378

JOURNAL OF SOFTWARE, VOL. 5, NO. 4, APRIL 2010

Conceptual Modeling for Competitive Intelligence Hiding in the Internet
Jie Zhao
School of Management, University of Science and Technology of China, Hefei, China School of Business Administration, Anhui University, Hefei, China Email: zjpq@mail.hf.ah.cn

Peiquan Jin
School of Computer Science and Technology, University of Science and Technology of China, Hefei, China Email: jpq@ustc.edu.cn
Abstract—Internet has been a major source to acquire enterprise competitive intelligence. But what types of competitive intelligence are hidden in the Internet? This paper presents an answer to this question. In this article, a conceptual framework to represent the competitive intelligence in the Internet is introduced. The basic idea is to use an entity-oriented method to capture the semantics of the competitive intelligence in the Internet. Traditional approaches regard the competitive intelligence in the Internet as a set of Web pages, which leads to a large amount of manual processing work. Oppositely, we represent them as a set of entities, and further design two types of entities, namely the entities about competitors and those about competition environment. Moreover, the detailed semantics of each entity are discussed. We develop formal approaches to represent these types of competitive intelligence. As a result, a conceptual model for the competitive intelligence in the Internet is proposed, which gives an explicit description for the competitive intelligence in the Internet. Finally, a case study to demonstrate the feasibility of our conceptual model is explained. The results show that it is feasible to use the proposed conceptual model to capture the semantics of the competitive intelligence in the Internet. Index Terms—competitive intelligence; Internet; conceptual model; entity-based representation

I.

INTRODUCTION

Nowadays, enterprise competitive intelligence has become one of the major factors which determine the core competitive power of companies in the worldwide market [1]. How to efficiently and effectively obtain competitive intelligence is an urgent issue to most companies. With the development and applications of Web technologies, it is possible to obtain a large amount of competitive intelligence from the Web. Previous survey shows that about 90% of competitive intelligence can be acquired from the Internet [2]. This is a challenge and also a chance for enterprises to enhance their competition powers. If we can build a system to acquire competitive intelligence from the Web automatically, it is expected that such a system will bring more and more

effective competitive intelligence for enterprises and thus makes it advanced in the market competition. However, according to the state-of-the-art of the research on competitive intelligence, a lot of issues are needed to be further studied [3]. Currently, most Competitive Intelligence Systems (CIS) are only able to search pieces of information from the Web, typically using a search engine like Google and Yahoo. It is usually inefficient and ineffective because search engines will return much useless information, e.g., advertisings. Recently, the text mining techniques are introduced into CIS and some tools are developed [4, 5]. Unfortunately, while these tools are capable of filtering non-related text blocks in Web pages, it divides a Web page into a set of text blocks. This eventually brings more information processing work into the analysis of competitive intelligence, because the information processing and analysis work is often required to be done manually in current CIS tools. Thus, many issues must be revised in order to develop a Web-based CIS system, among which the most important thing is to make it clear what types of competitive intelligence we can obtain from the Internet, and what details of those competitive intelligence we can extract, or in other words, to develop an ontology of the competitive intelligence in the Internet [6]. Traditional CIS tools regard the competitive intelligence in the Internet as a set of Web pages, which results in a large amount of information processing work for the generation of competitive intelligence. This paper aims at developing an ontology and a conceptual model for the competitive intelligence in the Internet. The major contributions of the paper are summarized as follows: 1. We introduce an entity-based representation method to capture the semantics of the competitive intelligence in the Internet, where each type of competitive intelligence is represented as a set of entities. Compared to the traditional Web-page-based viewpoint, our approach is more suitable for real applications, because applications always need a clear and structured competitive intelligence about competitors or competitive environment, but not just a set of Web pages.

© 2010 ACADEMY PUBLISHER doi:10.4304/jsw.5.4.378-386

news paper. suppose a company wants to get the competitive intelligence about one of its competitors.1. The Web-pagebased way has several shortcomings. number entities. Its main task is to recognize and classify the specific names and meaningful numeric words from the given texts. etc. employees. and Section 5 describes the conceptual model for the competitive intelligence in the Internet. Web pages are only the media that contains the needed information. and then transforming them into intelligence [1]. among which the most serious one is that search engines will return a huge amount of Web pages. e. namely time entities. rather than just the Web pages that contain the basic information. As shown in Fig.g. In this paper. Competitive intelligence is acquired. people can search any information in a real-time way.1 shows a high-level architecture of the entity-based competitive intelligence acquiring in the Internet. Furthermore. this method will consequently produce a large number of text blocks for each Web page. and so on. times. we focus on the entity-based competitive intelligence acquiring from the Internet and present an entity-based representation framework for competitive intelligence. the information about competitors or competition environment will spread among different clusters and introduce additional work for information analysis. and a set of text blocks are obtained and even © 2010 ACADEMY PUBLISHER clustered in some rules. produced and transmitted through competitive intelligence systems (CIS). 5] by using some fundamental technologies. 5. This means we should transfer the Web-page-based viewpoint into an entity-based viewpoint.” Are these information only Web pages? The answer is definitely “no”. We develop an ER (Entity-Relation)-based conceptual model for the competitive intelligence in the Web. typically using some keywords like “C Company”. the semantics of the competitive intelligence in the Internet is analyzed. the entity-based representation framework is applied in this system to extract and produce different granularities of entities. most people prefer to look up information by competitor. statistical analysis. addresses. you will get billions of Web pages.g. Competitive intelligence serves companies and people. Section 4 presents the framework to represent the competitive intelligence in the Internet. researchers introduced Web-based text mining approach into CIS. For instance.JOURNAL OF SOFTWARE. which will bring much more work to experts. Section 2 briefly introduces the related work. statistical methods.. such as computing linguistics. For example. In Section 3. or other industry reports. Most of the previous research in this field focused on three types of named entities. first we should study what competitive intelligence companies need. The major methods used in named entity extraction include rule-based approaches. when you search in Google using the keywords “Microsoft Office 2008”. VOL. thus it has become an important way to obtain competitive intelligence from the Web [2]. so in order to make competitive intelligence systems more effective. they will first search the information about the company C through some search engines. managers. and the results show it is practical and useful to represent the needed competitive intelligence for enterprises. NO. person names. but not just deliver Web pages or the text blocks in them. The detailed procedure to produce competitive intelligence from the Web can be described as follows. analyzing and delivering information about competition environment and competitors. Traditionally. RELATED WORK Competitive intelligence refers to the process that gathering. but note they are NOT competitive intelligence. Typical named entities are company names. Through the text-miningbased approaches. products. The following of the paper is structured as follows. All the competitive intelligence needed for a specific company can be described in such a conceptual model. people will utilize some publications to acquire competitive intelligence. as well as hybrid methods [10]. the noisy data in Web pages can be eliminated. In this paper. and information retrieval. When we further ask one more question: “What is the competitive intelligence about competitors?”. The term is called Named Entity Recognition. With the rapid development of the Web. In other words. Then some experts will analyze the gathered Web pages to make out a report about the company C. their history. Named entity recognition was first introduced as a subtask in the Message Understanding Conference (MUC) [11]. Recently. 4. little has been done for competitive intelligence acquiring. magazines. However. Fig. CIS should output competitive intelligence about entities such as competitors (or sub-entities such as the products of a specific competitor). . II.. it is ultimately not feasible for experts to analyze all the searched results and produce valuable competitive intelligence. Some specific methods have been proposed to process Web pages so far. However. although there are some existing work in named entity extraction from text (or Web pages). most companies will present the answer: “We want to know everything about our competitors. APRIL 2010 379 2. if the text blocks are clustered under specific rules. machine learning. we call this type of intelligence acquiring “Web-page-based Competitive Intelligence Acquiring”. which are corresponding to different types of competitive intelligence. This approach aims at finding implicit knowledge from a lot of text data [4. and organization entities [12]. We describe a case study to demonstrate the feasibility of the model. As a survey reported [9]. a company C. CIS is expected to produce competitive intelligence about competitors or competition environment from a large set of Web pages. Conclusions and future work are in Section 6. including extracting text blocks from Web pages [7] and detecting changes of Web pages [8]. e. Extracting entities from Web pages is one of the hottest issues in Web information extraction and retrieval [10].

According to the traditional theories about competitive intelligence. etc. A.3 is an example about IBM’s recent events. etc. we need a filtering mechanism to distinguish true information from those false ones [13]. Example of profile intelligence extracted from the Web 2. Being served. its data volume is too large to be processed efficiently. The disadvantages of Web data source should also be mentioned here. some internal information is usually not posted in the Web. listed stock. Competitors In business areas. more fresh means more valuable. APRIL 2010 TABLE I.g. Examples Business Relations relations with time-related facts other about a company organizations establishment of suppliers of the the company. VOL. release of new investors. etc. Many websites such as Wikipedia provide some general information about companies. this topic is out of the scope of this article. According to the time requirement of competitive intelligence. employee counts. Events 1. 4. SEMANTICS OF THE COMPETITIVE INTELLIGENCE IN THE INTERNET Compared to the traditional sources of competitive intelligence. some competitor) may post false information in the Web. Through the events expressed in the news. competition environment. we are able to know the recent development of competitors. as shown in Table 1. only the first two types of competitive intelligence may be available in the Web. Since the competition strategies are usually not posted in the Web due to the security reason. which are related with competitors. company. 5. which is the general information of the Microsoft Corporation in Wikipedia. Hence. telephone number. Fig. which are extracted from its Website. © 2010 ACADEMY PUBLISHER . products. the progress of some specific project. Many websites provide news which is updated frequently. there are three types of competitive intelligence. NO. However. Figure 1. etc. such as names. The architecture of the entity-based competitive intelligence acquiring in the Internet III. TYPES OF COMPETITIVE INTELLIGENCE ABOUT COMPETITORS Profile Definition the basic information about a company company name. and meanwhile the Web data is updated frequently. managers’ names. the Web has a larger volume of data. competitors refer to other companies that may contest with you in some business fields. the listed-in-stock. Competitive information about competitors contains three aspects. Fig. products set.2 shows an example. Figure 2. Profile The profile is the general information about competitor. staff customers reduction. managers’ names.380 JOURNAL OF SOFTWARE. etc. Events Events about competitor usually refer to the news about it. Second. Typical events are the establishment of a competitor company. and competition strategies. And third. First. which means they usually contain fresh information. due to the security reason. address. the data in the Web are usually less confidential than printed documents since some one (e.

you may have some countermeasures to control those suppliers so as to leave the competitor in a passive situation. economics. Example of business relations in the Web A. laws.2. FAX. the policies issued by the government. we will discuss the detailed sublevels of entities about the 2nd level of entities. business relations are usually more implicit. a specific competitor or a specific business area. the LOCATION is the site that the event takes place. This is inspired by the fact that users need competitive intelligence about entities. the habits about the resistance in a specific area or city. Example of events in the Web 3. the Web page in Fig. society. TYPES OF COMPETITIVE INTELLIGENCE ABOUT COMPETITION ENVIRONMENT Macro Environment the environmental information about the politics. this type of competitive intelligence may be more useful than others. e. The high levels of entities for competitive intelligence Entities for Competitors A competitor consists of three sub-levels of entities. However. customers. PROFILE (C) :: = {NAME. VOL. PRODUCTS. Business Environment Definition the business or industry state that a company belongs to Examples the companies set. traditions of the nation. An upper-level entity consists of all the sub-levels of entities. Figure 5. Figure 3. where the AREA indicates the business area of the company C. IV. local policies. Thus the representation of competitive intelligence should fulfill such requirements because it acts as the foundation of the extraction of competitive intelligence (see Fig. and the TIME is the happened instant or the lasting time period of the event. ADDR. government laws. 1. To obtain the business relations about competitor. from which we get to know that Oracle Corporation is the major shareholder of the Virtual Iron Software Corporation. NO. B. For example. Profile The profile of a competitor is represented as the entity “Profile”.1 and 4. 4. and nature. e. state-of-the-art of specific technology. suppliers. © 2010 ACADEMY PUBLISHER . alternatives of products. TEL.1). ENTITY-BASED REPRESENTATION OF THE COMPETITIVE INTELLIGENCE IN THE INTERNET The general idea of the entity-based representation of the competitive intelligence in the Internet is to define different sets of entities for the elements listed in Table 1 and Table 2. namely profile. EMAIL. and relations.JOURNAL OF SOFTWARE. 5. economic plans. potential new comers in a specific area. It is structured as follows. There is also 3rd level or even 4th level of entities if necessary. This is because most companies do not want their competitors know their suppliers or customers. science & technologies. Fig. if you know exactly the suppliers of some competitor. Competition Environment The competition environment refers to many aspects of surrounding information that may have impacts on the business development of the company. we must perform an intelligent analysis on the contents of Web pages. Business relations Differing from profile and events. where the TOPIC is the event’s topic. Figure 4. 4. It is structured as follows. etc.4 shows the companies acquired by Oracle Corporation during the last three years. events. cultures. For example. Events The events of a company is represented as the entity “EVENTS”. APRIL 2010 381 TABLE II.5 shows the high three levels of entities for competitive intelligence. Table 2 lists the two types of competition environment as well as their explanations that are possible to be extracted in the Web.g. In the following sections.. AREA} 2.g.

