Modeling Relevance Ranking under the Pre-training and Fine-tuning Paradigm
School of Information, Renmin University of China（中国人民大学信息学院）；Gaoling School of Artificial Intelligence, Renmin University of China（中国人民大学高瓴人工智能学院）；Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）；Huawei Noah’s Ark Lab（华为诺亚方舟实验室）
Lin Bo, Liang Pang, Gang Wang, Jun Xu, XiuQiang He, Ji-Rong Wen
Recently, pre-trained language models such as BERT have been applied to document ranking for information retrieval, which first pre-train a general language model on an unlabeled large corpus and then conduct ranking-specific fine-tuning on expert-labeled relevance datasets. Ideally, an IR system would model relevance from a user-system dualism: the user's view and the system's view. User's view judges the relevance based on the activities of "real users" while the system's view focuses on the relevance signals from the system side, e.g., from the experts or algorithms, etc. Inspired by the user-system relevance views and the success of pre-trained language models, in this paper we propose a novel ranking framework called Pre-Rank that takes both user's view and system's view into consideration, under the pre-training and fine-tuning paradigm. Specifically, to model the user's view of relevance, Pre-Rank pre-trains the initial query-document representations based on large-scale user activities data such as the click log. To model the system's view of relevance, Pre-Rank further fine-tunes the model on expert-labeled relevance data. More importantly, the pre-trained representations, are fine-tuned together with handcrafted learning-to-rank features under a wide and deep network architecture. In this way, Pre-Rank can model the relevance by incorporating the relevant knowledge and signals from both real search users and the IR experts. To verify the effectiveness of Pre-Rank, we showed two implementations by using BERT and SetRank as the underlying ranking model, respectively. Experimental results base on three publicly available benchmarks showed that in both of the implementations, Pre-Rank can respectively outperform the underlying ranking models and achieved state-of-the-art performances.
Beyond Relevance-Centric Retrieval: Rubric-Oriented Document Set Selection and Ranking
University of Science and Technology of China（中国科学技术大学）；Yuanbao Team, Tencent（腾讯元宝团队）；University of Chinese Academy of Sciences（中国科学院大学）；Shandong University（山东大学）
In this paper, we investigate the integration of sentence position and semantic role of words in a PageRank system to build a key phrase ranking method. We present the evaluation results of our approach on three scientific articles. We show that semantic role information, when integrated with a PageRank system, can become a new lexical feature. Our approach had an overall improvement on all the data sets over the state-of-art baseline approaches.
There are many existing retrieval and question answering datasets. However, most of them either focus on ranked list evaluation or single-candidate question answering. This divide makes it challenging to properly evaluate approaches concerned with ranking documents and providing snippets or answers for a given query. In this work, we present FiRA: a novel dataset of Fine-Grained Relevance Annotations. We extend the ranked retrieval annotations of the Deep Learning track of TREC 2019 with passage and word level graded relevance annotations for all relevant documents. We use our newly created da…
Pairing a lexical retriever with a neural re-ranking model has set state-of-the-art performance on large-scale information retrieval datasets. This pipeline covers scenarios like question answering or navigational queries, however, for information-seeking scenarios, users often provide information on whether a document is relevant to their query in form of clicks or explicit feedback. Therefore, in this work, we explore how relevance feedback can be directly integrated into neural re-ranking models by adopting few-shot and parameter-efficient learning techniques. Specifically, we introduce a k…
Athens University of Economics and Business（雅典经济与商业大学）；Google AI（谷歌人工智能）
We explore several new models for document relevance ranking, building upon the Deep Relevance Matching Model (DRMM) of Guo et al. (2016). Unlike DRMM, which uses context-insensitive encodings of terms and query-document term interactions, we inject rich context-sensitive encodings throughout our models, inspired by PACRR's (Hui et al., 2017) convolutional n-gram matching features, but extended in several ways including multiple views of query and document inputs. We test our models on datasets from the BIOASQ question answering challenge (Tsatsaronis et al., 2015) and TREC ROBUST 2004 (Voorhe…
Web search engines focus on serving highly relevant results within hundreds of milliseconds. Pre-trained language transformer models such as BERT are therefore hard to use in this scenario due to their high computational demands. We present our real-time approach to the document ranking problem leveraging a BERT-based siamese architecture. The model is already deployed in a commercial search engine and it improves production performance by more than 3%. For further research and evaluation, we release DaReCzech, a unique data set of 1.6 million Czech user query-document pairs with manually assi…
In this paper, we try to answer the question of how to improve the state-of-the-art methods for relevance ranking in web search by query segmentation. Here, by query segmentation it is meant to segment the input query into segments, typically natural language phrases, so that the performance of relevance ranking in search is increased. We propose employing the re-ranking approach in query segmentation, which first employs a generative model to create top $k$ candidates and then employs a discriminative model to re-rank the candidates to obtain the final segmentation result. The method has been…
University of Maryland, College Park（马里兰大学帕克分校）；Facebook Conversational AI（Facebook 对话人工智能）；University of Waterloo（滑铁卢大学）；Stanford University（斯坦福大学）；Comcast Applied AI Research（康卡斯特应用人工智能研究）
Despite substantial interest in applications of neural networks to information retrieval, neural ranking models have only been applied to standard ad hoc retrieval tasks over web pages and newswire documents. This paper proposes MP-HCNN (Multi-Perspective Hierarchical Convolutional Neural Network) a novel neural ranking model specifically designed for ranking short social media posts. We identify document length, informal language, and heterogeneous relevance signals as features that distinguish documents in our domain, and present a model specifically designed with these characteristics in mi…
This paper concerns a deep learning approach to relevance ranking in information retrieval (IR). Existing deep IR models such as DSSM and CDSSM directly apply neural networks to generate ranking scores, without explicit understandings of the relevance. According to the human judgement process, a relevance label is generated by the following three steps: 1) relevant locations are detected, 2) local relevances are determined, 3) local relevances are aggregated to output the relevance label. In this paper we propose a new deep learning architecture, namely DeepRank, to simulate the above human ju…
Recent work has proposed stochastic Plackett-Luce (PL) ranking models as a robust choice for optimizing relevance and fairness metrics. Unlike their deterministic counterparts that require heuristic optimization algorithms, PL models are fully differentiable. Theoretically, they can be used to optimize ranking metrics via stochastic gradient descent. However, in practice, the computation of the gradient is infeasible because it requires one to iterate over all possible permutations of items. Consequently, actual applications rely on approximating the gradient via sampling techniques. In this p…
In this paper, we propose a salient-context based semantic matching method to improve relevance ranking in information retrieval. We first propose a new notion of salient context and then define how to measure it. Then we show how the most salient context can be located with a sliding window technique. Finally, we use the semantic similarity between a query term and the most salient context terms in a corpus of documents to rank those documents. Experiments on various collections from TREC show the effectiveness of our model compared to the state-of-the-art methods.
Ranking systems are ubiquitous in modern Internet services, including online marketplaces, social media, and search engines. Traditionally, ranking systems only focus on how to get better relevance estimation. When relevance estimation is available, they usually adopt a user-centric optimization strategy where ranked lists are generated by sorting items according to their estimated relevance. However, such user-centric optimization ignores the fact that item providers also draw utility from ranking systems. It has been shown in existing research that such user-centric optimization will cause m…
Ranking evaluation metrics are a fundamental element of design and improvement efforts in information retrieval. We observe that most popular metrics disregard information portrayed in the scores used to derive rankings, when available. This may pose a numerical scaling problem, causing an under- or over-estimation of the evaluation depending on the degree of divergence between the scores of ranked items. The purpose of this work is to propose a principled way of quantifying multi-graded relevance judgments of items and enable a more accurate penalization of ordering errors in rankings. We pro…
The problem of proximity full-text search is considered. If a search query contains high-frequently occurring words, then multi-component key indexes deliver an improvement in the search speed compared with ordinary inverted indexes. It was shown that we can increase the search speed by up to 130 times in cases when queries consist of high-frequently occurring words. In this paper, we investigate how the multi-component key index architecture affects the quality of the search. We consider several well-known methods of relevance ranking, where these methods are of different authors. Using these…
The increasing popularity of Twitter and other microblogs makes improved trustworthiness and relevance assessment of microblogs evermore important. We propose a method of ranking of tweets considering trustworthiness and content based popularity. The analysis of trustworthiness and popularity exploits the implicit relationships between the tweets. We model microblog ecosystem as a three-layer graph consisting of : (i) users (ii) tweets and (iii) web pages. We propose to derive trust and popularity scores of entities in these three layers, and propagate the scores to tweets considering the inte…
Optimizing Group-Fair Plackett-Luce Ranking Models for Relevance and Ex-Post Fairness
Indian Institute of Science（印度科学研究所）；University of California, Berkeley（加利福尼亚大学伯克利分校）；Microsoft Research, Bengaluru（微软研究院班加罗尔分院）；Microsoft Research（微软研究院）
Transformer based re-ranking models can achieve high search relevance through context-aware soft matching of query tokens with document tokens. To alleviate runtime complexity of such inference, previous work has adopted a late interaction architecture with pre-computed contextual token representations at the cost of a large online storage. This paper proposes contextual quantization of token embeddings by decoupling document-specific and document-independent ranking contributions during codebook-based compression. This allows effective online decompression and embedding composition for better…
Rankings, especially those in search and recommendation systems, often determine how people access information and how information is exposed to people. Therefore, how to balance the relevance and fairness of information exposure is considered as one of the key problems for modern IR systems. As conventional ranking frameworks that myopically sorts documents with their relevance will inevitably introduce unfair result exposure, recent studies on ranking fairness mostly focus on dynamic ranking paradigms where result rankings can be adapted in real-time to support fairness in groups (i.e., race…
Invenio is a comprehensive web-based free digital library software suite originally developed at CERN. In order to improve its information retrieval and word similarity ranking capabilities, the goal of this thesis is to enhance Invenio by bridging it with modern external information retrieval systems. In the first part a comparison of various information retrieval systems such as Solr and Xapian is made. In the second part a system-independent bridge for word similarity ranking is designed and implemented. Subsequently, Solr and Xapian are integrated in Invenio via adapters to the bridge. In …
Georgetown University（乔治城大学）；Max Planck Institute for Informatics（马克斯·普朗克信息学研究所）；Amazon（亚马逊）
One challenge with neural ranking is the need for a large amount of manually-labeled relevance judgments for training. In contrast with prior work, we examine the use of weak supervision sources for training that yield pseudo query-document pairs that already exhibit relevance (e.g., newswire headline-content pairs and encyclopedic heading-paragraph pairs). We also propose filtering techniques to eliminate training samples that are too far out of domain using two techniques: a heuristic-based approach and novel supervised filter that re-purposes a neural ranker. Using several leading neural ra…
Siemens AG（西门子股份公司）；University of Munich (LMU)（慕尼黑大学（LMU））
This paper presents our system details and results of participation in the RDoC Tasks of BioNLP-OST 2019. Research Domain Criteria (RDoC) construct is a multi-dimensional and broad framework to describe mental health disorders by combining knowledge from genomics to behaviour. Non-availability of RDoC labelled dataset and tedious labelling process hinders the use of RDoC framework to reach its full potential in Biomedical research community and Healthcare industry. Therefore, Task-1 aims at retrieval and ranking of PubMed abstracts relevant to a given RDoC construct and Task-2 aims at extracti…
Learning to Rank has traditionally considered settings where given the relevance information of objects, the desired order in which to rank the objects is clear. However, with today's large variety of users and layouts this is not always the case. In this paper, we consider so-called complex ranking settings where it is not clear what should be displayed, that is, what the relevant items are, and how they should be displayed, that is, where the most relevant items should be placed. These ranking settings are complex as they involve both traditional ranking and inferring the best display order.…
In information retrieval (IR), candidate set pruning has been commonly used to speed up two-stage relevance ranking. However, such an approach lacks accurate error control and often trades accuracy off against computational efficiency in an empirical fashion, lacking theoretical guarantees. In this paper, we propose the concept of certified error control of candidate set pruning for relevance ranking, which means that the test error after pruning is guaranteed to be controlled under a user-specified threshold with high probability. Both in-domain and out-of-domain experiments show that our met…
Although considerable efforts have been devoted to transformer-based ranking models for document search, the relevance-efficiency tradeoff remains a critical problem for ad-hoc ranking. To overcome this challenge, this paper presents BECR (BERT-based Composite Re-Ranking), a composite re-ranking scheme that combines deep contextual token interactions and traditional lexical term-matching features. In particular, BECR exploits a token encoding mechanism to decompose the query representations into pre-computable uni-grams and skip-n-grams. By applying token encoding on top of a dual-encoder arch…
In this paper, an Eliteness Hypothesis for information retrieval is proposed, where we define two generative processes to create information items and queries. By assuming the deterministic relationships between the eliteness of terms and relevance, we obtain a new theoretical retrieval framework. The resulting ranking function is a unified one as it is capable of using available relevance information on both the document and the query, which is otherwise unachievable by existing retrieval models. Our preliminary experiment on a simple ranking function has demonstrated the potential of the app…
State-of-the-art neural models typically encode document-query pairs using cross-attention for re-ranking. To this end, models generally utilize an encoder-only (like BERT) paradigm or an encoder-decoder (like T5) approach. These paradigms, however, are not without flaws, i.e., running the model on all query-document pairs at inference-time incurs a significant computational cost. This paper proposes a new training and inference paradigm for re-ranking. We propose to finetune a pretrained encoder-decoder model using in the form of document to query generation. Subsequently, we show that this e…
Tsinghua University（清华大学）；Beijing National Research Center for Information Science and Technology（北京信息科学与技术国家研究中心）；Microsoft Research（微软研究院）
The effectiveness of Neural Information Retrieval (Neu-IR) often depends on a large scale of in-domain relevance training signals, which are not always available in real-world ranking scenarios. To democratize the benefits of Neu-IR, this paper presents MetaAdaptRank, a domain adaptive learning method that generalizes Neu-IR models from label-rich source domains to few-shot target domains. Drawing on source-domain massive relevance supervision, MetaAdaptRank contrastively synthesizes a large number of weak supervision signals for target domains and meta-learns to reweight these synthetic "weak…
