II.
Scientific and Technological Key Points of the Novelty Retrieval Project
The “Interactive Transaction Supporting Platform of Business Information Matching Oriented to Small- and Medium-sized Enterprises” developed by this project aims to provide platform customers, especially Small- and Medium-sized enterprises, rapid business information matching services, multichannel communication services and safe and reliable electronic transaction services. The specific research contents include:
(1)
On account of the cross-media property of text, audio, image and video containing the business information of Small- and Medium-sized enterprises, carry out feature extraction of the massive information on the platform and establish a feature base connecting goods information with business information; then based on OEM model, describe the feature relationship of and organize the storage of such cross-media data, thus supporting the multiple and comprehensive services of the massive goods resources data and business information in the platform.
(2)
On the basis of the above-mentioned methods, construct goods knowledge base, feature base and professional database oriented to different markets and different industries. According to the features of industries and the categories of goods, separately build corresponding knowledge base, feature base and professional knowledge base.
(3)
On the basis of the relationship description and organization of the OEM model-based goods feature data, build a knowledge base associated to XML-based goods resources features and semantic relationship sheet of goods resources, realize the semantic representation of goods and business information of the platform and then propose the automatic semantic matching technologies of Agent-based supply and demand information, thus supporting the rapid and accurate matching of demand and supply information of goods of the platform.
2
(4)
On account of the hysteresis quality and non-real-time quality existed in the transaction communication, design and develop the voice call transfer and callback technologies and bring forward the multichannel interactive way that integrates “telephone network, internet and text message network”; bind the key words, including “company name”, “brand”, “trademark”, “product/service name”, “industry name” and “personal name”, to the telephone of the merchant. All paging requests through key words in telephone communication network, text message network and internet can be directly led to the telephone number of the merchant who registers such key words, thus enabling the real-time interaction and communication between the calling party and the merchant. On account of the heterogeneous structure of the online information communication, develop self-adaption model oriented to the information exchange in different environments; provide customers with all-day and multichannel real-time communication services.
III.
Novelty Retrieval Points and Requirements
Novelty Retrieval Points
1.
Adopt OEM model-based relationship description and organizational mechanism of cross-media business information feature data and support the comprehensive service mode of text, audio, image and video containing massive goods resources data and business information.
2.
Build a semantic relationship sheet between goods resources and business information based on XML and Ontology technology, and then through the Agent-based supply and demand semantic matching technology, support the rapid and accurate matching of the goods supply and demand information on the platform.
3.
On the basis of voice call transfer and callback technology, bind the key words of business information to the business contact phone number, absorb the advantages of voice search and web text search, realize the unique “key words paging” technology and business pattern, and support the cross-network business opportunity search that integrates “telephone network, internet and mobile network” and multichannel business real-time communication service.
Novelty Retrieval Requirements:By novelty retrieval, prove there are no same or similar products in China within the retrieved scope.
IV.
Document Retrieval Scope and Strategies
Retrieval Word or Classification Number:
OEM/Object Exchange Model; Feature; Date; Extraction; Abstraction; Description; Search; Organization; Business; XML; Ontology/Ontology; Agent/Agent; Semantic; Matching; Integration of Multiple Networks; Cross-media Three Network Integration; Three-in-one Network; Telephone Network; Internet; Key Words; Telephone
Retrieval Strategies:
CS1:
CS2 * Business * Cross-media
3
CS2:
(OEM + Object Exchange Model) * (Feature + Data) * (Extraction + Abstraction + Description + Search + Organization)
CS3:
XML * (Ontology + Ontology) * (Agent + Agent) * Semantic * Matching
CS4:
XML * (Ontology + Ontology) *Semantic * Business + (Agent + Agent) * Semantic * Matching
CS5
CS6 * Multiple Network Integration + Multiple Networks in One + Three Network Integration + Three-in-one Network + Telephone Network * Internet + Mobile Network)
CS6:
Key Words * Telephone
Applied Database or Retrieval Tool
Chinese Database:
China Patent Information Web: http://www.patent.com.cn/
China State Intellectual Property Office: http://www.sipo.gov.cn/sipo/zljs/default.htm
VIP Company Chinese Sci-Tech Journal Database: 1989~
China Sci-Tech Paper Online (http://www.paper.edu.cn/index.html): 2004~
Wanfang Company Database System:
China Enterprises, Companies and Products Database: 1988~
China Sci-Tech Results Database: 1986~
China Academic Dissertation Database: 1995~
China Academic Conference Paper Database: 1997~
Digital Journal Full-text Database:
China Patent Database: 1989~
4
CNKI China Knowledge Resource Integrated Database (http://ckrd2.cnki.net/grid20/index.aspx)
China Journal Full-text Database: 1979~
China Excellent Master Dissertation Full-text Database: 1999~
China Doctor Dissertation Full-text Database: 1999~
China Core Newspapers Full-text Database: 2000~
China Proceedings of Conference Full-text Database: 1999~
Search Engines:
Google Search Engine: http://www.google.com
Baidu Search Engine: http://www.baidu.com
Google Scholar Search Engine: http://scholar.google.com
V. Search Results
According to the research contents and novelty retrieval requirements of this project entrusted for novelty retrieval, by referring to the retrieval words and standard words of relevant database word list and making use of the retrieval strategy CS1-CS6, retrieve Chinese literature concerning “Interactive Transaction Supporting Platform of Business Information Matching Oriented to Small- and Medium-sized Enterprises” on the above-mentioned 15 Chinese literature database and 3 search engines. The relevant articles are respectively described as below according to the strategies of retrieval” (for details, see Annex I):
1.
According to retrieval strategy CS1, retrieve the Chinese literature and reports related to the abstraction, relationship description and organizational mechanism of OEM model-based cross-media business information feature data. No relevant literature is retrieved.
By using the literature and reports related to the abstraction, relationship description and organizational mechanism of OEM model-based information feature data in CS2 extended retrieval Chinese literature, about 30 relevant articles are retrieved. According to the analysis of the results of retrieval, there are literature and reports concerning the abstraction, model discovery, query and description of OEM model-based semi-structured data. The bibliographic records and abstracts of relevant literature are listed as below:
1)
Lu Mingyu, Lu Yuchang (Department of Computer Science and Technology, Tsinghua University),Model Abstraction of OEM Model-based Semi-structured Data. Journal of Tsinghua University (Science and Technology), 2004, 44(9):1264-1267.
As typical semi-structured data, Web data lack explicit and prescient external form, the storage of which is separated from data, so the query, browse and integration of Web data have extremely low efficiency. This paper proposes a model abstraction algorithm of semi-structured data based on OEM (object exchange model) and adopts the top-down pruning strategy, which can rapidly discover the frequent and simple path set and is applied in the integration of semi-structured data and the query response and improvement. It features reducing the scale of target model and effectively improving the efficiency of abstraction.
5
2)
Xu Dezhi, Luo Fang, Wu Min, et al (Information Science and Engineering College of Zhongnan University), Realization of Storage of XML Date Query. Computing Technology and Automation, 2003, 22(3):98-101.
This paper brings forward the concept of discrete XML data model based on OEM data model and develops a model matching-based data query concept based on this philosophy; then, it discusses the storage methods to ensure the high-efficiency query of XML data; finally, it conducts the experimental analysis.
3)
Nie Peiyao (Department of Computer and information Engineering, Shandong University of Finance), Li Zhanhuai (Department of Computer Science and Technology, Northwestern Polytechnical University),Methods of Model Description of OEM-based XML Semi-structured Data. Computer Engineering and Design, 2003, 24(1)”9-12, 29.
This paper introduces that the type and model of semi-structured data are the key technologies to improve the processing efficiency of semi-structured data. Firstly, it discusses the features of semi-structured data and the features of semi-structured data models. And then, it conducts research on the description methods of XML-model and brings forward a definition of OEM-based XMLDTD model and methods of formalized description.
4)
Chen Enhong, Shi Zhu, et al (Computer Department, University of Science and Technology of China),Research on the Representation and Query Methods of Semi-structured Data. Computer Engineering, 2001, 27(5):5-7.
This paper introduces the ways of abstracting useful information out of WWW webpage, organizing and storing them via OEM-based data model, and querying semi-structured date based on such storage model.
5)
Huang Yuqing, Qi Guangzhi.Technologies of Constructing Semi-structured Information in Web Documents. Journal of Computer-Aided Design & Computer Graphics, 2000, 12(3):230-234.
In order to integrate and query the irregular and dynamic information on Web in the way of database, this paper constructs a Web-based information model by using OEM. For the purpose of presenting each part on the page as applied OEM object, it designs an abstraction algorithm of semi-structured information and provides the results of tests. Such method can be used to abstract structured and semi-structured information and is more universally applicable than the existing methods of abstraction.
6)
Tao Chun, Wang Wei, Shi Bole (Department of Computer and Information Technology, Fudan University).MiniCon Algorithm of Semi-structured Querying and Rewriting. Journal of Software, 2004, 15(11):1641-1647.
This paper conducts research on the query and rewriting based on TSL (Tree Specification Language), a semi-structured data query language, brings forward an algorithm of semi-structured querying and rewriting and solves the problem of finding the maximum inclusion rewriting under the given circumstance of a semi-structure query and a group of semi-structure views. The algorithm makes use of the philosophy of MiniCon algorithm of flexible relationship querying and rewriting, solves some new problems of querying and rewriting under semi-structured data model (such as dependence on identifiers, set value variable mapping, etc.) and proves the correctness of the algorithm.
7)
Tao Chun. Research on the Querying and Processing in the Integration System of Semi-structured Data [doctoral dissertation]. Name of the supervisor: Shi Bole, Zhang Liang. Awarded by: Fudan University. Awarded on: April 1st, 2004.
6
This paper mainly focuses on the querying and processing of the integration system of semi-structured data, two issues especially: 1. the maximum querying and rewriting based on TSL (Tree Specification Language) in the integration system of OEM semi-structured data; 2. Generation and optimization of maximum query implementation plan in XML integration system based on the ontology. In terms of the first issue, in the OEM semi-structured data integration system based on TSL querying language, it gives a formalized definition of the inclusion, equivalence and maximum inclusion of the query; under such formalized framework, it brings forward the algorithm of semi-structured data querying and rewriting based on TSL query, which applies the philosophy of MiniCon algorithm of flexible relationship querying and rewriting; besides, it theoretically proves the correctness of the algorithm. In terms of the second issue, it perfectly formalizes the XML data integration system based on the ontology; under such formalized framework, it brings forward the generation algorithm of maximum query implementation plans in XML integration system based on the ontology; it introduces the probability of incomplete roles and optimizes the incomplete roles’ generation algorithm of maximum query implementation plans; besides, the paper also brings forward the query implementation plan network cost optimization algorithm.
2.
According to retrieval strategy CS3, there is semantic relationship sheet concerning constructing goods resources and business information based on XML and ontology technologies. Then, no relevant literature and reports concerning supporting the rapid and accurate matching of the goods demands and supply information on the platform through the supply and demand semantic matching technologies based on Agent is in retrieved.
By using CS4 extension, retrieve the semantic relationship sheet concerning the construction of goods resources and business information based on XML and Ontology technologies, or the literature and reports concerning the demand and supply semantic matching technology based on Agent, nearly 30 relevant articles are retrieved. Through the analysis of the retrieval results, it is found that there are literature and reports concerning the Web service matching technology based on XML and Ontology technologies and automatic matching technologies based on Agent. The bibliographic records and abstracts of relevant literature are listed as below:
8)
Chen Wei.Research on XML Information Integration System Design based on Ontology and Its Key Technologies [master dissertation]. Name of supervisor: Jin Yuanping. Awarded by: Southeast University. Award on: March 1st, 2005
On the basis of the analysis of the basic features of XM, this paper attaches importance to the research on the concept, function, construction methods and description languages of the ontology, designs a XML data integration system based on ontology, namely OB-XIIS and proposes using the ontology as the overall model to obtain and integrate isomeric XML information data so as to realize the multisource XML data resources share and information integration. The system uses OWL, which is latest proposed by W3C, as the language to construct the ontology, a language based on XQquery Improvement as the ontology, and XQuery as the XML data query language. It interprets the detailed working processes of the system with an example of operation and provides a feasible method and thought of obtaining and integrating multisource XML information.
9)
Zhang Zhe (College of Computer and Information Technology, Liaoning Normal University).Similarity of XML Model Elements based on Domain Ontology. Microelectronics and Computer, 2007, 24(4):220-224.
XML model and domain ontology are the two key features of semantic Web. The similarity between evaluation concepts is becoming the basis of the processing of e-commerce and affairs. By using the domain ontology, it brings forward the methods to evaluate the similarity between XML model elements.
7
10)
Dai Weiping (College of Computer, Southwest University).Methods and Realization of Agent-based E-commerce. Journal of Changshu Institute of Technology, 007, 21(4).-121-124.
This paper brings forward a many-to-many multi-issue e-commerce method, which takes the e-market as a neutral third party, finds a proposal with relatively high similarity among many buyer proposals and seller proposals through market matching, and then generates one-to-one cooperative consultation with mutual benefits based on Agent application fuzzy theory so as to increase the satisfaction of both trading parties and the possibility of the conclusion of the business.
11)
Chen Donglin, Nie Guihua, Liu Pingfeng (Department of E-Business, College of Economics, Wuhan University of Technology).Ontology-based B2B E-Business MAS Model and Goods Matching Algorithm. Computer Engineering and Application, 07, 3(10):199-201, 218.
This paper constructs an ontology-based multiple Agent model, adds GMAg (Goods Matching Agent) to goods identification and automatic matching, takes into comprehensive consideration the semantic similarity between the concept of goods set and concepts of property set, and brings forwards the semantic similarity matching algorithm based on the ontology structure of goods: solving the problem of automatic matching of goods information between business systems of different structures. By using an example in the automobile field, it interprets the matching accuracy and efficiency.
12)
Zhou Qihai, Zhang Yuanxin, Wu Hongyu (College of Economic Information Engineering, Southwestern University of Finance and Economics).A Two-way Intelligent Automatic Matching System Model based on Multiple Agents. Computer Application, 2006, 26(7); 1713-1714, 1729.
This paper conducts research on the two-way intelligent matching of information in computer network; it provides a two-way intelligent automatic matching system model based on multiple agents, clarifies its structure, working principle and process, and proposes the cooperative mechanism of multiple Agents and the fuzzy matching of information; it realizes the high-speed accurate two-way intelligent matching of information during the information search process and indicates the developmental direction of its parallelism.
13)
Yuan Yang, Li Shanping (College of Computer, Zhejiang University).Summary of Semantic Web-based Ontology Mapping Methods. Computer Science, 2004, 31(5):5-8.
This paper discusses three system structures of ontology mapping and summarizes the existing main ontology mapping methods based on the classification standards proposed by E. Rahm. And finally, it conducts the comparative analysis of four methods.
3. According to retrieval strategy CS5, retrieve the Chinese literature and reports concerning binding the key words of business information to business contact telephone and supporting the convergence of “telephone network, internet and mobile network” and multi-channel instant commercial communication services, and no relevant literature has been found.
CS6 extension is used to retrieve the Chinese literature concerning binding commercial information keywords to business contact telephone, and over 10 articles have been found. The retrieval results have been analyzed, and the literatures on keyword recognition system for speech based on streaming media and telephone voice keyword retrieval services have been found. The bibliographic records and abstracts of related literatures are as follows:
8
14) Lv Shuqin (Communication and Information Engineering Department, Beijing Information Technology Institute) and Sun Chengli (Information Technology School, Shijiazhuang University of Economics, Information Engineering College, Beijing University of Posts and Telecommunications).Study of Keyword Recognition System for Speech Based on Streaming Media. Journal of Beijing Institute of Machinery, 2006, 21(4): 47-50.
The paper introduces that, in recent years, the keywords detection technology has been developed greatly in speech voice and telephone voice field, but the literatures on the keyword detection for speech of streaming media are scarce; therefore, a system program for keyword detection in streaming media has been developed. WMFSDK adopted in the system is used to extract the decoded voice data from streaming media. To discriminate out-of-vocabulary words and keywords, an online garbage model is used to reject out-of-vocabulary words and get several candidates of keywords. In the keyword confirmation stage, the word confidences based on MAP and N-best characteristics obtained in decoding are used as eigenvectors, and a classifier of support vector machine (SVM) has been designed. Comparison has been conducted for the SVM method and traditional Fisher method by experiments, and the results suggest that the overall application results of the former are better than the latter.
15) Li Yi (Beijing Third Network Communication Co., Ltd.).Insist on Innovation and Be a Top Runner of the Industry-Third Network Introduced a New Network Marketing Tool “Free Communication-Call by Real Name. China Multi-media Communications, 2006 (11): 72-73.
The paper introduces that: instant voice communication, a new internet service, is changing the living styles of people quietly: when you input your telephone number on an enterprise website, in a short time, enterprises will telephone you and provide services actively, and introduce the features and price of the products and services you want to learn. Even more, when you input the keyword of the products and services you want to learn in any telephone booth of “free communication” on the website, or input the keyword to an accessing number by short message, you can get the services you desire and learn the required information immediately.
16) Li Xuesong (Nanjing University of Posts and Telecommunications).An HMM Keyword Spotting System. Telecommunications Information, 2005 (6): 26-28.
The paper introduces that: keyword spotting is a valued field of speech recognition, which can be widely applied in the automatic answer of telephone and conversation monitoring, voice input and retrieval. In the paper, the HMM program of keyword spotting and the research results are introduced, the simulation experiments have been performed, and the problems to be resolved and research directions of keyword spotting have been illustrated.
17) GOTOTELFirst Telephone Search. Telecommunications Technology, 2005, (2): 80-80.
The paper introduces GOTOTEL, a telephone search service provider cooperates with China Mobile, to provide telephone search services to 200 million users of China Mobile, namely, to transplant the speech recognition technology to telephone service field, and the original intention is to use the increasing widespread telephones to search the information and services needed by people, so as to make the lives of people simpler and more convenient. The telephone search user can carry out the search by “pick up telephone - call 1259004999 - speak out keyword - return search results - select search results - use service”.
9
VI. Conclusion of Novelty Retrieval
In the Chinese literature published in China:
1.
The literature on the pattern extraction, pattern discovery, enquiry, description for semi-structured data based on OEM model has been found. The literature on comprehensive service modes adopting relationship description and organizational mechanism of inter-media commercial information characteristics based on OEM model and supporting mass commodity resource data and text, voice, image and video of commercial information has not been found.
2.
The literature on Web service matching technology based on XML and Ontoloty and automatic matching technology based on Agent has been found. The literature on constructing semantic relationship table for commodity resources and commercial information based on XML and Ontoloty and supporting the rapid and accurate matching of supply and demand information of commodity in the platform with supply and demand semantic matching technology based on Agent has not been found.
3.
The literature on keyword recognition system for speech based on streaming media and telephone voice keyword retrieval services has been found. The literature on inter-network business opportunity search binding commercial information keywords and commercial telephone and supporting the convergence of “telephone network, internet and mobile network” and multi-channel instant commercial communication services have not been found.
![[dwarfs1ex992001.jpg]](https://capedge.com/proxy/S-1/0001078782-11-002241/dwarfs1ex992001.jpg)