment of the present application provides a kind of method for matching resource for text message, it is therefore intended that according to the spy of text message
Levy, more accurately, more fully determine the resource with text information match, meet diversified need of the user to the form of expression
Ask.
The embodiment of the present application also provides a kind of device for matching resource for text message, it is therefore intended that according to text message
Feature, more accurately, more fully determines the resource with text information match, meets diversified need of the user to the form of expression
Ask.
The embodiment of the present application adopts following technical proposals:
A kind of method for matching resource for text message that the embodiment of the present application is provided, including:
According to the text message, text keyword is obtained;
According to the text keyword and the similarity of resource tag, it is determined that the resource matched with the text message;
Wherein, the resource tag is corresponding with the resource.
Alternatively, the embodiment of the present application provide for text message match resource method in, the text keyword with
The determination of the similarity of resource tag includes:
Obtain the second word of first term vector and the resource tag of the text keyword from term vector storehouse respectively
Vector;Term vector comprising at least one participle in the term vector storehouse;
The cosine similarity of first term vector and second term vector is calculated, as the text keyword and institute
State the similarity of resource tag.
Alternatively, what the embodiment of the present application was provided matches in the method for resource for text message, word in the term vector storehouse
Vectorial is included:
Obtain the corpus for training term vector;
The corpus are carried out with participle, participle set is obtained;
Cooccurrence relation according to the frequency of occurrence of each participle in the participle set and/or with other participles, determines each point
The term vector of word;The frequency of occurrence of each element representation correspondence participle and/or the co-occurrence with other participles in the term vector
The frequency.
Alternatively, what the embodiment of the present application was provided matches in the method for resource for text message, crucial according to the text
The similarity of word and resource tag, it is determined that the resource matched with the text message, including:
Similarity of the text keyword with the resource tag is ranked up;
By similarity highest default resource tag output, to determine the resource matched with the text message.
Alternatively, what the embodiment of the present application was provided matches in the method for resource for text message, crucial according to the text
The similarity of word and resource tag, it is determined that the resource matched with the text message, including:
According to the similarity and corresponding resource tag of the text keyword and the resource tag in affiliated money
Weight in source item, calculates the matching degree of the text keyword and the resource items;Wherein, comprising extremely in the resource items
A few resource tag;
According to the matching degree of the text keyword and each resource items, it is determined that matching with the text message
Resource;Wherein, resource described at least one is included in the resource items.
Alternatively, what the embodiment of the present application was provided matches in the method for resource for text message, calculates the text crucial
The matching degree of word and the resource items, including:
The sum of products of the similarity and the weight is calculated, as the text keyword and corresponding resource mark
The matching degree of resource items belonging to signing.
Alternatively, what the embodiment of the present application was provided matches in the method for resource for text message, crucial according to the text
The matching degree of word and each resource items, it is determined that the resource matched with the text message, including:
Matching degree of the text keyword with the resource items is ranked up;
By the default resource items output for most matching, to determine the resource matched with the text message.
Alternatively, what the embodiment of the present application was provided matches in the method for resource for text message, by default most matched
Resource items are exported, to determine the resource matched with the text message, specially:
By the default resource items output for most matching, alternately resource items are to user's recommendation;
The resource corresponding to resource items that user is chosen in the alternative resource item is defined as and the text message
The resource for matching.
Alternatively, what the embodiment of the present application was provided matches in the method for resource for text message, and the resource tag is in institute
The determination of the weight in category resource items includes:
Weight setting initial value to the resource tag in affiliated resource items;
The resource items chosen in the alternative resource item according to user, adjust resource tag in the resource items and/
Or weight of the adjustresources label in affiliated resource items.
Alternatively, what the embodiment of the present application was provided matches in the method for resource for text message, and methods described also includes:
Obtain the use information of user, the use information includes the history preference information of user, personal information and/or same
The label information of class user;
Then after matching degree of the text keyword with the resource items is calculated, according to the text keyword
With the matching degree of each resource items, it is determined that before the resource matched with the text message, also including:
According to the use information of user, matching degree of the text keyword with the resource items is adjusted
Whole, after being updated matching degree.
Alternatively, what the embodiment of the present application was provided matches in the method for resource, according to the text message for text message,
Text keyword is obtained, including:
Obtain the text message;
Participle is carried out to the text message, the text keyword is obtained.
Alternatively, what the embodiment of the present application was provided matches in the method for resource for text message, the classification bag of the resource
Include picture, photo, music, voice, audio, action, expression, animation and/or video.
A kind of device for matching resource for text message that the embodiment of the present application is provided, including:
Acquisition module, for according to the text message, obtaining text keyword;
Matching module, for the similarity according to the text keyword and resource tag, it is determined that with the text message
The resource for matching;Wherein, the resource tag is corresponding with the resource.
Alternatively, what the embodiment of the present application was provided matches in the device of resource for text message, and the matching module includes:
Computing module, for similarity and corresponding money according to the text keyword and the resource tag
Weight of the source label in affiliated resource items, calculates the matching degree of the text keyword and the resource items;Wherein, it is described
Resource tag described at least one is included in resource items;
Resource determination module, for the matching degree according to the text keyword and each resource items, it is determined that and institute
State the resource that text message matches;Wherein, resource described at least one is included in the resource items.
Above-mentioned at least one technical scheme that the embodiment of the present application is adopted can reach following beneficial effect:
(1) text keyword obtained according to text message can show the feature of text message, resource tag energy table
Reveal the feature of resource, text keyword and resource tag are all extract by text message or resource in itself, do not had
Synon step is extended in carrying out similar prior art, therefore, be not in meaning deviation, the matching result for obtaining is more accurate
Really.
(2) concept of similarity is introduced, text message and money is embodied by similarity of the text keyword with resource tag
The matching degree in source, it is to avoid the money for matching cannot be found as on all four keyword cannot be found in prior art
The problem in source.
(3) similarity between word is calculated by corpus, and from the frequency of occurrence of word itself and/or
Similarity from the point of view of cooccurrence relation in same sentence with other words between word, can more comprehensively, more accurately
Similitude semantically between word is embodied, and is not limited to literal similitude.
