CN106446025A - Method and device for standardizing text information - Google Patents
Method and device for standardizing text information Download PDFInfo
- Publication number
- CN106446025A CN106446025A CN201610766697.2A CN201610766697A CN106446025A CN 106446025 A CN106446025 A CN 106446025A CN 201610766697 A CN201610766697 A CN 201610766697A CN 106446025 A CN106446025 A CN 106446025A
- Authority
- CN
- China
- Prior art keywords
- text message
- standard
- information
- alternative criteria
- standard text
- 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.)
- Granted
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
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
-
- 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
- G06F16/353—Clustering; Classification into predefined classes
Abstract
The embodiment of the invention discloses a method and a device for standardizing text information. In the embodiment of the invention, for nonstandard text information used for describing medical relevant affairs in a medical system, a possibility that the nonstandard text information and each piece of alternative standard text information are both used for describing the same medical relevant affair can be obtained through a Bayes classification algorithm model according to the associated information of the nonstandard text information; and then, on the basis of the probability, one part of standard text information can be selected from all alternative standard text information, and therefore, target standard text information used for standardizing the nonstandard text information can be searched only from the part of standard text information. Therefore, a range for searching the target standard text information for the nonstandard text information is greatly reduced, so that manual work is simplified, and manual burdens are lightened.
Description
Technical field
The present invention relates to technical field of data processing, particularly relate to the method and apparatus of a kind of standardized text information.
Background technology
At present, same medical treatment correlate is often used different text messages to be described by each medical system.
For example, for " internal medicine " this medical concept, some medical systems may use text message " internal medicine ", other medical systems
Text message " internal medicine I " and " internal medicine II " may be used.And for example, for " infection of the upper respiratory tract " this medical concept, some doctors
Treatment system may use text message " infection of the upper respiratory tract ", and other medical systems may use text message " to catch a cold ", again
Some medical systems may use text message " to have a fever ".The inconsistent situation of this text message can make each medical system
Data message be difficult to share, thus be unfavorable for the development of medical industry.
The situation inconsistent in order to eliminate text message, can arrange received text information for each medical treatment relevant issues and be
For the same medical treatment received text information of correlate described and the non-standard text message that each medical system uses set up right
Should be related to.According to this corresponding relation, each medical system is for describing the different non-standard text envelope of same medical treatment correlate
Breath can be unified into the received text information of this medical treatment correlate, it is achieved thereby that the standardization of text message, eliminates
The inconsistent situation of text message, thus be conducive to the data information sharing of each medical system.But, inventor is through research
Discovery, owing to there is substantial amounts of medical treatment correlate, the quantity of usual received text information is very huge.And for each medical department
When the non-standard text message that system uses sets up corresponding relation, need manually to go out non-with this at substantial amounts of received text information searching
Received text information is used for describing the received text information of same medical treatment correlate, thus causes manual working loaded down with trivial details, complicated.
Content of the invention
Technical problems to be solved in this application are to provide the method and apparatus of a kind of standardized text information, to solve
Manually go out with this non-standard text message for describing same doctor at substantial amounts of received text information searching according in prior art
The numerous and diverse technical problem of the manual working treating the received text information of correlate and cause.
First aspect, provides a kind of method of standardized text information, and the method includes：
Obtaining non-standard text message and the related information of described non-standard text message, described non-standard text is medical treatment
For describing the text message of medical treatment correlate in system；
Based on described related information and the first alternative criteria text message, set up Bayesian Classification Arithmetic model, wherein, institute
The input node stating Bayesian Classification Arithmetic model corresponds to described related information, the output of described Bayesian Classification Arithmetic model
Node corresponds to described first alternative criteria text message；
By historical context information and the historical standard text message of historical standard text message, to described Bayes's classification
Algorithm model is trained；
In the described Bayesian Classification Arithmetic model that training completes, input described related information, export described non-standard
Correlation information between text message and described first alternative criteria text message, described correlation information represents described nonstandard
Quasi-text and described first alternative criteria text message are used to describe the possibility of same medical treatment relevant issues；
According to the correlation information between described non-standard text message and described first alternative criteria text message, in institute
State and the first alternative criteria text message selects the second alternative criteria text；
Target criteria text message, wherein, described target criteria literary composition is determined in described second alternative criteria text message
This information is used for standardized corresponding relation for setting up with described non-standard text message.
Optionally,
Dependency relation between described Bayesian Classification Arithmetic model interior joint is embodied by oriented vector, each
Oriented vector has a corresponding weight；
The described related information by historical standard text message and historical standard text message, to described Bayes's classification
Algorithm model is trained, and is specially：By related information and the historical standard text message of historical standard text message, calculate
The weight of each oriented vector in described Bayesian Classification Arithmetic model；
Described in the described Bayesian Classification Arithmetic model that training completes, input described related information, export described pass
Correlation information between connection information and described first alternative criteria text message, is specially：Corresponding according to described related information
Information in each input node of described Bayesian Classification Arithmetic model, is respectively arranged with in described Bayesian Classification Arithmetic model
To the weight of vector, calculate the information on each output node of described Bayesian Classification Arithmetic model, as each output node pair
Correlation information between the first alternative criteria text message answered and described non-standard text message.
Optionally, described determination target criteria text message in described second alternative criteria text message, including：
Feed back described second alternative criteria text message；
In response to selection operation, by corresponding for described selection operation standard literary composition in described second alternative criteria text message
This information is defined as described target criteria text message.
Optionally, after the related information of the non-standard text message of described acquisition and described non-standard text message, also
Including：
Described non-standard text message and the 3rd alternative criteria text message are carried out text matches；
Choose, in described 3rd alternative criteria text message, the received text information that described text matches obtains, as institute
State the first alternative criteria text message.
Optionally, the received text information that described text matches obtains includes the standard with described non-standard text message
Text message.
Optionally, described described non-standard text message and the 3rd alternative criteria text message are carried out text matches, bag
Include：
Word segmentation processing is carried out to described non-standard text message, and based on the result of described word segmentation processing, determines described non-
The corresponding lemma of received text information；
Corresponding for described non-standard text message lemma and described 3rd alternative criteria text message are carried out text matches；
Wherein, the received text information that described text matches obtains includes having any one of described non-standard text message
The received text information of individual lemma.
Optionally, the described result based on described word segmentation processing, determines the corresponding lemma of described non-standard text message, bag
Include：
The result of described word segmentation processing is got rid of and belongs to the default lemma stopping in word dictionary, and by true for remaining lemma
It is set to the corresponding lemma of described non-standard text message.
Optionally, the described result based on described word segmentation processing, determines the corresponding lemma of described non-standard text message, bag
Include：
The synonym corresponding with the result of described word segmentation processing is searched in default synonym dictionary, and by described point
The result that word is processed is defined as the corresponding lemma of described non-standard text message with the synonym finding.
Optionally, described method also includes：
Determine the corresponding classification of described non-standard text message, and the received text information in described classification is defined as institute
State the 3rd alternative criteria text message.
Second aspect, provides the device of a kind of standardized text information, including：
Acquiring unit, for obtaining non-standard text message and the related information of described non-standard text message, described non-
Received text is for describing the text message of medical treatment correlate in medical system；
Set up unit, for based on described related information and the first alternative criteria text message, set up Bayes's classification and calculate
Method model, wherein, the input node of described Bayesian Classification Arithmetic model corresponds to described related information, described Bayes's classification
The output node of algorithm model corresponds to described first alternative criteria text message；
Training unit, for by the related information of historical standard text message and historical standard text message, to described
Bayesian Classification Arithmetic model is trained；
Computing unit, for, in the described Bayesian Classification Arithmetic model that training completes, inputting described related information, defeated
Go out the correlation information between described non-standard text message and described first alternative criteria text message, described correlation information
Represent that described non-standard text and described first alternative criteria text message are used to describe the possibility of same medical treatment relevant issues
Property；
First chooses unit, for according between described non-standard text message and described first alternative criteria text message
Correlation information, in described first alternative criteria text message, select the second alternative criteria text；
First determining unit, for determining target criteria text message in described second alternative criteria text message, its
In, described target criteria text message is used for standardized corresponding relation for setting up with described non-standard text message.
Optionally, the dependency relation between described Bayesian Classification Arithmetic model interior joint can be come by oriented vector
Embodying, each oriented vector has a corresponding weight.
Optionally, described training unit, specifically for：By related information and the historical standard of historical standard text message
Text message, calculates the weight of each oriented vector in described Bayesian Classification Arithmetic model.
Optionally, described computing unit, specifically for：Corresponding at described Bayesian Classification Arithmetic according to described related information
Information in each input node of model, by the weight of each oriented vector in described Bayesian Classification Arithmetic model, calculates institute
State the information on each output node of Bayesian Classification Arithmetic model, as the corresponding first alternative criteria text of each output node
Correlation information between information and described non-standard text message.
Optionally, described first determining unit specifically can include：
Feedback subelement, is used for feeding back described second alternative criteria text message；
First determination subelement, in response to selection operation, by described in described second alternative criteria text message
Selection operation corresponding received text information is defined as described target criteria text message.
Optionally, described device also includes：
Matching unit, after obtaining non-standard text message in described acquiring unit, by described non-standard text envelope
Breath and the 3rd alternative criteria text message carry out text matches；
Second chooses unit, for choosing, in described 3rd alternative criteria text message, the mark that described text matches obtains
Quasi-text message, as described first alternative criteria text message.
Optionally, the received text information that described text matches obtains includes the standard with described non-standard text message
Text message.
Optionally, described matching unit specifically includes：
Participle subelement, for carrying out word segmentation processing to described non-standard text message；
Second determination subelement, for the result based on described word segmentation processing, determines that described non-standard text message is corresponding
Lemma；
Coupling subelement, for by corresponding for described non-standard text message lemma and described 3rd alternative criteria text envelope
Breath carries out text matches；
Wherein, the received text information that described text matches obtains can include having appointing of described non-standard text message
The received text information of one lemma of meaning.
Optionally, described determination subelement specifically for：
The result of described word segmentation processing is got rid of and belongs to the default lemma stopping in word dictionary, and by true for remaining lemma
It is set to the corresponding lemma of described non-standard text message.
Optionally, described determination subelement specifically for：
The synonym corresponding with the result of described word segmentation processing is searched in default synonym dictionary, and by described point
The result that word is processed is defined as the corresponding lemma of described non-standard text message with the synonym finding.
Optionally, described device also includes：
Second determining unit, is used for determining the corresponding classification of described non-standard text message；
3rd determining unit, for being defined as described 3rd alternative criteria text by the received text information in described classification
Information.
Compared with prior art, the application has the following advantages：
In this application, it is used for describing the text message of medical treatment correlate as non-standard text in certain medical system
Information, when this non-standard text message is standardized by needs, obtains the related information of this non-standard text message, based on
This related information and the first alternative criteria text message set up Bayesian Classification Arithmetic model.For known, there is corresponding relation
The related information of historical standard text message and historical standard text message, by the historical context letter of historical standard text message
Breath and historical standard text message, be trained to Bayesian Classification Arithmetic model.Then, the Bayes's classification completing in training
In algorithm model, input the related information of this non-standard text message, export described non-standard text message first alternative with each
Correlation information between received text information, wherein, it is first standby with each that this correlation information identifies this non-standard text message
Received text information is selected to be used to describe the probability of same medical treatment relevant issues.Finally, according to this non-standard text message with each
Correlation information between first alternative criteria text message, selects the second alternative mark in the first alternative criteria text message
Quasi-text message, thus just can determine in the second alternative criteria text message for standardizing this non-standard text message
Received text information.As can be seen here, the correlation information being calculated by Bayesian Classification Arithmetic model, is non-standard text envelope
The scope that breath searches target criteria text message can narrow down to quantity more from a fairly large number of first alternative criteria text message
The second few alternative criteria text message, therefore, manual working is simplified, and artificial burden is mitigated.
Brief description
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
In having technology to describe, the accompanying drawing of required use is briefly described, it should be apparent that, the accompanying drawing in describing below is only this
Some embodiments described in application, for those of ordinary skill in the art, on the premise of not paying creative work,
Other accompanying drawing can also be obtained according to these accompanying drawings.
Fig. 1 is the schematic flow sheet of a kind of method of standardized text information in the embodiment of the present invention；
Fig. 2 is the schematic flow sheet of a kind of method of standardized text information in the embodiment of the present invention；
Fig. 3 is the structural representation of the device of a kind of standardized text information in the embodiment of the present invention.
Detailed description of the invention
In order to make those skilled in the art be more fully understood that the application scheme, below in conjunction with in the embodiment of the present application
Accompanying drawing, is clearly and completely described to the technical scheme in the embodiment of the present application, it is clear that described embodiment is only this
Apply for a part of embodiment, rather than whole embodiments.Based on the embodiment in the application, those of ordinary skill in the art exist
The every other embodiment being obtained under the premise of not making creative work, broadly falls into the scope of the application protection.
