CN103838883A - Intelligent SKU matching method - Google Patents
Intelligent SKU matching method Download PDFInfo
- Publication number
- CN103838883A CN103838883A CN201410125592.XA CN201410125592A CN103838883A CN 103838883 A CN103838883 A CN 103838883A CN 201410125592 A CN201410125592 A CN 201410125592A CN 103838883 A CN103838883 A CN 103838883A
- Authority
- CN
- China
- Prior art keywords
- data
- client
- keyword
- trade name
- sku
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
Abstract
The invention discloses an intelligent SKU matching method. The intelligent SKU matching method includes the steps that client SKU data are input and stored in a database unit, wherein the client SKU data comprise one or more client commodity name data; E-business SKU data are input and stored in the database unit, wherein the E-business SKU data comprise a plurality of E-business commodity name data and relevant data of E-business commodities; key word sequences of all the client commodity name data are extracted, and grading indexes of the client SKU data are generated according to the key word sequences; interference data of all the E-business commodity name data are removed to form commodity indexes; all the client commodity name data are matched with the commodity indexes, the highest matched-degree E-business commodity name datum corresponding to each client commodity name datum and the relevant data of the E-business commodities are searched for according to the grading indexes. By adopting the intelligent SKU matching method, the advantages of high intelligent performance and high matching precision are achieved.
Description
Technical field
The present invention relates to a kind of data matching method, relate in particular to a kind of intelligent SKU matching process.
Background technology
The problem that existing SKU matching process ubiquity coupling degree of accuracy is low, positional accuracy is poor.
Summary of the invention
The object of the invention is to overcome the defect of prior art, and a kind of intelligent SKU matching process be provided, have advantages of intelligent strong, coupling degree of accuracy is high.
The technical scheme that realizes above-mentioned purpose is:
The intelligent SKU matching process of one of the present invention, comprises step:
Input client's SKU data are also stored in a Database Unit, described client SKU data comprise one or a plurality of client's trade name data and and client's commodity related data of client's trade name data correlation described in each;
Input electric business SKU data and be stored in described Database Unit, described electric business SKU data comprise a plurality of electric business's trade name data and the electric business commodity related data associated with each electric business's trade name;
Extract the keyword sequence of client's trade name data described in each and generate the scoring index of described client SKU data according to described keyword sequence;
Remove the interfering data of electric business's trade name data described in each and form commodity index;
Client's trade name data described in each are mated with described commodity index, and utilize described scoring index to search out electric business's trade name data and the electric business's commodity related data that the matching degree corresponding with client's trade name data described in each is the highest.
Further improvement of the present invention is, described extraction described in each the keyword sequence step of client's trade name data further comprise step:
A, remove the space symbol of special symbol both sides in current described client's trade name data, obtain one first character string;
B, described the first character string is carried out to the long cutting of major term, and extract and form one first vocabulary array;
C, filter the repetitor in described vocabulary array, obtain one second vocabulary array;
D, described the second vocabulary array is carried out to participle, obtain one the 3rd vocabulary array;
E, by the filtering profile setting in advance, described the 3rd vocabulary array is filtered the keyword array that obtains the keyword of current described client's trade name data and formed by described keyword;
F, generate described keyword sequence according to described keyword array.
Further improvement of the present invention is, describedly generates described keyword sequence step according to described keyword array and further comprises step:
In the time that described client's trade name data amount check is one, generate described keyword sequence according to the described keyword array of current acquisition;
Otherwise repeating step a-e, obtains the keyword array of all described client's trade name data, and generates described keyword sequence according to described keyword array.
Further improvement of the present invention is, the scoring index step of the described client SKU of described generation data further comprises step:
Filter out the public word in described keyword sequence, form public word list;
By an algorithm and described public word list, calculate public word scoring and the scoring of individual character word of each keyword in described keyword sequence, and the score data of described public word scoring and the scoring of individual character word is deposited in described client SKU data, form described scoring index.
Further improvement of the present invention is, described in filter out the public word in described keyword sequence, form public word listings step and further comprise step:
Keyword in described keyword sequence is added to an initiation sequence,
The totally occurrence number of keyword in described keyword sequence described in each;
If the occurrence number of described keyword in described keyword sequence is less than a fixed percentage of keyword sum in described keyword sequence, current described keyword is removed in described initiation sequence, form described public word list.
Further improvement of the present invention is, described fixed percentage is 25%.
Further improvement of the present invention is, describedly passes through an algorithm and described public word list, calculates the public word scoring of each keyword in described keyword sequence and the individual character word step of marking and further comprises step:
Set a public word score value and an individual character word score value, and respectively to described public word score value and the attached initial value of described individual character word score value, and described public word score value and described individual character word score value and be certain value;
Keyword in described public word and described lists of keywords is got to common factor, and obtain the common factor length value of described common factor;
As described in common factor length value be 0, described individual character word score value is made as to 100;
As described in common factor length value equal as described in length keywords value, described public word score value is made as to 100;
Public word scoring=public word score value/common factor length value;
Individual character word scoring=individual character word score value/(length keywords value-common factor length value).
Further improvement of the present invention is, the interfering data of described removal electric business's trade name data described in each also forms commodity index step and comprises:
A default measure word template;
By described measure word template to described described in each electric business's trade name data carry out measure word filtration, obtain the second character string;
Described the second character string is carried out participle and filtered repetitor, obtain three-character doctrine string;
Form described commodity index by described three-character doctrine string.
Further improvement of the present invention is, describedly utilizes described scoring index to search out electric business's trade name data and the electric business's commodity related data step that the matching degree corresponding with client's trade name data described in each is the highest further to comprise step:
The three-character doctrine matching in described commodity index is serially added in a results list;
The public word scoring summation comprising in three-character doctrine string described in each according to described scoring index calculation;
Sort according to described public word scoring sum value size;
Using electric business's trade name data corresponding the three-character doctrine string of the scoring of public word described in described the results list sum value maximum and electric business's commodity related data as the matching degree corresponding with current described client's trade name data the highest electric business's trade name data and electric business's commodity related data.
Further improvement of the present invention is, described client's commodity related data comprises: client's goods number data and client's descriptive labelling data;
Described electric business's commodity related data comprises: electric business's goods number data, electric business's descriptive labelling data and Commercial goods labels data.
The present invention has been owing to having adopted above technical scheme, makes it have following beneficial effect to be:
Remove the interfering data of electric business's trade name data described in each, thereby reduced the complexity of follow-up matching process, the accuracy that has strengthened follow-up coupling and efficiency; By client's trade name data described in each are mated with described commodity index, and utilize described scoring index to search out electric business's trade name data that the matching degree corresponding with client's trade name data described in each is the highest and the step of electric business's commodity related data has realized high-precision the mating of client's trade name data with electric business's trade name data and electric business's commodity related data.
Brief description of the drawings
Fig. 1 is the process flow diagram of the present invention's intelligence SKU matching process.
Embodiment
Below in conjunction with specific embodiment, the invention will be further described.
Refer to Fig. 1, the intelligent SKU matching process of one of the present invention, comprises step:
First, input client's SKU data are also stored in a Database Unit, client SKU data comprise one or a plurality of client's trade name data and and client's commodity related data of each client's trade name data correlation, wherein, client's commodity related data comprises: client's goods number data and client's descriptive labelling data;
Then, input electric business SKU data and be stored in Database Unit, electricity business SKU data comprise a plurality of electric business's trade name data and the electric business commodity related data associated with each electric business's trade name, and wherein electric business's commodity related data comprises: electric business's goods number data, electric business's descriptive labelling data and Commercial goods labels data;
Then, extract the keyword sequence of each client's trade name data and generate the scoring index of client SKU data according to keyword sequence;
The keyword sequence step of wherein extracting each client's trade name data further comprises step:
A, remove the space symbol of special symbol both sides in current client's trade name data, obtain one first character string;
B, the first character string is carried out to the long cutting of major term, and extract and form one first vocabulary array;
Repetitor in c, filtration vocabulary array, obtains one second vocabulary array;
D, the second vocabulary array is carried out to participle, obtain one the 3rd vocabulary array;
E, by the filtering profile setting in advance, the 3rd vocabulary array is filtered, obtain the keyword of current client's trade name data and the keyword array being formed by keyword;
F, according to keyword array generate keyword sequence.
In the time that client's trade name data amount check is one, generate keyword sequence according to the keyword array of current acquisition;
Otherwise repeating step a-e, obtains the keyword array of all client's trade name data, and generates keyword sequence according to keyword array.
For example: client's trade name data are " 200 milliliters of combination TM scalp moisturization shampoo are brought back to life in HS SHM200ml HAIR & SCALP CARE MOIST Head&Shoulders silk source ";
First, remove the space symbol of special symbol both sides in current client's trade name data by el expression formula, acquisition one first character string is:
" 200 milliliters of combination TM scalp moisturization shampoo are brought back to life in HS SHM200ml HAIR & SCALP CARE MOIST Head&Shoulders silk source ";
Then, utilize lucene technology to carry out the long cutting of major term to the first character string, and extract and form one first vocabulary array, now the first vocabulary array is:
[hs, shm, 200ml, hair & scalp, care, moist, Head&Shoulders, silk, source, brings back to life, combination, tm, scalp, moisturizing, washes, shampoo, 200, milliliter];
Then, filter the repetitor in vocabulary array, obtain one second vocabulary array, now the second vocabulary array is:
[hs, shm, 200ml, hair & scalp, care, moist, Head&Shoulders, source, brings back to life, combination, tm, scalp, moisturizing, shampoo, 200, milliliter];
Then, the second vocabulary array is carried out to participle, obtains one the 3rd vocabulary array:
[hs, shm, 200, ml, hair & scalp, care, moist, Head&Shoulders, source, brings back to life, combination, tm, scalp, moisturizing, shampoo, 200, milliliter], in this step, can judge whether it is Chinese operation to the vocabulary of the extraction in the second vocabulary array, Chinese can be put into a list by current vocabulary in this way, facilitates subsequent operation; The filtration that also can utilize regular expression as required the vocabulary in the second vocabulary array to be done to need is replaced;
Then, by the filtering profile setting in advance, the 3rd vocabulary array is filtered, remove unwanted word, for example: " sheet ", " bag ", " measuring many ", " ml ", "+", " ", " " waits word.Now, obtain the keyword of current client's trade name data and keyword array and the keyword sequence being formed by keyword:
[hs, shm, 200, hair & scalp, care, moist, Head&Shoulders, source, brings back to life, combination, tm, scalp, moisturizing, shampoo, 200, milliliter].
Wherein, the scoring index step of generation client SKU data further comprises step:
A, filter out the public word in keyword sequence, form public word list, further comprise step:
Keyword in keyword sequence is added to an initiation sequence,
The occurrence number of each keyword of accumulative total in keyword sequence;
If the occurrence number of b keyword in keyword sequence is less than a fixed percentage of keyword sum in keyword sequence, current keyword is removed in initiation sequence, form public word list.In the present embodiment, fixed percentage is 25%.
Then, by an algorithm and public word list, calculate public word scoring and the scoring of individual character word of each keyword in keyword sequence, and the score data of public word scoring and the scoring of individual character word is deposited in client SKU data, form scoring index, specifically comprise step:
Set a public word score value and an individual character word score value, and respectively to public word score value and the attached initial value of individual character word score value, and public word score value and individual character word score value and be certain value, in the present embodiment, setting public word score value initial value is 30.0; Setting individual character word score value initial value is 70.0;
Keyword in public word and lists of keywords is got to common factor, and obtain the common factor length value of common factor;
As the length value that occurs simultaneously is 0, individual character word score value is made as to 100;
As the length value that occurs simultaneously equals length keywords value, public word score value is made as to 100;
Public word scoring=public word score value/common factor length value;
Individual character word scoring=individual character word score value/(length keywords value-common factor length value).
Afterwards, remove the interfering data of each electric business's trade name data and form commodity index, specifically comprising step:
A default measure word template;
By measure word template, each electric business's trade name data is carried out to measure word filtration, obtain the second character string;
The second character string is carried out participle and filtered repetitor, obtain three-character doctrine string.
Form commodity index by three-character doctrine string.
For example: electric business's trade name data are " Olay multiple-effect maintenance suncream is bought one and sent three luxurious gift box dresses ";
By measure word template, each electric business's trade name data is carried out to measure word filtration, obtain the second character string: " Olay multiple-effect maintenance suncream buys one ";
The second character string is carried out participle and filtered repetitor, obtain three-character doctrine string: [Olay, multiple-effect, maintenance, suncream, buys one].
Finally, each client's trade name data is mated with commodity index, and utilizes scoring index to search out electric business's trade name data and the electric business's commodity related data that the matching degree corresponding with each client's trade name data is the highest, specifically comprise step:
The three-character doctrine matching in commodity index is serially added in a results list;
According to the public word scoring summation comprising in the every three-character doctrine string of scoring index calculation;
Sort according to public word scoring sum value size;
Using electric business's trade name data corresponding the three-character doctrine string of public word scoring sum value maximum in the results list and electric business's commodity related data as the highest electric business's trade name data of the matching degree corresponding with current client's trade name data and electric business's commodity related data.
The present invention provides a large amount of indexes by the analysis of the merchandise news to client, then make index of reference carry out the matching analysis to go out the merchandise news of electric business's platform, adopt intelligence accumulation dictionary, optimum indexing, accumulation commodity data and sku allocation map relation, extract not matched data reason, incorrect matched data reason, Optimized model, allows coupling can reach the higher data of scoring.Can solve the coupling of the sku on client's goods number and title and each large electric business website by the present invention, (for example as visible client on electric business's platforms such as sky cat, Jingdone district, Amazon goods number and title), solves client and can not see more intuitively the sale effect of own commodity on the electric business's platform of correspondence and the problem of review information.
Below embodiment has been described in detail the present invention by reference to the accompanying drawings, and those skilled in the art can make many variations example to the present invention according to the above description.Thereby some details in embodiment should not form limitation of the invention, the present invention by the scope defining using appended claims as protection scope of the present invention.
Claims (10)
1. an intelligent SKU matching process, is characterized in that, comprises step:
Input client's SKU data are also stored in a Database Unit, described client SKU data comprise one or a plurality of client's trade name data and and client's commodity related data of client's trade name data correlation described in each;
Input electric business SKU data and be stored in described Database Unit, described electric business SKU data comprise a plurality of electric business's trade name data and the electric business commodity related data associated with each electric business's trade name;
Extract the keyword sequence of client's trade name data described in each and generate the scoring index of described client SKU data according to described keyword sequence;
Remove the interfering data of electric business's trade name data described in each and form commodity index;
Client's trade name data described in each are mated with described commodity index, and utilize described scoring index to search out electric business's trade name data and the electric business's commodity related data that the matching degree corresponding with client's trade name data described in each is the highest.
2. intelligent SKU matching process according to claim 1, is characterized in that: described extraction described in each the keyword sequence step of client's trade name data further comprise step:
A, remove the space symbol of special symbol both sides in current described client's trade name data, obtain one first character string;
B, described the first character string is carried out to the long cutting of major term, and extract and form one first vocabulary array;
C, filter the repetitor in described vocabulary array, obtain one second vocabulary array;
D, described the second vocabulary array is carried out to participle, obtain one the 3rd vocabulary array;
E, by the filtering profile setting in advance, described the 3rd vocabulary array is filtered the keyword array that obtains the keyword of current described client's trade name data and formed by described keyword;
F, generate described keyword sequence according to described keyword array.
3. intelligent SKU matching process according to claim 2, is characterized in that: describedly generate described keyword sequence step according to described keyword array and further comprise step:
In the time that described client's trade name data amount check is one, generate described keyword sequence according to the described keyword array of current acquisition;
Otherwise repeating step a-e, obtains the keyword array of all described client's trade name data, and generates described keyword sequence according to described keyword array.
4. intelligent SKU matching process according to claim 3, is characterized in that: the scoring index step of the described client SKU of described generation data further comprises step:
Filter out the public word in described keyword sequence, form public word list;
By an algorithm and described public word list, calculate public word scoring and the scoring of individual character word of each keyword in described keyword sequence, and the score data of described public word scoring and the scoring of individual character word is deposited in described client SKU data, form described scoring index.
5. intelligent SKU matching process according to claim 4, is characterized in that: described in filter out the public word in described keyword sequence, form public word listings step and further comprise step:
Keyword in described keyword sequence is added to an initiation sequence,
The totally occurrence number of keyword in described keyword sequence described in each;
If the occurrence number of described keyword in described keyword sequence is less than a fixed percentage of keyword sum in described keyword sequence, current described keyword is removed in described initiation sequence, form described public word list.
6. intelligent SKU matching process according to claim 4, is characterized in that: described fixed percentage is 25%.
7. intelligent SKU matching process according to claim 5, is characterized in that: describedly pass through an algorithm and described public word list, calculate the public word scoring of each keyword in described keyword sequence and the individual character word step of marking and further comprise step:
Set a public word score value and an individual character word score value, and respectively to described public word score value and the attached initial value of described individual character word score value, and described public word score value and described individual character word score value and be certain value;
Keyword in described public word and described lists of keywords is got to common factor, and obtain the common factor length value of described common factor;
As described in common factor length value be 0, described individual character word score value is made as to 100;
As described in common factor length value equal as described in length keywords value, described public word score value is made as to 100;
Public word scoring=public word score value/common factor length value;
Individual character word scoring=individual character word score value/(length keywords value-common factor length value).
8. intelligent SKU matching process according to claim 7, is characterized in that: the interfering data of described removal electric business's trade name data described in each also forms commodity index step and comprises:
A default measure word template;
By described measure word template to described described in each electric business's trade name data carry out measure word filtration, obtain the second character string;
Described the second character string is carried out participle and filtered repetitor, obtain three-character doctrine string;
Form described commodity index by described three-character doctrine string.
9. intelligent SKU matching process according to claim 8, is characterized in that: describedly utilize described scoring index to search out electric business's trade name data and the electric business's commodity related data step that the matching degree corresponding with client's trade name data described in each is the highest further to comprise step:
The three-character doctrine matching in described commodity index is serially added in a results list;
The public word scoring summation comprising in three-character doctrine string described in each according to described scoring index calculation;
Sort according to described public word scoring sum value size;
Using electric business's trade name data corresponding the three-character doctrine string of the scoring of public word described in described the results list sum value maximum and electric business's commodity related data as the matching degree corresponding with current described client's trade name data the highest electric business's trade name data and electric business's commodity related data.
10. intelligent SKU matching process according to claim 9, is characterized in that:
Described client's commodity related data comprises: client's goods number data and client's descriptive labelling data;
Described electric business's commodity related data comprises: electric business's goods number data, electric business's descriptive labelling data and Commercial goods labels data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410125592.XA CN103838883A (en) | 2014-03-31 | 2014-03-31 | Intelligent SKU matching method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410125592.XA CN103838883A (en) | 2014-03-31 | 2014-03-31 | Intelligent SKU matching method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103838883A true CN103838883A (en) | 2014-06-04 |
Family
ID=50802379
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410125592.XA Pending CN103838883A (en) | 2014-03-31 | 2014-03-31 | Intelligent SKU matching method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103838883A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105354194A (en) * | 2014-08-19 | 2016-02-24 | 上海中怡通信息科技有限公司 | Intelligent commodity classifying method and system |
| CN106095759A (en) * | 2016-06-20 | 2016-11-09 | 西安交通大学 | A kind of invoice goods classifying method based on heuristic rule |
| CN109766491A (en) * | 2018-12-18 | 2019-05-17 | 深圳壹账通智能科技有限公司 | Product search method, device, computer equipment and storage medium |
| CN112906761A (en) * | 2021-01-30 | 2021-06-04 | 广西机电职业技术学院 | E-commerce product classification method and system |
| CN115357766A (en) * | 2022-10-17 | 2022-11-18 | 深圳美云集网络科技有限责任公司 | Data matching method and related device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005022408A1 (en) * | 2003-08-28 | 2005-03-10 | British Telecommunications Public Limited Company | Method and apparatus for storing and retrieving data using ontologies |
| CN102289436A (en) * | 2010-06-18 | 2011-12-21 | 阿里巴巴集团控股有限公司 | Method and device for determining weighted value of search term and method and device for generating search results |
| CN103136191A (en) * | 2013-03-14 | 2013-06-05 | 姚明东 | Automatic extracting method of word with single character in electronic commerce dictionary |
| CN103218364A (en) * | 2012-01-19 | 2013-07-24 | 阿里巴巴集团控股有限公司 | Searching method and system |
-
2014
- 2014-03-31 CN CN201410125592.XA patent/CN103838883A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005022408A1 (en) * | 2003-08-28 | 2005-03-10 | British Telecommunications Public Limited Company | Method and apparatus for storing and retrieving data using ontologies |
| CN102289436A (en) * | 2010-06-18 | 2011-12-21 | 阿里巴巴集团控股有限公司 | Method and device for determining weighted value of search term and method and device for generating search results |
| CN103218364A (en) * | 2012-01-19 | 2013-07-24 | 阿里巴巴集团控股有限公司 | Searching method and system |
| CN103136191A (en) * | 2013-03-14 | 2013-06-05 | 姚明东 | Automatic extracting method of word with single character in electronic commerce dictionary |
Non-Patent Citations (1)
| Title |
|---|
| 孙文礼: ""电子商务系统中的全文检索及排序优化算法"", 《中国优秀硕士学位论文全文数据库-信息科技辑》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105354194A (en) * | 2014-08-19 | 2016-02-24 | 上海中怡通信息科技有限公司 | Intelligent commodity classifying method and system |
| CN106095759A (en) * | 2016-06-20 | 2016-11-09 | 西安交通大学 | A kind of invoice goods classifying method based on heuristic rule |
| CN106095759B (en) * | 2016-06-20 | 2019-05-24 | 西安交通大学 | A kind of invoice cargo classifying method based on heuristic rule |
| CN109766491A (en) * | 2018-12-18 | 2019-05-17 | 深圳壹账通智能科技有限公司 | Product search method, device, computer equipment and storage medium |
| CN112906761A (en) * | 2021-01-30 | 2021-06-04 | 广西机电职业技术学院 | E-commerce product classification method and system |
| CN115357766A (en) * | 2022-10-17 | 2022-11-18 | 深圳美云集网络科技有限责任公司 | Data matching method and related device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103838883A (en) | Intelligent SKU matching method | |
| KR102190897B1 (en) | Method and Apparatus for analyzing fashion trend based on big data | |
| CN106529968A (en) | Customer classification method and system thereof based on transaction data | |
| CN110287329B (en) | E-commerce category attribute mining method based on commodity text classification | |
| CN105868185A (en) | Part-of-speech-tagging-based dictionary construction method applied in shopping comment emotion analysis | |
| CN102542061B (en) | Intelligent product classification method | |
| CN107291895B (en) | Quick hierarchical document query method | |
| CN107145523A (en) | Large-scale Heterogeneous Knowledge storehouse alignment schemes based on Iterative matching | |
| CN103150359B (en) | Micro-blog information display packing and device | |
| CN105930362A (en) | Search target recognition method and device, and terminal | |
| CN111914163A (en) | Medicine combination recommendation method and device, electronic equipment and storage medium | |
| CN105022740A (en) | Processing method and device of unstructured data | |
| CN105187242A (en) | Method for detecting abnormal user behaviours mined on the basis of variable-length sequence mode | |
| CN105023178A (en) | Main body-based electronic commercere commendation method | |
| CN104915420A (en) | Knowledge base data processing method and knowledge base data processing system | |
| KR20150018880A (en) | Information aggregation, classification and display method and system | |
| CN115293332A (en) | Method, device and equipment for training graph neural network and storage medium | |
| CN110750599A (en) | Associated information extraction and display method based on entity modeling | |
| CN102402535A (en) | Method and system for constructing product library | |
| CN106250402A (en) | A kind of Website classification method and device | |
| CN107153697A (en) | Product search method and device in a kind of commodity transaction website | |
| CN105678244A (en) | Approximate video retrieval method based on improvement of editing distance | |
| CN101256594A (en) | Method and system for measuring graph structure similarity | |
| Nguyen et al. | Fast tensor decompositions for big data processing | |
| CN105718457B (en) | Information pushing method and system based on electronic bill |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140604 |