Description of the drawings
Accompanying drawing described herein is used for providing further understanding of the present application, constitutes the part of the application, this Shen
Schematic description and description please does not constitute the improper restriction to the application for explaining the application.In the accompanying drawings:
The schematic flow sheet of methods of the Fig. 1 for matching resource for text message in the embodiment of the present application;
Fig. 2 is the schematic flow sheet of second method for text message matching resource in the embodiment of the present application;
Fig. 3 is the third schematic flow sheet for the method for text message matching resource in the embodiment of the present application;
The structural representation of devices of the Fig. 4 for matching resource for text message in the embodiment of the present application;
Fig. 5 is the structural representation of second device for text message matching resource in the embodiment of the present application.
Specific embodiment
For making purpose, technical scheme and the advantage of the application clearer, below in conjunction with the application specific embodiment and
Corresponding accompanying drawing is clearly and completely described to technical scheme.Obviously, described embodiment is only the application one
Section Example, rather than the embodiment of whole.Based on the embodiment in the application, those of ordinary skill in the art are not doing
The every other embodiment obtained under the premise of going out creative work, belongs to the scope of the application protection.
Below in conjunction with accompanying drawing, the technical scheme that each embodiment of the application is provided is described in detail.
In present embodiments, the classification of resource includes picture, photo, music, voice, audio, action, table
Feelings, animation and/or video.
Embodiment 1
Fig. 1 gives a kind of schematic flow sheet of the method for matching resource for text message, including:
S101:According to text message, text keyword is obtained;
S102:According to text keyword and the similarity of resource tag, it is determined that the resource matched with text message;Its
In, resource tag is corresponding with resource.
In above-described embodiment, the text keyword obtained using text message shows the feature of text message, using money
Source label shows the feature of resource, and text keyword and resource tag are all to be extract by text message or resource in itself
, synon step is extended in not carrying out similar prior art, therefore, be not in meaning deviation, the matching knot for obtaining
Fruit is more accurate.In addition, the concept of similarity in above-described embodiment, is introduced, by text keyword and the similarity of resource tag
Embody the matching degree of text message and resource, it is to avoid as on all four keyword cannot be found and nothing in prior art
Method finds the problem of the resource for matching.
Above-mentioned steps S101 obtain text keyword according to text message, can be specially:Text message is obtained first;It is right again
Text message carries out participle, obtains text keyword.
The method for obtaining text message has a lot, can be directly inputted by user, it is also possible to turned by the text that user imports
Change is obtained.When carrying out participle to text message, using conventional segmentation methods, it is not construed as limiting herein.
After getting the text message of user input, carry out participle and screen the word of specific part of speech, form the word of J word
List, is designated as w={ w1,w2,w3,...wJ}.The word of Modern Chinese can be divided into 12 classes according to part of speech.Notional word include noun,
Verb, adjective, number, measure word and pronoun.Function word includes adverbial word, preposition, conjunction, auxiliary word, interjection, onomatopoeia.Forming word
During list w, function word is screened out, notional word is carried out into cutting, w is obtained1、w2、w3Deng participle.
In above-mentioned embodiment illustrated in fig. 1, during execution step S102, the text keyword of institute's foundation is similar to resource tag
Degree, its determination process may include:
Obtain the second term vector of first term vector and resource tag of text keyword from term vector storehouse respectively;Word to
Term vector comprising at least one participle in amount storehouse;
The cosine similarity of the first term vector and the second term vector is calculated, it is similar to resource tag as text keyword
Degree.
Foundation carries out the word list w={ w that participle is obtained to text message1,w2,w3,...wJ, to the word list in it is every
One participle wj(j=1,2,3 ... J), its term vector (i.e. the first term vector) is obtained, vec (w are designated asj), j=1,2,3,
... J, J represent the quantity of word in the quantity of the corresponding text keyword of text message, i.e. word list.The word of resource tag to
Amount (i.e. the second term vector) is designated as vec (tk), k=1,2,3 ... K, tkK-th resource tag is represented, K represents resource tag
Total quantity.Then text keyword wjTerm vector and resource tag tkTerm vector cosine similarity using below equation calculate:
simjk=cos (vec (wj),vec(tk)) (wherein, j=1,2,3 ... J, k=1,2,3 ... K)
Obtaining term vector has various methods, for example can be using a for train open in 2013 of Google companies
The word2vec instruments of term vector, it would however also be possible to employ GloVe algorithms carry out the training of term vector, can also be obtained in the following ways
The term vector in term vector storehouse is obtained, specifically be may include:
Obtain the corpus for training term vector;
The corpus are carried out with participle, participle set is obtained;
Cooccurrence relation according to the frequency of occurrence of each participle in the participle set and/or with other participles, determines each point
The term vector of word;The frequency of occurrence of each element representation correspondence participle and/or the co-occurrence with other participles in the term vector
The frequency.
Corpus for training term vector can have various sources, capture from webpage for example with spiders
Web page text etc..Corpus are carried out before participle, capital and small letter unification can also be carried out and be converted to small letter, idle character and be converted to
Space, Chinese-traditional between space, deletion Chinese character is converted to the pretreatment of the aspects such as simplified form of Chinese Character.In the base of pretreatment
On plinth, participle set, the then appearance to each participle in participle set is obtained after the processes such as participle, duplicate removal, part of speech screening
Frequency and/or number of times and the frequency and/or number of times that are occurred in a word jointly with other participles are counted, will system
The result that meter is obtained constitutes the term vector of the participle as the component in the term vector of the participle.Each component is in term vector
Order need not be particularly limited to, as long as the term vector of each participle is corresponding on each component.It is illustrated below acquisition word to
The detailed process of amount.
Hypothesis has three corpus:
Language material one:The Dragon Boat Festival is happy!
Language material two:The Zongzi of the Dragon Boat Festival is fond of eating.
Language material three:The custom of the Dragon Boat Festival eats Zongzi and dragon-boat racing.
For above language material carries out participle, participle set is obtained:
{ the Dragon Boat Festival;It is happy;Zongzi;It is nice;Custom;Eat;Match;Dragon boat }
Assume the component composition rule of term vector of each participle for (frequency that participle occurs, participle are total to the first participle
The co-occurrence number of times of the co-occurrence number of times of occurrence number, participle and the second participle, participle and the n-th participle), to obtain " Dragon Boat Festival "
Term vector as a example by, then the term vector of participle " Dragon Boat Festival " is represented by:(1,3,1,2,1,1,1,1,1)
First component " 1 " expression " Dragon Boat Festival " probability of three language material appearance in corpus is 100%;
Second component " 3 " expression " Dragon Boat Festival " occurs in that 3 times in corpus, and (participle " Dragon Boat Festival " is first point
Word, now can also represent the total degree that participle " Dragon Boat Festival " occurs in corpus);
Three-component " 1 " expression " Dragon Boat Festival " occurs in that 1 time (in language material one) jointly with participle " happy ";
4th component " 2 " expression " Dragon Boat Festival " occurs in that 2 times (in language material two and language material three) jointly with participle " Zongzi ";
5th component " 1 " expression " Dragon Boat Festival " occurs in that 1 time (in language material two) jointly with participle " being fond of eating ";
6th component " 1 " expression " Dragon Boat Festival " occurs in that 1 time (in language material three) jointly with participle " custom ";
7th component " 1 " expression " Dragon Boat Festival " occurs in that 1 time (in language material three) jointly with participle " eating ";
8th component " 1 " expression " Dragon Boat Festival " occurs in that 1 time (in language material three) jointly with participle " match ";
9th component " 1 " expression " Dragon Boat Festival " occurs in that 1 time (in language material three) jointly with participle " dragon boat ".
Using the method for above example, the term vector of each word can be obtained, and then text keyword can be calculated
With the similarity of resource tag, such that it is able to the similarity according to text keyword and resource tag, it is determined that with text envelope manner of breathing
The resource of matching, specifically may include:
Similarity of the text keyword with resource tag is ranked up;
By similarity highest default resource tag output, to determine the resource matched with text message.
The high text keyword for representing text message of similarity is closely similar with resource tag, and relevance is strong, therefore, it can
The resource that the resource corresponding with the resource tag is defined as matching with text message, such that it is able to the resource for matching
Diversely to show the content of text message, diversified demand of the user to the form of expression is met.
Embodiment 2
On the basis of embodiment 1, execution step S101 is according to text message, after acquisition text keyword, step
S102 is according to text keyword and the similarity of resource tag, it is determined that the resource matched with text message, can specifically include,
It is shown in Figure 2:
S1021:According to text keyword with the similarity of resource tag and corresponding resource tag in affiliated resource
Weight in, calculates the matching degree of text keyword and resource items;Wherein, at least one resource mark is included in resource items
Sign;
S1022:According to the matching degree of text keyword and each resource items, it is determined that the resource matched with text message;
Wherein, at least one resource is included in resource items.
In example 2, text keyword is same as Example 1 with the acquisition modes of the similarity of resource tag, here
Repeat no more.
For same resource, there may be many features, characterize such that it is able to multiple resource tags, and
Each resource tag is not quite similar with the degree of association of respective resources.For example, the dynamic of birthday candle is lighted on one section of birthday cake
Draw, can be characterized with resource tag " birthday ", " cake ", " happy ", " blessing " etc., it is in these resource tags, " raw
Day " maximum with the relevance of this section of animation, and the relevance of " blessing " is minimum.Therefore, in the embodiment depicted in figure 2, will be same
The collection of the corresponding resource tag of resource is collectively referred to as resource items, so as to be introduced into weight of the resource tag in affiliated resource items this without exception
Read, with reference to text keyword and the similarity of resource tag, calculate jointly the matching degree of text keyword and resource items, enter
And the resource matched with text message is determined on this basis, believe with the text so as to be more beneficial for accurately and comprehensively determining
The resource of manner of breathing matching.
As described in example 1 above, text keyword wjTerm vector and resource tag tkThe cosine similarity of term vector adopt
Calculated with below equation:
simjk=cos (vec (wj),vec(tk)) (wherein, j=1,2,3 ... J, k=1,2,3 ... K)
Assume to have N number of resource items, resource items Tn(n=1,2,3 ... N) in have K resource tag, be expressed as Tn=
{t1,t2,t3...tK, resource tag tkIn affiliated resource items TnIn weight be expressed as qnk(n=1,2,3 ... N, k=1,2,
3 ... K), in resource items, the weight sum of each resource tag can be taken as 1, i.e.,The power of each resource tag in resource items
Weight can system it is set in advance, or based on neural metwork training study obtain;Can be it is changeless,
Can constantly be adjusted according to the self study process of the use of user and system.
Further, the sum of products of similarity and weight is calculated, i.e., according to corresponding resource tag in affiliated resource
Weight in, calculates the weighted sum of the similarity of text keyword and resource tag, as text keyword with it is corresponding
The matching degree of the affiliated resource items of resource tag.By text keyword wjWith resource items TnMatching degree be designated as Sjn, then according to
Below equation is calculated Sjn:
(wherein, j=1,2,3 ... J, n=1,2,
3,...N)
On this basis, matching degree that can be further to text keyword with resource items is ranked up, and most will match
Default resource items output, to determine the resource matched with text message.Now, big according to numerical value to matching degree
After little sequence, can directly by matching degree highest (i.e. SjnValue it is maximum) resource items as resource corresponding with text message
, and then the resource that the resource items corresponding resource is defined as matching with text message;Can also be default by what is most matched
Individual resource items are exported, and alternately resource items are recommended to choose for user to user;The money that user is chosen in alternative resource item
Resource corresponding to source item is defined as the resource matched with text message.
The determination of weight of the above-mentioned resource tag in affiliated resource items can include procedure below:
Weight setting initial value first to resource tag in affiliated resource items;This initial value, can be rule of thumb
Or the result setting of machine learning, it is also possible to (having K money in resource items is assumed to each resource tag setting identical weight
Source label, then the initial weight of each resource tag may be set to)。
Then, after user have chosen certain resource items y in alternative resource item, the resource items that can be chosen according to user
The text message of y and user input, the resource tag and/or adjustresources label in adjustresources item y is in affiliated resource items y
In weight.Specifically, the text keyword most matched with resource items y in the text message that can first find user input
Matching degree highest text keyword with resource items y is foundMeet:If text is crucial
WordIn the resource tag included by resource items y, then update weight of the resource tag in resource items y (renewal
Mode can be the weight for increasing the resource tag), otherwise by text keywordIncrease to the money as new resource tag
In source item y.For newly-increased resource tagInitial weight can be set as by the initial weight in new resources item y
Will text keywordWith the matching degree of former resource items yAs newly-increased resource tagIt is first in new resources item y
Beginning weight;Text keyword can also be calculatedWith each resource tag similarity in former resource items y and, as newly-increased money
Source labelInitial weight in new resources item y.Further the resource items that the resource tag that have updated weight is located are entered
Row weight is normalized.
