CN103136191A - Automatic extracting method of word with single character in electronic commerce dictionary - Google Patents
Automatic extracting method of word with single character in electronic commerce dictionary Download PDFInfo
- Publication number
- CN103136191A CN103136191A CN2013100798089A CN201310079808A CN103136191A CN 103136191 A CN103136191 A CN 103136191A CN 2013100798089 A CN2013100798089 A CN 2013100798089A CN 201310079808 A CN201310079808 A CN 201310079808A CN 103136191 A CN103136191 A CN 103136191A
- Authority
- CN
- China
- Prior art keywords
- word
- potential
- words
- count
- effective
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Machine Translation (AREA)
Abstract
The invention discloses an automatic extracting method of a word with a single character in an electronic commerce dictionary. The method comprises the following steps that A1, linguistic data preparation and pretreatment are carried out; A2, combination of all possible potential words is obtained; A3, filtration of invalid words based on distributed dispersion is carried out; A4, the filtration of invalid words is carried out; A5, regular filtration is carried out; A6, compensation of significant words is carried out; A7, frequency statistic of the single character serving as a substring and appearing in a longer word is carried out; A8, overlap type statistical error correction is compensated, and repeatedly recorded words with single characters serving as substrings and appearing in longer words are subtracted; A9, frequency statistic calculation of the words with the single characters independently appearing is carried out; and A10, filtration is finished, and low frequency with little appearing times in all linguistic data are gotten rid of. The dictionary generated by the method can be widely used for searching, semanteme word segmentation, recommendation, weight calculation and the like in electronic commerce application.
Description
Technical field
The main Electronic Commerce of the present invention field.Relate in particular to the Automatic Extraction method of monosyllabic word in a kind of ecommerce dictionary
Background technology
Along with the develop rapidly of infotech, various fresh words emerge in an endless stream, and especially in e-commerce field, various new terms occur more frequent.Mainstream applications adopts manual the generation or the simple statistics generation mostly at present, also has part to adopt the method for machine learning to remove to collect entry and forms dictionary.
The ecommerce dictionary is the basis that e-commerce website is used, monosyllabic word refer to the word that independent meaning is arranged that consisted of by a word as " water ", " tin ", " with ", " very little " etc., be indispensable composition at many-sided monosyllabic words such as e-commerce website search engine, commending system, Words partition systems.Yet, because the monosyllabic word character length is short, thereby easily be left in the basket in Automatic Extraction and participle process, how accurately extract monosyllabic word and be one from long section text than the challenging work of tool.
Present obtaining for monosyllabic word, can't distinguish word based on the method for machine learning merely is autonomous word or the substring of other word, therefore be all by collecting some sample files, then removing by hand to obtain the word of single character and the probability that single character occurs as autonomous word.The shortcoming of this method mainly comprises: the one, and the artificial treatment workload is large, and efficient is low; The 2nd, the artificial treatment subjectivity is strong, and standard is difficult to unified; The 3rd, whether single character can independently exist and domain-specific, and artificial judgment has no basis.Such as " bear ", be word in general dictionary, but be not that (based on the mass data of our statistics, bear occurred in the e-commerce platform commodity, such as bear gall, but never independently occurred in ecommerce.)
Summary of the invention
The present invention is directed to the e-commerce field characteristics, propose the automatic obtaining method of monosyllabic word in a kind of e-commerce field.
