Diese Datenbank enthält über 40.000 Dokumente zu Themen aus den Bereichen Formalerschließung – Inhaltserschließung – Information Retrieval.
© 2015 W. Gödert, TH Köln, Institut für Informationswissenschaft / Powered by litecat, BIS Oldenburg (Stand: 28. April 2022)
1Croft, W.B. ; Metzler, D. ; Strohman, T.: Search engines : information retrieval in practice.
Boston : Addison-Wesley, 2010. xxv, 524 S.
Abstract: For introductory information retrieval courses at the undergraduate and graduate level in computer science, information science and computer engineering departments. Written by a leader in the field of information retrieval, Search Engines: Information Retrieval in Practice, is designed to give undergraduate students the understanding and tools they need to evaluate, compare and modify search engines. Coverage of the underlying IR and mathematical models reinforce key concepts. The book's numerous programming exercises make extensive use of Galago, a Java-based open source search engine. SUPPLEMENTS / Extensive lecture slides (in PDF and PPT format) / Solutions to selected end of chapter problems (Instructors only) / Test collections for exercises / Galago search engine
LCSH: Search engines / Programming ; Information retrieval ; Information Storage and Retrieval ; Knowledge Bases
RSWK: Suchmaschine / Information Retrieval
BK: 54.64 / Datenbanken ; 54.75 / Sprachverarbeitung
; 06.74 / Informationssysteme
DDC: 025.04 ; 005.75/8
GHBS: TWX (FH K)
LCC: TK 5105.884 .C765 2010
RVK: ST 205 ; ST 270
2Kolluri, V. ; Metzler, D.P.: Knowledge guided rule learning.
In: Knowledge: creation, organization and use. Proceedings of the 62nd Annual Meeting of the American Society for Information Science, 31.10.-4.11.1999. Ed.: L. Woods. Medford, NJ : Information Today, 1999. S.146-160.
(Proceedings of the American Society for Information Science; vol.36)
Abstract: Rule learning algorithms, developed by traditional supervised machine learning research community, are being used as data analysis tools for generating accurate concept definitions, given a set of instances (pre-classified) and a goal-task (concept class). Most rule learners use straightforward data driven approaches using information theoretic principles to search for statistically defined "interesting" patterns in the data sets. There are two main drawbacks with such purely data driven approaches. First, they perform poorly when insufficient data is available. Second, when large training data sets are available they tend to generate many uninteresting patterns from data sets, and usually it is left to the domain expert to distinguish the "useful" pieces of information from the rest. The size of this problem (a data mining issue onto itself) suggests the need to guide the learning system's search to relevant sub spaces within the space of all possible hypotheses. This paper explores the utility of using prior domain knowledge (in the form of taxonomies over attributes, attribute values and concept classes) to constrain the rule learner's search by requiring it to be consistent with what is already known about the domain. Spreading Activation Learning (SAL) using marker propagation techniques introduced by Aronis and Provost (1994) is used to efficiently learn over taxonomically structured attributes and attribute values. An extension to the SAL methodology to handle rule learning over concept class values is presented. By representing the range of numeric (continuous) values for attributes in the form of simplified IS A taxonomies, the SAL methodology is shown to be capable of handling numeric (continuous) attribute values. Large taxonomies over value sets (especially taxonomies over numeric value sets) usually result in too many redundant rules. This problem can be addressed by pruning the rule set using "rule interest" measures. The focus of this study is to explore the utility of taxonomic structures in rule learning and, in particular the use of taxonomic structures as a way of incorporating background knowledge in the rule learning process. Initial results obtained from an ongoing research work are presented
3Metzler, D.P.: Connectionist and symbolic information processing : a critical analysis and suggested research agenda for connectionism from the symbolic perspective.
In: ASIS'90: Information in the year 2000, from research to applications. Proc. of the 53rd Annual Meeting of the American Society for Information Science, Toronto, Canada, 4.-8.11.1990. Ed. by Diana Henderson. Medford, NJ : Learned Information Inc., 1990. S.261-278.
Abstract: Connectionism (also known as parallel distributed processing), has generated a great deal of interest and has even been presented as a potential replacement for the symbolic approach to artificial intelligence and/or for the entire cognitive paradigm of cognitive science. Reviews some of the reasons for this interest
4Metzler, D.P. ; Haas, S.W. ; Cosic, C.L. ; Wheeler, L.H.: Constituent object parsing for information retrieval and similar text processing problems.
In: Journal of the American Society for Information Science. 40(1989) no.6, S.398-423.
Abstract: Describes the architecture and functioning of the Constituent Object Parser. This system has been developed specially for text processing applications such as information retrieval, which can benefit from structural comparisons between elements of text such as a query and a potentially relevant abstract. Describes the general way in which this objective influenced the design of the system.
5Metzler, D.P. ; Haas, S.W.: ¬The constituent object parser : syntactic structure matching for information retrieval.
In: ACM transactions on information systems. 7(1989) no.3, S.292-316.
Abstract: The constituent object parser is designed to improve the precision and recall performance of information retrieval by providing more powerful matching procedures. Describes the dependency tree representations and the relationship between the intended use of the parser and its design.