Query Context Expansion for Open-Domain Question Answering

Clinical decision support (CDS) search is performed to retrieve key medical literature that can assist the practice of medical experts by offering appropriate medical information relevant to the medical case in hand. In this paper, we present a novel CDS search framework designed for passage retrieval from biomedical textbooks in order to support clinical decision-making using laboratory test results. The framework utilizes two unique characteristics of the textual reports derived from the test results, which are syntax variation and negation information. The proposed framework consists of three components: domain ontology, index repository, and query processing engine. We first created a domain ontology to resolve syntax variation by applying the ontology to detect medical concepts from the test results with language translation. We then preprocessed and performed indexing of biomedical textbooks recommended by clinicians for passage retrieval. We finally built the query-processing engine tailored for CDS, including translation, concept detection, query expansion, pseudo-relevance feedback at the local and global levels, and ranking with differential weighting of negation information. To evaluate the effectiveness of the proposed framework, we followed the standard information retrieval evaluation procedure. An evaluation dataset was created, including 28,581 textual reports for 30 laboratory test results and 56,228 passages from widely used biomedical textbooks, recommended by clinicians. Overall, our proposed passage retrieval framework, GPRF-NEG, outperforms the baseline by 36.2, 100.5, and 69.7 percent for MRR, R-precision, and Precision at 5, respectively. Our study results indicate that the proposed CDS search framework specifically designed for passage retrieval of biomedical literature represents a practically viable choice for clinicians as it supports their decision-making processes by providing relevant passages extracted from the sources that they prefer to refer to, with improved performances.

Query Expansion using Pseudo Relevance Feedback is a useful and a popular technique for reformulating the query. In our proposed query expansion method, we assume that relevant information can be found within a document near the central idea. The document is normally divided into sections, paragraphs and lines. The proposed method tries to extract keywords that are closer to the central theme of the document. The expansion terms are obtained by equi-frequency partition of the documents obtained from pseudo relevance feedback and by using tf-idf scores. The idf factor is calculated for number of partitions in documents. The group of words for query expansion is selected using the following approaches: the highest score, average score and a group of words that has maximum number of keywords. As each query behaved differently for different methods, the effect of these methods in selecting the words for query expansion is investigated. From this initial study, we extend the experiment to develop a rule-based statistical model that automatically selects the best group of words incorporating the tf-idf scoring and the 3 approaches explained here, in the future. The experiments were performed on FIRE 2011 Adhoc Hindi and English test collections on 50 queries each, using Terrier as retrieval engine.

Vocabulary mismatch in Information retrieval can be solved by Query Expansion (QE) techniques. Relevance feedback is a prominent solution to improve recall of retrieval system. Sometimes user may be reluctant and novice in providing feedback to improve the retrieval performance. Pseudo Relevance Feedback (PRF) automates the process. PRF treats top ranked resultant items are relevant and uses them to expand the query, which is not always correct. PRF by local analysis does not give guarantee to feedback positive terms to the system. Use of global analysis to capture the positive feedback is a regular practice in information retrieval process. This paper addresses the limitations of local analysis and global analysis alone by a novel approach that integrates both PRF and WordNet to select good expansion terms. The proposed solution filters the expansion terms and optimizes the expanded query. The proposed work is carried out on a huge Telugu text corpus collected from Wikipedia and other Telugu daily news portals.

Microblog retrieval has received much attention in recent years due to the wide spread of social microblogging platforms such as Twitter. Many research studies investigated different approaches for microblog retrieval. Query expansion is one of the approaches that showed stable performance for improving microblog retrieval effectiveness. In this paper we comprehensively investigate query expansion for microblog retrieval using pseudo relevance feedback (PRF) with different configurations. In addition, we propose a hyperlink-extended PRF by expanding queries using embedded hyperlinks in retrieved microblogs. Our experimental results on TREC microblog data showed that PRF alone could outperform many retrieval approaches if configured properly. Results also showed that utilizing hyperlinked documents in tweets for query expansion improves results over traditional PRF. Our best results achieved using hyperlink-extended PRF outperformed most of the reported results using other approaches.

Vision-Language Pretrained (VLP) models have achieved impressive performance on multimodal tasks, including text-image retrieval, based on dense representations. Meanwhile, Learned Sparse Retrieval (LSR) has gained traction in text-only settings due to its interpretability and efficiency with fast term-based lookup via inverted indexes. Inspired by these advantages, recent work has extended LSR to the multimodal domain. However, these methods often rely on computationally expensive contrastive pre-training, or distillation from a frozen dense model, which limits the potential for mutual enhancement. To address these limitations, we propose a simple yet effective framework that enables bi-directional learning between dense and sparse representations through Self-Knowledge Distillation. This bi-directional learning is achieved using an integrated similarity score-a weighted sum of dense and sparse similarities-which serves as a shared teacher signal for both representations. To ensure efficiency, we fine-tune the final layer of the dense encoder and the sparse projection head, enabling easy adaptation of any existing VLP model. Experiments on MSCOCO and Flickr30k demonstrate that our sparse retriever not only outperforms existing sparse baselines, but also achieves performance comparable to-or even surpassing-its dense counterparts, while retaining the benefits of sparse models.

At the core of any inference procedure, deep neural networks are dot product operations, which are the component that requires the highest computational resources. For instance, deep neural networks, such as VGG-16, require up to 15-G operations in order to perform the dot products present in a single forward pass, which results in significant energy consumption and thus limits their use in resource-limited environments, e.g., on embedded devices or smartphones. One common approach to reduce the complexity of the inference is to prune and quantize the weight matrices of the neural network. Usually, this results in matrices whose entropy values are low, as measured relative to the empirical probability mass distribution of its elements. In order to efficiently exploit such matrices, one usually relies on, inter alia, sparse matrix representations. However, most of these common matrix storage formats make strong statistical assumptions about the distribution of the elements; therefore, cannot efficiently represent the entire set of matrices that exhibit low-entropy statistics (thus, the entire set of compressed neural network weight matrices). In this paper, we address this issue and present new efficient representations for matrices with low-entropy statistics. Alike sparse matrix data structures, these formats exploit the statistical properties of the data in order to reduce the size and execution complexity. Moreover, we show that the proposed data structures can not only be regarded as a generalization of sparse formats but are also more energy and time efficient under practically relevant assumptions. Finally, we test the storage requirements and execution performance of the proposed formats on compressed neural networks and compare them to dense and sparse representations. We experimentally show that we are able to attain up to ×42 compression ratios, ×5 speed ups, and ×90 energy savings when we lossless convert the state-of-the-art networks, such as AlexNet, VGG-16, ResNet152, and DenseNet, into the new data structures and benchmark their respective dot product.

In this paper, we propose an effective saliency detection method based on dense and sparse representation in-terms of an optimized background template. Firstly, the input image is divided into compact and uniform super-pixels. Then, the optimized background template is produced by introducing boundary conductivity measurement to improve the dense and sparse representation of the image's super-pixels in terms of the optimized background, where the reconstruction error represents a saliency measure. Based on the optimized template, two saliency maps are generated by dense and sparse representation. Finally, the Bayesian framework used to integrate the two saliency maps to obtain the final one. Experimental results show that the proposed method performs favorably against eight state-of-the-art methods. In addition, the proposed method is shown to be more effective in highlighting the challenging salient objects that touch the image boundary.

Query-document vocabulary mismatch, the lack of query expressiveness for user needs and the phenomenon of short queries are the main issues associated with information retrieval systems. Query Expansion (QE) is one of the well-known alternative for overcoming these problems. It mainly involves finding synonyms or related words for the query terms. There are several approaches in the query expansion field such as statistical and semantic approaches; they focus on expanding the individual query terms rather than the entire query during the expansion process. An other category of approaches deals with the whole query by using a neural approach based on Pseudo Relevance feedback (PRF) documents. In this work, we carried out an ablation study to measure the impact of the classical and semantic (word embedding, order, context) based query expansion on the retrieval performance. The experiments conducted on the Arabic EveTAR dataset reveal that our hybrid proposed approach combining classical (PRF) and transformer (AraBERT) is competitive with the state-of-the-art methods. In fact, the obtained result in terms of the Mean Average Precision (MAP) is up to 0.72.

In microblog retrieval, query expansion can be essential to obtain good search results due to the short size of queries and posts. Since information in microblogs is highly dynamic, an up-to-date index coupled with pseudo-relevance feedback (PRF) with an external corpus has a higher chance of retrieving more relevant documents and improving ranking. In this paper, we focus on the research question:how can we reduce the query expansion computational cost while maintaining the same retrieval precision as standard PRF? Therefore, we propose to accelerate the query expansion step of pseudo-relevance feedback. The hypothesis is that using an expansion corpus organized into verticals for expanding the query, will lead to a more efficient query expansion process and improved retrieval effectiveness. Thus, the proposed query expansion method uses a distributed search architecture and resource selection algorithms to provide an efficient query expansion process. Experiments on the TREC Microblog datasets show that the proposed approach can match or outperform standard PRF in MAP and NDCG@30, with a computational cost that is three orders of magnitude lower.

Dense feature representations are recently popular for human action recognition based on video. They achieve better performance than sparse local representations. However, dense representations usually lead to high-dimensional data which contain plenty of redundancy information. In order to solve this issue, a novel compact representation based on global Gist feature and 2DPCA-2DLDA is proposed. Firstly Global grid based Gist features are extracted from image sequences and described as Gist matrices. The Gist matrices reflect global structure distribution of local grids, and at the same time depict frequency characteristics of the local grids. Secondly, bilateral dimensionality reduction is developed to our proposed Gist matrices. The most discriminative spatial structure and frequency characteristic of local grids are separately selected with 2DPCA and 2DLDA. Finally, the derived structure and frequency properties are simultaneously embedded into a low dimensional compact global descriptor. Experimental results on UCF Sports dataset validate that the proposed method makes a good compromise between feature discrimination and dimensions. And the action recognition accuracy is promising.

Emotion detection from text has become a popular task due to the key role of emotions in human-machine interaction. Current approaches represent text as a sparse bag-of-words vector. In this work, we propose a new approach that utilizes pre-trained, dense word embedding representations. We introduce an ensemble approach combining both sparse and dense representations. Our experiments include five datasets for emotion detection from different domains and show an average improvement of 11.6% in macro average F1-score.

Pseudo-Relevance Feedback (PRF) utilises the relevance signals from the top-k passages from the first round of retrieval to perform a second round of retrieval aiming to improve search effectiveness. A recent research direction has been the study and development of PRF methods for deep language model based rankers, and in particular in the context of dense retrievers. Dense retrievers provide a trade off between effectiveness, which is often reduced compared to more complex neural rankers, and query latency, which also is reduced making the retrieval pipeline more efficient. The introduction of PRF methods for dense retrievers has been motivated as an attempt to further improve their effectiveness. In this paper, we reproduce and study a recent method for PRF with dense retrievers, called ANCE-PRF. This method concatenates the query text and that of the top-k feedback passages to form a new query input, which is then encoded into a dense representation using a newly trained query encoder based on the original dense retriever used for the first round of retrieval. While the method can potentially be applied to any of the existing dense retrievers, prior work has studied it only in the context of the ANCE dense retriever. We study the reproducibility of ANCE-PRF in terms of both its training (encoding of the PRF signal) and inference (ranking) steps. We further extend the empirical analysis provided in the original work to investigate the effect of the hyper-parameters that govern the training process and the robustness of the method across these different settings. Finally, we contribute a study of the generalisability of the ANCE-PRF method when dense retrievers other than ANCE are used for the first round of retrieval and for encoding the PRF signal.

Open-domain question answering relies on efficient passage retrieval to select candidate contexts, where traditional sparse vector space models, such as TF-IDF or BM25, are the de facto method. In this work, we show that retrieval can be practically implemented using dense representations alone, where embeddings are learned from a small number of questions and passages by a simple dual-encoder framework. When evaluated on a wide range of open-domain QA datasets, our dense retriever outperforms a strong Lucene-BM25 system largely by 9%-19% absolute in terms of top-20 passage retrieval accuracy, and helps our end-to-end QA system establish new state-of-the-art on multiple open-domain QA benchmarks.

This paper presents our participation at the ImageCLEF Photo 2008 task. We submitted six runs, experimenting with our own block-based visual retrieval as well as with query expansion. The results we obtained show that despite the poor performance of the visual and text retrieval components, better results can be obtained through pseudorelevance feedback and the inter-media fusion of the results.

The objective of the paper is to amalgamate theories of text retrieval from various research traditions into a cognitive theory for information retrieval interaction. Set in a cognitive framework, the paper outlines the concept of polyrepresentation applied to both the user's cognitive space and the information space of IR systems. The concept seeks to represent the current user's information need, problem state, and domain work task or interest in a structure of causality. Further, it implies that we should apply different methods of representation and a variety of IR techniques of different cognitive and functional origin simultaneously to each semantic full‐text entity in the information space. The cognitive differences imply that by applying cognitive overlaps of information objects, originating from different interpretations of such objects through time and by type, the degree of uncertainty inherent in IR is decreased. Polyrepresentation and the use of cognitive overlaps are associated with, but not identical to, data fusion in IR. By explicitly incorporating all the cognitive structures participating in the interactive communication processes during IR, the cognitive theory provides a comprehensive view of these processes. It encompasses the ad hoc theories of text retrieval and IR techniques hitherto developed in mainstream retrieval research. It has elements in common with van Rijsbergen and Lalmas' logical uncertainty theory and may be regarded as compatible with that conception of IR. Epistemologically speaking, the theory views IR interaction as processes of cognition, potentially occurring in all the information processing components of IR, that may be applied, in particular, to the user in a situational context. The theory draws upon basic empirical results from information seeking investigations in the operational online environment, and from mainstream IR research on partial matching techniques and relevance feedback. By viewing users, source systems, intermediary mechanisms and information in a global context, the cognitive perspective attempts a comprehensive understanding of essential IR phenomena and concepts, such as the nature of information needs, cognitive inconsistency and retrieval overlaps, logical uncertainty, the concept of ‘document’, relevance measures and experimental settings. An inescapable consequence of this approach is to rely more on sociological and psychological investigative methods when evaluating systems and to view relevance in IR as situational, relative, partial, differentiated and non‐linear. The lack of consistency among authors, indexers, evaluators or users is of an identical cognitive nature. It is unavoidable, and indeed favourable to IR. In particular, for full‐text retrieval, alternative semantic entities, including Salton et al.'s ‘passage retrieval’, are proposed to replace the traditional document record as the basic retrieval entity. These empirically observed phenomena of inconsistency and of semantic entities and values associated with data interpretation support strongly a cognitive approach to IR and the logical use of polyrepresentation, cognitive overlaps, and both data fusion and data diffusion.