Technical scheme of the present invention is as follows:
A kind of Automatic Extraction method of monosyllabic word in ecommerce dictionary comprises the following steps:
A1, language material are prepared and pre-service;
A2, language material is carried out with going forward one by one of redundant data exhaustive, obtain all possible potential word combination; Employing is gone forward one by one exhaustive method by the exhaustive various minutes word combinations in effective word maximum length+1, adds up simultaneously the frequency that the combination of various individual characters and multiword occurs, and forms a complete possible potential set of words that comprises.Introduction is used for filtering the cutting data boundary greater than the invalid potential word of effective word length;
A3, be that the dispersion that distributes in the shortest length potential word of this word as prefix/postfix at a group based on shorter potential word more than 2 is carried out the filtration of invalid word for length;
A4, be that probability that more than 2, potential word independently occurs carries out invalid word and filters based on length;
Shorter potential word in comprising its potential word of the shortest length occurrence number greater than certain threshold values, and long potential word does not meet the canonical filtercondition, short potential word count subtracts the difference of long potential word count, be that 0 short potential word is directly deleted for difference, otherwise short potential word count is described difference;
A5, be that potential word 2 or more carries out canonical and filters for remaining length after filtering through two steps of A3, A4, prefix/postfix/tundish is contained in predefined set, and remainder be all effective word after above filtration, deletes this potential word; Be included in predefined set for prefix/postfix simultaneously, and not in the set of the word of exception;
A6, appear in unique context mistake and delete the compensation that length is effective word 2 or more; At first appear at the potential word of mistake deletion in unique context based on following condition judgment:
(1) this potential word does not meet the canonical filtercondition;
(2) all potential words that comprise this potential word all have been filtered, and how many length of tube is not;
(3) count of this potential word is identical with the potential word count that all comprise it;
Secondly, find the longest potential word that comprises this mistake stop word; In the situation that a potential word appears at the longest a plurality of potential words, do splicing, again reduce the cutting border; Then the cutting unit of the longest above potential word/reduction just carried out/reverse maximum coupling participle based on existing effective word, if the combination of cutting appears in effective word dictionary as a potential word, continue to scan backward character string, be not less than 2 the longest cutting combination and join in effective potential word dictionary for not appearing at length in potential word dictionary, frequency is the original frequency of the full cutting of this combination; At last for not comprising effective word in the longest potential word, keep the longest potential word, add in effective word dictionary;
The frequency statistics that A7, single character occur as substring in long word more: find all other longer effective words that comprise this word, longly process successively from being short to, delete the more long word that all comprise current word;
A8, compensation overlap type mistake statistical correction cut the frequency that the monosyllabic word that repeats to add up occurs as substring in long word more;
1) for the original count that remains in all steps A 7 after word obtains full cutting end, as the current count of word;
2) find all with the potential word of current word as prefix and suffix for residue word in steps A 7, as be divided into two groups of prefix and suffix, respectively get a combination in twos in two groups;
3) to 2) in the combined result that generates coupling one by one in the result of steps A 7, for the combination that the match is successful, deduct the original count of combination with one of them current count of two words of this combination of composition;
A9, the statistical computation of the monosyllabic word independence frequency of occurrences, the sum frequency of the monosyllabic word that obtains exhaustive from going forward one by one cuts the count that step 8 is finally obtained;
A10, filtration finish, and reject occurrence number word low-frequency word seldom in all language materials.
E-commerce field new term renewal frequency is high, and not only workload is large and renewal speed is slow to adopt traditional manual mode to remove to extract monosyllabic word.The one, can the Obtaining Accurate Electronic Commerce field can self-existent monosyllabic word, the 2nd, analyze the monosyllabic word self-existent probability in e-commerce website based on statistics, for the participle disambiguation provides important reference frame.The dictionary that the method generates can be widely used in the E-business applications such as search, semantic participle, recommendation, weight calculation.
Description of drawings
Fig. 1 is the process flow diagram of the inventive method.
Embodiment
Below in conjunction with Fig. 1 and specific embodiment, the present invention is described in detail.
A1, language material are prepared and pre-service
Gather a large amount of ecommerce language materials from e-commerce platform, network search engines etc.Language material is carried out pre-service, namely filter the junk information such as picture, link.
A2, carry out with going forward one by one of redundant data exhaustive to language material, (for example the institute of " associate professor of Hunan University " might potential word to obtain all possible potential word combination, comprise " lake ", " south ", " greatly ", " ", " pair ", " religion ", " award ", " Hunan ", " south is large ", " university ", " learn secondary ", " secondary religion ", " professor ", " Hunan is large ", " southern university ", " university is secondary ", " learn secondary religion ", " associate professor ", " Hunan University ", " southern university is secondary ", " the secondary religion of university ", " learn associate professor ", " Hunan University is secondary ", " the secondary religion of southern university ", " associate professor of university ", " the secondary religion of Hunan University ", " associate professor Nan Daxue ", " associate professor of Hunan University "),
Employing is gone forward one by one exhaustive method by effective word maximum length+exhaustive various minutes of 1 (capable of regulating) word combination, add up simultaneously the frequency that the combination of various individual characters and multiword occurs, form complete comprise might potential word (potential word is exactly the combination of various possible words, may be that significant word may be also nonsensical rubbish) set.Introduction is used for filtering the cutting data boundary greater than the invalid potential word of effective word length.Such as introducing " stainless steel Foldable mosquito-net " filtration " rust steel Foldable mosquito-net " and " stainless steel folds mosquito ", otherwise the boundary error data can't not processed because filter according to.
A3, be that the dispersion that distributes in the shortest length potential word of this word as prefix/postfix at a group based on shorter potential word more than 2 is carried out invalid word (effectively word is exactly through filtering, the word that praises for length; Invalid word is exactly to think wrong) filtration.For concentrating the short potential word that appears in long potential word to filter.For short potential word (for " sheet shooting " and " short-movie shooting ", " sheet shooting " is short potential word, and " short-movie shooting " is long potential word), the shortest potential word that only comprises it satisfies following condition and all satisfies and just can be filtered:
1) count is the number of times that potential word occurs in all language materials, and the potential word proportion of the length of count maximum need to reach certain threshold values (be generally 50%, namely surpass half), is mainly used in filtering the lower short potential word of the frequency of occurrences;
2) front n long potential word proportion reaches certain threshold values;
3) comprising the long potential word number of short potential word must be less than certain threshold values;
4) grow potential word and must not be based on the invalid word of canonical filtration;
5) all satisfy above condition in the situation that short potential word is sewed as front and back, during filtration based on larger one group of the cumulative sum of long potential word count, otherwise based on satisfying a group of filtercondition.
6) short potential word count subtracts above cumulative count (being exactly to satisfy above 1.1)-1.5) with the count sum of this word as the potential word of the shortest length of prefix/postfix), be that 0 short potential word is directly deleted for difference, otherwise short potential word count is above difference.
Because short potential word is concentrated to be disperseed, and be filtered (numeral after potential word is the number of times count that potential word occurs, lower with) such as following data in all language materials in long potential word:
Because too disperse, do not satisfy filtercondition such as following data
Canonical to long potential word is filtered, such as following data:
The time 76
The time 68
When wait 3
...
Filter based on canonical, " time " be invalid, therefore " time " not can based on " time " go judgement and be filtered
A4, be that probability that more than 2, potential word independently occurs carries out invalid word and filters based on length
Shorter potential word in comprising its potential word of the shortest length occurrence number (threshold values can arrange 70% greater than certain threshold values, in fact as long as greater than 50%, the possibility that is to say independent appearance is just passable less than the possibility that occurs as other substring, can do as the case may be repeatedly and regulate), and long potential word does not meet the canonical filtercondition, short potential word count subtracts long potential word count, be that 0 short potential word is directly deleted for difference, otherwise short potential word count is above difference (short potential word count subtracts long potential word count).
Such as following data:
But long potential word satisfies can not filtering of canonical filtercondition, such as:
The time 76
The time 68
A5, be that potential word 2 or more carries out canonical and filters for remaining length after filtering through two steps of A3, A4, prefix/postfix/tundish is contained in predefined set, and remainder be all effective word after above filtration, deletes this potential word; Be included in predefined set for prefix/postfix simultaneously, and not in the set of word of exception, such as for " " " taxi ", " taxi driver brother ", " really " should be in the exception set for the word of doing prefix, for " " do suffix should put into also all filtering of exception set such as " beautiful ".
1) such as " China's Mainland and TaiWan, China ", " China and U.S.A " is because medium term (namely " reach " and be connected and be connected so two-part word in connection front and back) satisfies filtercondition, and " China's Mainland ", " TaiWan, China ", " China ", " U.S. " etc. all are significant word independently
2) " time ", prefix satisfies filtercondition, and " time " be significant word.
3) " during use ", suffix satisfies filtercondition, and " use " is significant word
4) " product ", " look " etc. comprise the word of two words, and are not included in the set that can not filter of exception, such as " taxi ", " beautiful " etc.
A6, appear in unique context mistake and delete the compensation that length is effective word 2 or more
At first appear at the potential word of mistake deletion in unique context based on following condition judgment:
(1) this potential word does not meet the canonical filtercondition
(2) all potential words that comprise this potential word all have been filtered, and how many length of tube is not
(3) count of this potential word is identical with the potential word count that all comprise it
Secondly, find the longest potential word that comprises this mistake stop word.In the situation that a potential word appears at the longest a plurality of potential words, do splicing, again reduce the cutting border.Then the cutting unit of the longest above potential word/reduction just carried out/reverse maximum coupling participle based on existing effective word, if the combination of cutting appears in effective word dictionary as a potential word, continue to scan backward character string, be not less than 2 the longest cutting combination and join in effective potential word dictionary for not appearing at length in potential word dictionary, frequency is the original frequency of the full cutting of this combination.
Such as following this group:
Cutting is: machine/product/common/question and answer
Product and common be all effective word, " question and answer " add in effective word dictionary, count is 6
At last for not comprising effective word in the longest potential word, keep the longest potential word, add in effective word dictionary, such as " cartoonist's Takehiko Inoue ".
The frequency statistics that A7, single character occur as substring (substring is the part of other character strings) in long word more: find all other longer effective words that comprise this word, from being short to long processing successively, delete all more long words that comprise current word (filtering the number of times that is comprised by other word).
A8, compensation overlap type mistake statistical correction cut the frequency that the monosyllabic word that repeats to add up occurs as substring in long word more, such as " want " and " requirements " when the number of times of statistics " wanting ", need to the filtration by mistake of consideration overlap type.
1) for the original count that remains in all steps A 7 after word obtains full cutting end, as the current count of word;
2) find all with the potential word of current word as prefix and suffix for residue word in steps A 7, as be divided into two groups of prefix and suffix, respectively get a combination in twos in two groups;
3) to 2) in the combined result that generates coupling one by one in the result of steps A 7, for the combination that the match is successful, deduct the original count of combination with one of them current count of two words of this combination of composition;
A9, the statistical computation of the monosyllabic word independence frequency of occurrences, the sum frequency of the monosyllabic word that obtains exhaustive from going forward one by one cuts the count that step 8 is finally obtained
A10, filtration finish, and reject occurrence number word low-frequency word seldom in all language materials.
Should be understood that, for those of ordinary skills, can be improved according to the above description or conversion, and all these improve and conversion all should belong to the protection domain of claims of the present invention.
Claims (1)
1. the Automatic Extraction method of monosyllabic word in an ecommerce dictionary, is characterized in that, comprises the following steps:
A1, language material are prepared and pre-service;
A2, language material is carried out with going forward one by one of redundant data exhaustive, obtain all possible potential word combination; Employing is gone forward one by one exhaustive method by the exhaustive various minutes word combinations in effective word maximum length+1, adds up simultaneously the frequency that the combination of various individual characters and multiword occurs, and forms a complete possible potential set of words that comprises.Introduction is used for filtering the cutting data boundary greater than the invalid potential word of effective word length;
A3, be that the dispersion that distributes in the shortest length potential word of this word as prefix/postfix at a group based on shorter potential word more than 2 is carried out the filtration of invalid word for length;
A4, be that probability that more than 2, potential word independently occurs carries out invalid word and filters based on length;
Shorter potential word in comprising its potential word of the shortest length occurrence number greater than certain threshold values, and long potential word does not meet the canonical filtercondition, short potential word count subtracts the difference of long potential word count, be that 0 short potential word is directly deleted for difference, otherwise short potential word count is described difference;
A5, be that potential word 2 or more carries out canonical and filters for remaining length after filtering through two steps of A3, A4, prefix/postfix/tundish is contained in predefined set, and remainder be all effective word after above filtration, deletes this potential word; Be included in predefined set for prefix/postfix simultaneously, and not in the set of the word of exception;
A6, appear in unique context mistake and delete the compensation that length is effective word 2 or more; At first appear at the potential word of mistake deletion in unique context based on following condition judgment:
(1) this potential word does not meet the canonical filtercondition;
(2) all potential words that comprise this potential word all have been filtered, and how many length of tube is not;
(3) count of this potential word is identical with the potential word count that all comprise it;
Secondly, find the longest potential word that comprises this mistake stop word; In the situation that a potential word appears at the longest a plurality of potential words, do splicing, again reduce the cutting border; Then the cutting unit of the longest above potential word/reduction just carried out/reverse maximum coupling participle based on existing effective word, if the combination of cutting appears in effective word dictionary as a potential word, continue to scan backward character string, be not less than 2 the longest cutting combination and join in effective potential word dictionary for not appearing at length in potential word dictionary, frequency is the original frequency of the full cutting of this combination; At last for not comprising effective word in the longest potential word, keep the longest potential word, add in effective word dictionary;
The frequency statistics that A7, single character occur as substring in long word more: find all other longer effective words that comprise this word, longly process successively from being short to, delete the more long word that all comprise current word;
A8, compensation overlap type mistake statistical correction cut the frequency that the monosyllabic word that repeats to add up occurs as substring in long word more;
1) for the original count that remains in all steps A 7 after word obtains full cutting end, as the current count of word;
2) find all with the potential word of current word as prefix and suffix for residue word in steps A 7, as be divided into two groups of prefix and suffix, respectively get a combination in twos in two groups;
3) to 2) in the combined result that generates coupling one by one in the result of steps A 7, for the combination that the match is successful, deduct the original count of combination with one of them current count of two words of this combination of composition;
A9, the statistical computation of the monosyllabic word independence frequency of occurrences, the sum frequency of the monosyllabic word that obtains exhaustive from going forward one by one cuts the count that step 8 is finally obtained;
A10, filtration finish, and reject occurrence number word low-frequency word seldom in all language materials.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100798089A CN103136191A (en) | 2013-03-14 | 2013-03-14 | Automatic extracting method of word with single character in electronic commerce dictionary |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100798089A CN103136191A (en) | 2013-03-14 | 2013-03-14 | Automatic extracting method of word with single character in electronic commerce dictionary |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103136191A true CN103136191A (en) | 2013-06-05 |
Family
ID=48496030
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013100798089A Pending CN103136191A (en) | 2013-03-14 | 2013-03-14 | Automatic extracting method of word with single character in electronic commerce dictionary |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103136191A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103500214A (en) * | 2013-09-30 | 2014-01-08 | 北京奇虎科技有限公司 | Word segmentation information pushing method and device based on video searching |
| CN103838883A (en) * | 2014-03-31 | 2014-06-04 | 上海久科信息技术有限公司 | Intelligent SKU matching method |
| CN104899190A (en) * | 2015-06-04 | 2015-09-09 | 百度在线网络技术（北京）有限公司 | Generation method and device for word segmentation dictionary and word segmentation processing method and device |
| CN109582962A (en) * | 2018-11-28 | 2019-04-05 | 北京创鑫旅程网络技术有限公司 | Segmenting method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030233235A1 (en) * | 2002-06-17 | 2003-12-18 | International Business Machines Corporation | System, method, program product, and networking use for recognizing words and their parts of speech in one or more natural languages |
| CN1629836A (en) * | 2003-12-17 | 2005-06-22 | 北京大学 | Method and apparatus for learning Chinese new words |
| CN1664818A (en) * | 2004-03-03 | 2005-09-07 | 微软公司 | Word collection method and system for use in word-breaking |
| CN101131705A (en) * | 2007-09-27 | 2008-02-27 | 中国科学院计算技术研究所 | New word discovering method and system thereof |
| CN102902757A (en) * | 2012-09-25 | 2013-01-30 | 姚明东 | Automatic generation method of e-commerce dictionary |
-
2013
- 2013-03-14 CN CN2013100798089A patent/CN103136191A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030233235A1 (en) * | 2002-06-17 | 2003-12-18 | International Business Machines Corporation | System, method, program product, and networking use for recognizing words and their parts of speech in one or more natural languages |
| CN1629836A (en) * | 2003-12-17 | 2005-06-22 | 北京大学 | Method and apparatus for learning Chinese new words |
| CN1664818A (en) * | 2004-03-03 | 2005-09-07 | 微软公司 | Word collection method and system for use in word-breaking |
| CN101131705A (en) * | 2007-09-27 | 2008-02-27 | 中国科学院计算技术研究所 | New word discovering method and system thereof |
| CN102902757A (en) * | 2012-09-25 | 2013-01-30 | 姚明东 | Automatic generation method of e-commerce dictionary |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103500214A (en) * | 2013-09-30 | 2014-01-08 | 北京奇虎科技有限公司 | Word segmentation information pushing method and device based on video searching |
| CN103500214B (en) * | 2013-09-30 | 2017-04-19 | 北京奇虎科技有限公司 | Word segmentation information pushing method and device based on video searching |
| CN103838883A (en) * | 2014-03-31 | 2014-06-04 | 上海久科信息技术有限公司 | Intelligent SKU matching method |
| CN104899190A (en) * | 2015-06-04 | 2015-09-09 | 百度在线网络技术（北京）有限公司 | Generation method and device for word segmentation dictionary and word segmentation processing method and device |
| CN109582962A (en) * | 2018-11-28 | 2019-04-05 | 北京创鑫旅程网络技术有限公司 | Segmenting method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104239539B (en) | A kind of micro-blog information filter method merged based on much information | |
| CN103226576A (en) | Comment spam filtering method based on semantic similarity | |
| CN108829658B (en) | Method and device for discovering new words | |
| CN105426539A (en) | Dictionary-based lucene Chinese word segmentation method | |
| CN103136191A (en) | Automatic extracting method of word with single character in electronic commerce dictionary | |
| CN102207961B (en) | Automatic web page classification method and device | |
| CN101727464A (en) | Method and device for acquiring alternative name matched pair | |
| CN105843795A (en) | Topic model based document keyword extraction method and system | |
| CN103778243A (en) | Domain term extraction method | |
| CN101901235A (en) | Method and system for document processing | |
| CN106407484A (en) | Video tag extraction method based on semantic association of barrages | |
| WO2011072125A3 (en) | Method and apparatus for real time semantic filtering of posts to an internet social network | |
| CN105893414A (en) | Method and apparatus for screening valid term of a pronunciation lexicon | |
| CN103488635A (en) | Method and device for acquiring product information | |
| CN108845982A (en) | A kind of Chinese word cutting method of word-based linked character | |
| CN104102658A (en) | Method and device for mining text contents | |
| CN108563667A (en) | Hot issue acquisition system based on new word identification and its method | |
| CN103955453A (en) | Method and device for automatically discovering new words from document set | |
| CN104090961B (en) | A kind of social networks junk user filter method based on machine learning | |
| CN102902757B (en) | A kind of Automatic generation method of e-commerce dictionary | |
| CN105224604A (en) | A kind of microblogging incident detection method based on heap optimization and pick-up unit thereof | |
| CN105718584A (en) | Web page content extracting method and device | |
| CN103218368A (en) | Method and device for discovering hot words | |
| CN104572736A (en) | Keyword extraction method and device based on social networking services | |
| CN103942226A (en) | Method and device for obtaining hot content |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C53 | Correction of patent of invention or patent application | ||
| CB03 | Change of inventor or designer information |
Inventor after: Yao Mingdong Inventor after: Fan Yinglei Inventor after: Chen Hao Inventor before: Yao Mingdong Inventor before: Chen Hao Inventor before: Fan Yinglei |
|
| COR | Change of bibliographic data |
Free format text: CORRECT: INVENTOR; FROM: YAO MINGDONG CHEN HAO FAN YINGLEI TO: YAO MINGDONG FAN YINGLEI CHEN HAO |
|
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130605 |