A survey on the use of relevance feedback for information access systems

By Ian Ruthven, Mounia Lalmas. Knowledge engineering review (Print), 2003

493 citations recorded by the source corpus.

Subfields: Computer Networks and Communications, Information Systems

View Paper: https://www.semanticscholar.org/paper/c94aa9a88bfdb93df87d6a62a356e06535ae1709

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

In Information Systems

Citations in the corpus

Evaluating sentence-level relevance feedback for high-recall information retrieval (2018, methodology): “…implementation of CAL, which is eﬀectively a relevance-feedback (RF) method, albeit with a diﬀerent objective and implementation than to construct the ultimate query by selecting and weighting search terms, as typically re-ported in the RF literature (Aalbersberg, 1992; Ruthven and Lalmas, 2003).”
Personalizing search via automated analysis of interests and activities (2005, methodology): “This is a kind of blind or pseudorelevance feedback in which the top-k documents are considered relevant [19].”
A review of ontology based query expansion (2007, background): “Ruthven and Lalmas (2003) state that domain specific collections perform better with relevance feedback than domain independent collections because it is easier to select good expansion terms from this type of a collection or because the ambiguity of search terms is less significant.”
Spoken Content Retrieval—Beyond Cascading Speech Recognition with Text Retrieval (2015, methodology): “Feedback Scenario: Relevance feedback [128] well used in text retrieval is useful to integrate the ASR and retrieval modules as a whole and optimize the overall retrieval performance, rather than considering them as two cascaded independent components [129]–[132], as shown in Fig.”
Retrieval from software libraries for bug localization: a comparative study of generic and composite text models (2011, methodology): “The deterministic models use the cosine similarity measure on the normalized representation of the documents and the queries [23].”
Query Expansion Techniques for Information Retrieval: a Survey (2017, unknown): “Some of the research works based on relevance feedback are [51,236,231,177]; these have been discussed earlier in Sec.”
Concept-Based Information Retrieval Using Explicit Semantic Analysis (2011, methodology): “Most of the work on PRF used the top ranked documents or passages [Ruthven and Lalmas 2003; Xu and Croft 2000] as pseudo-relevant documents (or positive examples).”
Pseudo-Relevance Feedback Based on Matrix Factorization (2016, methodology): “To do so, we compare the proposed method with two baselines: (1) document retrieval without feedback (NoPRF) and (2) the modified Rocchio algorithm [31, 32] which has been extensively used as a state-of-the-art PRF method in the vector space model framework for many years.”
Task-aware Retrieval with Instructions (2022, background): “, Figure 1 bottom left), the difference is the explicit definition of relevance in the instruction t as additional input to the system, which can be diverse and may not be fully defined by the query only (Ruthven and Lalmas, 2003).”
Find-similar: similarity browsing as a search tool (2006, background): “We constructed our runs’ results lists for evaluation in the same manner as Aalbersberg [1].”
The Simplest Thing That Can Possibly Work: Pseudo-Relevance Feedback Using Text Classification (2019, background): “Relevance feedback in IR systems dates back to the 1960s [20] and the idea of using pseudo labels for relevance feedback dates back to at least the late 1970s [7]; a nice historical overview is offered by Ruthven and Lalmas [21].”