In the text message of the resource items y and user input chosen according to user, the resource tag in adjustresources item y
And/or during weight of the adjustresources label in affiliated resource items y, it is also possible to include in directly judging the resource items that user chooses
Have with whether text keyword is completely the same.If including and the on all four money of text keyword in the resource items that user chooses
Source label (can get highest equivalent to similarity of the resource tag with text keyword), then can increase the resource tag
Weight;If user choose resource items in not comprising with the on all four resource tag of text keyword, can be by text
Keyword is added in the resource items as resource tag, can also set weight accordingly.Resource mark in adjustresources item
Sign and/or adjustresources label is after the weight in affiliated resource items, can also further to each resource tag in resource items
Weight be normalized, i.e., be adjusted according to original ratio so that the weight sum of each resource tag remains 1.
Embodiment 3
It is on the basis of embodiment 2, shown in Figure 3, can with execution step S103 obtain user use information,
On the basis of use information includes the label information of history preference information, personal information and/or fellow users of user, performing
After step S1021 calculates the matching degree of text keyword and resource items, execution step S1022 according to text keyword with
The matching degree of each resource items, it is determined that before the resource matched with text message, go back execution step S104 making according to user
With information, matching degree of the text keyword with resource items is adjusted, the matching degree after being updated.
The history preference information of user, the personal information (such as age, sex, region etc.) of user, and similar users
Label information, tendency Sexual behavior mode of the user to resource can be reflected.Therefore, according to these use informations to text keyword
Be adjusted with the matching degree of resource items, for example, can by the use information with user associate more close resource items
Amplify with degree, using the coefficient being multiplied by more than 1, or increase the modes such as certain numerical value;Again for example, will can make with user
Reduced with the matching degree of the weaker resource items of information association.It is ranked up for the matching degree after adjustment, and then to user
Carry out recommendation of personalized information or matching degree highest resource items are recommended into user directly.
It should be noted that the executive agent of each step of 1~embodiment of embodiment, 3 provided method may each be same
Device, or, the method also can be by different device as executive agent.
Embodiment 4
Present invention also provides a kind of device 100 for matching resource for text message, shown in Figure 4, including:
Acquisition module 101, for according to text message, obtaining text keyword;
Matching module 102, for the similarity according to text keyword and resource tag, it is determined that match with text message
Resource;Wherein, resource tag is corresponding with resource.
As shown in figure 5, above-mentioned matching module 102 can be further included:
Computing module 1021, for according to text keyword and the similarity of resource tag and corresponding resource mark
The weight in affiliated resource items is signed, the matching degree of text keyword and resource items is calculated;Wherein, comprising at least in resource items
One resource tag;
Resource determination module 1022, for the matching degree according to text keyword and each resource items, it is determined that and text envelope
The resource of manner of breathing matching;Wherein, at least one resource is included in resource items.
The device for matching resource for text message provided by the present embodiment, with the method described in 1~embodiment of embodiment 3
It is corresponding, will not be described here.
Embodiment 5
Based on the statement of 1~embodiment of embodiment 4, below with regard to the provided herein side for matching resource for text message
Method and device provide specific application example.
Hypothesis has four kinds of resource items:
T1:Happy birthday!(resource tag and its weight are:Cake 0.2, song 0.15, the birthday 0.5, it is happy 0.15)
T2:(resource tag and its weight are for dragon boat match:Dragon boat 0.8, match is 0.2)
T3:(resource tag and its weight are Zongzi:Zongzi 0.7, the Dragon Boat Festival is 0.3)
T4:Happy Mid-autumn Festival, and (resource tag and its weight are:Moon cake 0.6, reunion 0.3, it is happy 0.1)
Include at least one section music sources corresponding with the resource items in every kind of resource items.
The text message for getting user input is:The Dragon Boat Festival is happy!
Participle is then first carried out to text message, obtains text keyword { Dragon Boat Festival, happy }.
According to the similarity of each word in term vector storehouse, it is assumed that " Dragon Boat Festival " is 0.8 with the similarity of " dragon boat ", " Dragon Boat Festival
Section " is 0.9 with the similarity of " Zongzi ".In this example, the similarity very little of " Dragon Boat Festival " and other resource tags, " happy " with
The similarity of other resource tags in addition to its own also very little, therefore, ignore in the follow-up calculating of this example.In reality
When border is implemented, a similarity numerical threshold can be set, when the similarity of two words it is very little, during less than the threshold value, can be with
The numerical value of the similarity of two word is set to into 0, to simplify calculating.
Then text keyword " Dragon Boat Festival " and resource items T2And T3Matching degree be respectively:0.8*0.8=0.64,0.7*
0.9=0.63.
Calculate by similar approach, it is assumed that obtain text keyword " happy " and resource items T1And T4Matching degree be respectively:
0.15,0.1.
Then obtain to being ranked up to resource items according to matching degree order from high to low:T2, T3, T1, T4.
Further, it is contemplated that the attribute such as the historical record of user, personal region/sex, above-mentioned resource items can be entered
Row rearrangement.Show as here:Find that user is northerner according to the geography information of user, the northern Dragon Boat Festival does not have dragon boat
Match, then, text keyword and T can be reduced accordingly2Matching degree, if drop weight factor is 0.9, then T2New matching
Degree is 0.64*0.9=0.576.So as to the resource items after being resequenced are:T3, T2, T1, T4。
Yet further, if user clicks T3, then T can be increased3In " Dragon Boat Festival " weight.If user clicks on
T2, then can be in T2Resource tag in increase " Dragon Boat Festival " this label.If user selects T1Or T4, then corresponding
Resource items in increase correspondence resource tag " happy " weight.
Those skilled in the art are it should be appreciated that embodiments of the invention can be provided as method, system or computer program
Product.Therefore, the present invention can adopt complete hardware embodiment, complete software embodiment or with reference to the reality in terms of software and hardware
Apply the form of example.And, the present invention can be using the computer for wherein including computer usable program code at one or more
The computer program implemented in usable storage medium (including but not limited to magnetic disc store, CD-ROM, optical memory etc.) is produced
The form of product.
