The widespread adoption of Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG) systems is reshaping the landscape of information retrieval. However, the long-term effects of LLM-generated texts on retrieval systems remain underexplored, creating challenges for mitigating their impact. The effects are examined in this study, with a particular focus on the “Spiral of Silence” phenomenon, which refers to the marginalization of diverse information as certain types of content dominate, leading to a homogenized information ecosystem. To investigate this, a simulation pipeline is constructed to model the iterative introduction of LLM-generated texts into retrieval systems. Experimental results across multiple iterations reveal that as the presence of LLM-generated texts within the system grows, retrieval systems exhibit a stronger tendency to retrieve these texts. This trend, in turn, reduces the visibility of human-generated content, diminishes diversity, propagates errors, and results in a notable decline in retrieval performance. To address these challenges, we propose a Utility-Driven Multi-Objective Optimization (UMO) framework to effectively mitigate the “Spiral of Silence.” This framework employs a two-phase approach: an optimization phase, leveraging the NSGA-II algorithm to derive optimal preference weights for multiple objectives, and a memorization phase, which directly integrates these weights into the retrieval vector space without requiring additional model retraining. Experimental results demonstrate that this framework maintains stable retrieval effectiveness, improves the retrieval proportion of human-generated content, reduces the excessive influence of LLM-generated texts, and preserves information diversity, effectively mitigating the “Spiral of Silence.”

The proliferation of complex non-factoid questions in modern information seeking (IS) systems exposes critical limitations in conventional Retrieval-Augmented Generation (RAG) approaches, particularly their static search strategies and the lack of systematic multi-source information integration capabilities. Facing these limitations, we present PASS (Proactive Agent-driven Search System), a novel multi-agent framework that operationalizes human-like proactive search strategies through five specialized agents: Revealer for intent analysis, Navigator for search planning, Seeker/Reader for adaptive retrieval, and Writer for response synthesis, systematically expanding the search space through iterative query refinement and multi-perspective knowledge integration. Crucially, our framework demonstrates remarkable adaptability to mid-sized LLMs, demonstrating its scalability in resource-constrained environments. To comprehensively assess the effectiveness of the proposed framework, we carry out extensive experiments on both mid-sized and proprietary large-scale LLMs, evaluating response quality for complex non-factoid questions using a newly introduced nugget-based assessment. Experimental results from offline nugget-based evaluation and online A/B Tests confirm substantial improvements in answer quality, advancing proactive information seeking methodologies and offering practical pathways for democratizing complex reasoning capabilities to resource-constrained environments.

Pre-trained transformer models, such as BERT and T5, have shown to be highly effective at ad hoc passage and document ranking. Due to the inherent sequence length limits of these models, they need to process document passages one at a time rather than processing the entire document sequence at once. Although several approaches for aggregating passage-level signals into a document-level relevance score have been proposed, there has yet to be an extensive comparison of these techniques. In this work, we explore strategies for aggregating relevance signals from a document’s passages into a final ranking score. We find that passage representation aggregation techniques can significantly improve over score aggregation techniques proposed in prior work, such as taking the maximum passage score. We call this new approach PARADE. In particular, PARADE can significantly improve results on collections with broad information needs where relevance signals can be spread throughout the document (such as TREC Robust04 and GOV2). Meanwhile, less complex aggregation techniques may work better on collections with an information need that can often be pinpointed to a single passage (such as TREC DL and TREC Genomics). We also conduct efficiency analyses and highlight several strategies for improving transformer-based aggregation.

Knowledge graph completion (KGC) aims to infer missing knowledge triples based on known facts in a knowledge graph. Current KGC research mostly follows an entity ranking protocol, wherein the effectiveness is measured by the predicted rank of a masked entity in a test triple. The overall performance is then given by a micro(-average) metric over all individual answer entities. Due to the incomplete nature of the large-scale knowledge bases, such an entity ranking setting is likely affected by unlabelled top-ranked positive examples, raising questions on whether the current evaluation protocol is sufficient to guarantee a fair comparison of KGC systems. To this end, this paper presents a systematic study on whether and how the label sparsity affects the current KGC evaluation with the popular micro metrics. Specifically, inspired by the TREC paradigm for large-scale information retrieval (IR) experimentation, we create a relatively "complete" judgment set based on a sample from the popular FB15k-237 dataset following the TREC pooling method. According to our analysis, it comes as a surprise that switching from the original labels to our "complete" labels results in a drastic change of system ranking of a variety of 13 popular KGC models in terms of micro metrics. Further investigation indicates that the IR-like macro(-average) metrics are more stable and discriminative under different settings, meanwhile, less affected by label sparsity. Thus, for KGC evaluation, we recommend conducting TREC-style pooling to balance between human efforts and label completeness, and reporting also the IR-like macro metrics to reflect the ranking nature of the KGC task.

Online search latency is a major bottleneck in deploying large-scale pre-trained language models, e.g. BERT, in retrieval applications. Inspired by the recent advances in transformer-based document expansion technique, we propose to trade offline relevance weighting for online retrieval efficiency by utilizing the powerful BERT ranker to weight the neighbour documents collected by generated pseudo-queries for each document. In the online retrieval stage, the traditional query-document matching is reduced to the much less expensive query to pseudo-query matching, and a document rank list is quickly recalled according to the pre-computed neighbour documents. Extensive experiments on the standard MS MARCO dataset with both passage and document ranking tasks demonstrate promising results of our method in terms of both online efficiency and effectiveness.

Entity Resolution (ER) identifies records from different data sources that refer to the same real-world entity. Conventional ER approaches usually employ a structure matching mechanism, where attributes are aligned, compared and aggregated for ER decision. The structure matching approaches, unfortunately, often suffer from heterogeneous and dirty ER problems. That is, entities from different data sources are described using different schemas, and attribute values may be misplaced, missing, or noisy. In this paper, we propose a deep sequence-to-sequence entity matching model, denoted Seq2SeqMatcher, which can effectively solve the heterogeneous and dirty problems by modeling ER as a token-level sequence-to-sequence matching task. Specifically, we propose an align-compare-aggregate neural network for Seq2Seq entity matching, which can learn the representations of tokens, capture the semantic relevance between tokens, and aggregate matching evidence for accurate ER decisions in an end-to-end manner. Experimental results show that, by comparing entity records in token level and learning all components in an end-to-end manner, our Seq2Seq entity matching model can achieve remarkable performance improvements on 9 standard entity resolution benchmarks.

Clinical Decision Support (CDS) is widely seen as an information retrieval (IR) application in the medical domain. The goal of CDS is to help physicians find useful information from a collection of medical articles with respect to the given patient records, in order to take the best care of their patients. Most of the existing CDS methods do not sufficiently consider the semantic relation between texts, hence the potential in improving the performance in biomedical articles retrieval. This paper proposes a novel feedback-based approach which considers the semantic association between a retrieved biomedical article and a pseudo feedback set. Evaluation results show that our method outperforms the strong baselines and is able to improve over the best runs in the TREC CDS tasks.

Transductive learning is a semi-supervised learning paradigm that can leverage unlabeled data by creating pseudo labels for learning a ranking model, when there is only limited or no training examples available. However, the effectiveness of transductive learning in information retrieval (IR) can be hindered by the low quality pseudo labels. To this end, we propose to incorporate a two-step k-means clustering algorithm to select the high quality training queries for generating the pseudo labels. In particular, the first step selects the high-quality queries for which the relevant documents are highly coherent as indicated by the clustering results. The second step then selects the initial training examples for the transductive learning that iteratively aggregating the pseudo examples. Finally, the learning to rank (LTR) algorithms are applied to learn the ranking model using the pseudo training examples created by the transductive learning process. Our proposed approach is particularly suitable for applications where there is only little or no human labels available as it does not necessarily involve the use of relevance assessments information or human efforts. Experimental results on the standard TREC Tweets11 collection show that our proposed approach outperforms strong baselines, namely the conventional applications of learning to rank algorithms using human labels for the training and transductive learning using all the queries available.

Term proximity is effective for many information retrieval (IR) research fields yet remains unexplored in blogosphere IR. The blogosphere is characterized by large amounts of noise, including incohesive, off‐topic content and spam. Consequently, the classical bag‐of‐words unigram IR models are not reliable enough to provide robust and effective retrieval performance. In this article, we propose to boost the blog postretrieval performance by employing term proximity information. We investigate a variety of popular and state‐of‐the‐art proximity‐based statistical IR models, including a proximity‐based counting model, the Markov random field (MRF) model, and the divergence from randomness (DFR) multinomial model. Extensive experimentation on the standard TREC Blog06 test dataset demonstrates that the introduction of term proximity information is indeed beneficial to retrieval from the blogosphere. Results also indicate the superiority of the unordered bi‐gram model with the sequential‐dependence phrases over other variants of the proximity‐based models. Finally, inspired by the effectiveness of proximity models, we extend our study by exploring the proximity evidence between uery terms and opinionated terms. The consequent opinionated proximity model shows promising performance in the experiments.

The major challenge for Question Retrieval (QR) in Community Question Answering (CQA) is the lexical gap between the queried question and the historical questions. This paper proposes a novel Question-Answer Topic Model (QATM) to learn the latent topics aligned across the question-answer pairs to alleviate the lexical gap problem, with the assumption that a question and its paired answer share the same topic distribution. Experiments conducted on a real world CQA dataset from Yahoo! Answers show that combining both parts properly can get more knowledge than each part or both parts in a simple mixing way and combining our QATM with the state-of-the-art translation-based language model, where the topic and translation information is learned from the question-answer pairs at two different grained semantic levels respectively, can significantly improve the QR performance.

By incorporating diverse sources of evidence of relevance, learning to rank has been widely applied to real-time Twitter search, where users are interested in fresh relevant messages. Such approaches usually rely on a set of training queries to learn a general ranking model, which we believe that the benefits brought by learning to rank may not have been fully exploited as the characteristics and aspects unique to the given target queries are ignored. In this paper, we propose to further improve the retrieval performance of learning to rank for real-time Twitter search, by taking the difference between queries into consideration. In particular, we learn a query-biased ranking model with a semi-supervised transductive learning algorithm so that the query-specific features, e.g. the unique expansion terms, are utilized to capture the characteristics of the target query. This query-biased ranking model is combined with the general ranking model to produce the final ranked list of tweets in response to the given target query. Extensive experiments on the standard TREC Tweets11 collection show that our proposed query-biased learning to rank approach outperforms strong baseline, namely the conventional application of the state-of-the-art learning to rank algorithms.

Wikipedia is characterized by its dense link structure and a large number of articles in different languages, which make it a notableWeb corpus for knowledge extraction and mining, in particular for mining the multilingual associations. In this paper, motivated by a psychological theory of word meaning, we propose a graph‐based approach to constructing a cross‐language association dictionary (CLAD) fromWikipedia, which can be used in a variety of cross‐language accessing and processing applications. In order to evaluate the quality of the minedCLAD, and to demonstrate how the minedCLADcan be used in practice, we explore two different applications of the minedCLADto cross‐language information retrieval (CLIR). First, we use the minedCLADto conduct cross‐language query expansion; and, second, we use it to filter out translation candidates with low translation probabilities. Experimental results on a variety of standardCLIRtest collections show that theCLIRretrieval performance can be substantially improved with the above two applications ofCLAD, which indicates that the minedCLADis of sound quality.

With the rapid development of the information technology, there exists the difficulty in deploying state-of-the-art retrieval models in environments such as peer-to-peer networks and pervasive computing, where it is expensive or even infeasible to maintain the global statistics. To this end, this paper presents an investigation in the validity of different statistical assumptions of term distributions. Based on the findings in this investigation, a variety of weighting models, called NG (standing for "no global statistics") models, are derived from the Divergence from Randomness framework, in which only the within-document statistics are used in the relevance weighting. Compared to the state-of-the-art weighting models in extensive experiments on various standard TREC test collections, our proposed NG models can provide acceptable retrieval performance in ad-hoc search, without the use of global statistics.

User generated social annotations provide extra information for describing document contents. In this paper, we propose an effective method to model the categorization property of social annotations and explore the potential of combining it with classical language models for improving retrieval performance. Specifically, a novel TR-LDA model is presented to take annotations as an additional source for generating document contents apart from the document itself. We provide strategies for representing and weighting the categorization property and develop an efficient inference algorithm, where space saving is taken into account. Experiments are carried out on synthetic datasets, where documents and queries come from the standard evaluation conference TREC and annotations come from the website Delicious.com. Our results demonstrate the effectiveness of the proposed method on the ad-hoc retrieval task, which significantly outperforms state-of-art baselines.

Term proximity retrieval rewards a document where the matched query terms occur close to each other. Although term proximity is known to be effective in many Information Retrieval (IR) applications, the within-document distribution of each individual query term and how the query terms associate with each other, are not fully considered. In this paper, we introduce a pseudo term, namely Cross Term, to model term proximity for boosting retrieval performance. An occurrence of a query term is assumed to have an impact towards its neighboring text, which gradually weakens with the increase of the distance to the place of occurrence. We use a shape function to characterize such an impact. A Cross Term occurs when two query terms appear close to each other and their impact shape functions have an intersection. We propose a Cross Term Retrieval (CRTER) model that combines the Cross Terms' information with basic probabilistic weighting models to rank the retrieved documents. Extensive experiments on standard TREC collections illustrate the effectiveness of our proposed CRTER model.

Classical probabilistic information retrieval (IR) models, e.g. BM25, deal with document length based on a trade-off between the Verbosity hypothesis, which assumes the independence of a document's relevance of its length, and the Scope hypothesis, which assumes the opposite. Despite the effectiveness of the classical probabilistic models, the potential relationship between document length and relevance is not fully explored to improve retrieval performance. In this paper, we conduct an in-depth study of this relationship based on the Scope hypothesis that document length does have its impact on relevance. We study a list of probability density functions and examine which of the density functions fits the best to the actual distribution of the document length. Based on the studied probability density functions, we propose a length-based BM25 relevance weighting model, called BM25L, which incorporates document length as a substantial weighting factor. Extensive experiments conducted on standard TREC collections show that our proposed BM25L markedly outperforms the original BM25 model, even if the latter is optimized.