Inventor is through research discovery, in prior art, is using for describing medical treatment relevant issues for each medical system
Non-standard text message when being standardized, need manually to find out and non-standard text in all of received text information
Information is used for describing the received text information of same medical treatment correlate, and for describing the standard of substantial amounts of medical treatment correlate
Text message substantial amounts again itself, therefore, manual working is very loaded down with trivial details, complicated, thus causes and is related to thing for medical treatment
The text message standardisation process inefficiency of thing.
In order to solve this problem, in embodiments of the present invention, for medical system is used for describing medical treatment correlate
Non-standard text message, according to the related information of non-standard text message, by Bayesian Classification Arithmetic model, it can be deduced that
Non-standard text message and each alternative received text information are used to describe the possibility of same medical treatment correlate, then,
A part of received text information can be selected based on possibility in all alternative received text information, thus only at this
The received text information of part just can find out the target criteria text message for standardizing this non-standard text message.
Therefore, the scope for non-standard text message lookup target criteria text message greatly reduces, and therefore, manual working is able to
Simplifying, artificial burden is mitigated.
For example, one of scene of the embodiment of the present invention, can be applied in any one computer system.This meter
Calculation machine system can obtain non-standard text message and the related information of described non-standard text message, wherein, described non-standard
Text is for describing the text message of medical treatment correlate in medical system.Then, this computer system can be based on described
Related information and the first alternative criteria text message, set up Bayesian Classification Arithmetic model, wherein, described Bayesian Classification Arithmetic
The input node of model corresponds to described related information, and the output node of described Bayesian Classification Arithmetic model is corresponding to described the
One alternative criteria text message.After again, this computer system can by the historical context information of historical standard text message and
Historical standard text message, is trained to described Bayesian Classification Arithmetic model.Again again after, this computer system can instruction
Practice in the described Bayesian Classification Arithmetic model completing, input described related information, export described non-standard text message and institute
State the correlation information between the first alternative criteria text message, described correlation information represent described non-standard text with described
First alternative criteria text message is used to describe the possibility of same medical treatment relevant issues.Again again after, this computer system can
With according to the correlation information between described non-standard text message and described first alternative criteria text message, described first
Alternative criteria text message selects the second alternative criteria text.Again again after, this computer system can be described second standby
Selecting determination target criteria text message in received text, wherein, described target criteria text message is used for and described non-standard literary composition
This information is set up and is used for standardized corresponding relation.
It is understood that above-mentioned scene is only the Sample Scenario that the embodiment of the present invention provides, the embodiment of the present invention
It is not limited to this scene.
Below in conjunction with the accompanying drawings, by embodiment describe in detail embodiment of the present invention Plays text message method and
The specific implementation of device.
See Fig. 1, show the schematic flow sheet of a kind of method of standardized text information in the embodiment of the present invention.At this
In embodiment, described method for example specifically may comprise steps of：
Step 101, obtain non-standard text message and the related information of described non-standard text message, described non-standard literary composition
Originally it is for describing the text message of medical treatment correlate in medical system.
When implementing, each medical system for describe medical treatment correlate each text message in, obtain a literary composition
This information is as the non-standard text message being currently standardized.Meanwhile, the association letter of this non-standard text message is also obtained
Breath.
In the present embodiment, non-standard text message represents the text that medical system uses for describing medical treatment correlate
Information, received text information table shows in normalisation rule for describing the text message of medical treatment correlate.It is understood that
Non-standard text message is generally not belonging to received text information, i.e. this non-standard text message and any one received text information
It is different from, also accordingly, it would be desirable to find out with non-standard text message for describing the received text of same medical treatment correlate
Information, to realize the standardization to non-standard text message.Certainly, in some cases, non-standard text message is likely to itself
Be exactly a received text information, i.e. this medical system is to use this received text information itself to go to describe medical treatment correlate
's.
It it should be noted that the related information of the non-standard text message involved by the present embodiment, is determined for non-
Received text information and each received text information are for describing the possibility of same medical treatment correlate.Specifically, non-standard literary composition
The related information of this information, can be with relevant other information of the medical correlate described by right and wrong received text information.Example
If, it is assumed that non-standard text message is " flu ", the related information of non-standard text message can be diagnostic result be " flu "
Prescription information, more specifically, related information can include the illness in the prescription that diagnostic result is " flu " describe information and/
Or therapeutic scheme information.
Step 102, based on described related information and the first alternative criteria text message, set up Bayesian Classification Arithmetic mould
Type, wherein, the input node of described Bayesian Classification Arithmetic model corresponds to described related information, described Bayesian Classification Arithmetic
The output node of model corresponds to described first alternative criteria text message.
When implementing, if desired determine in the first alternative criteria text message and be used for standardizing non-standard text message
Target criteria text message, Bayesian Classification Arithmetic can be set up based on the target criteria text message of non-standard text message
Model.
It is understood that Bayesian Classification Arithmetic model has input node and output node.Wherein, for one
For calculating process, each input node is to the independent variable that should calculate process, and each output node was to should calculate
One dependent variable of journey.In the present embodiment, Bayesian Classification Arithmetic model is for the association letter according to non-standard text message
Breath calculates non-standard text message and each first alternative criteria text message is used to describe the possibility of same medical treatment correlate
Property degree, therefore, when setting up Bayesian Classification Arithmetic model, the input node of Bayesian Classification Arithmetic model can correspond to
The related information of non-standard text message, the output node of Bayesian Classification Arithmetic model can correspond to the first alternative criteria literary composition
This information.Specifically, in the Bayesian Classification Arithmetic model set up, each input node corresponds to non-standard text envelope
One related information of breath, each output node corresponds to a first alternative criteria text message.For an input joint
For Dian, this input node is input in Bayesian Classification Arithmetic model for the occurrence of a corresponding related information
Position.For an output node, this output node be the first corresponding alternative criteria text message with non-standard
Text message is used for describing outgoing position in Bayesian Classification Arithmetic model for the possibility degree of same medical treatment correlate.
It should be noted that in Bayesian Classification Arithmetic model, the relation between input node and output node meets
Bayesian Classification Arithmetic.In the present embodiment, the implementation of any one Bayesian Classification Arithmetic may be incorporated for tissue shellfish
Node in this sorting algorithm model of leaf.For example, in some embodiments, at described Bayesian Classification Arithmetic model interior joint
Between dependency relation can be embodied by oriented vector.Wherein, each oriented vector can have a corresponding power
Weight.By the corresponding weight of oriented vector between each node, in Bayesian Classification Arithmetic model input node and output node it
Between dependency relation can be defined out, therefore, input node inputs every pass of non-standard text message accordingly
Connection information, calculates through the corresponding weight of each oriented vector, it is possible to export non-standard literary composition on output node accordingly
This information and each first alternative criteria text message are for describing the possibility degree of same medical treatment correlate.
Step 103, the historical context information passing through historical standard text message and historical standard text message, to described shellfish
This sorting algorithm model of leaf is trained.
It is understood that for the Bayesian Classification Arithmetic model set up, this model is used for limit each node it
Between dependency relation parameter can by training determine.For training the data of Bayesian Classification Arithmetic model can include going through
The historical context information of history received text information and historical standard text message.Wherein, the history of historical standard text message is closed
Having known corresponding relation between connection information and historical standard text message, this known corresponding relation represents historical standard literary composition
The historical context information of this information and historical standard text message are the text envelope becoming known for describing same medical treatment correlate
Breath.
When implementing, can believe based on the historical context of the historical standard text message in a large number with known corresponding relation
Breath and historical standard text message, with historical context information and these historical standard texts of these historical standard text messages
Information forms training dataset.By training dataset, Bayesian Classification Arithmetic model is trained, thus after training
Can determine in Bayesian Classification Arithmetic model for limiting the parameter of dependency relation between each node.That is, complete in training
Bayesian Classification Arithmetic model in, input node, output node and for limiting the parameter of dependency relation between each node
Be all it has been determined that.
For example, in some embodiments, the dependency relation between described Bayesian Classification Arithmetic model interior joint can
To be embodied by oriented vector, each oriented vector can have a corresponding weight, and now, each oriented vector is corresponding
Weight be i.e. the parameter for limiting dependency relation between each node.Visible, in this embodiment, step 103 is for example
Specifically can include：By related information and the historical standard text message of historical standard text message, calculate described Bayes
The weight of each oriented vector in sorting algorithm model.
Step 104, in the described Bayesian Classification Arithmetic model that completes of training, input described related information, export institute
Stating the correlation information between non-standard text message and described first alternative criteria text message, described correlation information represents
Described non-standard text and described first alternative criteria text message are used to describe the possibility of same medical treatment correlate.
When implementing, in training the Bayesian Classification Arithmetic model completing, every by non-standard text message
Related information is input in each input node accordingly, based on being used in Bayesian Classification Arithmetic model limiting phase between each node
The parameter of pass relation calculates, can obtain accordingly on each output node each first alternative criteria text message with nonstandard
Correlation information between quasi-text message.
For example, in some embodiments, the dependency relation between described Bayesian Classification Arithmetic model interior joint can
To be embodied by oriented vector, each oriented vector can have a corresponding weight, and now, each oriented vector is corresponding
Weight be i.e. the parameter for limiting dependency relation between each node.Visible, in this embodiment, step 104 is for example
Specifically can include：According to the corresponding letter in each input node of described Bayesian Classification Arithmetic model of described related information
Breath, by the weight of each oriented vector in described Bayesian Classification Arithmetic model, calculates described Bayesian Classification Arithmetic model
Information on each output node, as the corresponding first alternative criteria text message of each output node and described non-standard text envelope
Correlation information between breath.
In the present embodiment, the correlation information between the first alternative criteria text message and non-standard text message, tool
Body surface shows, the first alternative criteria text message and non-standard text message are for describing the possibility of same medical treatment correlate
Property size.For example, in some embodiments, the correlation between the first alternative criteria text message and non-standard text message
Information is concrete it may be that the first alternative criteria text message and non-standard text message are used for describing same medical treatment correlate
Probability.
Step 105, according to the correlation between described non-standard text message and described first alternative criteria text message
Information, selects the second alternative criteria text in described first alternative criteria text message.
When implementing, according to the correlation letter between non-standard text message and each first alternative criteria text message
Breath, can select most possible and non-standard text message for describing same doctor from each first alternative criteria text message
Treat one or more received text information of correlate as the second alternative criteria text message.
For example, in some embodiments, if the phase between the first alternative criteria text message and non-standard text message
Closing property information is specially the first alternative criteria text message with non-standard text message for describing same medical treatment correlate
Probability, then can choose with non-standard text message for describing the general of correlate from each first alternative criteria text message
Maximum front n the standard alternate text of rate is as the second alternative criteria text.Wherein, n belongs to positive integer.
Step 106, in described second alternative criteria text message determine target criteria text message, wherein, described mesh
Mark received text information is used for standardized corresponding relation for setting up with described non-standard text message.
It is understood that target criteria text message can be determined same for describing with non-standard text message
The text message of medical treatment correlate.
In some embodiments of the present embodiment, can be based on the operation of user in the second alternative criteria text message
Determine target criteria text message.Specifically, step 106 for example can include：Feed back described second alternative criteria text message；
In response to selection operation, by true for corresponding for described selection operation received text information in described second alternative criteria text message
It is set to described target criteria text message.More specifically, by not feeding back all of second alternative criteria text message,
After user performs selection operation to certain received text information in these the second alternative criteria text messages, in response to user
This selection operation, the received text information can chosen user is as target criteria text message.
It should be noted that the second alternative criteria text message is probably a received text information, it is also possible to multiple
Received text information.If the second alternative criteria text message is only a received text information, can be directly second alternative by this
Received text information is defined as target criteria text message, or also can should after the confirmation being responsive to user operates again
Second alternative criteria text message is defined as target criteria text message.If the second alternative criteria text message is multiple standard literary composition
This information, the standard literary composition that based on the selection operation of user, user can be selected in multiple second alternative criteria text messages
This information is as target criteria text message.
The various embodiments being provided by the present embodiment, in medical system for describe medical treatment correlate non-
Received text information, according to the related information of non-standard text message, by Bayesian Classification Arithmetic model, it can be deduced that nonstandard
Quasi-text message and each alternative received text information are used to describe the possibility of same medical treatment correlate, it is then possible to
Select a part of received text information based on possibility in all alternative received text information, thus only in this part
Received text information in just can find out the target criteria text message for standardizing this non-standard text message.Cause
This, be that the scope of non-standard text message lookup target criteria text message greatly reduces, and therefore, manual working is able to letter
Changing, artificial burden is mitigated.
It is understood that computer system can be with number of nodes to the processing load of Bayesian Classification Arithmetic model
Increase and be increased dramatically.In order to reduce the processing load of Bayesian Classification Arithmetic model, can by with non-standard text message
Carry out the mode of text matches, all alternative received text information filter out a part of received text information as first
Alternative criteria text message, then again by the process of Bayesian Classification Arithmetic model, is that non-standard text message determines target
Received text information.So not only may reduce artificial for non-standard text message lookup target criteria text message further
Scope, and the processing load to Bayesian Classification Arithmetic model for the computer system can be reduced, thus save process resource,
Improve processing speed and reduce processing delay.
Specifically, see Fig. 2, show the flow process signal of a kind of method of standardized text information in the embodiment of the present invention
Figure.In the present embodiment, described method for example specifically may comprise steps of：
Step 201, obtain non-standard text message and the related information of described non-standard text message, described non-standard literary composition
Originally it is for describing the text message of medical treatment correlate in medical system.
Step 202, described non-standard text message and the 3rd alternative criteria text message are carried out text matches.
Step 203, in described 3rd alternative criteria text message, choose the received text letter that described text matches obtains
Breath, as described first alternative criteria text message.
It is understood that before by Bayesian Classification Arithmetic model treatment, by non-standard text message and the 3rd
Alternative criteria text message carries out text matches and chooses the received text that coupling obtains from the 3rd alternative criteria text message
Information is as the first alternative criteria text message, such that it is able to carry out Bayes's classification calculation based on the first alternative criteria text message
Method model treatment.Visible, in medical system for describe medical treatment correlate non-standard text message, can first based on
Text matches is chosen a part of received text information in all alternative received text information and is carried out Bayesian Classification Arithmetic mould
Type process, is then processed by Bayesian Classification Arithmetic again and therefrom chooses a part of received text information further, so that
Processing of Bayesian Classification Arithmetic model determines target criteria in obtained received text information.So, not only contract further
The little scope manually searching target criteria text message for non-standard text message, and greatly reduce and need to pass through shellfish
The quantity of the received text information of this sorting algorithm model treatment of leaf.
In the present embodiment, described text matches has multiple possible embodiment.
For example, in some embodiments, complete non-standard text message can be standby as one coupling word and the 3rd
Received text information is selected to carry out text matches.Specifically, the received text information that described text matches obtains can include having
The received text information of described non-standard text message.For example, if not received text information is for " male ", received text information is
" male sex ", then this non-standard text message and this received text information meet the condition of text matches, that is, for non-standard literary composition
For this information " male ", received text information " male sex " belongs to the received text information that text matches obtains.
And for example, in some embodiments, by carrying out word segmentation processing, non-standard text message to non-standard text message
Corresponding lemma can carry out text matches as another coupling word and the 3rd alternative criteria text message.Specifically, step
203 for example can include：Word segmentation processing is carried out to described non-standard text message, and based on the result of described word segmentation processing, really
The fixed corresponding lemma of described non-standard text message；By corresponding for described non-standard text message lemma and described 3rd alternative mark
Quasi-text message carries out text matches；Wherein, the received text information that described text matches obtains includes having described non-standard
The received text information of any one lemma of text message.For example, it is assumed that non-standard text message is " internal medicine III ", this is nonstandard
The corresponding lemma of quasi-text message can include " internal medicine ", then received text information " Neurology " and this non-standard text message
Meet the condition of text matches, that is, for non-standard text message " internal medicine III ", received text information " Neurology "
Belong to the received text information that text matches obtains.
Furthermore, it is contemplated that some lemmas obtained by non-standard text message participle may be not easy to directly participate in
Text matches, therefore can first the lemma being obtained by non-standard text message participle be adjusted, then the word to obtain after adjustment
Unit carries out text matches as the corresponding lemma of non-standard text message.
For example, in several embodiments it is contemplated that potentially include to the lemma obtaining after non-standard text message participle
" ", " " etc. not there is the lemma of physical meaning, these can not had the lemma of physical meaning and be preset in and stop word dictionary
In, after each non-standard text message participle, first pass through and stop word dictionary to the word obtaining after non-standard text message participle
Unit carries out filtration and re-forms the corresponding lemma of non-standard text message, thus avoids wrapping in the corresponding lemma of non-standard text message
Include the lemma without physical meaning.Specifically, based on the result of word segmentation processing to non-standard text message corresponding lemma
Determine mode, for example, can include：The result of described word segmentation processing is got rid of and belongs to the default lemma stopping in word dictionary, and
Remaining lemma is defined as the corresponding lemma of described non-standard text message.
And for example, in several embodiments it is contemplated that potentially include difficulty to the lemma obtaining after non-standard text message participle
To match the lemma of received text information, if these lemmas have synonym and its synonym can match received text
Information, then can be difficult to these match the lemma of received text information and corresponding synonym is preset in synonym dictionary
In, after each non-standard text message participle, first pass through synonym dictionary to obtaining after non-standard text message participle
Lemma carries out synonym conversion and re-forms the corresponding lemma of non-standard text message, so that alternative received text information is more
Mate the corresponding lemma of non-standard text message all sidedly.Specifically, based on the result of word segmentation processing to non-standard text message
The determination mode of corresponding lemma, for example, can include：The knot with described word segmentation processing is searched in default synonym dictionary
Really corresponding synonym, and the result of described word segmentation processing and the synonym finding are defined as described non-standard text envelope
Cease corresponding lemma.
It is understood that above two embodiment can realize to non-standard text message in the way of in any combination
Corresponding lemma is determined.Stop word dictionary determine the corresponding lemma of non-standard text message for example, it is possible to be based only upon, now,
The corresponding lemma of non-standard text message does not includes the lemma not having physical meaning.And for example, synonymicon can be based only upon
Determine the corresponding lemma of non-standard text message, now, the corresponding lemma of non-standard text message not only includes non-standard literary composition
The lemma that this information participle obtains also includes the synonym of the lemma that non-standard text message participle obtains.For another example, can be simultaneously
Based on stopping word dictionary and synonymicon determines the corresponding lemma of non-standard text message, now, at non-standard text message pair
In the lemma answered, on the one hand do not include the lemma not having physical meaning, on the other hand not only include non-standard text message participle
The lemma obtaining also includes the synonym of the lemma that non-standard text message participle obtains.
In some embodiments of the present embodiment, in order to mitigate the processing load to text matches for the computer system, can
To filter out a part of received text information according to being sorted in all alternative received text information of non-standard text message
Carry out text matches.Specifically, before step 202, the method for the present embodiment can also include：Determine described non-standard text
The corresponding classification of information, and the received text information in described classification is defined as described 3rd alternative criteria text message.?
After determining the 3rd alternative criteria text message, step 202 can be entered.
More specifically, all alternative received text information can be preset in database based on classified index.Obtaining
After being currently needed for standardized non-standard text message, target can be determined based on the corresponding classification of non-standard text message
Classified index, then in database, find out the corresponding received text information of target classification index as the 3rd alternative criteria text
Information carries out text matches.
Step 204, based on described related information and the first alternative criteria text message, set up Bayesian Classification Arithmetic mould
Type, wherein, the input node of described Bayesian Classification Arithmetic model corresponds to described related information, described Bayesian Classification Arithmetic
The output node of model corresponds to described first alternative criteria text message.
Step 205, the historical context information passing through historical standard text message and historical standard text message, to described shellfish
This sorting algorithm model of leaf is trained.
Step 206, in the described Bayesian Classification Arithmetic model that completes of training, input described related information, export institute
Stating the correlation information between non-standard text message and described first alternative criteria text message, described correlation information represents
Described non-standard text and described first alternative criteria text message are used to describe the possibility of same medical treatment relevant issues.
Step 207, according to the correlation between described non-standard text message and described first alternative criteria text message
Information, selects the second alternative criteria text in described first alternative criteria text message.
Step 208, in described second alternative criteria text message determine target criteria text message, wherein, described mesh
Mark received text information is used for standardized corresponding relation for setting up with described non-standard text message.
It is understood that the step 201 mentioned by the present embodiment is corresponding to mentioned by the embodiment shown in earlier figures 1
Step 101, the step 204 mentioned by the present embodiment corresponds to the step 102 mentioned by embodiment shown in earlier figures 1, this reality
Execute the step 205 mentioned by example and correspond to the step 103 mentioned by embodiment shown in earlier figures 1, mentioned by the present embodiment
Step 206 corresponds to the step 104 mentioned by embodiment shown in earlier figures 1, and the step 207 mentioned by the present embodiment corresponds to
The step 105 mentioned by embodiment shown in earlier figures 1, the step 208 mentioned by the present embodiment is corresponding to shown in earlier figures 1
Step 106 mentioned by embodiment.The specific implementation of above-mentioned steps may refer to the introduction of the embodiment shown in Fig. 1, this
Embodiment does not repeats them here.
The embodiment being provided by the present embodiment, in medical system for describe medical treatment correlate non-standard
Text message, first can choose a part of received text information based on text matches in all alternative received text information and enter
Row Bayesian Classification Arithmetic model treatment, then processes, by Bayesian Classification Arithmetic, a part of standard of therefrom choosing further again
Text message, so that determining target criteria in processing of Bayesian Classification Arithmetic model in obtained received text information.
So, not only reduce the artificial scope searching target criteria text message for non-standard text message further, thus enter one
Step simplifies manual working and alleviates artificial burden, and greatly reduces and need by Bayesian Classification Arithmetic model treatment
The quantity of received text information, thus reduce the processing load of computer system, saved process resource, improve process
Speed simultaneously reduces processing delay.
See Fig. 3, show the structural representation of the device of a kind of standardized text information in the embodiment of the present invention.At this
In embodiment, described device for example can include：
Acquiring unit 301, for obtaining non-standard text message and the related information of described non-standard text message, described
Non-standard text is for describing the text message of medical treatment correlate in medical system；
Set up unit 302, for based on described related information and the first alternative criteria text message, set up Bayes's classification
Algorithm model, wherein, the input node of described Bayesian Classification Arithmetic model corresponds to described related information, and described Bayes divides
The output node of class algorithm model corresponds to described first alternative criteria text message；
Training unit 303, is used for the related information by historical standard text message and historical standard text message, to institute
State Bayesian Classification Arithmetic model to be trained；
Computing unit 304, in the described Bayesian Classification Arithmetic model that training completes, inputs described association letter
Breath, exports the correlation information between described non-standard text message and described first alternative criteria text message, described related
Property information represent that described non-standard text and described first alternative criteria text message are used to describe same medical treatment relevant issues
Probability；
First chooses unit 305, for according to described non-standard text message and described first alternative criteria text message
Between correlation information, in described first alternative criteria text message, select the second alternative criteria text；
First determining unit 306, for determining target criteria text message in described second alternative criteria text message,
Wherein, described target criteria text message is used for standardized corresponding relation for setting up with described non-standard text message.
Optionally, the dependency relation between described Bayesian Classification Arithmetic model interior joint can be come by oriented vector
Embodying, each oriented vector has a corresponding weight.
Furthermore, described training unit, specifically may be used for：By the related information of historical standard text message and
Historical standard text message, calculates the weight of each oriented vector in described Bayesian Classification Arithmetic model.
Furthermore, described computing unit, specifically may be used for：Corresponding described Bayes according to described related information
Information in each input node of sorting algorithm model, by the power of each oriented vector in described Bayesian Classification Arithmetic model
Weight, calculates the information on each output node of described Bayesian Classification Arithmetic model, corresponding first standby as each output node
Select the correlation information between received text information and described non-standard text message.
Optionally, described first determining unit specifically can include：
Feedback subelement, is used for feeding back described second alternative criteria text message；
First determination subelement, in response to selection operation, by described in described second alternative criteria text message
Selection operation corresponding received text information is defined as described target criteria text message.
Optionally, described device can also include：
Matching unit, after obtaining non-standard text message in described acquiring unit, by described non-standard text envelope
Breath and the 3rd alternative criteria text message carry out text matches；
Second chooses unit, for choosing, in described 3rd alternative criteria text message, the mark that described text matches obtains
Quasi-text message, as described first alternative criteria text message.
Optionally, the received text information that described text matches obtains can include having described non-standard text message
Received text information.
Optionally, described matching unit specifically can include：
Participle subelement, for carrying out word segmentation processing to described non-standard text message；
Second determination subelement, for the result based on described word segmentation processing, determines that described non-standard text message is corresponding
Lemma；
Coupling subelement, for by corresponding for described non-standard text message lemma and described 3rd alternative criteria text envelope
Breath carries out text matches；
Wherein, the received text information that described text matches obtains can include having appointing of described non-standard text message
The received text information of one lemma of meaning.
Optionally, described determination subelement specifically may be used for：
The result of described word segmentation processing is got rid of and belongs to the default lemma stopping in word dictionary, and by true for remaining lemma
It is set to the corresponding lemma of described non-standard text message.
Optionally, described determination subelement specifically may be used for：
The synonym corresponding with the result of described word segmentation processing is searched in default synonym dictionary, and by described point
The result that word is processed is defined as the corresponding lemma of described non-standard text message with the synonym finding.
Optionally, described device can also include：
Second determining unit, is used for determining the corresponding classification of described non-standard text message；
3rd determining unit, for being defined as described 3rd alternative criteria text by the received text information in described classification
Information.
The various embodiments being provided by the present embodiment, in medical system for describe medical treatment correlate non-
Received text information, according to the related information of non-standard text message, by Bayesian Classification Arithmetic model, it can be deduced that nonstandard
Quasi-text message and each alternative received text information are used to describe the possibility of same medical treatment correlate, it is then possible to
Select a part of received text information based on possibility in all alternative received text information, thus only in this part
Received text information in just can find out the target criteria text message for standardizing this non-standard text message.Cause
This, be that the scope of non-standard text message lookup target criteria text message greatly reduces, and therefore, manual working is able to letter
Changing, artificial burden is mitigated.
It should be noted that herein, the relational terms of such as first and second or the like is used merely to a reality
Body or operation separate with another entity or operating space, and deposit between not necessarily requiring or imply these entities or operating
Relation or order in any this reality.Term " includes ", "comprising" or its any other variant are intended to non-row
Comprising of his property, so that include that the process of a series of key element, method, article or equipment not only include those key elements, and
And also include other key elements being not expressly set out, or also include intrinsic for this process, method, article or equipment
Key element.In the case of there is no more restriction, the key element being limited by statement " including ... ", it is not excluded that including
Process, method, article or the equipment of stating key element there is also other identical element.
For system embodiment, owing to it corresponds essentially to embodiment of the method, so related part sees method in fact
The part executing example illustrates.System embodiment described above is only schematically, wherein said as separating component
The unit illustrating can be or may not be physically separate, as the parts that unit shows can be or also permissible
It not physical location, i.e. may be located at a place, or can be distributed on multiple NE yet.Can be according to actual
Need to select some or all of module therein to realize the purpose of the present embodiment scheme.Those of ordinary skill in the art are not
It in the case of paying creative work, is i.e. appreciated that and implements.
The above is only the detailed description of the invention of the application, it is noted that for the ordinary skill people of the art
For Yuan, on the premise of without departing from the application principle, can also make some improvements and modifications, these improvements and modifications also should
It is considered as the protection domain of the application.
Claims (10)
1. the method for a standardized text information, it is characterised in that include：
Obtaining non-standard text message and the related information of described non-standard text message, described non-standard text is medical system
In for describe medical treatment correlate text message；
Based on described related information and the first alternative criteria text message, set up Bayesian Classification Arithmetic model, wherein, described shellfish
The input node of this sorting algorithm model of leaf corresponds to described related information, the output node of described Bayesian Classification Arithmetic model
Corresponding to described first alternative criteria text message；
By historical context information and the historical standard text message of historical standard text message, to described Bayesian Classification Arithmetic
Model is trained；
In the described Bayesian Classification Arithmetic model that training completes, input described related information, export described non-standard text
Correlation information between information and described first alternative criteria text message, described correlation information represents described non-standard literary composition
This is used to describe the possibility of same medical treatment relevant issues with described first alternative criteria text message；
According to the correlation information between described non-standard text message and described first alternative criteria text message, described
One alternative criteria text message selects the second alternative criteria text；
Target criteria text message, wherein, described target criteria text envelope is determined in described second alternative criteria text message
Breath is used for standardized corresponding relation for setting up with described non-standard text message.
2. method according to claim 1, it is characterised in that
Dependency relation between described Bayesian Classification Arithmetic model interior joint is embodied by oriented vector, and each is oriented
Vector has a corresponding weight；
The described related information by historical standard text message and historical standard text message, to described Bayesian Classification Arithmetic
Model is trained, and is specially：By related information and the historical standard text message of historical standard text message, calculate described
The weight of each oriented vector in Bayesian Classification Arithmetic model；
Described in the described Bayesian Classification Arithmetic model that completes of training, input described related information, export described association letter
Correlation information between breath and described first alternative criteria text message, is specially：Corresponding in institute according to described related information
State the information in each input node of Bayesian Classification Arithmetic model, by each oriented arrow in described Bayesian Classification Arithmetic model
The weight of amount, calculates the information on each output node of described Bayesian Classification Arithmetic model, corresponding as each output node
Correlation information between first alternative criteria text message and described non-standard text message.
3. method according to claim 1, it is characterised in that described determination in described second alternative criteria text message
Target criteria text message, including：
Feed back described second alternative criteria text message；
In response to selection operation, in described second alternative criteria text message, corresponding for described selection operation received text is believed
Breath is defined as described target criteria text message.
4. method according to claim 1, it is characterised in that at the described non-standard text message and described non-standard of obtaining
After the related information of text message, also include：
Described non-standard text message and the 3rd alternative criteria text message are carried out text matches；
Choose, in described 3rd alternative criteria text message, the received text information that described text matches obtains, as described
One alternative criteria text message.
5. method according to claim 4, it is characterised in that the received text information that described text matches obtains includes tool
There is the received text information of described non-standard text message.
6. method according to claim 4, it is characterised in that described by described non-standard text message and the 3rd alternative mark
Quasi-text message carries out text matches, including：
Word segmentation processing is carried out to described non-standard text message, and based on the result of described word segmentation processing, determines described non-standard
The corresponding lemma of text message；
Corresponding for described non-standard text message lemma and described 3rd alternative criteria text message are carried out text matches；
Wherein, the received text information that described text matches obtains includes any one word with described non-standard text message
The received text information of unit.
7. method according to claim 6, it is characterised in that the described result based on described word segmentation processing, determines described
The corresponding lemma of non-standard text message, including：
The result of described word segmentation processing is got rid of and belongs to the default lemma stopping in word dictionary, and remaining lemma is defined as
The corresponding lemma of described non-standard text message.
8. method according to claim 6, it is characterised in that the described result based on described word segmentation processing, determines described
The corresponding lemma of non-standard text message, including：
The synonym corresponding with the result of described word segmentation processing is searched in default synonym dictionary, and by described participle
The result of reason is defined as the corresponding lemma of described non-standard text message with the synonym finding.
9. method according to claim 4, it is characterised in that also include：
Determine the corresponding classification of described non-standard text message, and the received text information in described classification is defined as described the
Three alternative criteria text messages.
10. the device of a standardized text information, it is characterised in that include：
Acquiring unit, for obtaining non-standard text message and the related information of described non-standard text message, described non-standard
Text is for describing the text message of medical treatment correlate in medical system；
Set up unit, for based on described related information and the first alternative criteria text message, set up Bayesian Classification Arithmetic mould
Type, wherein, the input node of described Bayesian Classification Arithmetic model corresponds to described related information, described Bayesian Classification Arithmetic
The output node of model corresponds to described first alternative criteria text message；
Training unit, is used for the related information by historical standard text message and historical standard text message, to described pattra leaves
This sorting algorithm model is trained；
Computing unit, in the described Bayesian Classification Arithmetic model that training completes, inputs described related information, exports institute
Stating the correlation information between non-standard text message and described first alternative criteria text message, described correlation information represents
Described non-standard text and described first alternative criteria text message are used to describe the possibility of same medical treatment relevant issues；
First chooses unit, for according to the phase between described non-standard text message with described first alternative criteria text message
Closing property information, selects the second alternative criteria text in described first alternative criteria text message；
First determining unit, for determining target criteria text message, wherein, institute in described second alternative criteria text message
State target criteria text message for setting up for standardized corresponding relation with described non-standard text message.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610766697.2A CN106446025B (en) | 2016-08-30 | 2016-08-30 | A kind of method and apparatus of standardized text information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610766697.2A CN106446025B (en) | 2016-08-30 | 2016-08-30 | A kind of method and apparatus of standardized text information |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106446025A true CN106446025A (en) | 2017-02-22 |
| CN106446025B CN106446025B (en) | 2019-10-11 |
Family
ID=58090430
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610766697.2A Active CN106446025B (en) | 2016-08-30 | 2016-08-30 | A kind of method and apparatus of standardized text information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106446025B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784057A (en) * | 2017-03-03 | 2018-03-09 | 平安医疗健康管理股份有限公司 | Medical data matching process and device |
| CN109766440A (en) * | 2018-12-17 | 2019-05-17 | 航天信息股份有限公司 | A kind of method and system for for the determining default categories information of object text description |
| CN111046657A (en) * | 2019-12-04 | 2020-04-21 | 东软集团股份有限公司 | Method, device and equipment for realizing text information standardization |
| CN112700881A (en) * | 2020-12-29 | 2021-04-23 | 医渡云（北京）技术有限公司 | Text standardization processing method and device, electronic equipment and computer medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007179181A (en) * | 2005-12-27 | 2007-07-12 | Oki Electric Ind Co Ltd | Slip classification device |
| CN102521517A (en) * | 2011-12-20 | 2012-06-27 | 深圳市人民医院 | System and method for discriminatory analysis of breast tumors |
| CN103514375A (en) * | 2013-10-10 | 2014-01-15 | 中国中医科学院 | Electronic medical record rapid recording system based on standard terminology |
-
2016
- 2016-08-30 CN CN201610766697.2A patent/CN106446025B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007179181A (en) * | 2005-12-27 | 2007-07-12 | Oki Electric Ind Co Ltd | Slip classification device |
| CN102521517A (en) * | 2011-12-20 | 2012-06-27 | 深圳市人民医院 | System and method for discriminatory analysis of breast tumors |
| CN103514375A (en) * | 2013-10-10 | 2014-01-15 | 中国中医科学院 | Electronic medical record rapid recording system based on standard terminology |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784057A (en) * | 2017-03-03 | 2018-03-09 | 平安医疗健康管理股份有限公司 | Medical data matching process and device |
| CN107784057B (en) * | 2017-03-03 | 2020-07-28 | 平安医疗健康管理股份有限公司 | Medical data matching method and device |
| CN109766440A (en) * | 2018-12-17 | 2019-05-17 | 航天信息股份有限公司 | A kind of method and system for for the determining default categories information of object text description |
| CN109766440B (en) * | 2018-12-17 | 2023-09-01 | 航天信息股份有限公司 | Method and system for determining default classification information for object text description |
| CN111046657A (en) * | 2019-12-04 | 2020-04-21 | 东软集团股份有限公司 | Method, device and equipment for realizing text information standardization |
| CN111046657B (en) * | 2019-12-04 | 2023-10-13 | 东软集团股份有限公司 | Method, device and equipment for realizing text information standardization |
| CN112700881A (en) * | 2020-12-29 | 2021-04-23 | 医渡云（北京）技术有限公司 | Text standardization processing method and device, electronic equipment and computer medium |
| CN112700881B (en) * | 2020-12-29 | 2022-04-08 | 医渡云（北京）技术有限公司 | Text standardization processing method and device, electronic equipment and computer medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106446025B (en) | 2019-10-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106446025A (en) | Method and device for standardizing text information | |
| Qin et al. | Frank aggregation operators and their application to hesitant fuzzy multiple attribute decision making | |
| CN104899267B (en) | A kind of integrated data method for digging of social network sites account similarity | |
| WO2020124856A1 (en) | Diagnosis standardization method and device based on word vectors | |
| Silvapulle | On the existence of maximum likelihood estimators for the binomial response models | |
| CN103530334B (en) | Based on the data matching system and method for comparing template | |
| Lindley | The Bayesian analysis of contingency tables | |
| CN110459321A (en) | A kind of aided diagnosis of traditional Chinese medicine system based on card element | |
| CN104008106B (en) | A kind of method and device obtaining much-talked-about topic | |
| WO2017066202A1 (en) | Systems and method for dynamic autonomous transactional identity management | |
| Zhao et al. | Dual hesitant fuzzy information aggregation with Einstein t-conorm and t-norm | |
| CN104572733B (en) | The method and device of user interest labeling | |
| CN109657116A (en) | A kind of public sentiment searching method, searcher, storage medium and terminal device | |
| CN105975794A (en) | Breast cancer chemotherapy regimen recommendation method based on weighted KNN | |
| Joshi et al. | Exponential Jensen intuitionistic fuzzy divergence measure with applications in medical investigation and pattern recognition | |
| CN108304509A (en) | A kind of comment spam filter method for indicating mutually to learn based on the multidirectional amount of text | |
| CN111831871A (en) | Method, device and equipment for recommending air conditioner working mode | |
| Zhao et al. | Model for multiple attribute decision making Based on the Einstein correlated Information Fusion with Hesitant Fuzzy Information | |
| Meng et al. | Induced Atanassov’s interval-valued intuitionistic fuzzy hybrid Choquet integral operators and their application in decision making | |
| CN109902157A (en) | A kind of training sample validation checking method and device | |
| CN104820775A (en) | Discovery method of core drug of traditional Chinese medicine prescription | |
| Kuelbs et al. | Concerns with functional depth | |
| CN107992563A (en) | A kind of recommendation method and system of user's browsing content | |
| CN110516078A (en) | Alignment schemes and device | |
| Yu et al. | Empirical likelihood methods based on influence functions |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |