CN104133868B - A kind of strategy integrated for the classification of vertical reptile data - Google Patents
A kind of strategy integrated for the classification of vertical reptile data Download PDFInfo
- Publication number
- CN104133868B CN104133868B CN201410347438.7A CN201410347438A CN104133868B CN 104133868 B CN104133868 B CN 104133868B CN 201410347438 A CN201410347438 A CN 201410347438A CN 104133868 B CN104133868 B CN 104133868B
- Authority
- CN
- China
- Prior art keywords
- classification
- reptile
- category
- data
- mapping
- 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.)
- Expired - Fee Related
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
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A kind of strategy integrated for the classification of vertical reptile data, including taxonomic hierarchies and its mapping mechanism and dynamic cataloging Data Integration mechanism two parts, taxonomic hierarchies and its mapping mechanism therein comprise the following steps：1) structure of reference category system；2) structure of the classification system of reptile targeted website；3) structure of classification system mapping mechanism.The strategy can effectively integrate carries out data grabber and the classification system for data obtain after unstructured parsing by vertical reptile, and keeps the integrality of source taxonomic hierarchies, while can also realize and enter Mobile state tracking to source taxonomic hierarchies.
Description
Technical field
The present invention relates to vertical search engine technical field, is used for the classification of vertical reptile data more particularly, to one kind and integrates
Strategy.
Background technology
With the explosive increase of info web, the use value of search engine also more and more higher, turning into the network user must
Indispensable instrument, provides the user information navigation and inquiry services.It incorporates web page resources numerous on internet, according to
The keyword of user's inquiry provides correlation webpage, and is the entrance of whole internet according to relevance ranking.At present, it is comprehensive
Search engine is the main force for providing the user inquiry service, but its is comprehensive, determines that it can not meet professional crowd couple
The precision information requirement service of specialized field.User is diversification to the demand of information, therefore the service mould of search engine
Formula will also be segmented, and provide more accurate trade information for different field, this promotes the flourishing of vertical search
Development.
Once searched for using comprehensive search engine, return to the Query Result of user may have it is up to ten thousand or even more than one hundred million
Bar, although relative to immense incomparable Web information, the filtering of very limits has been carried out, the information returned is still use
The magnanimity information that family can not possibly all browse, so as to huge waste caused by between user's inquiry request and network search service
And contrast.And vertical search engine is occurred for this problem, as more there is specialty targetedly search engine, it is only
The information searched in specific area.Because the ambit of covering is single-minded, information content is relative to be greatly reduced, and this reduces adopt
Collect the difficulty of information, while improve the quality of information.
Vertical search engine is the professional search engine for some field, and it passes through to certain class specialty on internet
Information be acquired, integrate, then extraction is oriented according to the corresponding structure of specialized information, at non-structured information
The information of structuring is managed into, then user is returned in the form of information list.
A part for most critical is web page searcher in search engine, and by a reptile, (Crawler, also known as webpage are grabbed for it
Take device, network robot, Web Spider) program composition.So-called crawlers refer to can automatically, never stoppingly on network
The program of search and webpage.During operational network reptile, as long as providing minimal amount of start page, information acquisition device just can be by certain
Rule roamed along the hyperlink on webpage on network, resource information is collected, until traveling through whole website.Its performance
It largely have impact on the scale of search engine site.
For the scale of current internet, the web crawlers of single machine operation far can not be completed in the effective time
The task of the interior whole web networks of search, therefore the web crawlers used now all distributions are run parallel on multimachine, are claimed
For distributed reptile.Reptile controller plays a part of concentration control, its all reptile end of management, the equilibrium of guarantee web resource energy,
It is unduplicated to be crawled by each reptile.Page analyzer carries out analysis filtering to the web page resources crawled, filters out substantial amounts of html marks
Label and junk information, finally give valuable web page content information.
The strategy of existing universal search is to try to obtain data, but the processing level to data, than relatively low, protrusion is asked
Topic is exactly：Invalid information excessive (noise data is more), effective information deficiency, effective information are unstructured, returning result is without individual character
Change Optimization Mechanism.
