CN106021239B - A kind of translation quality real-time estimating method - Google Patents
A kind of translation quality real-time estimating method Download PDFInfo
- Publication number
- CN106021239B CN106021239B CN201610272936.9A CN201610272936A CN106021239B CN 106021239 B CN106021239 B CN 106021239B CN 201610272936 A CN201610272936 A CN 201610272936A CN 106021239 B CN106021239 B CN 106021239B
- Authority
- CN
- China
- Prior art keywords
- translation
- user
- sentence
- engine
- quality
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/40—Processing or translation of natural language
- G06F40/51—Translation evaluation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/40—Processing or translation of natural language
- G06F40/58—Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation
Abstract
The invention discloses a kind of translation quality real-time estimating methods, include the following steps:According to the translation request of user, the sentence that user is initiated is received;Multiple translation engines receive the Client-initiated sentence, the frequency collected the sentence according to the respective total translation number and user that count different engines, play the sentence, it calculates the weighting frequency of usage of the translation result of translation engine and calculates frequency of usage per capita and divide as quality, and the ratio that the quality point for accounting for all engines using quality point sums it up is probability, the highest translation engine of probability is then as the optimal translation engine of this statement requests;Using above-mentioned obtained optimal translation engine as asking the preferential of this sentence to use engine, the translation engine is dispatched to translate the requested sentence of user, obtains translation result.Through the invention, to the translation quality of multiple engines, the real-time quality evaluation automated provides optimal translation engine result for different translation sentences.
Description
Technical field
The present invention relates to translation quality to evaluate field, specifically a kind of translation quality real-time estimating method.
Background technology
People using translation tool when, typically realized by a kind of translation engine, and manual identified this translation knot
The accuracy of fruit, if feeling, translation result is not ideal enough, can attempt to change a kind of translation engine and be attempted again, until that can not obtain
Until more excellent result.The mode of this manual switching translation engine, the efficiency for obtaining optimal translation result is low, and time-consuming, especially
It cannot be satisfied the demand of real time translation.
So, multiple translation engines are integrated, the higher translation result of coverage is provided to the user, can partly solve above-mentioned
Technical problem is particularly suitable for overseas trip user.In order to provide the translation result of high-quality to overseas trip user, promotes translation and cover
Cover degree can integrate multiple translation engines, it is ensured that the coverage of translation, but the same sentence to be translated, for limited common-use words
The use of sentence, can ask expert to evaluate, and to select and determine optimal translation result, and this mode obviously can not be coped with again
Miscellaneous translation demand.The mode of existing experts' evaluation, it is maximum the disadvantage is that can not accomplish quick real-time response, for magnanimity
Sentence is translated, manpower evaluation is of high cost, can not accomplish the degree for automating and quantifying, also can not sustained improvement evaluation method.Separately
A kind of mode is to be sampled quality evaluation to each translation engine, evaluates the synthesized translation quality of some engine.But it can not do
To the translation quality for being accurate to different sentences.
It is more intractable to be, for real-time translation on line, how the translation result of the multiple engines of quantitative evaluation, be to obtain
The optimal necessary technical problems to be solved of translation result institute, this is just at a very urgent technical problem.The prior art is often
This technical problem is solved by metalanguage structure, such as number of patent application disclosed in State Intellectual Property Office is
201410461334.9 application for a patent for invention machine translation method and system, but this can not equally adapt to different contexts, field
The speech habits etc. of conjunction and user, therefore above-mentioned technical problem can not be solved.
Invention content
In view of this, the present invention existing technology that can not efficiently obtain optimal translation result for the above-mentioned prior art
Problem provides a kind of translation quality real-time estimating method integrating multiple translation engines and real-time online assessment, to obtain most
Excellent translation result.
Technical solution of the invention is to provide a kind of translation quality real-time estimating method of following steps, including with
Lower step:
Step 1:According to the translation request of user, the sentence that user is initiated is received;
Step 2:Integrated multiple translation engines receive the Client-initiated sentence, according to the different engines of statistics
The frequency that respective total translation number and user collect the sentence, play the sentence, calculates the translation knot of corresponding translation engine
Frequency of usage divides the quality for accounting for all engines to divide adduction as quality point, and with quality per capita for the weighting frequency of usage of fruit and calculating
Ratio be probability, the highest translation engine of probability is then as the optimal translation engine of this statement requests;
Step 3:Using above-mentioned obtained optimal translation engine as asking the preferential of this sentence to use engine, dispatches this and turn over
Engine is translated to translate the requested sentence of user, translation result is obtained, returns to user.
Above step is used, compared with prior art, the present invention haing the following advantages:Using the present invention, sent out for user
The particular statement risen, each translation engine are not quite similar, and frequency is used by calculating the weighting of translation result of corresponding translation engine
It is secondary and calculate per capita that frequency of usage is as quality point, as the evaluation to different translation engines for the translation result of the sentence,
The translation engine used as priority with the person of having higher rating completes the selection to the optimal translation engine of the sentence, to obtain most
Excellent translation result, returns to user.Invention introduces the frequencys as Appreciation gist, is not necessarily to the language construction of anolytic sentence, energy
It is enough fast and effeciently to obtain optimal translation result, the translation quality of multiple engines is carried out certainly by this systematic method
The real-time quality evaluation of dynamicization provides optimal translation engine result for different translation sentences.
Preferably, after completing step 3, turning over for the sentence initiated user under the optimal translation engine is updated
The frequency is translated, lifts the user of statement translation request in the follow-up behavior using translation result, is sent to quality as daily record and comments
Database is estimated, for subsequent evaluation and the dynamic corrections to selecting optimal translation engine.
Preferably, quality assessment data library is according to the log information reported, and the location information of user is acquired to judge
User location and whether at abroad, in conjunction with user's geographical location circumstances as weigh translation quality cofactor, Bian Keshe
Determine regional extent, obtains the optimal translation engine of the regional extent.After being superimposed location information, it more can accurately reflect translation engine
Translation quality because translation quality it is often related with the usage scenario of sentence.
Preferably, for the identical translation request of different user, different user obtains translation result using different engines,
And by analyzing processing subsequent user to translation result in different time, the occupation mode and the frequency of different geographical, quantitative evaluation
Different translation engines are to the translation quality of this sentence, to provide the best translation result of this sentence for other users.Pass through
The mode of this comprehensive analysis and assessment can obtain more preferably translation result.
Description of the drawings
Fig. 1 is the step schematic diagram of translation quality real-time estimating method;
Specific implementation mode
The invention will be further described in the following with reference to the drawings and specific embodiments, but the present invention is not restricted to these
Embodiment.
The present invention covers any replacement, modification, equivalent method and scheme made in the spirit and scope of the present invention.For
So that the public is had thorough understanding to the present invention, is described in detail concrete details in following present invention preferred embodiment, and
Description without these details can also understand the present invention completely for a person skilled in the art.
As shown in Figure 1, a kind of translation quality real-time estimating method of the present invention, includes the following steps:
Step 1:According to the translation request of user, the sentence that user is initiated is received;
Step 2:Integrated multiple translation engines receive the Client-initiated sentence, according to the different engines of statistics
The frequency that respective total translation number and user collect the sentence, play the sentence, calculates the translation knot of corresponding translation engine
Frequency of usage divides the quality for accounting for all engines to divide adduction as quality point, and with quality per capita for the weighting frequency of usage of fruit and calculating
Ratio be probability, the highest translation engine of probability is then as the optimal translation engine of this statement requests;
In step 2, data volume, the threshold value of quality can be set, after reaching, optimal translation is not returned to further according to probability and draws
It holds up, because having obtained reliable as a result, need not then repeat every time, improves response speed.
Step 3:Using above-mentioned obtained optimal translation engine as asking the preferential of this sentence to use engine, dispatches this and turn over
Engine is translated to translate the requested sentence of user, translation result is obtained, returns to user.
After completing step 3, the translation frequency for the sentence initiated user under the optimal translation engine is updated, is carried
The user of statement translation request is played in the follow-up behavior using translation result, quality assessment data library is sent to as daily record,
For subsequent evaluation and the dynamic corrections to selecting optimal translation engine.
Quality assessment data library acquires the location information of user to judge user location according to the log information reported
And whether at abroad, in conjunction with user's geographical location circumstances as weigh translation quality cofactor, can setting regions range,
Obtain the optimal translation engine of the regional extent.After being superimposed location information, it more can accurately reflect the translation quality of translation engine,
Because the quality of translation is often related with the usage scenario of sentence, usage scenario and geographical location often determine use habit,
Also different communicative habits be there is.
For the identical translation request of different user, different user obtains translation result using different engines, and by dividing
Analysis processing subsequent user is to translation result in different time, the occupation mode and the frequency of different geographical, the translation of quantitative evaluation difference
Engine is to the translation quality of this sentence, to provide the best translation result of this sentence for other users.Pass through this synthesis
The mode of analysis and assessment, can obtain more preferably translation result.
It should be noted that quality assessment data library mentioned above refers to the quality evaluation library in Fig. 1.
Steps are as follows for the concrete principle of the present invention:In particular for the people for being in overseas travelling, sentence is initiated to client
The request of translation obtains its geographical location information, according to the use habit in this area, and combines and is turned over to respective request sentence
Translate the frequency and the follow-up behavior of user, provide optimal translation engine to the user, obtain optimal translation result, and mostly because
Under the synthesis of element, dynamic corrections can be carried out to the selection of translation engine.
The more specific application scenarios of the present invention are following (being realized by mobile terminal APP):
Step 1:Client initiates the translation request of sentence.
Step 2:This sentence of translation engine scheduling request preferentially uses engine.
Step 3:Quality evaluation library is carried out according to the respective total translation number and subsequent user foreign countries that count different engines
Collection sentence, play sentence the frequency, calculate this translation result weighting frequency of usage calculate per capita frequency of usage as matter
Amount point, and the ratio for accounting for the quality point adduction of all engines using quality point returns to the translation engine scheduling person of dividing and corresponds to as probability
Engine, and update the translation frequency of this lower engine to this sentence.
Step 4:According to engine is returned, the translation result of corresponding engine is asked.
Step 5:Corresponding engine returns to translation result and is dispatched to translation engine.
Step 6:Translation result is returned to client by translation engine scheduling.
Step 7:User uses the follow-up behavior of translation result in client, and quality evaluation engine is sent to as daily record.
Step 8:The log information that quality evaluation engine is reported according to client calculates user location and whether in state
Outside, and this daily record is added to quality evaluation library, and calculate user's frequency of usage data
Only preferred embodiments of the present invention are described above, but are not to be construed as limiting the scope of the invention.This
Invention is not only limited to above example, and concrete structure is allowed to vary.In short, all guarantors in independent claims of the present invention
Various change is within the scope of the invention made by shield range.
Claims (3)
1. a kind of translation quality real-time estimating method, includes the following steps:
Step 1:According to the translation request of user, the sentence that user is initiated is received;
Step 2:Integrated multiple translation engines receive the Client-initiated sentence, according to the respective of the different engines of statistics
Total translation number and user's frequency for collecting the sentence, playing the sentence, calculate the translation result of corresponding translation engine
It weights frequency of usage and calculates frequency of usage per capita and divide as quality, and divide the ratio for the quality point adduction for accounting for all engines with quality
Example is probability, and the highest translation engine of probability is then as the optimal translation engine of this statement requests;
Step 3:Using the optimal translation engine as asking the preferential of this sentence to use engine, the translation engine is dispatched to translate
The requested sentence of user, obtains translation result, returns to user;
After completing step 3, the translation frequency for the sentence initiated user under the optimal translation engine is updated, this is lifted
The user of statement translation request is sent to quality assessment data library as daily record, is used in the follow-up behavior using translation result
Subsequent evaluation and the dynamic corrections to selecting optimal translation engine.
2. a kind of translation quality real-time estimating method according to claim 1, it is characterised in that:Quality assessment data library root
According to the log information reported, and the location information of user is acquired to judge user location and whether at abroad, in conjunction with user
Manage situation as measurement translation quality cofactor, can setting regions range, obtain the optimal of the regional extent and turn over
Translate engine.
3. a kind of translation quality real-time estimating method according to claim 1, it is characterised in that:For the phase of different user
Same translation request, different user obtain translation result using different engines, and by analyzing processing subsequent user to translation result
In different time, the occupation mode and the frequency of different geographical, quantitative evaluation difference translation engine to the translation quality of this sentence, from
And the best translation result of this sentence is provided for other users;By way of this comprehensive analysis and assessment, it can obtain
More preferably translation result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610272936.9A CN106021239B (en) | 2016-04-29 | 2016-04-29 | A kind of translation quality real-time estimating method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610272936.9A CN106021239B (en) | 2016-04-29 | 2016-04-29 | A kind of translation quality real-time estimating method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106021239A CN106021239A (en) | 2016-10-12 |
CN106021239B true CN106021239B (en) | 2018-10-26 |
Family
ID=57081012
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610272936.9A Active CN106021239B (en) | 2016-04-29 | 2016-04-29 | A kind of translation quality real-time estimating method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106021239B (en) |
Families Citing this family (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107894983A (en) * | 2017-11-08 | 2018-04-10 | 深圳市沃特沃德股份有限公司 | The switching method and device of engine |
CN107979856B (en) * | 2017-11-22 | 2020-10-27 | 深圳市沃特沃德股份有限公司 | Method and device for connecting engines |
CN108319591A (en) * | 2018-02-05 | 2018-07-24 | 深圳市沃特沃德股份有限公司 | Realize the method, apparatus and speech translation apparatus of voiced translation |
CN109118109B (en) * | 2018-08-31 | 2021-06-01 | 传神语联网网络科技股份有限公司 | ETM-based quality assessment |
CN109299737B (en) * | 2018-09-19 | 2021-10-26 | 语联网(武汉)信息技术有限公司 | Translator gene selection method and device and electronic equipment |
CN109214011A (en) * | 2018-09-19 | 2019-01-15 | 深圳市合言信息科技有限公司 | It is a kind of by user feedback come the cognitive engine selection strategy of self-perfection |
US11074420B2 (en) | 2019-06-25 | 2021-07-27 | International Business Machines Corporation | Self-learning input-driven brokering of language-translation engines |
CN110532574A (en) * | 2019-08-20 | 2019-12-03 | 语联网(武汉)信息技术有限公司 | MT engine selection method and device |
CN110502762B (en) * | 2019-08-27 | 2023-07-28 | 北京金山数字娱乐科技有限公司 | Translation platform and management method thereof |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1465018A (en) * | 2000-05-11 | 2003-12-31 | 南加利福尼亚大学 | Machine translation mothod |
CN1530861A (en) * | 2003-02-24 | 2004-09-22 | Language translating method and system | |
WO2010052696A1 (en) * | 2008-11-05 | 2010-05-14 | Ubstream Ltd. | Session initiation protocol in multi-cast/unicast system for connection between users behind nat and connection between server and users behind nat |
CN102253930A (en) * | 2010-05-18 | 2011-11-23 | 腾讯科技(深圳)有限公司 | Method and device for translating text |
CN102446169A (en) * | 2010-10-13 | 2012-05-09 | 张龙哺 | Translation system by utilizing online translation services |
CN103914446A (en) * | 2013-01-08 | 2014-07-09 | 崴强科技股份有限公司 | Translation method and system |
CN104011712A (en) * | 2011-06-24 | 2014-08-27 | 谷歌公司 | Evaluating query translations for cross-language query suggestion |
CN104090870A (en) * | 2014-06-26 | 2014-10-08 | 武汉传神信息技术有限公司 | Pushing method of online translation engines |
CN105051728A (en) * | 2013-01-03 | 2015-11-11 | 时新公司 | Data base query translation system |
-
2016
- 2016-04-29 CN CN201610272936.9A patent/CN106021239B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1465018A (en) * | 2000-05-11 | 2003-12-31 | 南加利福尼亚大学 | Machine translation mothod |
CN1530861A (en) * | 2003-02-24 | 2004-09-22 | Language translating method and system | |
WO2010052696A1 (en) * | 2008-11-05 | 2010-05-14 | Ubstream Ltd. | Session initiation protocol in multi-cast/unicast system for connection between users behind nat and connection between server and users behind nat |
CN102253930A (en) * | 2010-05-18 | 2011-11-23 | 腾讯科技(深圳)有限公司 | Method and device for translating text |
CN102446169A (en) * | 2010-10-13 | 2012-05-09 | 张龙哺 | Translation system by utilizing online translation services |
CN104011712A (en) * | 2011-06-24 | 2014-08-27 | 谷歌公司 | Evaluating query translations for cross-language query suggestion |
CN105051728A (en) * | 2013-01-03 | 2015-11-11 | 时新公司 | Data base query translation system |
CN103914446A (en) * | 2013-01-08 | 2014-07-09 | 崴强科技股份有限公司 | Translation method and system |
CN104090870A (en) * | 2014-06-26 | 2014-10-08 | 武汉传神信息技术有限公司 | Pushing method of online translation engines |
Also Published As
Publication number | Publication date |
---|---|
CN106021239A (en) | 2016-10-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106021239B (en) | A kind of translation quality real-time estimating method | |
KR102097426B1 (en) | Analysis method of fluidized population information capable of providing real-time fluidized population data by pcell algorithm | |
Beige et al. | Interdependencies between turning points in life and long-term mobility decisions | |
De Beer et al. | Overcoming the problems of inconsistent international migration data: A new method applied to flows in Europe | |
CN104980950B (en) | A kind of network optimization service device, the mobile device and system for realizing the network optimization | |
CN108874959B (en) | User dynamic interest model building method based on big data technology | |
CN105007171A (en) | User data analysis system and method based on big data in communication field | |
CN106487664B (en) | Information recommendation method and device and mobile terminal | |
JP2012198870A (en) | Device and method for recognizing user behavior | |
CN109816226B (en) | Method and system for evaluating disaster prevention and risk avoidance capabilities of park green land | |
CN107026750B (en) | User Internet QoE evaluation method and device | |
CN111866776A (en) | Population measurement and calculation method and device based on mobile phone signaling data | |
KR20140017489A (en) | Computer system for calculating country-specific fees | |
CN110727752B (en) | Position fingerprint database processing method, device and computer readable storage medium | |
CN109995549B (en) | Method and device for evaluating flow value | |
CN108319591A (en) | Realize the method, apparatus and speech translation apparatus of voiced translation | |
CN106604297A (en) | Method and equipment for optimizing longitude and latitude data in center of sector of base station | |
WO2017128502A1 (en) | Method and device for collecting fingerprint positioning data | |
JP5331719B2 (en) | Method, apparatus and program for determining action situation | |
Chung et al. | Identification of a city‟ s soundscape using soundwalks | |
CN110335626A (en) | Age recognition methods and device, storage medium based on audio | |
CN101620597A (en) | Method for analyzing product association of data service in mobile communication industry | |
CN103188629B (en) | Flow bootstrap technique between a kind of networks with different systems and device | |
CN109005552A (en) | The method for accurately assessing wireless network based on LTE MR data | |
CN110400160B (en) | Method and device for identifying competitive product user, electronic equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
CB03 | Change of inventor or designer information | ||
CB03 | Change of inventor or designer information |
Inventor after: Han Xin Inventor after: Zhang Chu Inventor after: Chen Gang Inventor before: Chen Gang |
|
GR01 | Patent grant | ||
GR01 | Patent grant |