Agrawal R and Wimmers EL. A framework for expressing and combining preferences. In 2000 ACM SIGMOD International Conference on Management of Data, pages 297--306, Dallas, USA, 2000. ACM Press.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....that o#er ranked queries. A speculative version of the pipelining algorithm is described. Key words Ranked queries merge ranked views materialization 1 Introduction An increasing number of Web applications allow queries that rank the source objects according to a function of their attributes [7,6,3]. For example consider a database containing houses available for sale. The properties have attributes such as price, number of bedrooms, number of bathrooms, square feet, etc. For a user, the price of a property and the square feet area may be the most important issues, equally weighted in the ....

....on the problem of answering a ranked query using a view. Second, the work relevant to PREFER and MERGE is presented. Personalization and customization of software components (e.g. myexcite.com) can be thought of as simple expressions of preferences. Agrawal and Wimmers in their pioneering work [3] put the notion on preferences into perspective and introduce a framework for their expression and combination. Our work, essentially deals with the algorithmic issues associated with the implementation of specific features of this framework. We adopt terminology in alignment with the framework of ....

[Article contains additional citation context not shown here]

R. Agrawal and E. Wimmers. A Framework For Expressing and Combining Preferences. ACM SIGMOD, 2000.

....model similarity, relevance, or preference: A multimedia database may rank objects by their similarity to an example image. A text search engine orders documents by their relevance to query terms. An e commerce service may sort their products according to a user s preference criteria [1] to facilitate purchase decisions. For these applications, Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profitor commercial advantage and that copies bear this notice ....

....queries with nonstandard predicates. While user defined functions are now commonly supported in Boolean systems, we believe such functions will be even more important for ranked queries, because they are mainly intended for data retrieval based on similarity, relevance, and preference (e.g. as in [1]) As these concepts are inherently imprecise and user (or application) specific, a practical system should support ad hoc criteria to be specifically defined (by users or application programmers) Consider our real estate example. Although the system may provide 4 as built in, users will ....

R. Agrawal and E. Wimmers. A framework for expressing and combining preferences. SIGMOD 2000.

....XFilter system [3] is a recent example of such a filtering system. XFilter, however, has no notion of variable utility as exists in our profile model. One recent effort that has investigated the representation of variable utility is the user preference framework proposed by Agrawal and Wimmers [1]. This model allows users to specify numeric weights for entities or sets of entities and provides a well defined mechanism for combining sets of preferences. This model, however, provides no notion of thresholds or dependencies and this work did not address the performance and scalabablity of ....

R. Agrawal and E. L. Wimmers. A framework for expressing and combining preferences. In W. Chen, J. F. Naughton, and P. A. Bernstein, editors, Proceedings of the 2000.

....the Pareto optimal tuples with respect to criteria. Thus, this is also called the Pareto operator. Skyline has prompted much interest recently, as people see its potential as a basis for or at least a major component of efforts to provide preference queries over relational database systems [1, 3, 6, 7, 8]. Skyline queries can be translated into current SQL, but the resulting queries are inefficient to execute. To handle skyline queries well, the skyline operator would have to be built into the query optimizer, and good algorithms for skyline implemented. We are seeking to devise algorithms for ....

R. Agrawal and E. L. Wimmers. A framework for expressing and combining preferences. In SIGMOD, pages 297--306, 2000.

....the need for significant modifications to the underlying relational engine. Indeed, over the years, several proposals enhancing the query capabilities of relational systems have been made. Recent examples include preference queries, which incorporate qualitative and quantitative user preferences [1, 3, 13, 8, 17] and top k queries [10, 9, 2] In this paper, we initiate the study of a new class of queries that we refer to as OPAC (optimization under parametric aggregation constraints) queries. Such queries aim to identify sets of database tuples that constitute solutions of a large class of optimization ....

....are not aware of work directly related to the work presented herein on OPAC Queries. Such queries are introduced as a novel query type seeking to provide greater query flexibility on top of relational data sources. Recently proposed, but not directly related, query types include preference queries [1, 12, 3, 13, 8, 17] and top k queries [10, 9, 2] The notion of Pareto optimality is discussed in the context of preference queries in database systems in [11] A specialized form of Pareto optimality is introduced in which a user seeks the tuple with the highest values in a collection of select attributes, among ....

R. Agrawal and E. Wimmers. A Framework For Expressing and Combining Preferences. Proceedings of ACM SIGMOD, pages 297-306, June 2000.

....experimental evaluation using real and synthetic datasets, presenting the performance advantages of our approach when compared with other applicable approaches. Finally, section 9 concludes the paper and points to problems of interest for further study. 2 Related Work Agrawal and Wimmers [1] proposed a framework for preference based query processing. Various works considered realizations of a specific instance of this framework, namely top k selection queries, that is, quickly identifying k tuples that optimize scores assigned by monotone linear scoring functions on a variety of ....

R. Agrawal and E. Wimmers. A Framework For Expressing and Combining Preferences. Proceedings of ACM SIGMOD, pages 297--306, June 2000.

....system would do. We can also specify complex conditions. For instance, that the score of node 4 is 0 unless the term search engine occurs at least once, in which case the score is calculated using the weightedsum function. In many IR systems, the range of a scoring function is restricted to be [0, 1]. We could certainly do the same, and obtain all the attendant benefits. We have chosen a scoring function with range [0, #) just to make the point that any such range restrictions on the scoring functions are not required for our algebra. Sometime a scoring function can also be defined on an ....

....than Comp3. The reason for this is the extra work done at the filter level in the Access Methods to check o#sets especially if the result of the intersection is big. 7. RELATED WORK A number of studies focusing on supporting ranked or preference queries have been done in the relational context [1, 16, 8]. In the framework proposed in [1] a preference function is a mapping from a record (tuple) of a given record type (relation) to a score based on user preferences and a meta combine function is a way of combining such functions to compute a new score based on those original scores. The scoring ....

[Article contains additional citation context not shown here]

R. Agrawal and E. L. Wimmers. A framework for expressing and combining preferences. In SIGMOD, 2000.

.... be a subjective one, thus the system has to learn it and then to exploit the acquired knowledge about user preferences to retrieve the most relevant objects [3] In the database field, the problem of expressing and managing user preferences has received growing attention in the last few years [1, 4, 6, 8, 9, 11 15]. In many approaches [1, 7 9, 12] preferences are expressed quantitatively by defining a scoring function that is a weighted linear combination of attributes values (which have therefore to be numeric) Since the scoring function associates each tuple with a numeric score, tuple t 1 is preferred ....

.... learn it and then to exploit the acquired knowledge about user preferences to retrieve the most relevant objects [3] In the database field, the problem of expressing and managing user preferences has received growing attention in the last few years [1, 4, 6, 8, 9, 11 15] In many approaches [1, 7 9, 12] preferences are expressed quantitatively by defining a scoring function that is a weighted linear combination of attributes values (which have therefore to be numeric) Since the scoring function associates each tuple with a numeric score, tuple t 1 is preferred to t 2 if the score of t 1 is ....

R. Agrawal and E. L. Wimmers. A Framework for Expressing and Combining Preferences. In Proc. of the 2000 ACM SIGMOD International Conference on Management of Data, Dallas, USA, pp. 297--306, 2000.

....database applications on the web have necessitated the attachment of (a) functions to relations of object relational databases, and (b) function evaluations to tuples of object relations. Using functions in various forms, recent publications have assigned scores [Coh98] preference values [AW00, HKP01] and probabilistic values [BMP92] to object relational database tuples. We refer to these values and the functions that generate them as sideway values and sideway value functions, respectively. Sideway functions and sideway values represent the recommendations of data creators. We illustrate ....

....queries the importance values of two topics. Sideway values are not necessarily maintained as a column of base relations; sometimes, they can be defined by functions (e.g. preference functions [HKP01] and attached to the base relations of the database in different forms: a) Open form [Coh98, AW00] where, for each tuple, a value is specified, i.e. sideway values are stored in a column of the base relation) b) Closed form [HKP01] where each tuple s sideway value is derived from a closed function. e.g. for a relation R with attributes X and Y, we may have f (X,Y) a.X b.Y where a ....

Agrawal, R., Wimmers, E.L., "A Framework for Expressing and Combining Preferences", SIGMOD'00.

....imprecise queries. The Eureka [20] browser is a user interface that lets users interactively manipulate the conditions of a query with instantly update the results. The goal of the CoBase [8] cooperative database system is to avoid empty answers to queries through condition relaxation. Agrawal [2] proposes a framework where users submit preferences (feedback) and explains how they may be composed into compound preferences. 7 Conclusions In this paper, we concentrated on developing a data access model based on similarity retrieval and query re nement. We presented a framework to integrate ....

Rakesh Agrawal and Edward L. Wimmers. A framework for expressing and combining preferences. In ACM SIGMOD, 2000.

....via TAHs roughly corresponds to decreasing the similarity threshold in our framework and using query point movement. In [19] the authors consider a succession of manually modi ed precise queries to be a browsing session (session query) and focus on optimizing the computation of results. Agrawal [2] proposes a framework where users submit preferences (feedback) and explains how they may be composed into compound preferences. 7 Conclusions In this paper, we concentrated on developing a data access model based on similarity retrieval and query re nement. We presented a framework to ....

Rakesh Agrawal and Edward L. Wimmers. A framework for expressing and combining preferences. In ACM SIGMOD, 2000.

....3 ; 2 0:9 1 0:4 3 ) The QPM model computes the intra feature weights for F (i.e. the F j s in Equation 1) as follows. Let F j denote the standard deviation of the feature values of the relevant examples for feature F along the 5 The initial weights can be chosen more intelligently. [1] proposes to collect feedback information provided by different users over time and use it to determine the best initial weights for the particular user. This technique reduces the number of feedback iterations, especially for already seen queries. 6 Besides these 2 steps, some researchers have ....

....the answers returned by the system. 8 We use the L1 metric (i.e. Manhattan distance) as the distance function DF for the color feature since it corresponds to the histogram intersection similarity measure, the most commonly used similarity measure for color histograms [14, 15] 12 References [1] I. Bartolini, P. Ciaccia, and F. Waas. Feedbackbypass: A new approach to interactive similarity query processing. Proceedings of VLDB Conference, 2001. 2] K. Chakrabarti and S. Mehrotra. The hybrid tree: An index structure for high dimensional feature spaces. Proceedings of the IEEE ....

[Article contains additional citation context not shown here]

R. Agrawal and E. Wimmers. A Framework for Expressing and Combining Preferences. In Proc. of the ACMSIGMOD

No context found.

Agrawal R and Wimmers EL. A framework for expressing and combining preferences. In 2000 ACM SIGMOD International Conference on Management of Data, pages 297--306, Dallas, USA, 2000. ACM Press.

No context found.

R. Agrawal and E. Wimmers. A Framework For Expressing and Combining Preferences. Proceedings of ACM SIGMOD, pages 297--306, June 2000.

No context found.

Agrawal, R., Wimmers, E.L.: A Framework for Expressing and Combining Preferences. In: Proc. of SIGMOD. (2000)

No context found.

R. Agrawal and E.L. Wimmers. A Framework for Expressing and Combining Preferences. In Proceedings of the ACM-SIGMOD International Conference on Management of Data, Dallas, USA, pages 297--306, 2000.

No context found.

R. Agrawal and E. L. Wimmers, A Framework for Expressing and Combining Preferences. ACM SIGMOD Conference on Management of Data, Dallas, May, 2000. pp. 297-306.

No context found.

Agrawal R., Wimmers E. L. (2000): A Framework for Expressing and Combining Preferences. Proceedings of the 2000 ACM SIGMOD International Conference on Management of Data. Dallas, USA, 297 - 306, ACM Press.

No context found.

R. Agrawal and E. L. Wimmers. A framework for expressing and combining preferences. In SIGMOD'00, pages 297--306, 2000.

No context found.

R. Agrawal and E. L. Wimmers. A Framework for Expressing and Combining Preferences. In Proc. of the Intl. Conf. on Very Large Data Bases (VLDB), pages 297--306, 2002.

No context found.

R. Agrawal and E. Wimmers. A Framework For Expressing and Combining Preferences. Proceedings of ACM SIGMOD, pages 297--306, June 2000.

No context found.

R. Agrawal and E.L. Wimmers, "A framework for Expressing and Combining Preferences", In Proc' ACM SIGMOD, May 2000, pp.297-306

No context found.

R. Agrawal and E. Wimmers. A Framework For Expressing and Combining Preferences. Proceedings of ACM SIGMOD, pages 297--306, June 2000.

No context found.

Rakesh Agrawal and Edward Wimmers. A framework for expressing and combining preferences. SIGMOD 2000.

No context found.

Rakesh Agrawal and Edward L. Wimmers. A framework for expressing and combining preferences. In International Conference on Management of Data (SIGMOD), 2000.

First 50 documents

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