The present invention be with reference to method according to embodiments of the present invention, equipment (system), and computer program flow process
Figure and/or block diagram are describing.It should be understood that can be by computer program instructions flowchart and/or each stream in block diagram
The combination of journey and/or square frame and flow chart and/or the flow process in block diagram and/or square frame.These computer programs can be provided
The processor of all-purpose computer, special-purpose computer, Embedded Processor or other programmable data processing devices is instructed to produce
A raw machine so that produced for reality by the instruction of computer or the computing device of other programmable data processing devices
The device of the function of specifying in present one flow process of flow chart or one square frame of multiple flow processs and/or block diagram or multiple square frames.
These computer program instructions may be alternatively stored in and can guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works so that the instruction being stored in the computer-readable memory is produced to be included referring to
Make the manufacture of device, the command device realize in one flow process of flow chart or one square frame of multiple flow processs and/or block diagram or
The function of specifying in multiple square frames.
These computer program instructions can be also loaded in computer or other programmable data processing devices so that in meter
Series of operation steps is performed on calculation machine or other programmable devices to produce computer implemented process, so as in computer or
The instruction performed on other programmable devices is provided for realizing in one flow process of flow chart or multiple flow processs and/or block diagram one
The step of function of specifying in individual square frame or multiple square frames.
In a typical configuration, computing device includes one or more processors (CPU), input/output interface, net
Network interface and internal memory.
Internal memory potentially includes the volatile memory in computer-readable medium, random access memory (RAM) and/or
The forms such as Nonvolatile memory, such as read-only storage (ROM) or flash memory (flash RAM).Internal memory is computer-readable medium
Example.
Computer-readable medium includes that permanent and non-permanent, removable and non-removable media can be by any method
Or technology is realizing information Store.Information can be computer-readable instruction, data structure, the module of program or other data.
The example of the storage medium of computer includes, but are not limited to phase transition internal memory (PRAM), static RAM (SRAM), moves
State random access memory (DRAM), other kinds of random access memory (RAM), read-only storage (ROM), electric erasable
Programmable read only memory (EEPROM), fast flash memory bank or other memory techniques, read-only optical disc read-only storage (CD-ROM),
Digital versatile disc (DVD) or other optical storages, magnetic cassette tape, the storage of tape magnetic rigid disk or other magnetic storage apparatus
Or any other non-transmission medium, can be used to store the information that can be accessed by a computing device.Define according to herein, calculate
Machine computer-readable recording medium does not include temporary computer readable media (transitory media), the such as data-signal and carrier wave of modulation.
Also, it should be noted that term " including ", "comprising" or its any other variant are intended to nonexcludability
Comprising so that a series of process, method, commodity or equipment including key elements not only includes those key elements, but also wrapping
Other key elements being not expressly set out are included, or also includes intrinsic for this process, method, commodity or equipment wanting
Element.In the absence of more restrictions, the key element for being limited by sentence "including a ...", it is not excluded that wanting including described
Also there is other identical element in the process of element, method, commodity or equipment.
It will be understood by those skilled in the art that embodiments herein can be provided as method, system or computer program.
Therefore, the application can adopt complete hardware embodiment, complete software embodiment or with reference to the embodiment in terms of software and hardware
Form.And, the application can be deposited using the computer for wherein including computer usable program code at one or more is available
The shape of the computer program implemented on storage media (including but not limited to magnetic disc store, CD-ROM, optical memory etc.)
Formula.
Embodiments herein is the foregoing is only, the application is not limited to.For those skilled in the art
For, the application can have various modifications and variations.All any modifications made within spirit herein and principle, equivalent
Replace, improve etc., within the scope of should be included in claims hereof.
Claims (14)
1. it is a kind of for text message match resource method, it is characterised in that include:
According to the text message, text keyword is obtained;
According to the text keyword and the similarity of resource tag, it is determined that the resource matched with the text message;Wherein,
The resource tag is corresponding with the resource.
2. according to claim 1 methods described, it is characterised in that the determination of the similarity of the text keyword and resource tag
Including:
Obtain the second term vector of first term vector and the resource tag of the text keyword from term vector storehouse respectively;
Term vector comprising at least one participle in the term vector storehouse;
The cosine similarity of first term vector and second term vector is calculated, as the text keyword and the money
The similarity of source label.
3. according to claim 2 methods described, it is characterised in that in the term vector storehouse, term vector is included:
Obtain the corpus for training term vector;
The corpus are carried out with participle, participle set is obtained;
Cooccurrence relation according to the frequency of occurrence of each participle in the participle set and/or with other participles, determines each participle
Term vector;The frequency of occurrence of each element representation correspondence participle and/or the co-occurrence frequency with other participles in the term vector.
4. according to claim 1 methods described, it is characterised in that according to the text keyword and the similarity of resource tag,
It is determined that the resource matched with the text message, including:
Similarity of the text keyword with the resource tag is ranked up;
By similarity highest default resource tag output, to determine the resource matched with the text message.
5. according to claim 1 methods described, it is characterised in that according to the text keyword and the similarity of resource tag,
It is determined that the resource matched with the text message, including:
According to the similarity and corresponding resource tag of the text keyword and the resource tag in affiliated resource items
In weight, calculate the matching degree of the text keyword and the resource items;Wherein, at least one is included in the resource items
The individual resource tag;
According to the matching degree of the text keyword and each resource items, it is determined that the money matched with the text message
Source;Wherein, resource described at least one is included in the resource items.
6. according to claim 5 methods described, it is characterised in that the calculating text keyword matches journey with the resource items
Degree, including:
The sum of products of the similarity and the weight is calculated, as the text keyword and corresponding resource tag institute
The matching degree of category resource items.
7. according to claim 5 methods described, it is characterised in that according to matching for the text keyword and each resource items
Degree, it is determined that the resource matched with the text message, including:
Matching degree of the text keyword with the resource items is ranked up;
By the default resource items output for most matching, to determine the resource matched with the text message.
8. according to claim 7 methods described, it is characterised in that by the default resource items output for most matching, to determine with
The resource that the text message matches, specially:
By the default resource items output for most matching, alternately resource items are to user's recommendation;
The resource corresponding to resource items that user is chosen in the alternative resource item is defined as and the text envelope manner of breathing
The resource matched somebody with somebody.
9. according to claim 8 methods described, it is characterised in that the determination of weight of the resource tag in affiliated resource items
Including:
Weight setting initial value to the resource tag in affiliated resource items;
The resource items chosen in the alternative resource item according to user, adjust the resource tag and/or tune in the resource items
Weight of the whole resource tag in affiliated resource items.
10. according to claim 5 methods described, it is characterised in that methods described also includes:
The use information of user is obtained, the use information includes history preference information, personal information and/or the similar use of user
The label information at family;
Then after the matching degree of the text keyword with the resource items is calculated, according to the text keyword with it is each
The matching degree of the resource items, it is determined that before the resource matched with the text message, also include:
According to the use information of user, matching degree of the text keyword with the resource items is adjusted, is obtained
Matching degree to after renewal.
11. according to claim 1 methods described, it is characterised in that according to the text message, obtains text keyword, including:
Obtain the text message;
Participle is carried out to the text message, the text keyword is obtained.
12. according to claim 1~11 arbitrary methods described, it is characterised in that the classification of the resource includes picture, photograph
Piece, music, voice, audio, action, expression, animation and/or video.
13. a kind of devices for matching resource for text message, it is characterised in that include:
Acquisition module, for according to the text message, obtaining text keyword;
Matching module, for the similarity according to the text keyword and resource tag, it is determined that with the text envelope manner of breathing
The resource matched somebody with somebody;Wherein, the resource tag is corresponding with the resource.
14. according to claim 13 described device, it is characterised in that the matching module includes:
Computing module, for according to the text keyword and the similarity of the resource tag and corresponding resource mark
The weight in affiliated resource items is signed, the matching degree of the text keyword and the resource items is calculated;Wherein, the resource
Resource tag described at least one is included in;
Resource determination module, for the matching degree according to the text keyword and each resource items, it is determined that with the text
The resource of this information match;Wherein, resource described at least one is included in the resource items.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610824080.1A CN106528588A (en) | 2016-09-14 | 2016-09-14 | Method and apparatus for matching resources for text information |
PCT/CN2017/097611 WO2018049960A1 (en) | 2016-09-14 | 2017-08-16 | Method and apparatus for matching resource for text information |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610824080.1A CN106528588A (en) | 2016-09-14 | 2016-09-14 | Method and apparatus for matching resources for text information |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106528588A true CN106528588A (en) | 2017-03-22 |
Family
ID=58343954
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610824080.1A Pending CN106528588A (en) | 2016-09-14 | 2016-09-14 | Method and apparatus for matching resources for text information |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN106528588A (en) |
WO (1) | WO2018049960A1 (en) |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107193962A (en) * | 2017-05-24 | 2017-09-22 | 百度在线网络技术(北京)有限公司 | A kind of intelligent figure method and device of internet promotion message |
WO2018049960A1 (en) * | 2016-09-14 | 2018-03-22 | 厦门幻世网络科技有限公司 | Method and apparatus for matching resource for text information |
CN107895200A (en) * | 2017-11-15 | 2018-04-10 | 广东天泽阳光康众医疗投资管理有限公司 | A kind of intelligence reservation hospital register method, device, terminal device and storage medium |
CN108776901A (en) * | 2018-04-27 | 2018-11-09 | 微梦创科网络科技(中国)有限公司 | Method and system for advertisement recommendation based on search term |
CN109144954A (en) * | 2018-09-18 | 2019-01-04 | 天津字节跳动科技有限公司 | Edit resource recommendation method, device and the electronic equipment of document |
CN109376352A (en) * | 2018-08-28 | 2019-02-22 | 中山大学 | A kind of patent text modeling method based on word2vec and semantic similarity |
CN109657710A (en) * | 2018-12-06 | 2019-04-19 | 北京达佳互联信息技术有限公司 | Data screening method, apparatus, server and storage medium |
CN109756751A (en) * | 2017-11-07 | 2019-05-14 | 腾讯科技(深圳)有限公司 | Multimedia data processing method and device, electronic equipment, storage medium |
CN110020411A (en) * | 2019-03-29 | 2019-07-16 | 上海掌门科技有限公司 | Graph-text content generation method and equipment |
CN110377558A (en) * | 2019-06-14 | 2019-10-25 | 平安科技(深圳)有限公司 | Document searching method, device, computer equipment and storage medium |
CN110516061A (en) * | 2019-07-24 | 2019-11-29 | 视联动力信息技术股份有限公司 | A kind of data processing method, device and computer readable storage medium |
CN111126054A (en) * | 2019-12-03 | 2020-05-08 | 东软集团股份有限公司 | Method, device, storage medium and electronic equipment for determining similar texts |
CN111178064A (en) * | 2019-12-13 | 2020-05-19 | 平安医疗健康管理股份有限公司 | Information pushing method and device based on field word segmentation processing and computer equipment |
CN111191454A (en) * | 2020-01-06 | 2020-05-22 | 精硕科技(北京)股份有限公司 | Entity matching method and device |
CN111460177A (en) * | 2020-03-27 | 2020-07-28 | 北京奇艺世纪科技有限公司 | Method and device for searching film and television expression, storage medium and computer equipment |
CN112148904A (en) * | 2020-07-31 | 2020-12-29 | 中标慧安信息技术股份有限公司 | Indexing method and device for alarm linkage video |
CN113342179A (en) * | 2021-05-26 | 2021-09-03 | 北京百度网讯科技有限公司 | Input text processing method and device, electronic equipment and storage medium |
CN114040142A (en) * | 2021-11-05 | 2022-02-11 | 深圳壹账通智能科技有限公司 | Video call method, device, equipment and storage medium of intelligent outbound robot |
WO2023045710A1 (en) * | 2021-09-27 | 2023-03-30 | 北京字节跳动网络技术有限公司 | Multimedia display and matching methods and apparatuses, device and medium |
Families Citing this family (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108829669A (en) * | 2018-06-01 | 2018-11-16 | 北京玄科技有限公司 | Support the term vector generation method and device of polarity differentiation and ambiguity |