For example, there is a respective network address taxonomic hierarchies different classification Web side navigation website at present, such as hao123.com,
2345.com wait.When we want to integrate the resource of these websites, data grabber is carried out and to data by vertical reptile
A series of categorical data can be obtained after carrying out unstructured parsing, at this moment we face some such problems：It is how effective
Integrate these classification systemsHow the integrality of source taxonomic hierarchies is keptHow Mobile state tracking is entered to source taxonomic hierarchiesSolve
These problems are the emphasis and difficult point studied at this stage.
The content of the invention
It is a primary object of the present invention to overcome drawbacks described above of the prior art, propose that one kind is used for vertical reptile data
The strategy that classification is integrated.
The present invention adopts the following technical scheme that：
A kind of strategy integrated for the classification of vertical reptile data, it is characterised in that：Including taxonomic hierarchies and its mapping machine
System and dynamic cataloging Data Integration mechanism two parts, taxonomic hierarchies and its mapping mechanism therein comprise the following steps：
1) structure of reference category system；
2) structure of the classification system of reptile targeted website；
3) structure of classification system mapping mechanism.
Preferably, the step 1) uses tri-layer classification application build reference category system, the system include category IDs,
Big classification, middle classification and small classification, the benchmark analogy system have its corresponding category content.
Preferably, the step 2) uses the classification system of the multiple reptile targeted websites of tri-layer classification application build, often
The classification system of individual reptile targeted website includes category IDs, big classification, middle classification and small classification, and each reptile targeted website
Classification system has its corresponding category content.
Preferably, in step 2), for each targeted website for treating reptile, a corresponding reptile mesh is all each created
Mark the classification system of website；During reptile, the grouped data that unstructured parsing obtains is put into respective reptile target
Categories of websites system, and be then put into each self-corresponding category content for the details of particular content.
Preferably, described category content includes the specifying information and its source web of website.
Preferably, in step 3), the structure of classification system mapping mechanism refers to that reference category system provides its infima species
Other specific name, the classification system of reptile targeted website also provide its infima species other specific name, pass through the two infima species
Other specific name is matched to establish mapping relations.
Preferably, described dynamic cataloging Data Integration mechanism refers to, by the classification that reptile obtains according to following several feelings
Condition carries out integrated operation：
Processing in the presence of what A was newly added be sorted in reference category system：In reptile targeted website
Corresponding mapping position in the map architecture option for classifying and insert obtained new category IDs structure is added in classification system；
What B was newly added is sorted in the processing in the case of being not present in reference category system：Classification is added first and is obtained
New category IDs, semantic similar classification is found according to item name, if the classification of high similar semantic can be obtained, by new class
Other ID inserts corresponding mapping position in the map architecture option of structure；Otherwise, new classification is added, and in the map architecture option of structure
Add new mapping relations；
The processing in the case of categories combination in C reference category systems：Merge phase in the map architecture option of corresponding structure
The classification answered；
Processing in D reference category systems in the case of classification division：D1. the classification in reference category system is split into two
Individual classification, one type do not replace original classification, add a classification again again in addition；D2. by each reptile targeted website
Manual sort is re-started with the classification of former reference category mapping, is mapped in two new classifications；
Processing in the case of classification is added and deleted in E reference category systems：New category is added in reference category system
Afterwards, new category ID and its minimum classification after having added are put into corresponding map architecture option；And for deleting situation, then only need
Related category is deleted in reference category system.
Preferably, in the situation B, the determination methods of described high similar semantic classification are mainly according to two class names
The Similarity Measure of title obtains：For two item name W1And W2, W1Including concept set { S11,S12,…,S1n, W2Including
Concept set { S21,S22,…,S2m, then W1And W2Similarity it is as follows：M, n represent two classifications to be compared respectively
The number of concept notional word corresponding to title；Each concept notional word of concept set includes the former feature of following four justice：First is basic
Adopted former description, the former description of other basic meanings, the former description of relation justice, relational symbol description；The similarity of two concept notional words is designated as
The weighted sum of the part similarity of the former feature of aforementioned four justice, i.e., two concept realities are calculated according to the path distance between adopted original
The similarity of word：Wherein, βiConcept reality is represented respectively
Weight coefficient i=1,2,3,4 corresponding to four features of word, it is adjustable parameter and satisfaction：
From the above-mentioned description of this invention, compared with prior art, the present invention has the advantages that：
By means of the invention it is also possible to provide accurately and effectively information for Internet user, user is avoided to carry out multiple
Search, consumes more energy.It is to solve web crawlers collection that the present invention, which is used for the new method that the classification of vertical reptile data is integrated,
The disorderly and unsystematic shortage level of data, the problem of gatherer process is unstable, less efficient provides new strategy.
Brief description of the drawings
Fig. 1 is strategy system structure chart of the present invention.
Embodiment
Below by way of embodiment, the invention will be further described.
Reference picture 1, a kind of strategy integrated for the classification of vertical reptile data, including taxonomic hierarchies and its mapping mechanism and
Dynamic cataloging Data Integration mechanism two parts, taxonomic hierarchies and its mapping mechanism therein comprise the following steps：
1) structure of reference category system；
2) structure of the classification system of reptile targeted website；
3) structure of classification system mapping mechanism.
In step 1).Reference category system is the classification system of web station system after integration, has benchmark effect, other websites
Classification system and its snap.Adoptable tri-layer classification application is built, and its architecture is as shown in table 1 below.Mainly
There are four dimension attributes：Category IDs, big classification, middle classification and small classification.Wherein, category IDs automatically generate when classification is added, big/
In/small classification uses M respectivelyAi, MBj, MCkRepresent, i herein, j, k represent respectively it is big/in/the automatic increase serial number of group.
Greatly/in/the other structure of group can manually input realization, can also using the classification of a certain band reptile website as reference,
Then modify on this basis.
Under reference category system, there is its corresponding category content information, it is specific as shown in table 2.Category content will be made
To integrate the source of website specifying information.X in table 2, Y, Z etc. represent source web.Here category content information also includes being somebody's turn to do
The source web of specifying information, this is beneficial to tracing to the source for information later.
For step 2) for multiple targeted websites for treating reptile, we keep its complete classification system structure as far as possible.
Here the structure similar with Benchmark System is used, as shown in table 3.Mainly also there are four dimension attributes：Category IDs, big classification, middle classification
With small classification.Wherein, category IDs automatically generate when classification is added, big/in/small classification uses M respectivelyAi, MBj, MCkRepresent, herein
I, j, k represent respectively it is big/in/the automatic increase serial number of group.For each targeted website for treating reptile, all each create
One corresponding reptile targeted website classification system.During reptile, the grouped data obtained to unstructured parsing is first
Respective reptile targeted website classification system is put into, and is then put into corresponding classification for details such as particular contents
In appearance as shown in table 4.Category content includes the specifying information and its source web of website.
In step 3), on the basis of structure reference category system and targeted website classification system, classification is finally built
System mapping relations.As shown in table 5, classification system mapping relations are primarily directed to the minimum classification in system of all categories, benchmark
Classification system provides its category IDs and its minimum classification, and for each reptile targeted website class complicated variant for treating reptile targeted website
System, such as X, Y, Z websites, also all provides its category IDs and its minimum classification, finally matches (equal or similar according to item name
With) structure mapping relations.
Pass through above-mentioned taxonomic hierarchies and its mapping mechanism, you can completely to preserve all data that reptile parses to obtain,
Ensure that data do not lack；Also the information content and classification is caused all to trace source simultaneously, this is the maintenance of later stage system
Great convenience is provided with management.
The classification system mapping relations of table 5
The dynamic cataloging Data Integration mechanism of the present invention refers to, the classification that reptile obtains is carried out according to following several situations
Integrated operation：
Processing in the presence of what A was newly added be sorted in reference category system：In reptile targeted website
Corresponding mapping position in the map architecture option for classifying and insert obtained new category IDs structure is added in classification system.As
For the classification that reptile obtains in reference category system existing situation, classification and general need to only adds in the structure of table 3
Obtained new category ID inserts corresponding mapping position in the structure of table 5.
