Chen H., Schatz, B., Ng, T., Martinez, J., Kirchhoff, A., & Lin, C. (1996). A parallel computing approach to creating engineering concept spaces for semantic retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence Special Section on Digital Libraries: Representation and Retrieval 18(8): 771-782.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....experience and to allow sharing of users knowledge about resources, it is often desirable to allow users to annotate the resources with their own knowledge, store these annotations in the DLs and make them available to other users. Nevertheless, in most existing georeferenced and geospatial DLs [8, 2, 9], users cannot easily place comments on geospatial and georeferenced resources or share their knowledge about these resources due to the lack of direct support for annotation services in these systems. Users contribution of contents, including annotations, are usually not permitted or only ....

H. Chen, B.R. Schatz, T.D. Ng, J.P. Martinez, A.J. Kirchho#, and C. Lin. A parallel computing approach to creating engineering concept spaces for semantic retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(8):771--782, August 1996.

....resources such as thesauri. While in the first case the results reported in the literature are mixed, showing that only experienced users tend to benefit from these techniques [11] in the latter case there have been promising results, that have prompted a flurry of work in the area (see e.g. [4, 19]) especially within the context of the Digital Libraries Initiative. Querying from within categories also requires that query refinement too be performed in a category specific way. In the rest of the paper we discuss work in progress within Eurosearch aimed at allowing the user to interactively ....

....extraneous to the domain specific terminology of the category of interest. For instance, from a training sample of documents previously categorised under the BusinessAndEconomy label, the two words banana 6 Associative thesauri are also called term term relationship matrices [15] concept spaces [3, 4], or similarity thesauri [13, 20, 21] 5 and coconut might (correctly ) emerge as strongly related, simply because they co occur in a few documents (e.g. related to stock prices of exotic fruits) and they are co absent in most of the others. In order to avoid this, before proceeding to term term ....

[Article contains additional citation context not shown here]

H. Chen, B. R. Schatz, T. D. Ng, J. Martinez, A. Kircho#, and C. Lin. A parallel computing approach to creating engineering concept spaces for semantic retrieval: the Illinois Digital Library Initiative project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(8):771--782, 1996.

....procedures. Similarly, a concept space approach has been proposed to create meaningful and understandable domainspecific networks of terms and weighted associations, which are used to represent the underlying information spaces, i.e. documents in different domain specific databases [3]. The concept space approach consists of (1) acquiring complete and recent collections of documents as the sources of vocabularies, 2) automatically indexing all terms in the documents, 3) clustering the documents based on term frequency and document frequency, and (4) organizing documents based ....

....on term frequency and document frequency, and (4) organizing documents based on multi term associations. The research on conceptual clustering and concept spaces indicates that techniques exist that are capable of automatically creating conceptual networks of millions of domain specific terms [3]. 2.4. Knowledge Management Systems Knowledge Management Systems are based on the idea that knowledge can be made explicit, stored in a database and distributed on demand to users [12] A major problem in such systems is to maintain and communicate the relative quality and worth of the stored ....

Chen, H. et al., "A parallel computing approach to creating engineering concept spaces for semantic retrieval: the Illinois digital library initiative project", IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 18, No. 8, August 1996. Pp. 771-782.

....to search images through their perceptual characteristics. These databases should operate directly on the image data, and require no encoding or labeling by an operator. Databases based on this idea, or on similar ideas, which we call perceptual databases, have received a lot of attention lately [35, 45, 21, 23, 12, 9, 32, 20, 10, 41]. This is, of course, a somewhat ideal situation, and in some practical applications we can expect that some information encoding will still be necessary. This is particularly true when images have a cultural, rather than perceptual, content. For instance, one of the figures in this paper contains ....

H. CHen, B. Schatz, T. Ng, J. Martinez, A. Kircho#, and C. Lin. A parallel computing approach to creating engineering concept spaces for semantic retrieval: The illinois digital library initiative project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(8), August 1996.

....the user, who may then decide which to include and which not to include in the revised query. There have been promising results in the use of IQE techniques in which the suggested terms come from lexical resources such as thesauri; these results have prompted a flurry of work in the area (see e.g. [3,6]) especially within the context of the Digital Libraries Initiative. It is our contention that querying from within categories requires that query refinement too be performed in a category specific way. In the rest of the paper we discuss work in progress within Eurosearch aimed at allowing the ....

H. Chen, B. R. Schatz, T. D. Ng, J. Martinez, A. Kircho#, and C. Lin. A parallel computing approach to creating engineering concept spaces for semantic retrieval: the Illinois Digital Library Initiative project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(8):771--782, 1996.

.... domains systems design engineering has been appeared (a need of several systems persons is changed to thousands of required system engineers in all basic engineering domains) 3 Basic Design Flow Let us examine the following basic stages of design activity [Bahill and Gissing, 1998; Chen et al. 1996; Goel and Singh, 1998; Kuppuraju, et al. 1985; Levin, 1998; Prasad, 1996; Sage and Rouse, 1999; Sause and Powell, 1991; Van Gigch, 1978; etc. Phase 1. Conceptual design of new systems and or system architecture (i.e. system structure) design of system specifications, requirements ....

H. Chen et al., A Parallel Computing Approach to Creating Engineering Concept Spaces for Semantic Retrieval: The Illinois Digital Library Initiative Project. IEEE Trans. on Pattern Analysis and Mach. Intel., 18(8): 771-782, 1996.

....to search images through their perceptual characteristics. These databases should operate directly on the image data, and require no encoding or labeling by an operator. Databases based on this idea, or on similar ideas, which we call perceptual databases, have received a lot of attention lately [35, 45, 21, 23, 12, 9, 32, 20, 10, 41]. This is, of course, a somewhat ideal situation, and in some practical applications we can expect that some information encoding will still be necessary. This is particularly true when images have a cultural, rather than perceptual, content. For instance, one of the figures in this paper contains ....

H. CHen, B. Schatz, T. Ng, J. Martinez, A. Kirchoff, and C. Lin. A parallel computing approach to creating engineering concept spaces for semantic retrieval: The illinois digital library initiative project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(8), August 1996.

....number of unique terms in the entire collection. In our previous experiments, we have found that the number of unique terms for mid scale collections (10 50 MBs) after applying thresholds (for example, a term appearing more than 2 times) can be as high as tens or hundreds of thousands [1 12 2] [1 11 1]. By observing the document collections that we have analyzed and that have been mentioned in the literature, for example, Grefenstette s works on automatic thesauri [20] it is possible to conclude that the number of unique terms in a collection is typically proportional to the size of a ....

....parallel supercomputers) will not be able to build SOMs using traditional approach. The conventional SOM s time complexity of square of the size of collection is deemed unscaleable. We believe that even parallel processing on the prevailing supercomputers will not resolve this algorithmic problem [1 11 1]. Based on extensive experimentation [3 39 9] 1 13 3] and recent analyses of the characteristics of textual classification, we noticed that sparsity of the input vector stood out as a prime candidate for SOM optimization. It is evident that for large scale textual collections a vast majority of ....

H. Chen, B. R. Schatz, T. D. Ng, J. P. Martinez, A. J. Kirchhoff, and C. Lin. A parallel computing approach to creating engineering concept spaces for semantic retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(8):771-782, August 1996.

No context found.

Chen H., Schatz, B., Ng, T., Martinez, J., Kirchhoff, A., & Lin, C. (1996). A parallel computing approach to creating engineering concept spaces for semantic retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence Special Section on Digital Libraries: Representation and Retrieval 18(8): 771-782.

No context found.

Chen H., Schatz, B., Ng, T., Martinez, J., Kirchhoff, A., & Lin, C. (1996). A parallel computing approach to creating engineering concept spaces for semantic retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence Special Section on Digital Libraries: Representation and Retrieval 18(8): 771-782.

No context found.

H. Chen et al., "A parallel computing approach to creating engineering concept spaces for semantic retrieval: The Illinois digital library initiative project," IEEE Trans. Pattern Anal. Machine Intell., vol. 18, pp. 771--78 2, 1996.

....properties. A focus on to semantic interoperability has prompted several of the NSF DARPA NASA funded large scale digital library initiative (DLI) projects to explore various statistical, and pattern recognition techniques, e.g. concept spaces and category maps in the Illinois project [46] [5], textile and word sense dis ambiguiation in the Berkeley project [53] voice recognition in the CMU project [50] and image segmentation and clustering in the UCSB project [29] Definition and use of metadata and clustering and automatic hierarchical organization of information, which require ....

....The most fundamental techniques in IR involve identifying key features in objects. For example, automatic indexing and natural language processing (e.g. noun phrase extraction or object type tagging) are frequently used to extract meaningful keywords or phrases from texts automatically [43] [5]. Texture, color, or shape based indexing and segmentation techniques are often used to identify images [29] For audio and video applications, voice recognition, speech recognition, and scene segmentation techniques can be used to identify meaningful descriptors in audio or video streams [50] ....

[Article contains additional citation context not shown here]

H. Chen, B. R. Schatz, T. D. Ng, J. P. Martinez, A. J. Kirchhoff, and C. Lin. A parallel computing approach to creating engineering concept spaces for semantic retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(8):771-782, August 1996.

.... extracting noun phrases using statistical parsers and computing contextual cooccurrence of phrases within the documents and documents within the collection [11] Recently, we have computed large collections from ten of thousands to tens of millions of documents using high end supercomputers [3][12] A combination of efficient algorithms and much faster computers makes it appear that scalable semantics is now becoming feasible. To appear in Eighth International Conference on Information and Knowledge Management (CIKM) Kansas City, November 2 6, 1999, Missouri, Sponsored by ACM ....

....books. The collection used in our experiments ranged from 100 to 10000 documents. In our experiments, we repeated the same computation using 100, 1000, and 10000 documents with 1 to 5 machines. Computation time for each experiment was recorded. Unlike other similar experiments on supercomputers [3][4] which include only the cooccurrence computation and exclude the time for things like recording the results to disk, we include the total time for the whole process from starting with a raw collection and ending with the results of the computation recorded on disk. The time reported in our ....

Chen, H., Schatz, B.R., Ng, D., Martinez, J., Kirchhoff, A., and Lin, C. A Parallel Computing Approach to Creating Engineering Concept Spaces for Semantic Retrieval: The Illinois Digital Library Project. IEEE Transactions Pattern Analysis Machine, 18(8):771-772, August 1996.

....up todate and fine grained subject categories and the requirement that an owner place a homepage under a proper subject category has significantly hampered Yahoo s success and popularity. 3 Research Design and Algorithms Funded by the ongoing Illinois Digital Library Initiative project [21] [7], this research aims to create intelligent Internet personal search agents that can be deployed on the Web for efficient, timely, and optimal searches. We planned to answer the following two general research questions: ffl Can an Internet spider be designed to take individual users requests and ....

....A and B based on links was computed as follows: J link (A; B) #(X Y ) #(X [ Y ) 1) where #(S) indicates the cardinality of set S. ffl Jaccard s Scores from Keywords: For a given homepage, terms were identified based on an automatic indexing procedure developed in our previous research [7]. Term frequency (tf) and inverse document frequency (idf) term weighting heuristics also adopted in such popular searchable databases as Lycos, were then computed, Term frequency, tf ij , represents the number of occurrences of term j in document (homepage) i. Homepage frequency, df j , ....

H. Chen, B. R. Schatz, T. D. Ng, J. P. Martinez, A. J. Kirchhoff, and C. Lin. A parallel computing approach to creating engineering concept spaces for semantic retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(8):771--782, August 1996.

No context found.

Hsinchun Chen, Bruce Schatz, Tobun Ng, Joanne Martinez, Amy Kirchhoff, and Chienting Lin. A Parallel Computing Approach to Creating Engineering Concept Spaces for Semantic Retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(8):771--772, August 1996.

.... In particular, the project has developed texture extraction and image segmentation techniques [29] for indexing aerial photos with the goal of creating a conceptual, iconic thesaurus that can assist users in image based, concept based browsing [29] ffl The Univeristy of Illinois DLI project [8], which encompasses all fields of engineering, is investigating the feasibility of creating textual geographic thesauri (called concept spaces) on the basis of textual analysis and clustering techniques. The resulting geographic concept spaces may be used to suggest subject descriptors and place ....

....is more meaningful to users (see [2] 14] 23] 18] for overviews of AI based learning techniques. Several recent large scale digital library experiments have adopted cluster analysis for creating domain specific concept spaces consisting of networks of terms and their weighted relationships [9] [8]. The Kohonen self organizing map algorithm has been extended to create textual graphical category maps [10] and visual geographic thesauri [29] In our proposed GKRS architecture, we suggest the use of concept spaces and category map techniques in order to create compatible network based ....

[Article contains additional citation context not shown here]

H. Chen, B. R. Schatz, T. D. Ng, J. P. Martinez, A. J. Kirchhoff, and C. Lin. A parallel computing approach to creating engineering concept spaces for semantic retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 18(8):771--782, August 1996.

No context found.

Hsinchun Chen, Bruce Schatz, Tobun Ng, Joanne Martinez, Amy Kirchhoff, and Chienting Lin. A Parallel Computing Approach to Creating Engineering Concept Spaces for Semantic Retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1996.

No context found.

Chen, H., Schatz, B., Ng, T., Martinez, J., Kirchhoff, A., and Lin, C. A parallel computing approach to creating engineering concept spaces for semantic retrieval: The Illinois Digital Library Initiative Project. IEEE Trans. Pattern Analysis and Machine Intelligence 18, 8 (1996), 771--782.

No context found.

Hsinchun Chen, Bruce Schatz, Tobun Ng, Joanne Martinez, AmyKirchhoff, and Chienting Lin. A Parallel Computing Approach to Creating Engineering Concept Spaces for Semantic Retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1996.

No context found.

Hsinchun Chen, Bruce Schatz, Tobun Ng, Joanne Martinez, AmyKirchhoff, and Chienting Lin. A Parallel Computing Approach to Creating Engineering Concept Spaces for Semantic Retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1996.

No context found.

Hsinchun Chen, Bruce Schatz, Tobun Ng, Joanne Martinez, Amy Kirchhoff, and Chienting Lin. A Parallel Computing Approach to Creating Engineering Concept Spaces for Semantic Retrieval: The Illinois Digital Library Initiative Project. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1996.

No context found.

Chen, H. et al., "A parallel computing approach to creating engineering concept spaces for semantic retrieval: the Illinois digital library initiative project", IEEE Trans on PAMI, Vol. 18, No. 8, August 1996. Pp. 771-782.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC