This work presents a two-stage neural architecture for learning and refining structural correspondences between graphs. First, we use localized node embeddings computed by a graph neural network to obtain an initial ranking of soft correspondences between nodes. Secondly, we employ synchronous message passing networks to iteratively re-rank the soft correspondences to reach a matching consensus in local neighborhoods between graphs. We show, theoretically and empirically, that our message passing scheme computes a well-founded measure of consensus for corresponding neighborhoods, which is then used to guide the iterative re-ranking process. Our purely local and sparsity-aware architecture scales well to large, real-world inputs while still being able to recover global correspondences consistently. We demonstrate the practical effectiveness of our method on real-world tasks from the fields of computer vision and entity alignment between knowledge graphs, on which we improve upon the current state-of-the-art. Our source code is available under https://github.com/rusty1s/ deep-graph-matching-consensus.
Citation recommendation systems for the scientific literature, to help authors find papers that should be cited, have the potential to speed up discoveries and uncover new routes for scientific exploration. We treat this task as a ranking problem, which we tackle with a two-stage approach: candidate generation followed by re-ranking. Within this framework, we adapt to the scientific domain a proven combination based on "bag of words" retrieval followed by re-scoring with a BERT model. We experimentally show the effects of domain adaptation, both in terms of pretraining on in-domain data and exploiting in-domain vocabulary. In addition, we introduce a novel navigation-based document expansion strategy to enrich the candidate documents processed by our neural models. On three different collections from different scientific disciplines, we achieve the best-reported results in the citation recommendation task.
Current benchmarks for optical flow algorithms evaluate the estimation either directly by comparing the predicted flow fields with the ground truth or indirectly by using the predicted flow fields for frame interpolation and then comparing the interpolated frames with the actual frames. In the latter case, objective quality measures such as the mean squared error are typically employed. However, it is well known that for image quality assessment, the actual quality experienced by the user cannot be fully deduced from such simple measures. Hence, we conducted a subjective quality assessment crowdscouring study for the interpolated frames provided by one of the optical flow benchmarks, the Middlebury benchmark. We collected forced-choice paired comparisons between interpolated images and corresponding ground truth. To increase the sensitivity of observers when judging minute difference in paired comparisons we introduced a new method to the field of full-reference quality assessment, called artefact amplification. From the crowdsourcing data, we reconstructed absolute quality scale values according to Thurstone's model. As a result, we obtained a re-ranking of the 155 participating algorithms w.r.t. the visual quality of the interpolated frames. This re-ranking not only shows the necessity of visual quality assessment as another evaluation metric for optical flow and frame interpolation benchmarks, the results also provide the ground truth for designing novel image quality assessment (IQA) methods dedicated to perceptual quality of interpolated images. As a first step, we proposed such a new full-reference method, called WAE-IQA. By weighing the local differences between an interpolated image and its ground truth WAE-IQA performed slightly better than the currently best FR-IQA approach from the literature.
This paper proposes Attribute Attention Network (AANet), a new architecture that integrates person attributes and attribute attention maps into a classification framework to solve the person re-identification (re-ID) problem. Many person re-ID models typically employ semantic cues such as body parts or human pose to improve the re-ID performance. Attribute information, however, is often not utilized. The proposed AANet leverages on a baseline model that uses body parts and integrates the key attribute information in an unified learning framework. The AANet consists of a global person ID task, a part detection task and a crucial attribute detection task. By estimating the class responses of individual attributes and combining them to form the attribute attention map (AAM), a very strong discriminatory representation is constructed. The proposed AANet outperforms the best state-of-the-art method arXiv:1711.09349v3 [cs.CV] using ResNet-50 by 3.36% in mAP and 3.12% in Rank-1 accuracy on DukeMTMC-reID dataset. On Market1501 dataset, AANet achieves 92.38% mAP and 95.10% Rank-1 accuracy with re-ranking, outperforming arXiv:1804.00216v1 [cs.CV], another state of the art method using ResNet-152, by 1.42% in mAP and 0.47% in Rank-1 accuracy. In addition, AANet can perform person attribute prediction (e.g., gender, hair length, clothing length etc.), and localize the attributes in the query image.
In this paper we look beyond metrics-based evaluation of Information Retrieval systems, to explore the reasons behind ranking results. We present the content-focused Neural-IR-Explorer, which empowers users to browse through retrieval results and inspect the inner workings and fine-grained results of neural re-ranking models. The explorer includes a categorized overview of the available queries, as well as an individual query result view with various options to highlight semantic connections between query-document pairs. The Neural-IR-Explorer is available at: https://neural-ir-explorer.ec.tuwien.ac.at/
The usage of neural network models puts multiple objectives in conflict with each other: Ideally we would like to create a neural model that is effective, efficient, and interpretable at the same time. However, in most instances we have to choose which property is most important to us. We used the opportunity of the TREC 2019 Deep Learning track to evaluate the effectiveness of a balanced neural re-ranking approach. We submitted results of the TK (Transformer-Kernel) model: a neural re-ranking model for ad-hoc search using an efficient contextualization mechanism. TK employs a very small number of lightweight Transformer layers to contextualize query and document word embeddings. To score individual term interactions, we use a document-length enhanced kernel-pooling, which enables users to gain insight into the model. Our best result for the passage ranking task is: 0.420 MAP, 0.671 nDCG, 0.598 P@10 (TUW19-p3 full). Our best result for the document ranking task is: 0.271 MAP, 0.465 nDCG, 0.730 P@10 (TUW19-d3 re-ranking).
Most syntactic dependency parsing models may fall into one of two categories: transition- and graph-based models. The former models enjoy high inference efficiency with linear time complexity, but they rely on the stacking or re-ranking of partially-built parse trees to build a complete parse tree and are stuck with slower training for the necessity of dynamic oracle training. The latter, graph-based models, may boast better performance but are unfortunately marred by polynomial time inference. In this paper, we propose a novel parsing order objective, resulting in a novel dependency parsing model capable of both global (in sentence scope) feature extraction as in graph models and linear time inference as in transitional models. The proposed global greedy parser only uses two arc-building actions, left and right arcs, for projective parsing. When equipped with two extra non-projective arc-building actions, the proposed parser may also smoothly support non-projective parsing. Using multiple benchmark treebanks, including the Penn Treebank (PTB), the CoNLL-X treebanks, and the Universal Dependency Treebanks, we evaluate our parser and demonstrate that the proposed novel parser achieves good performance with faster training and decoding.
We explore neural language modeling for speech recognition where the context spans multiple sentences. Rather than encode history beyond the current sentence using a cache of words or document-level features, we focus our study on the ability of LSTM and Transformer language models to implicitly learn to carry over context across sentence boundaries. We introduce a new architecture that incorporates an attention mechanism into LSTM to combine the benefits of recurrent and attention architectures. We conduct language modeling and speech recognition experiments on the publicly available LibriSpeech corpus. We show that conventional training on a paragraph-level corpus results in significant reductions in perplexity compared to training on a sentence-level corpus. We also describe speech recognition experiments using long-span language models in second-pass re-ranking, and provide insights into the ability of such models to take advantage of context beyond the current sentence.
The increasing availability of semantic data has substantially enhanced Web applications. Semantic data such as RDF data is commonly represented as entity-property-value triples. The magnitude of semantic data, in particular the large number of triples describing an entity, could overload users with excessive amounts of information. This has motivated fruitful research on automated generation of summaries for entity descriptions to satisfy users' information needs efficiently and effectively. We focus on this prominent topic of entity summarization, and our research objective is to present the first comprehensive survey of entity summarization research. Rather than separately reviewing each method, our contributions include (1) identifying and classifying technical features of existing methods to form a high-level overview, (2) identifying and classifying frameworks for combining multiple technical features adopted by existing methods, (3) collecting known benchmarks for intrinsic evaluation and efforts for extrinsic evaluation, and (4) suggesting research directions for future work. By investigating the literature, we synthesized two hierarchies of techniques. The first hierarchy categories generic technical features into several perspectives: frequency and centrality, informativeness, and diversity and coverage. In the second hierarchy we present domain-specific and task-specific technical features, including the use of domain knowledge, context awareness, and personalization. Our review demonstrated that existing methods are mainly unsupervised and they combine multiple technical features using various frameworks: random surfer models, similarity-based grouping, MMR-like re-ranking, or combinatorial optimization. We also found a few deep learning based methods in recent research.
This paper describes Facebook AI's submission to the WAT 2019 Myanmar-English translation task. Our baseline systems are BPE-based transformer models. We explore methods to leverage monolingual data to improve generalization, including self-training, back-translation and their combination. We further improve results by using noisy channel re-ranking and ensembling. We demonstrate that these techniques can significantly improve not only a system trained with additional monolingual data, but even the baseline system trained exclusively on the provided small parallel dataset. Our system ranks first in both directions according to human evaluation and BLEU, with a gain of over 8 BLEU points above the second best system.
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 extraction of the most relevant sentence from a given PubMed abstract. We investigate (1) attention based supervised neural topic model and SVM for retrieval and ranking of PubMed abstracts and, further utilize BM25 and other relevance measures for re-ranking, (2) supervised and unsupervised sentence ranking models utilizing multi-view representations comprising of query-aware attention-based sentence representation (QAR), bag-of-words (BoW) and TF-IDF. Our best systems achieved 1st rank and scored 0.86 mean average precision (mAP) and 0.58 macro average accuracy (MAA) in Task-1 and Task-2 respectively.
We present "AutoJudge", an automated evaluation method for conversational dialogue systems. The method works by first generating dialogues based on self-talk, i.e. dialogue systems talking to itself. Then, it uses human ratings on these dialogues to train an automated judgement model. Our experiments show that AutoJudge correlates well with the human ratings and can be used to automatically evaluate dialogue systems, even in deployed systems. In a second part, we attempt to apply AutoJudge to improve existing systems. This works well for re-ranking a set of candidate utterances. However, our experiments show that AutoJudge cannot be applied as reward for reinforcement learning, although the metric can distinguish good from bad dialogues. We discuss potential reasons, but state here already that this is still an open question for further research.
We present DeepPrint, a deep network, which learns to extract fixed-length fingerprint representations of only 200 bytes. DeepPrint incorporates fingerprint domain knowledge, including alignment and minutiae detection, into the deep network architecture to maximize the discriminative power of its representation. The compact, DeepPrint representation has several advantages over the prevailing variable length minutiae representation which (i) requires computationally expensive graph matching techniques, (ii) is difficult to secure using strong encryption schemes (e.g. homomorphic encryption), and (iii) has low discriminative power in poor quality fingerprints where minutiae extraction is unreliable. We benchmark DeepPrint against two top performing COTS SDKs (Verifinger and Innovatrics) from the NIST and FVC evaluations. Coupled with a re-ranking scheme, the DeepPrint rank-1 search accuracy on the NIST SD4 dataset against a gallery of 1.1 million fingerprints is comparable to the top COTS matcher, but it is significantly faster (DeepPrint: 98.80% in 0.3 seconds vs. COTS A: 98.85% in 27 seconds). To the best of our knowledge, the DeepPrint representation is the most compact and discriminative fixed-length fingerprint representation reported in the academic literature.
In this work we describe the system from Natural Language Processing group at Arizona State University for the TextGraphs 2019 Shared Task. The task focuses on Explanation Regeneration, an intermediate step towards general multi-hop inference on large graphs. Our approach consists of modeling the explanation regeneration task as a \textit{learning to rank} problem, for which we use state-of-the-art language models and explore dataset preparation techniques. We utilize an iterative re-ranking based approach to further improve the rankings. Our system secured 2nd rank in the task with a mean average precision (MAP) of 41.3\% on the test set.
Generating the Top-N recommendations from a large corpus is computationally expensive to perform at scale. Candidate generation and re-ranking based approaches are often adopted in industrial settings to alleviate efficiency problems. However it remains to be fully studied how well such schemes approximate complete rankings (or how many candidates are required to achieve a good approximation), or to develop systematic approaches to generate high-quality candidates efficiently. In this paper, we seek to investigate these questions via proposing a candidate generation and re-ranking based framework (CIGAR), which first learns a preference-preserving binary embedding for building a hash table to retrieve candidates, and then learns to re-rank the candidates using real-valued ranking models with a candidate-oriented objective. We perform a comprehensive study on several large-scale real-world datasets consisting of millions of users/items and hundreds of millions of interactions. Our results show that CIGAR significantly boosts the Top-N accuracy against state-of-the-art recommendation models, while reducing the query time by orders of magnitude. We hope that this work could draw more attention to the candidate generation problem in recommender systems.
Product search serves as an important entry point for online shopping. In contrast to web search, the retrieved results in product search not only need to be relevant but also should satisfy customers' preferences in order to elicit purchases. Previous work has shown the efficacy of purchase history in personalized product search. However, customers with little or no purchase history do not benefit from personalized product search. Furthermore, preferences extracted from a customer's purchase history are usually long-term and may not always align with her short-term interests. Hence, in this paper, we leverage clicks within a query session, as implicit feedback, to represent users' hidden intents, which further act as the basis for re-ranking subsequent result pages for the query. It has been studied extensively to model user preference with implicit feedback in recommendation tasks. However, there has been little research on modeling users' short-term interest in product search. We study whether short-term context could help promote users' ideal item in the following result pages for a query. Furthermore, we propose an end-to-end context-aware embedding model which can capture long-term and short-term context dependencies. Our experimental results on the datasets collected from the search log of a commercial product search engine show that short-term context leads to much better performance compared with long-term and no context. Our results also show that our proposed model is more effective than word-based context-aware models.
This paper proposes the Global-Local Temporal Representation (GLTR) to exploit the multi-scale temporal cues in video sequences for video person Re-Identification (ReID). GLTR is constructed by first modeling the short-term temporal cues among adjacent frames, then capturing the long-term relations among inconsecutive frames. Specifically, the short-term temporal cues are modeled by parallel dilated convolutions with different temporal dilation rates to represent the motion and appearance of pedestrian. The long-term relations are captured by a temporal self-attention model to alleviate the occlusions and noises in video sequences. The short and long-term temporal cues are aggregated as the final GLTR by a simple single-stream CNN. GLTR shows substantial superiority to existing features learned with body part cues or metric learning on four widely-used video ReID datasets. For instance, it achieves Rank-1 Accuracy of 87.02% on MARS dataset without re-ranking, better than current state-of-the art.
We implement a method for re-ranking top-10 results of a state-of-the-art question answering (QA) system. The goal of our re-ranking approach is to improve the answer selection given the user question and the top-10 candidates. We focus on improving deployed QA systems that do not allow re-training or re-training comes at a high cost. Our re-ranking approach learns a similarity function using n-gram based features using the query, the answer and the initial system confidence as input. Our contributions are: (1) we generate a QA training corpus starting from 877 answers from the customer care domain of T-Mobile Austria, (2) we implement a state-of-the-art QA pipeline using neural sentence embeddings that encode queries in the same space than the answer index, and (3) we evaluate the QA pipeline and our re-ranking approach using a separately provided test set. The test set can be considered to be available after deployment of the system, e.g., based on feedback of users. Our results show that the system performance, in terms of top-n accuracy and the mean reciprocal rank, benefits from re-ranking using gradient boosted regression trees. On average, the mean reciprocal rank improves by 9.15%.
Recent advances in Session-based recommender systems have gained attention due to their potential of providing real-time personalized recommendations with high recall, especially when compared to traditional methods like matrix factorization and item-based collaborative filtering. Nowadays, two of the most recent methods are Short-Term Attention/Memory Priority Model for Session-based Recommendation (STAMP) and Neural Attentive Session-based Recommendation (NARM). However, when these two methods were applied in the similar-item recommendation dataset of Zalando (Fashion-Similar), they did not work out-of-the-box compared to a simple Collaborative-Filtering approach. Aiming for improving the similar-item recommendation, we propose to concentrate efforts on enhancing the rank of the few most relevant items from the original recommendations, by employing the information of the session of the user encoded by an attention network. The efficacy of this strategy was confirmed when using a novel Candidate Rank Embedding that encodes the global ranking information of each candidate in the re-ranking process. Experimental results in Fashion-Similar show significant improvements over the baseline on Recall and MRR at 20, as well as improvements in Click Through Rate based on an online test. Additionally, it is important to point out from the evaluation that was performed the potential of this method on the next click prediction problem because when applied to STAMP and NARM, it improves the Recall and MRR at 20 on two publicly available real-world datasets.
De novo peptide sequencing algorithms have been widely used in proteomics to analyse tandem mass spectra (MS/MS) and assign them to peptides, but quality-control methods to evaluate the confidence of de novo peptide sequencing are lagging behind. A fundamental part of a quality-control method is the scoring function used to evaluate the quality of peptide-spectrum matches (PSMs). Here, we propose a genetic programming (GP) based method, called GP-PSM, to learn a PSM scoring function for improving the rate of confident peptide identification from MS/MS data. The GP method learns from thousands of MS/MS spectra. Important characteristics about goodness of the matches are extracted from the learning set and incorporated into the GP scoring functions. We compare GP-PSM with two methods including Support Vector Regression (SVR) and Random Forest (RF). The GP method along with RF and SVR, each is used for post-processing the results of peptide identification by PEAKS, a commonly used de novo sequencing method. The results show that GP-PSM outperforms RF and SVR and discriminates accurately between correct and incorrect PSMs. It correctly assigns peptides to 10% more spectra on an evaluation dataset containing 120 MS/MS spectra and decreases the false positive rate (FPR) of peptide identification.