D) :: = { C. ITEMS} Here. CUSTOMERS (A. D) :: = { C.7. while the AMOUNT is the number of customers corresponding to a given TYPE.5). APRIL 2010 EVENTS (C) :: = {TOPIC.8). The PRODUCTS entity is the set of different products. The sub-entities for the relations entity These relations are defined as follows. The ALTERNATIVES entity is the set of alternatives of products. Here we do not repeat their definitions. where the COMPETITOR refers to a competitor entity (see Fig. DEPT } Here. D) means that the company D is a supplier of the company C. The binary relation COOPERATION(C. The sub-entities for the macro environment entity The politics environment of a given business field A in a specific geographical area L is represented as the POLITICS entity. Thus in our representation framework. 5. which is defined as follows. D) refers to the historical cooperation of the company with another one. due to the unfeasibility of constructing a complete description of the macro environment. D.. The CUSTOMERS entity is the types of different customers in the given business filed and local area. as shown in Fig. D. AREA. the TOPIC refers to the detected new topic of the political information. which represent the potential products that may enter into the business environment. The binary relation SALES(C. we concentrate on representing and extracting the new coming information about the macro environment. PERIOD. The COMPANIES entity of a given business field A in a specific geographical area L is defined as follows. AREA. The COMPANIES entity refers to a set of company entities. 2. TYPE} Here. B. Macro Environment The macro environment contains many aspects of factors. VOL. Figure 7. Business Environment The business environment entity consists of six subentities (see Fig. such as the name of a political meeting. the BMW corporation is one of the IBM’s customers. because many laws or policies are not available in the Web. However. The definitions of the other six entities are much like that of the POLITICS entity. Thus D can be looked as one friend of C’s. 1. These companies are the possible new competitors in the business environment. D) :: = { C. TYPE. it is still possible to obtain some information about the changes of the macro environment. DATE. INVEST (C. the TYPE is the different customer types. whereas TYPE is the cooperation type. Entities for Competition Environment The entity of competition environment consists of two sub-leveled entities. PERIOD refers to the time duration of the cooperation. the ITEM is the set of the supplied item names. The binary relation SUPPLY(C. This relation is very important in business competition. The DEPT is the department name who issues the new topic. PERIOD. LOCATION. Relations The business relations of a company consist of a sublevel of entities. which are shown in Fig. AREA. Here. TIME} 3. The CUSTOMERS entity is defined as follows. For example. AREA means the business field. D. It is difficult to obtain a complete description of the macro environment. new laws or policies usually appear in news reports. it is very possible for C to cooperation with these companies in the future.382 JOURNAL OF SOFTWARE. L) :: = { COMPETITOR } The SUPLLIERS and NEWCOMERS entity has the same definition as the COMPANIES entity. There are also some sub-entities of the macro environment. The Figure 6. AMOUNT. POLITICS (A. the attributes have similar meanings as which are in the relation COOPERATION. e. AMOUNT is the total contracted currency amount. The business environment refers to the whole business situation about a specific business field in a given geographical area. COOPERATION (C. L) :: = { TOPIC. which enable us to extract at least the name and area of the new laws or policies. PERIOD } SUPPLY (C. 4. D) indicates that the company D is one of the customers of the company C. The NEWCOMERS represents those companies that may enter into the business field. PERIOD. D) refers to that the company D is one of the investors of the company C. COMPANIES (A.6. AMOUNT} The PRODUCTS entity and the ALTERNATIVES entity have the same representation structure. The SUPPLIERS entity is the set of the supplier entities. © 2010 ACADEMY PUBLISHER . NO.g. AMOUNT} Here. If a COOPERATION relation exists between C and D. The binary relation INVEST(C. The DATE is the issued date of the topic. D. L) :: = { TYPE. D) :: = { C. SALES (C. the macro environment entity and the business environment entity. since the customers of a company usually determine its incomes. The TYPE is the type of the information.

e. we introduce a conceptual model for the competitive intelligence in the Internet. There are three types of domains in the conceptual model. Many CASE (Computer Assisted Software Engineering) tools support the ER-based conceptual modeling procedure. we present a case study to demonstrate our representation framework. each of which is a set of entities. Set of <relation>.9.11 shows the single <Profile> entity about Oracle. The basic notations used in the model are described in Fig. the competitive intelligence about competitors consists of three sub-entities. The ER method is widely used in data engineering to conceptually model the application data in the real world. General types Entity references common types to represent an atomic value a reference to a specific entity a finite enumeration of some values a set of entities. we use the Sybase PowerDesigner to build the ER-based conceptual model for competitive intelligence. L) :: = { PRODUCT_NAME. Enumeration Entity set Figure 9. a case study is discussed at the end of this section. In this section. as listed in Table 3. TYPE} Figure 8.10.1 to generate the competitive intelligence about Oracle.JOURNAL OF SOFTWARE. the competitive intelligence about Oracle consists of a set of <Profile> entities. E. Suppose a company wants to find the competitive intelligence about one of its competitors. and Fig. In order to demonstrate the feasibility of the model. <competitor>. The “Relation” entity and “Environment” entity both have several sub-entities. string.7. CONCEPTUAL MODELLING OF THE COMPETITIVE INTELLIGENCE IN THE INTERNET The Conceptual Model The conceptual model for the competitive intelligence in the Internet is shown in Fig. A. C. Domain DOMAINS IN THE CONCEPTUAL MODEL Description Example B. VOL. etc. as shown in the figure. integer. Events. Each attribute is composed with a name and a domain. The sub-entities for the business environment entity V. N}. date. etc. Methods and Notations We use the ER (Entity-Relation) method [14] to build the conceptual model for the competitive intelligence. and Relations. a set of <Event> entities.5. Note an entity “Competitor” has exact one “Profile” sub-entity. we use fourteen entities to conceptually model the competitive intelligence in the Internet. Hence. We first define the notations. APRIL 2010 383 PRODUCT_NAME and TYPE denote the product name and product type respectively. and then present the conceptual model. the Oracle Corporation.. L. C.10. The intelligence extraction and generation approaches are out of the scope. The conceptual model is based on the representation framework defined in Section 4 and focuses on the graphical representation of the competitive intelligence. TABLE III. Each entity has an entity name and some attributes describing its characteristics. and it consists of a set of “Competitor” entities and a set of “Environment” entities. As shown in Fig. namely the Profile. The entity “Competitive Intelligence” represents the root entity in Fig.12 shows the set of <Event> entities about Oracle. representing the seven types of macro environment in Fig. 5. Fig. Set of <event>. as well as a set of <Relation> entities. As shown in Fig. It runs the whole process shown in Fig. real. In this paper. etc. PRODUCTS (A. while has a set of “Event” entities and a set of “Relation” entities. NO.g. © 2010 ACADEMY PUBLISHER . 4. Notations used in the conceptual model A Case Study In this section. and we will not discuss them here.10. <customer>. Now we just want to know whether our presented conceptual model is able to represent the extracted competitive intelligence about Oracle. S. {P. T.

APRIL 2010 Environment macro environment <macro environment> business environment <business environment> consists_of Competitive Intelligence Competitors Set of <competitor> Competition Environment Set of <environment> consists_of is_composed_with is_composed_with has_competitors Macro identifier topic type date dept Environement {P. 5. 4.T.S.N} string string date string Business Environment companies Set of <competitor> Customers Set of <customer> Suppliers Set of <competitor> Products Set of <product> Alternatives Set of <product> Newcomers Set of <competitor> Profile Events Relations Competitor <profile> Set of <event> Set of <relation> has_events has_profile has_relations Event event_ID string topic string location location time period Relation Set of <cooperation> Set of <invest> Set of <customer> Set of <competitor> has_customer Customer type amount string number has_product company_ID name addr tel fax email Products area Profile string string string string string string Set of <product> string Cooperation Invest Customer Supply Product product_name type string string producing company1 company2 period area amount Invest <competitor> <competitor> period string number company1 company2 period Sales <competitor> <competitor> period supplying Supply company1 <competitor> company2 <competitor> period period area string items Set of <product> Cooperation company1 <competitor> company2 <competitor> period period area string amount number topic string Figure 10.L. NO.384 JOURNAL OF SOFTWARE. The <Profile> entity about Oracle © 2010 ACADEMY PUBLISHER .C.E. The conceptual model for the competitive intelligence in the Internet Figure 11. VOL.

5. which are the <Invest> entity. The set of <Cooperation> entities about Oracle Figure 14. VOL. © 2010 ACADEMY PUBLISHER . the <Cooperation> entity. The set of <Sales> entities about Oracle Figure 15. The other three entities are shown in Fig. APRIL 2010 385 Figure 12.13. and the <Supply> entity. the <Sales> entity. The set of <Sales> entities about Oracle Figure 16.14 to Fig. which represents the investors to the Oracle Corporation. The set of <Event> entities about Oracle Figure 13. The acquired intelligence about the <Sales> entity is shown in Fig.16.JOURNAL OF SOFTWARE. The set of <Invest> entities about Oracle The <Relation> entity consists of four sub-entities. 4. NO.

19-62. [14] P. G. LaMar. Whitelaw. including DEXA'09-10. Nürnberger.123-132. Information & Management. in 2000 and 1997. Lecture Notes in Computer Science 4845. A. New York: Simon & Schuster. and e-Society 2007: Integration and Innovation Orient to E-Society. [4] J. 2005. we have presented an answer to this question. 1995. 2009. information retrieval. Wing. et al. Washington DC. We studied the semantics of competitive intelligence in the Internet. Dr. pp. et al. We also presented a case study for the proposed conceptual model. China. Heidelberg. Heidelberg. L. University of Science and Technology of China. [13] X. 9-36.1 (1). (ed. “An Ontology-Based Mining System for Competitive Intelligence in Neuroscience”.). C.39 (1). Zhao is a member of SCIP (Society of Competitive Intelligence Professionals). “An Exploratory Discuss of New Ways for Competitive Intelligence on WEB 2. [5] A. Vol. Jie Zhao was born in Hefei. El-Mawas. 2005.19(5). pp. such as ICCIT’09. Persidis. Springer. “Named Entity Task Definition-Version 2. Before the design of such systems. Li. G. Hefei. Vo. China. China. Jijkoun. knowledge management. (ed. 70803001 and 60776801. In this paper.). and International Journal of Digital Content Technology and its Applications. CIKM 2008: Proceedings of the 17th ACM Conference on Information and Knowledge Management. 2008. NISS'09. degree in computer science from University of Science and Technology of China.20 (1). pp. in 2003. Springer. Now she is a Ph. “A Brief Survey of Text Mining”. M. “Assessing the Impact of Using the Internet for Competitive Intelligence”. California. (eds. University of Science and Technology of China.1”.Y. VOL. Available at: http://joshlamar. ECIR 2008: Proceedings of the 30th European Conference on IR Research. Hefei. (ed. 2001. R. and IEEE ComSoc. e-Services. His research interests include Web intelligence. APRIL 2010 VI. a conceptual model for the competitive intelligence was built. 2006. Kahaner.10 (1).705-710. CONCLUSIONS It is the future trend to build competitive intelligence systems to automatically acquire competitive intelligence from the Web. Ananyan. Moreover. and information extraction. Springer. pp. Froelich. Yao. In MUC 2006: Proceedings of the Sixth Message Understanding Conference. Vol. [10] C. Bruza. Business Intelligence Journal. N. et al. In C.). Kehlenbeck. he received his master and bachelor degree in management science both from Tianjin University of Finance and Economics. 06120804). B. A. El-Rawas. ACM SIGMOD. Khoury. Rijke. 43-50. in 2003. ICCIT'09-08. V. and the Science Research Fund of MOE-Microsoft Key Laboratory of Multimedia Computing and Communication (grant no. “An Efficient Web Page Change Detection System Based on an Optimized Hungarian Algorithm”. [8] I.M.Towards a Unified View of Data”. He received his Ph. WI 2006: Proceedings of the 2006 IEEE/WIC/ACM International Conference on Web Intelligence. pp. “Business Intelligence through Text Mining”. Li. He is currently an associate professor in School of Computer Science and Technology. 4. X. China. pp.319-332. Wang (ed. P. Assoc. She received her master degree in management science and engineering from Hefei University of Technology. D. [12] M. REFERENCES [1] L. 2007. [2] S. [7] A. and databases.com/documents/CIT Survey Report. IEEE Transaction on Knowledge Data Engineering. © 2010 ACADEMY PUBLISHER . China. IEEE Computer Society. WI 2007: Proceedings of IEEE/WIC/ACM International Conference on Web Intelligence. IEEE Computer Society. Journal of Convergence Information Technology. 99-613. IEEE. pp. in October.0”. respectively. Boston. “Webscale Named Entity Recognition”. 2007. 2008. WimBI 2006: Proceedings of the First WICI International Workshop on Web Intelligence Meets Brain Informatics. Peiquan Jin was born in Zhejiang. Paass. and IMS’09. Shanahan.).). Zhou. pp. Volume 2. Lecture Notes in Computer Science 4956. ACKNOWLEDGEMENT This work was supported in part by the National Natural Science Foundation of China under the grant no. Thompson.. Zhu.). In J. which makes it easy to represent the competitive intelligence about competitors as well as competition environment. Deng. and is an editor of International Journal of Advancements in Computing Technology. “Using Information Filtering in Web Data Mining Process”. Seventh IFIP International Conference on e-Business. She is currently an associate professor in School of Business Administration. The results show that it is feasible and practical to represent competitive intelligence acquired from the Web based on our proposed approach. Petrovic. “PARMENIDES: Towards Business Intelligence Discovery from Web Data”.. et al. Napa Valley. [9] J. NCM'09-08. China. accessed in 11 May. “The Entity Relationship Model . 597604. “Competitive Intelligence Survey Report”. Vol. Anhui University. China. Before that. Hotho.386 JOURNAL OF SOFTWARE. San Francisco.. In J. (ed.pdf .. one of the important things is to make it clear what types of competitive intelligence are hided in the Internet. in August. A. Washington DC. candidate in School of Management. pp. [11] M. 1996. In N. Prof. 1974. Her research interests include Web intelligence. Macdonald. Theodoulidis. Tianjin. Sundheim. In W. Khalid. He serves as a PC member of many international conferences. Morgan Kaufmann. Huang.D. ACM Transactions on Database Systems. NCM’09. Competitive Intelligence. In Y. California. Zhong et al. 1975. LDV Forum. L.. Olson. 2006. O. 1057-1060. et al.163-169. ACM.. Vol. and NDBC'09. et al. Y. S. and serves as a PC member of several international conferences. NO. pp. Jin is a member of ACM. M. “The Impact of Named Entity Normalization on Information Retrieval for Question Answering”. pp. pp. and proposed a framework to represent the competitive intelligence in the Internet. 291-304. Mikroyannidis.). Chen. 1976. Luo. Liu et al. 67-83. [3] D. P.D. [6] J. (ed. 2007. 5.