CN108763566A (en) * | 2018-06-05 | 2018-11-06 | 北京玄科技有限公司 | Text similarity computing method and device, intelligent robot |
CN108920546B (en) * | 2018-06-13 | 2020-09-11 | 杭州数理大数据技术有限公司 | Steady-state label development method and system based on user requirements |
CN109033162A (en) * | 2018-06-19 | 2018-12-18 | 深圳市元征科技股份有限公司 | A kind of data processing method, server and computer-readable medium |
CN109582966A (en) * | 2018-12-03 | 2019-04-05 | 北京容联易通信息技术有限公司 | A kind of information matching method and device |
CN111488453B (en) * | 2019-01-25 | 2024-02-23 | 北京猎户星空科技有限公司 | Resource grading method, device, equipment and storage medium |
CN109857872A (en) * | 2019-02-18 | 2019-06-07 | 浪潮软件集团有限公司 | The information recommendation method and device of knowledge based map |
CN109871540B (en) * | 2019-02-21 | 2022-12-23 | 武汉斗鱼鱼乐网络科技有限公司 | Text similarity calculation method and related equipment |
CN110163121B (en) * | 2019-04-30 | 2023-09-05 | 腾讯科技(深圳)有限公司 | Image processing method, device, computer equipment and storage medium |
CN110263161B (en) * | 2019-05-29 | 2023-09-26 | 创新先进技术有限公司 | Information processing method, device and equipment |
CN110659419B (en) * | 2019-09-17 | 2023-09-05 | 平安科技(深圳)有限公司 | Method and related device for determining target user |
CN111259117B (en) * | 2020-01-16 | 2023-11-21 | 广州拉卡拉信息技术有限公司 | Short text batch matching method and device |
CN111400516B (en) * | 2020-03-16 | 2024-04-16 | 北京奇艺世纪科技有限公司 | Label determining method, electronic device and storage medium |
CN111708819B (en) * | 2020-05-28 | 2023-04-07 | 北京百度网讯科技有限公司 | Method, apparatus, electronic device, and storage medium for information processing |
CN112037792B (en) * | 2020-08-20 | 2022-06-17 | 北京字节跳动网络技术有限公司 | Voice recognition method and device, electronic equipment and storage medium |
CN112989022B (en) * | 2021-03-16 | 2022-11-25 | 中国平安人寿保险股份有限公司 | Intelligent virtual text selection method and device and computer equipment |
CN113377971B (en) * | 2021-05-31 | 2024-02-27 | 北京达佳互联信息技术有限公司 | Multimedia resource generation method and device, electronic equipment and storage medium |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102419773A (en) * | 2011-12-13 | 2012-04-18 | 百度在线网络技术(北京)有限公司 | Method, device and equipment used for sequencing resource items |
CN103425687A (en) * | 2012-05-21 | 2013-12-04 | 阿里巴巴集团控股有限公司 | Retrieval method and system based on queries |
CN104239535A (en) * | 2014-09-22 | 2014-12-24 | 重庆邮电大学 | Method and system for matching pictures with characters, server and terminal |
CN104636334A (en) * | 2013-11-06 | 2015-05-20 | 阿里巴巴集团控股有限公司 | Keyword recommending method and device |
CN105354182A (en) * | 2015-09-28 | 2016-02-24 | 北大方正集团有限公司 | Method for obtaining related digital resources and method and apparatus for generating special topic by using method |
JP2016081265A (en) * | 2014-10-16 | 2016-05-16 | 日本電気株式会社 | Picture selection device, picture selection method, picture selection program, characteristic-amount generation device, characteristic-amount generation method and characteristic-amount generation program |
CN105824904A (en) * | 2016-03-15 | 2016-08-03 | 浙江大学 | Chinese herbal medicine plant picture capturing method based on professional term vector of traditional Chinese medicine and pharmacy field |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101984437B (en) * | 2010-11-23 | 2012-08-08 | 亿览在线网络技术(北京)有限公司 | Music resource individual recommendation method and system thereof |
CN103412880B (en) * | 2013-07-17 | 2017-02-22 | 百度在线网络技术(北京)有限公司 | Method and device for determining implicit associated information between multimedia resources |
CN105893410A (en) * | 2015-11-18 | 2016-08-24 | 乐视网信息技术(北京)股份有限公司 | Keyword extraction method and apparatus |
CN106528588A (en) * | 2016-09-14 | 2017-03-22 | 厦门幻世网络科技有限公司 | Method and apparatus for matching resources for text information |
-
2016
- 2016-09-14 CN CN201610824080.1A patent/CN106528588A/en active Pending
-
2017
- 2017-08-16 WO PCT/CN2017/097611 patent/WO2018049960A1/en active Application Filing
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102419773A (en) * | 2011-12-13 | 2012-04-18 | 百度在线网络技术(北京)有限公司 | Method, device and equipment used for sequencing resource items |
CN103425687A (en) * | 2012-05-21 | 2013-12-04 | 阿里巴巴集团控股有限公司 | Retrieval method and system based on queries |
CN104636334A (en) * | 2013-11-06 | 2015-05-20 | 阿里巴巴集团控股有限公司 | Keyword recommending method and device |
CN104239535A (en) * | 2014-09-22 | 2014-12-24 | 重庆邮电大学 | Method and system for matching pictures with characters, server and terminal |
JP2016081265A (en) * | 2014-10-16 | 2016-05-16 | 日本電気株式会社 | Picture selection device, picture selection method, picture selection program, characteristic-amount generation device, characteristic-amount generation method and characteristic-amount generation program |
CN105354182A (en) * | 2015-09-28 | 2016-02-24 | 北大方正集团有限公司 | Method for obtaining related digital resources and method and apparatus for generating special topic by using method |
CN105824904A (en) * | 2016-03-15 | 2016-08-03 | 浙江大学 | Chinese herbal medicine plant picture capturing method based on professional term vector of traditional Chinese medicine and pharmacy field |
Cited By (29)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2018049960A1 (en) * | 2016-09-14 | 2018-03-22 | 厦门幻世网络科技有限公司 | Method and apparatus for matching resource for text information |
CN107193962A (en) * | 2017-05-24 | 2017-09-22 | 百度在线网络技术(北京)有限公司 | A kind of intelligent figure method and device of internet promotion message |
CN107193962B (en) * | 2017-05-24 | 2021-06-11 | 百度在线网络技术(北京)有限公司 | Intelligent map matching method and device for Internet promotion information |
CN109756751A (en) * | 2017-11-07 | 2019-05-14 | 腾讯科技(深圳)有限公司 | Multimedia data processing method and device, electronic equipment, storage medium |
CN107895200A (en) * | 2017-11-15 | 2018-04-10 | 广东天泽阳光康众医疗投资管理有限公司 | A kind of intelligence reservation hospital register method, device, terminal device and storage medium |
CN108776901B (en) * | 2018-04-27 | 2021-01-15 | 微梦创科网络科技(中国)有限公司 | Advertisement recommendation method and system based on search terms |
CN108776901A (en) * | 2018-04-27 | 2018-11-09 | 微梦创科网络科技(中国)有限公司 | Method and system for advertisement recommendation based on search term |
CN109376352A (en) * | 2018-08-28 | 2019-02-22 | 中山大学 | A kind of patent text modeling method based on word2vec and semantic similarity |
CN109376352B (en) * | 2018-08-28 | 2022-11-29 | 中山大学 | Patent text modeling method based on word2vec and semantic similarity |
CN109144954A (en) * | 2018-09-18 | 2019-01-04 | 天津字节跳动科技有限公司 | Edit resource recommendation method, device and the electronic equipment of document |
CN109144954B (en) * | 2018-09-18 | 2021-03-16 | 北京字节跳动网络技术有限公司 | Resource recommendation method and device for editing document and electronic equipment |
CN109657710B (en) * | 2018-12-06 | 2022-01-21 | 北京达佳互联信息技术有限公司 | Data screening method and device, server and storage medium |
CN109657710A (en) * | 2018-12-06 | 2019-04-19 | 北京达佳互联信息技术有限公司 | Data screening method, apparatus, server and storage medium |
CN110020411A (en) * | 2019-03-29 | 2019-07-16 | 上海掌门科技有限公司 | Graph-text content generation method and equipment |
CN110377558B (en) * | 2019-06-14 | 2023-06-20 | 平安科技(深圳)有限公司 | Document query method, device, computer equipment and storage medium |
CN110377558A (en) * | 2019-06-14 | 2019-10-25 | 平安科技(深圳)有限公司 | Document searching method, device, computer equipment and storage medium |
CN110516061A (en) * | 2019-07-24 | 2019-11-29 | 视联动力信息技术股份有限公司 | A kind of data processing method, device and computer readable storage medium |
CN111126054A (en) * | 2019-12-03 | 2020-05-08 | 东软集团股份有限公司 | Method, device, storage medium and electronic equipment for determining similar texts |
CN111126054B (en) * | 2019-12-03 | 2024-03-05 | 东软集团股份有限公司 | Method and device for determining similar text, storage medium and electronic equipment |
CN111178064A (en) * | 2019-12-13 | 2020-05-19 | 平安医疗健康管理股份有限公司 | Information pushing method and device based on field word segmentation processing and computer equipment |
CN111178064B (en) * | 2019-12-13 | 2022-11-29 | 深圳平安医疗健康科技服务有限公司 | Information pushing method and device based on field word segmentation processing and computer equipment |
CN111191454A (en) * | 2020-01-06 | 2020-05-22 | 精硕科技(北京)股份有限公司 | Entity matching method and device |
CN111460177A (en) * | 2020-03-27 | 2020-07-28 | 北京奇艺世纪科技有限公司 | Method and device for searching film and television expression, storage medium and computer equipment |
CN111460177B (en) * | 2020-03-27 | 2023-12-15 | 北京奇艺世纪科技有限公司 | Video expression search method and device, storage medium and computer equipment |
CN112148904B (en) * | 2020-07-31 | 2021-07-27 | 中标慧安信息技术股份有限公司 | Indexing method and device for alarm linkage video |
CN112148904A (en) * | 2020-07-31 | 2020-12-29 | 中标慧安信息技术股份有限公司 | Indexing method and device for alarm linkage video |
CN113342179A (en) * | 2021-05-26 | 2021-09-03 | 北京百度网讯科技有限公司 | Input text processing method and device, electronic equipment and storage medium |
WO2023045710A1 (en) * | 2021-09-27 | 2023-03-30 | 北京字节跳动网络技术有限公司 | Multimedia display and matching methods and apparatuses, device and medium |
CN114040142A (en) * | 2021-11-05 | 2022-02-11 | 深圳壹账通智能科技有限公司 | Video call method, device, equipment and storage medium of intelligent outbound robot |
Also Published As
Publication number | Publication date |
---|---|
WO2018049960A1 (en) | 2018-03-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106528588A (en) | Method and apparatus for matching resources for text information | |
US20210165955A1 (en) | Methods and systems for modeling complex taxonomies with natural language understanding | |
US11106983B2 (en) | Intelligent interaction method and intelligent interaction system | |
CN104252533B (en) | Searching method and searcher | |
WO2021164255A1 (en) | Presentation generation method and apparatus, computer device and storage medium | |
US10971133B2 (en) | Voice synthesis method, device and apparatus, as well as non-volatile storage medium | |
CN106503192A (en) | Name entity recognition method and device based on artificial intelligence | |
CN104484377B (en) | Replace dictionary generating method and device | |
US11636251B2 (en) | Content aware font recommendation | |
CN111582910A (en) | Method, system and equipment for generating advertisement case | |
CN109117470B (en) | Evaluation relation extraction method and device for evaluating text information | |
CN108984711B (en) | Personalized APP recommendation method based on hierarchical embedding | |
CN110532547A (en) | Building of corpus method, apparatus, electronic equipment and medium | |
CN104881428B (en) | A kind of hum pattern extraction, search method and the device of hum pattern webpage | |
CN110196910B (en) | Corpus classification method and apparatus | |
Ramina et al. | Topic level summary generation using BERT induced Abstractive Summarization Model | |
CN108920521A (en) | User's portrait-item recommendation system and method based on pseudo- ontology | |
CN103631874A (en) | UGC label classification determining method and device for social platform | |
JP2018005690A (en) | Information processing apparatus and program | |
CN106294358A (en) | The search method of a kind of information and system | |
CN116628198A (en) | Training method and device of text generation model, medium and electronic equipment | |
JP2020098592A (en) | Method, device and storage medium of extracting web page content | |
CN112527963B (en) | Dictionary-based multi-label emotion classification method and device, equipment and storage medium | |
CN103136221A (en) | Method capable of generating requirement template and requirement identification method and device | |
CN116882414A (en) | Automatic comment generation method and related device based on large-scale language model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
TA01 | Transfer of patent application right | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20190326 Address after: 361012 3F-A193, Innovation Building C, Software Park, Xiamen Torch High-tech Zone, Xiamen City, Fujian Province Applicant after: Xiamen Black Mirror Technology Co., Ltd. Address before: 9th Floor, Maritime Building, 16 Haishan Road, Huli District, Xiamen City, Fujian Province, 361000 Applicant before: XIAMEN HUANSHI NETWORK TECHNOLOGY CO., LTD. |
|
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170322 |