What B was newly added is sorted in the processing in the case of being not present in reference category system：The classification obtained for reptile
Situation about being not present in reference category, classification is added first in table 3 and obtains new category IDs, then according to class in table 1
Alias claims to find semantic similar classification, if the classification of high similar semantic can be obtained, new category IDs are inserted into what table 5 was built
Corresponding mapping position in map architecture option；Otherwise, new classification is added in table 1, and is added in the map architecture option built in table 5
Add new mapping relations.
The calculating of high similar semantic classification mainly obtains according to the Similarity Measure of two item names.Specific calculating process
It is described as follows.Calculating process mainly make use of the calculation that Hownet (HowNet) provides.In Hownet, " concept " and " justice is former "
It is two important ways of semantic meaning representation.Each word can be expressed as multiple concepts, and each concept is described using adopted original,
Adopted original is the least meaning unit for being used to describe concept that is most basic, can not splitting again.
For two item name W1And W2If W1By concept set { S11,S12,…,S1nComposition, W2By concept set
{S21,S22,…,S2mComposition, then W1And W2Similarity it is as follows：
So, the similarity problem just the similarity problem between two words being attributed between two concepts.For general
The description for reading notional word is represented by a feature structure, and this feature structure contains following four feature：The former description of first basic meaning,
The former description of other basic meanings, the former description of relation justice, relational symbol description.Then, the overall similarity of two concept notional words is designated as
The weighted sum of the part similarity of aforementioned four feature, i.e.,
Wherein, βiWeight coefficient i=1,2,3,4 corresponding to four features of concept notional word are represented respectively, and it is adjustable
Parameter and satisfaction：And all concepts were all described originally by justice, institute
Similarity problem between adopted original is finally attributed to the problem.Because all adopted primitive roots constitute one according to hyponymy
Tree-like hierarchy system, for tree, have between any two node and an only paths, therefore can be according to adopted original
Between path distance calculate both similarities.
The processing in the case of categories combination in C reference category systems：Two classifications in reference category system are carried out
Merge, only need to merge corresponding reference category in the structure in table 5 over there.Assuming that as described in Table 5 in structure, MC5With
MC6Categories combination is carried out, over there only need to be by MC5And MC6Give same title, ID5And ID6Give the new ID values after merging.
Processing in D reference category systems in the case of classification division：D1. the classification in reference category system is split into two
Individual classification, one type do not replace original classification, add a classification again again in addition；D2. by each reptile targeted website
Manual sort is re-started with the classification of former reference category mapping, is mapped in two new classifications.If necessary to basis point
Class in class enters line splitting processing, deals with this case comparatively laborious.Assuming that in such as structure of table 5, to MC8Carry out class point
Processing is split, needs to complete following two step over there：
D1. benchmark is split into two classes, one type does not replace original classification MC8, then, add one again in addition
Individual classification, such as MC9。
D2. the classification in each reptile targeted website with the mapping of former reference category is re-started into manual sort, and be mapped to
New MC8, MC9Classification.
Such case is more special, typically it is not recommended that doing division processing.Therefore need to refer to as far as possible when building benchmark classification
Existing taxonomic hierarchies, the refinement that granularity of classification is tried one's best.
Processing in the case of classification is added and deleted in E reference category systems：New category is added in reference category system
Afterwards, new category ID and its minimum classification after having added are put into map architecture option corresponding to table 5；And for deleting situation, then
Related category only need to be deleted in reference category system.
The embodiment of the present invention is above are only, but the design concept of the present invention is not limited thereto, it is all to utilize this
Conceive the change that unsubstantiality is carried out to the present invention, the behavior for invading the scope of the present invention all should be belonged to.
