CN105404693B - A kind of service clustering method based on demand semanteme - Google Patents
A kind of service clustering method based on demand semanteme Download PDFInfo
- Publication number
- CN105404693B CN105404693B CN201510961111.3A CN201510961111A CN105404693B CN 105404693 B CN105404693 B CN 105404693B CN 201510961111 A CN201510961111 A CN 201510961111A CN 105404693 B CN105404693 B CN 105404693B
- Authority
- CN
- China
- Prior art keywords
- service
- information
- function information
- similarity
- verb
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
- G06F16/355—Class or cluster creation or modification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/211—Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/253—Grammatical analysis; Style critique
Abstract
The invention discloses a kind of service clustering methods based on demand semanteme, comprising the following steps: obtains service description document, carries out subordinate sentence processing to document text;Syntax parsing is carried out using Stanford Parser as unit of sentence, obtains SD set;Basic service functional information collection is obtained according to SD set analysis;Semantic extension is carried out to basic service functional information collection;It removes basic service functional information to concentrate without practical semantic information, carries out speech reduction and generate service function information collection;Similarity value between servicing is calculated using service function information collection;It sets similarity between k value and service and carries out service cluster using K-means algorithm as input.The present invention extracts service function information collection using the method for natural language processing and carries out service cluster from demand for services semanteme angle, breaks through the limitation of service documents type, more precisely provides technical support for on-demand service discovery.
Description
Technical field
The present invention relates to field of service calculation more particularly to a kind of service clustering methods based on demand semanteme.
Background technique
Enterprise SOA (Service Oriented Architecture, SOA) is a kind of coarseness, loose coupling
Service system structure, for instructing service-oriented software system design, it has also become one of the research topic of software field hot topic.
As Enterprise SOA technology and software are the development for servicing (Software as a Service, SaaS), on internet
Web service quantity be always maintained at the trend of rapid growth, the type of Web service also becomes increasingly abundant, from traditional based on simple
The Web service of Object Accessing Protocol (Simple Object Access Protocol, SOAP) follows expression to currently a popular
Property state transition (representational state transfer, REST) agreement lightweight REST style Web clothes
Business, Web service have become a kind of important computing resource and software asset on internet, bring for enterprises and individuals user
Huge convenience.But with the sharp increase of quantity of service and the diversification of type, so that how accurately and efficiently to find to meet
The service of user demand becomes the problem in the field service-oriented computing (Service Oriented Computing, SOC).
Traditional service based on UDDI (Universal Description, Discovery and Integration)
Registration is only supported only to support the service based on keyword in the service discovery stage to the operation of service syntactic level with discovery mechanism
Matching is often unable to satisfy user in the case where quantity of service increases severely and requires.For this purpose, how to improve Web service discovery efficiency
The concern of domestic and international researcher is attracted.
Service cluster is a kind of technology for effectively improving service discovery.Usually it regard service cluster as other complex services
Preprocessing Algorithm with algorithm: firstly, by intimate service aggregating to together;Then, user's request is directly targeted to
Specific classification of service reduces the search space of service, improves service discovery efficiency.Studies have shown that based on functional similitude
Carrying out service cluster can be improved Web service recall precision.It is largely ground currently, the service cluster based on functional similarity is existing
Study carefully.For example, extracting the pass for embodying service function from WSDL (Web Service Description Language) document
Key feature is then based on these features and clusters service for intimate class cluster;Using vector space model to service source document
Descriptive text in part is indicated and handles, and then carries out service cluster using multiple Hybrid Clustering Algorithm MHC.But it is existing
There are still following two o'clock deficiencies for some service clustering methods:
1) the service documents type clustered has limitation.Existing majority clustering method is just for WSDL document or OWL-S
The service description document of the single types such as (Semantic Markup for Web Service) document, and to pass through nature language
Say that the services such as the REST style of description concern is less.
2) less to consider service characteristic from appellative function semanteme angle.Existing service clustering method service documents multipair greatly
Dimension-reduction treatment is carried out, document is indicated and is handled using the methods of vector space model.And from demand for services functional semantics
The angle research clustered of setting out is less.
Therefore, how for having deficiency present in service clustering method, carrying out efficiently and accurately service cluster becomes
One challenging critical issue.
Summary of the invention
In view of the above shortcomings of the prior art, the object of the present invention is to provide a kind of service cluster sides based on demand semanteme
Method, this method is from demand semanteme angle, using service describing text, extracts service function using natural language processing technique
Information collection is then based on service function information collection, carries out service similarity calculation, realizes and services finally by K-means algorithm
Cluster, obtains the service cluster result semantically even more like in demand.
Above-mentioned technical problem of the invention is mainly to be addressed by following technical proposals: one kind is semantic based on demand
Service clustering method, method includes the following steps:
(1) service description document is obtained, subordinate sentence processing is carried out to document text;
(2) syntax parsing is carried out using Stanford Parser as unit of sentence, obtains SD (Stanford
Dependencies) gather;
(3) it is analyzed based on the SD set that step (2) obtain and extracts the service for indicating demand for services functional semantics
Functional information collection;
(4) the service function information collection obtained based on step (3) calculates the similarity two-by-two between service.Phase between service
Like degree calculation formula are as follows:
Wherein, Ss(s1,s2) indicate service s1And s2Similarity;Service s1With s2The functional information number for including not phase
Together, wherein m is service s1, s2Functional information concentrates the functional information number less comprising functional information number, and n is that functional information is concentrated
Include the more functional information number of functional information number;Sf(fi1,fi2) indicate service s1In i-th functional information and service s2In
The similarity of i functional information, calculation formula are as follows:
Wherein, V1, V2Respectively indicate service function information fi1,fi2In verb, Ni1, Ni2Respectively indicate service function letter
Cease fi1,fi2In noun, t1, t2The weight of corresponding verb part and noun part, Sw(Ni1,Ni2) indicate the similar of two words
Degree, calculation formula are as follows:
Wherein, F (Ni1)、F(Ni2) respectively indicate word Ni1、Ni2Feature set, I (F (Ni1))、I(F(Ni2))、I(F(Ni1)
∩F(Ni2)) respectively indicate feature set F (Ni1)、F(Ni2)、F(Ni1)∩F(Ni2) Information Number that includes.
(5) similarity between the service obtained based on step (4), clusters service using K-means algorithm, k value is
It is manually set, using service similarity value as the distance value in K-means algorithm.Finally export k service cluster class.
Further, above-mentioned steps (3) specifically include following sub-step:
(3.1) basic service functional information extracts: in the SD collection basis that step (2) obtain, to following four situation
It is analyzed, extracts basic service functional information collection.
A) directly Transformational Grammar relationship dobj:
The part governor of dobj is converted directly into the part Verb in basic function information, and the part dependent is straight
Switch through as at the part Object in basic function information.
B) directly Transformational Grammar relationship nsubjpass:
The part governor of nsubjpass is converted directly into the part Verb in basic function information, the portion dependent
Divide the part Object directly switched into basic function information.
C) comprehensive analysis grammatical relation prep_p and nsubj:
It, can be with if the part governor in grammatical relation prep_p is identical as the part governor in nsubj
The part governor in prep_p is converted into the part Verb in basic function information, the part dependent is converted into base
The part Object in this functional information.
D) potential function information collection is excavated using grammatical relation conj:
The business movement { p (verb), p (object) } in sentence has been obtained,
If there is conj (p (verb), dependent), then there is its business movement { p (dependent), p arranged side by side
(object)};
If there is conj (governor, p (object)), then there is its business movement { p (verb), p arranged side by side
(governor)}。
(3.2) service function information semantic extends: may in the basic service functional semantics information obtained in step (3.1)
In the presence of semanteme missing the case where, in response to this, will continue to SD set analysis, and by the semantic extension of missing to service function
In energy information.The grammatical relation mainly considered is nn.
(3.3) it generates service function information collection: further place is needed to the service function information that step (3.2) obtains
Reason, mainly include that two parts work: 1) creation stops vocabulary, removes the contents such as some weak verbs for lacking semanteme;2) to service function
The word that energy information is concentrated carries out speech reduction, is convenient for follow-up service similarity calculation.
The prior art is compared, and the present invention has the advantages that:
(1) present invention using natural language processing the relevant technologies extracted from text can service function information collection,
This makes the type that service documents are no longer limited by when carrying out service cluster, and takes to the REST style of natural language description
Business cluster provides effective clustering method.
(2) present invention from demand for services angle consider service similitude carry out service cluster, be different from existing method only from
Server-side considers the defect of service characteristic, provides effective support for on-demand service discovery.
Detailed description of the invention
Service description document case Fig. 1 of the invention.
Specific embodiment
Further description of the technical solution of the present invention by way of example and in conjunction with the accompanying drawings.
A kind of service clustering method based on demand semanteme proposed by the present invention, the specific steps are as follows:
(1) service description document is obtained, subordinate sentence processing is carried out to document text:
Firstly the need of by page download tool by the related pages in service registration website under the format of static page
It is downloaded to local, these static pages are then parsed by script, service metamessage is extracted, these information is pre-processed, with
The problems such as removing spcial character, nonstandard grammatical representation, and it is saved in local database files.
Subordinate sentence processing is carried out to the service description document text that extracts, with " ", "? ", "!" it is that separator is divided
Sentence.
(2) subordinate sentence based on step (1) is unit as input, is parsed, is obtained to it using Stanford Parser
Obtain SD set:
Stanford Parser is the Open-Source Tools of Stanford NLP Group exploitation, is widely used for natural language
Processing, mainly realizes following function: 1) identifying and mark the part of speech of word in sentence;2) word two-by-two in a sentence is created
Between grammatical relation Stanford Dependencies, 3) obtain the syntactic structure of a sentence.The Stanford of latest edition
Parser includes 50 kinds of SD, and all SD may be expressed as a binary group, it may be assumed that sdType (governor,
Dependent), wherein governor indicates that leading word, dependent indicate attached word.Each sentence will parse
Gather to corresponding SD.Such as it can be with for sentence " This mashup allow you to upload your photo. "
Obtain SD set are as follows:
det(mashup,This)
nsubj(allow,mashup)
root(ROOT,allow)
dobj(allow,you)
nsubj(upload,you)
mark(upload,to)
xcomp(allow,upload)
nmod:poss(photo,your)
dobj(upload,photo)
(3) it is analyzed based on the SD set that step (2) obtain and extracts the service for indicating demand for services functional semantics
Functional information collection.
(3.1) firstly the need of extracting basic service functional information collection, after comprehensive analysis SD collection can induction and conclusion be 4 kinds of feelings
Condition:
A) directly Transformational Grammar relationship dobj:
The part governor of dobj is converted directly into the part Verb in basic function information, and the part dependent is straight
Switch through as at the part Object in basic function information.
Such as sentence " This mashup allow you to upload your photo. " includes grammatical relation dobj
(upload, photo) then can directly be converted to basic function information collection { upload photo }.
B) directly Transformational Grammar relationship nsubjpass:
The part governor of nsubjpass is converted directly into the part Verb in basic function information, the portion dependent
Divide the part Object directly switched into basic function information.
Such as sentence " Your photo will be uploaded to the server. " includes grammatical relation
Nsubjpassc) comprehensive analysis grammatical relation prep_p and nsubj:
It, can be with if the part governor in grammatical relation prep_p is identical as the part governor in nsubj
The part governor in prep_p is converted into the part Verb in basic function information, the part dependent is converted into base
The part Object in this functional information.
Such as example sentence " The user can search for music information. " includes grammatical relation nsubj
(search, user) and prep_p (search, information), therefore available basic service functional information collection
{search information}。
D) potential function information collection is excavated using grammatical relation conj:
The business movement { p (verb), p (object) } in sentence has been obtained,
If there is conj (p (verb), dependent), then there is its business movement { p (dependent), p arranged side by side
(object) },
If there is conj (governor, p (object)), then there is its business movement { p (verb), p arranged side by side
(governor)}。
Such as example sentence " This mashup allow you to upload and share your photo and
Video. " can directly extract basic service functional information { upload video }, then by it includes grammatical relation
Conj (upload, share) and conj (photo, video) can find potential service function information collection { (upload
video),(share photo),(share video)}。
(3.2) it is semantic there may be service function is lost to be based on the basic service functional information collection that step (3.1) extract
The case where, it is therefore desirable to the extension of service function information semantic is carried out to such case.Semantic extension is mainly for functional information
Noun part, and mainly for including noun, the noun qualifier of gerundial form.It is related to SD concentration, mainly considers language
Method relationship nn.Grammatical relation nn (governor, dependent) indicates that governor and dependent is noun, and
Dependent modifies governor as qualifier.It is therefore desirable to by the part dependent of grammatical relation nn as semantic
Extension is added in service function information.Such as sentence " The user can search for music information. "
Can directly extract basic service functional information { search information }, sentence include grammatical relation nn (music,
Information), by the way that service function information { search music information } will be obtained after semantic extension.
(3.3) the basic service functional information collection obtained based on step (3.2), it is still desirable to be further processed.Firstly, mentioning
In the functional information taken out, lack practical semantic service function information there are some, that is since these service functions are believed
Include some not semantic weak verbs in breath.Therefore, it is necessary to establish one to stop vocabulary, the service function for stopping vocabulary will be belonged to and believed
Breath is deleted.Then, speech reduction is carried out for obtained service function information collection, verb and noun is mainly reduced into lemma,
Convenient for follow-up service similarity calculation.Speech reduction method is realized using the Stemming algorithm based on WordNet.
As shown in Figure 1 it is a simple service describing text, is based on step (3.1) and (3.2) available basic service
Functional information collection be (is platform), (is network), (upload photos), (upload videos),
(share photos), (share videos), (search video information) }, lacked by stopping vocabulary removal
Semantic service function information collection { (is platform), (is network) }, will finally obtain service function by speech reduction
It can information collection { (upload photo), (upload video), (share photo), (share video), (search
video information)}。
(4) similarity between servicing is calculated.Service is described with the example of real data on the website ProgrammableWeb below
Between similarity calculation process.
ProgrammableWeb is famous at present API service and Mashup application service LOGIN directory website, it is provided
Based on the API information that Web and mobile phone are applied, a API service more than 10000 has been enumerated up to now, on ProgrammableWeb
With more than 6000 a Mashup application services.It is as shown in table 1 two of them service describing text and its service function extracted
Semanteme collection.Similarity calculation between service are as follows:
Wherein, m=3, n=4, and need to calculate service BestParking Parking Finder service function information
3 functional informations are concentrated to concentrate the similarity of 4 service function information with service Chart.FM service function information respectively.Service
The calculating formula of similarity of functional information are as follows:
Wherein, V1, V2Respectively indicate service function information fi1,fi2In verb, Ni1, Ni2Respectively indicate service function letter
Cease fi1,fi2In noun, t1, t2The weight of corresponding verb part and noun part, Sw(Ni1,Ni2) indicate the similar of two words
Degree, calculation formula are as follows:
The similarity calculation result of the two final services are as follows:
Since the two service generics are different, their similarity value is not high.Using same calculating side
Formula carries out service similarity value two-by-two to 311 services randomly selected and calculates.
1 service describing text of table and service function information collection
(5) similarity value between the service obtained based on step (4), clusters service using K-means algorithm.Cluster
Algorithm is as follows:
Input: services set NF, cluster class number k
Output: k service cluster result
1. initializing k cluster class set Ci (i=0 ... k)
2. from NFInitial center point service of the k service of middle random selection as k cluster;
Before and after 3.WHILE twice cluster result it is not identical
4. calculating each service s ∈ NFITo the similarity of these cluster centre point services, and it is stored in array similarity
[k];
5. taking the corresponding subscript of max (similarity [k]), service is divided into the corresponding cluster of the small tenon;
6. resetting the central point in each cluster
7.}
8. exporting k cluster
9.RETURN
Here k=3 is set, using similarity value between service as the distance between different elements.3 generated in algorithm execution
As shown in table 2, central point service represents the classification situation of the cluster class to a certain extent for central point service.Table 3 illustrates most
Whole cluster result.By table 2, table 3 is as can be seen that the functional information collection of classification 1 central point service Bitcoin-Contact embodies
Many related Email receive and dispatch correlation function, and the functional information collection for 2 central point services of classifying then is embodied and largely grasped about video
The information of work, the functional information collection for 3 central point services of classifying situation it does not appear that it is classified.After analysis cluster
Classified service can be found that most of service in classification 1 is closely related with mail transmission/reception, information exchange;Service in classification 2
It is then mostly social about video multimedia operation, internet;It is then mainly supported comprising medical information in classification 3, content provides etc.
Service.
The classification of 2 cluster centre point of table and functional information collection
3 cluster result of table
Claims (4)
1. a kind of service clustering method based on demand semanteme, which comprises the following steps:
(1) service description document is obtained, subordinate sentence processing is carried out to document text;
(2) syntax parsing is carried out using Stanford Parser as unit of sentence, obtains SD set;
(3) it is analyzed based on the SD set that step (2) obtain and extracts the service function for indicating demand for services functional semantics
Information collection;
(4) the service function information collection obtained based on step (3) calculates the similarity two-by-two between service;Similarity between service
Calculation formula are as follows:
Wherein, Ss(s1,s2) indicate service s1And s2Similarity;Service s1With s2The functional information number for including is not identical, wherein
M is service s1, s2Functional information concentrates the functional information number less comprising functional information number, and n is that functional information is concentrated comprising function
The more functional information number of Information Number;Sf(fi1,fi2) indicate service s1In i-th functional information and service s2In i-th of function
The similarity of information, calculation formula are as follows:
Wherein, V1, V2Respectively indicate service function information fi1,fi2In verb, Ni1, Ni2Respectively indicate service function information fi1,
fi2In noun, t1, t2The weight of corresponding verb part and noun part, Sw(Ni1,Ni2) indicate two words similarity,
Calculation formula is as follows:
Wherein, F (Ni1)、F(Ni2) respectively indicate word Ni1、Ni2Feature set, I (F (Ni1))、I(F(Ni2))、I(F(Ni1)∩F
(Ni2)) respectively indicate feature set F (Ni1)、F(Ni2)、F(Ni1)∩F(Ni2) Information Number that includes;
(5) similarity between the service obtained based on step (4), clusters service using K-means algorithm, k value is artificial
Setting, using service similarity value as the distance value in K-means algorithm;Finally export k service cluster class.
2. a kind of service clustering method based on demand semanteme according to claim 1, which is characterized in that the step
(3) further comprise following sub-step:
(3.1) basic service functional information extracts: in the SD collection basis that step (2) obtain, carrying out to following four situation
Basic service functional information collection is extracted in analysis;
A) directly Transformational Grammar relationship dobj:
The part governor of dobj is converted directly into the part Verb in basic function information, and the part dependent directly turns
For the part Object in basic function information;
B) directly Transformational Grammar relationship nsubjpass:
The part governor of nsubjpass is converted directly into the part Verb in basic function information, and the part dependent is straight
Switch through as the part Object in basic function information;
C) comprehensive analysis grammatical relation prep_p and nsubj:
If the part governor in grammatical relation prep_p is identical as the part governor in nsubj, can incite somebody to action
The part governor in prep_p is converted into the part Verb in basic function information, and the part dependent is converted into basic
The part Object in functional information;
D) potential function information collection is excavated using grammatical relation conj:
The business movement { p (verb), p (object) } in sentence has been obtained,
If there is conj (p (verb), dependent), then there is its business movement { p (dependent), p arranged side by side
(object)};
If there is conj (governor, p (object)), then there is its business movement { p (verb), p (governor) } arranged side by side;
(3.2) service function information semantic extends: in the basic service functional semantics information obtained in step (3.1) there may be
It the case where semanteme missing, in response to this, will continue to SD set analysis, and the semantic extension of missing to service function believed
In breath;Here it only needs to consider grammatical relation nn;
(3.3) it generates service function information collection: further processing is needed to the service function information that step (3.2) obtains, it is main
To work comprising two parts: 1) creation stops vocabulary, and removal lacks semantic weak verb;2) word that service function information is concentrated
Speech reduction is carried out, follow-up service similarity calculation is convenient for.
3. a kind of service clustering method based on demand semanteme according to claim 1, which is characterized in that the step
(2) the natural language processing tool in is Stanford Parser, and SD refers to the grammer that Stanford Parser is parsed
Relationship binary group.
4. a kind of service clustering method based on demand semanteme according to claim 1, which is characterized in that the step
(4) between word similarity calculation using Lin algorithm;Wherein feature set F (w) is based on WordNet dictionary.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510961111.3A CN105404693B (en) | 2015-12-18 | 2015-12-18 | A kind of service clustering method based on demand semanteme |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510961111.3A CN105404693B (en) | 2015-12-18 | 2015-12-18 | A kind of service clustering method based on demand semanteme |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105404693A CN105404693A (en) | 2016-03-16 |
CN105404693B true CN105404693B (en) | 2018-12-25 |
Family
ID=55470182
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510961111.3A Active CN105404693B (en) | 2015-12-18 | 2015-12-18 | A kind of service clustering method based on demand semanteme |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105404693B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105930443B (en) * | 2016-04-20 | 2017-06-27 | 武汉大学 | A kind of object-oriented RESTful Web services find method |
US10148740B2 (en) * | 2016-06-03 | 2018-12-04 | Microsoft Technology Licensing, Llc | Multi-service application fabric architecture |
CN109495556B (en) * | 2018-10-31 | 2021-04-16 | 深圳高灯计算机科技有限公司 | File processing method and device |
CN109992244B (en) * | 2018-12-06 | 2022-08-02 | 中国船舶重工集团公司第七0九研究所 | Discrete division method for requirements of service information system |
CN110457466A (en) * | 2019-06-28 | 2019-11-15 | 谭浩 | Generate method, computer readable storage medium and the terminal device of interview report |
CN111475608B (en) * | 2020-02-28 | 2022-06-17 | 浙江工业大学 | Mashup service characteristic representation method based on functional semantic correlation calculation |
CN114091473B (en) * | 2022-01-20 | 2022-05-03 | 北京建筑大学 | Web service discovery method based on comprehensive semantics |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101695082A (en) * | 2009-09-30 | 2010-04-14 | 北京航空航天大学 | Service organization method based on relation mining and device thereof |
CN103440343A (en) * | 2013-09-11 | 2013-12-11 | 武汉大学 | Knowledge base construction method facing domain service target |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9378202B2 (en) * | 2010-03-26 | 2016-06-28 | Virtuoz Sa | Semantic clustering |
KR20130060720A (en) * | 2011-11-30 | 2013-06-10 | 한국전자통신연구원 | Apparatus and method for interpreting service goal for goal-driven semantic service discovery |
-
2015
- 2015-12-18 CN CN201510961111.3A patent/CN105404693B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101695082A (en) * | 2009-09-30 | 2010-04-14 | 北京航空航天大学 | Service organization method based on relation mining and device thereof |
CN103440343A (en) * | 2013-09-11 | 2013-12-11 | 武汉大学 | Knowledge base construction method facing domain service target |
Non-Patent Citations (1)
Title |
---|
余海飞.基于服务关联的服务推荐和发现方法研究.《中国优秀硕士学位论文全文数据库》.2015,(第5期),第27-42页. * |
Also Published As
Publication number | Publication date |
---|---|
CN105404693A (en) | 2016-03-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105404693B (en) | A kind of service clustering method based on demand semanteme | |
CN109033387B (en) | Internet of things searching system and method fusing multi-source data and storage medium | |
CN102855312B (en) | Domain-and-theme-oriented Web service clustering method | |
US20170242934A1 (en) | Methods for integrating semantic search, query, and analysis and devices thereof | |
WO2016112679A1 (en) | Method, system and storage medium for realizing intelligent answering of questions | |
CN102831253B (en) | Distributed full-text retrieval system | |
CN102184227A (en) | General crawler engine system used for WEB service and working method thereof | |
CN103092943A (en) | Method of advertisement dispatch and advertisement dispatch server | |
Wu et al. | Searching services" on the web": A public web services discovery approach | |
KR20210002619A (en) | Creation of domain-specific models in network systems | |
CN109325173B (en) | Reading content personalized recommendation method and system based on AI open platform | |
JP6130270B2 (en) | Comment list public server, program and method for sorting and specifying comment sets corresponding to media contents | |
CN110990057A (en) | Extraction method, device, equipment and medium of small program sub-chain information | |
US10423678B1 (en) | Query language filter for cross-language information retrieval | |
US8954438B1 (en) | Structured metadata extraction | |
CN116628328A (en) | Web API recommendation method and device based on functional semantics and structural interaction | |
KR101931859B1 (en) | Method for selecting headword of electronic document, method for providing electronic document, and computing system performing the same | |
CN103095849A (en) | A method and a system of spervised web service finding based on attribution forecast and error correction of quality of service (QoS) | |
Jiang et al. | A semantic-based approach to service clustering from service documents | |
CN101853307A (en) | Note establishing method, corresponding network searching system and method thereof | |
CN107657060B (en) | Feature optimization method based on semi-structured text classification | |
KR20170136679A (en) | Method for recommending konwledge contents based on big data and system at the same | |
Kamath et al. | Similarity analysis of service descriptions for efficient Web service discovery | |
KR101746594B1 (en) | push message providing system based on web crawler by learning and following user search history | |
CN104598482A (en) | Method for updating book information based on depth-first search strategy |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
TR01 | Transfer of patent right | ||
TR01 | Transfer of patent right |
Effective date of registration: 20220516 Address after: 214101 Room 308, floor 3, Jizhi Plaza office building, No. 19, Erquan East Road, Xishan District, Wuxi City, Jiangsu Province Patentee after: Wuxi ruitesipu Intelligent Technology Co.,Ltd. Address before: 310018, No. 18 Jiao Tong Street, Xiasha Higher Education Park, Hangzhou, Zhejiang Patentee before: ZHEJIANG GONGSHANG University |