Topic
Ranking (information retrieval)
About: Ranking (information retrieval) is a research topic. Over the lifetime, 21109 publications have been published within this topic receiving 435130 citations.
Papers published on a yearly basis
Papers
More filters
••
28 Jul 2003TL;DR: A framework for evaluating subtopic retrieval is proposed which generalizes the traditional precision and recall metrics by accounting for intrinsic topic difficulty as well as redundancy in documents and a maximal marginal relevance (MMR) ranking strategy is proposed.
Abstract: We present a non-traditional retrieval problem we call subtopic retrieval. The subtopic retrieval problem is concerned with finding documents that cover many different subtopics of a query topic. In such a problem, the utility of a document in a ranking is dependent on other documents in the ranking, violating the assumption of independent relevance which is assumed in most traditional retrieval methods. Subtopic retrieval poses challenges for evaluating performance, as well as for developing effective algorithms. We propose a framework for evaluating subtopic retrieval which generalizes the traditional precision and recall metrics by accounting for intrinsic topic difficulty as well as redundancy in documents. We propose and systematically evaluate several methods for performing subtopic retrieval using statistical language models and a maximal marginal relevance (MMR) ranking strategy. A mixture model combined with query likelihood relevance ranking is shown to modestly outperform a baseline relevance ranking on a data set used in the TREC interactive track.
611 citations
•
14 Aug 1996TL;DR: In this article, the user enters a query and the system processes the query to generate an alternative representation, which includes conceptual-level abstraction and representations based on complex nominals (CNs), proper nouns (PNs), single terms, text structure, and logical make-up of the query, including mandatory terms.
Abstract: Techniques for generating sophisticated representations of the contents of both queries and documents in a retrieval system by using natural language processing (NLP) techniques to represent, index, and retrieve texts at the multiple levels (e.g., the morphological, lexical, syntactic, semantic, discourse, and pragmatic levels) at which humans construe meaning in writing. The user enters a query and the system processes the query to generate an alternative representation, which includes conceptual-level abstraction and representations based on complex nominals (CNs), proper nouns (PNs), single terms, text structure, and logical make-up of the query, including mandatory terms. After processing the query, the system displays query information to the user, indicating the system's interpretation and representation of the content of the query. The user is then given an opportunity to provide input, in response to which the system modifies the alternative representation of the query. Once the user has provided desired input, the possibly modified representation of the query is matched to the relevant document database, and measures of relevance generated for the documents. A set of documents is presented to the user, who is given an opportunity to select some or all of the documents, typically on the basis of such documents being of particular relevance. The user then initiates the generation of a query representation based on the alternative representations of the selected document(s).
605 citations
••
06 Sep 2014TL;DR: A novel model to automatically select the most discriminative video fragments from noisy image sequences of people where more reliable space-time features can be extracted, whilst simultaneously to learn a video ranking function for person re-id is presented.
Abstract: Current person re-identification (re-id) methods typically rely on single-frame imagery features, and ignore space-time information from image sequences. Single-frame (single-shot) visual appearance matching is inherently limited for person re-id in public spaces due to visual ambiguity arising from non-overlapping camera views where viewpoint and lighting changes can cause significant appearance variation. In this work, we present a novel model to automatically select the most discriminative video fragments from noisy image sequences of people where more reliable space-time features can be extracted, whilst simultaneously to learn a video ranking function for person re-id. Also, we introduce a new image sequence re-id dataset (iLIDS-VID) based on the i-LIDS MCT benchmark data. Using the iLIDS-VID and PRID 2011 sequence re-id datasets, we extensively conducted comparative evaluations to demonstrate the advantages of the proposed model over contemporary gait recognition, holistic image sequence matching and state-of-the-art single-shot/multi-shot based re-id methods.
600 citations
•
01 Apr 2002
TL;DR: In this paper, a modular intelligent personal agent system is presented for search, navigation, control, retrieval, analysis, and results reporting on networks and databases, where hypertext documents and associated content media are displayed as symbol or thumbnail web documents as nodes with connector lines representing links between the documents.
Abstract: A modular intelligent personal agent system is presented for search, navigation, control, retrieval, analysis, and results reporting on networks and databases. A client-side or server-side software application retrieves and interprets hypertext documents executing a search algorithm, which search results are displayed in alternate three-dimensional and two-dimensional graphical visualization formats. Hypertext documents and associated content media are displayed as symbol or thumbnail web documents as nodes with connector lines representing links between the documents. Nodes and connector lines are color-coded symbol form for the user according to truth of search terms, numeric data tested in hypertext documents, according to domain type, link density, and metric counts. Different symbols represent search and Boolean evaluation status, document type, and thumbnails represent whole or incremental portions of the document page or type documents found. The three-dimensional displayed nodes are visually navigated based on recency, chronology of discovery and metric information values. The result of searches performed by the system can retrieve user selected documents from a network and automatically format results of the search and content retrieval using a plurality of ranking methods. The system provides alerts and content delivery to users using email, instant messaging and audio. Multiple agents can operate to accomplish complex tasks a singular agent cannot. Agents are deployed
596 citations
••
TL;DR: A statistical ranking theory is introduced, which can describe different learning-to-rank algorithms, and be used to analyze their query-level generalization abilities.
Abstract: Learning to rank for Information Retrieval (IR) is a task to automatically construct a ranking model using training data, such that the model can sort new objects according to their degrees of relevance, preference, or importance. Many IR problems are by nature ranking problems, and many IR technologies can be potentially enhanced by using learning-to-rank techniques. The objective of this tutorial is to give an introduction to this research direction. Specifically, the existing learning-to-rank algorithms are reviewed and categorized into three approaches: the pointwise, pairwise, and listwise approaches. The advantages and disadvantages with each approach are analyzed, and the relationships between the loss functions used in these approaches and IR evaluation measures are discussed. Then the empirical evaluations on typical learning-to-rank methods are shown, with the LETOR collection as a benchmark dataset, which seems to suggest that the listwise approach be the most effective one among all the approaches. After that, a statistical ranking theory is introduced, which can describe different learning-to-rank algorithms, and be used to analyze their query-level generalization abilities. At the end of the tutorial, we provide a summary and discuss potential future work on learning to rank.
591 citations