Claims (6)
- A kind of 1. strategy integrated for the classification of vertical reptile data, it is characterised in that：Including taxonomic hierarchies and its mapping mechanism With dynamic cataloging Data Integration mechanism two parts, taxonomic hierarchies and its mapping mechanism therein comprise the following steps：1) tri-layer classification application build reference category system is used, the system includes category IDs, big classification, middle classification and group Not, the reference category system has its corresponding category content；2) structure of the classification system of reptile targeted website；3) structure of classification system mapping mechanism；Described dynamic cataloging Data Integration mechanism refers to, the classification that reptile obtains is carried out into integration behaviour according to following several situations Make：Processing in the presence of what A was newly added be sorted in reference category system：Classification in reptile targeted website Corresponding mapping position in the map architecture option for classifying and insert obtained new category IDs structure is added in system；What B was newly added is sorted in the processing in the case of being not present in reference category system：Classification is added first and is obtained new Category IDs, semantic similar classification is found according to item name, if the classification of high similar semantic can be obtained, by new category IDs Insert corresponding mapping position in the map architecture option of structure；Otherwise, new classification is added, and is added in the map architecture option of structure New mapping relations；The processing in the case of categories combination in C reference category systems：Merge in the map architecture option of corresponding structure corresponding Classification；Processing in D reference category systems in the case of classification division：D1. the classification in reference category system is split into two classes Not, one type does not replace original classification, adds a classification again again in addition；D2. by each reptile targeted website with original The classification of reference category mapping re-starts manual sort, is mapped in two new classifications；Processing in the case of classification is added and deleted in E reference category systems：, will after adding new category in reference category system New category ID and its minimum classification after having added are put into corresponding map architecture option；And for deleting situation, then only need to be in base Related category is deleted in quasi- classification system.
- A kind of 2. strategy integrated for the classification of vertical reptile data as claimed in claim 1, it is characterised in that：The step 2) the classification system of the multiple reptile targeted websites of tri-layer classification application build, the classification system of each reptile targeted website are used Including category IDs, big classification, middle classification and small classification, and the classification system of each reptile targeted website has in its corresponding classification Hold.
- A kind of 3. strategy integrated for the classification of vertical reptile data as claimed in claim 2, it is characterised in that：In step 2) In, for each targeted website for treating reptile, all respective classification system for creating a corresponding reptile targeted website；In reptile During, the grouped data that unstructured parsing obtains is put into respective reptile targeted website classification system, and for specific The details of content are then put into each self-corresponding category content.
- A kind of 4. strategy integrated for the classification of vertical reptile data as claimed in claim 1 or 2, it is characterised in that:It is described Category content include the specifying information and its source web of website.
- A kind of 5. strategy integrated for the classification of vertical reptile data as claimed in claim 1, it is characterised in that：In step 3) In, the structure of classification system mapping mechanism refers to, reference category system provides its infima species other specific name, reptile target network The classification system stood also provides its infima species other specific name, is matched by the other specific name of the two infima species to build Vertical mapping relations.
- A kind of 6. strategy integrated for the classification of vertical reptile data as claimed in claim 1, it is characterised in that：In the feelings In condition B, the determination methods of described high similar semantic classification mainly obtain according to the Similarity Measure of two item names：For Two item name W1And W2, W1Including concept set { S11,S12,…,S1n, W2Including concept set { S21,S22,…,S2m, Then W1And W2Similarity it is as follows：m,n The number of concept notional word corresponding to two item names to be compared is represented respectively；Each concept notional word of concept set include with The former feature of lower four justice：The former description of first basic meaning, the former description of other basic meanings, the former description of relation justice, relational symbol description；Two The similarity of individual concept notional word is designated as the weighted sum of the part similarity of the former feature of aforementioned four justice, i.e., according to the road between adopted original Footpath distance calculates the similarity of two concept notional words： Wherein, βiRepresent weight coefficient i=1,2,3,4 corresponding to four features of concept notional word respectively, its be adjustable parameter and Meet：
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410347438.7A CN104133868B (en) | 2014-07-21 | 2014-07-21 | A kind of strategy integrated for the classification of vertical reptile data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410347438.7A CN104133868B (en) | 2014-07-21 | 2014-07-21 | A kind of strategy integrated for the classification of vertical reptile data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104133868A CN104133868A (en) | 2014-11-05 |
| CN104133868B true CN104133868B (en) | 2018-01-05 |
Family
ID=51806546
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410347438.7A Expired - Fee Related CN104133868B (en) | 2014-07-21 | 2014-07-21 | A kind of strategy integrated for the classification of vertical reptile data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104133868B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106933962A (en) * | 2017-02-06 | 2017-07-07 | 涂正富 | A kind of film micro area network insertion and vertical search precise positioning obtain mesh calibration method |
| CN107436955B (en) * | 2017-08-17 | 2022-02-25 | 齐鲁工业大学 | English word correlation degree calculation method and device based on Wikipedia concept vector |
| CN107491524B (en) * | 2017-08-17 | 2022-02-25 | 齐鲁工业大学 | Method and device for calculating Chinese word relevance based on Wikipedia concept vector |
| CN107679121B (en) * | 2017-09-20 | 2020-10-20 | 晶赞广告（上海）有限公司 | Mapping method and device of classification system, storage medium and computing equipment |
| CN109101541B (en) * | 2018-07-02 | 2022-10-04 | 土巴兔集团股份有限公司 | Newly added index management method, device and computer readable storage medium |
| CN109033286B (en) * | 2018-07-12 | 2021-10-29 | 北京猫眼文化传媒有限公司 | Data statistical method and device |
| CN111008645A (en) * | 2019-11-05 | 2020-04-14 | 北京邮电大学 | Scientific and technological service resource classification system construction method and device based on coreference resolution |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1770159A (en) * | 2005-10-28 | 2006-05-10 | 北大方正集团有限公司 | Method for automatically finding network content quotation |
| CN102402539A (en) * | 2010-09-15 | 2012-04-04 | 倪毅 | Design technology for object-level personalized vertical search engine |
-
2014
- 2014-07-21 CN CN201410347438.7A patent/CN104133868B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1770159A (en) * | 2005-10-28 | 2006-05-10 | 北大方正集团有限公司 | Method for automatically finding network content quotation |
| CN102402539A (en) * | 2010-09-15 | 2012-04-04 | 倪毅 | Design technology for object-level personalized vertical search engine |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104133868A (en) | 2014-11-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104133868B (en) | A kind of strategy integrated for the classification of vertical reptile data | |
| CN101593200B (en) | Method for classifying Chinese webpages based on keyword frequency analysis | |
| CN103049575B (en) | A kind of academic conference search system of topic adaptation | |
| CN103116657B (en) | A kind of individuation search method of network teaching resource | |
| JP5092165B2 (en) | Data construction method and system | |
| CN101615178B (en) | Method and system for building object hierarchy | |
| CN105706078A (en) | Automatic definition of entity collections | |
| CN107885793A (en) | A kind of hot microblog topic analyzing and predicting method and system | |
| CN103838756A (en) | Method and device for determining pushed information | |
| CN102591992A (en) | Webpage classification identifying system and method based on vertical search and focused crawler technology | |
| CN103955529A (en) | Internet information searching and aggregating presentation method | |
| CN103902597A (en) | Method and device for determining search relevant categories corresponding to target keywords | |
| CN102411617B (en) | Method for storing and inquiring a large quantity of URLs | |
| CN106951526B (en) | Entity set extension method and device | |
| CN102122291A (en) | Blog friend recommendation method based on tree log pattern analysis | |
| CN104750704A (en) | Webpage uniform resource locator (URL) classification and identification method and device | |
| CN101393555A (en) | Rubbish blog detecting method | |
| CN105528422A (en) | Focused crawler processing method and apparatus | |
| CN105893611A (en) | Method for establishing interest theme semantic network facing to social networking services | |
| CN103324700A (en) | Noumenon concept attribute learning method based on Web information | |
| CN103838798A (en) | Page classification system and method | |
| CN105512333A (en) | Product comment theme searching method based on emotional tendency | |
| CN100447793C (en) | Method for extracting page query interface based on character of vision | |
| CN104834736A (en) | Method and device for establishing index database and retrieval method, device and system | |
| CN105550169A (en) | Method and device for identifying point of interest names based on character length |
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 |
Inventor after: Wu Meihong Inventor after: Hong Zhiling Inventor before: Hong Zhiling Inventor before: Wu Meihong |
|
| CB03 | Change of inventor or designer information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180105 Termination date: 20180721 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |