Search

The article focuses on addressing the challenge of enhancing recommendation diversity in sequential recommendation (SR) models, particularly for short interaction sequences where limited information hampers accurate modeling of user intents. It mentions the proposed TRIER model tackles this by retrospectively predicting historical interactions to augment sequences and prospectively capturing potential intents.

Conversational interfaces are increasingly popular as a way of connecting people to information. With the increased generative capacity of corpus‐based conversational agents comes the need to classify and filter out malevolent responses that are inappropriate in terms of content and dialogue acts. Previous studies on the topic of detecting and classifying inappropriate content are mostly focused on a specific category of malevolence or on single sentences instead of an entire dialogue. We make three contributions to advance research on the malevolent dialogue response detection and classification (MDRDC) task. First, we define the task and present a hierarchical malevolent dialogue taxonomy. Second, we create a labeled multiturn dialogue data set and formulate the MDRDC task as a hierarchical classification task. Last, we apply state‐of‐the‐art text classification methods to the MDRDC task, and report on experiments aimed at assessing the performance of these approaches. [ABSTRACT FROM AUTHOR]

Most previous work on outfit recommendation focuses on designing visual features to enhance recommendations. Existing work neglects user comments of fashion items, which have been proven to be effective in generating explanations along with better recommendation results. We propose a novel neural network framework, neural outfit recommendation (NOR), that simultaneously provides outfit recommendations and generates abstractive comments. Neural outfit recommendation (NOR) consists of two parts: outfit matching and comment generation. For outfit matching, we propose a convolutional neural network with a mutual attention mechanism to extract visual features. The visual features are then decoded into a rating score for the matching prediction. For abstractive comment generation, we propose a gated recurrent neural network with a cross-modality attention mechanism to transform visual features into a concise sentence. The two parts are jointly trained based on a multi-task learning framework in an end-to-end back-propagation paradigm. Extensive experiments conducted on an existing dataset and a collected real-world dataset show NOR achieves significant improvements over state-of-the-art baselines for outfit recommendation. Meanwhile, our generated comments achieve impressive ROUGE and BLEU scores in comparison to human-written comments. The generated comments can be regarded as explanations for the recommendation results. We release the dataset and code to facilitate future research. [ABSTRACT FROM AUTHOR]

Most Web search diversity approaches can be categorized as Document Level Diversification ( DocLD ), Topic Level Diversification ( TopicLD ) or Term Level Diversification ( TermLD ). DocLD selects the relevant documents with minimal content overlap to each other. It does not take the coverage of query subtopics into account. TopicLD solves this by modeling query subtopics explicitly. However, the automatic mining of query subtopics is difficult. TermLD tries to cover as many query topic terms as possible, which reduces the task of finding a query's subtopics into finding a set of representative topic terms. In this paper, we propose a novel User Session Level Diversification ( UserLD ) approach based on the observation that a query's subtopics are implicitly reflected by the search intents in different user sessions. Our approach consists of two phases: (I) Session Graph Construction and (II) Diversity Reranking . For a given query, phase (I) builds a Session Graph which considers relevant user sessions and preliminary retrieval results as nodes and the nodes' pairwise similarities as edge weights. Phase (II) reranks the preliminary retrieval results by minimizing a Session Graph based diversity loss function. Extensive experiments on two standard datasets of NACSIS Test Collections for IR (NTCIR) demonstrate the effectiveness of our approach. The advantage of our approach lies in its ability of avoiding mining the query subtopics in advance while achieving almost the same or better performances compared with previous approaches. [ABSTRACT FROM AUTHOR]

Query classification is an important part of exploring the characteristics of web queries. Existing studies are mainly based on Broder's classification scheme and classify user queries into navigational, informational, and transactional categories according to users' information needs. In this article, we present a novel classification scheme from the perspective of queries' temporal patterns. Queries' temporal patterns are inherent time series patterns of the search volumes of queries that reflect the evolution of the popularity of a query over time. By analyzing the temporal patterns of queries, search engines can more deeply understand the users' search intents and thus improve performance. Furthermore, we extract three groups of features based on the queries' search volume time series and use a support vector machine ( SVM) to automatically detect the temporal patterns of user queries. Extensive experiments on the Million Query Track data sets of the Text REtrieval Conference ( TREC) demonstrate the effectiveness of our approach. [ABSTRACT FROM AUTHOR]

How to understand intents behind user queries is crucial towards improving the performance of Web search systems. NTCIR-11 IMine task focuses on this problem. In this paper, we address the NTCIR-11 IMine task with two phases referred to as Query Intent Mining ( QIM) and Query Intent Ranking ( QIR). (I) QIM is intended to mine users' potential intents by clustering short text fragments related to the given query. (II) QIR focuses on ranking those mined intents in a proper way. Two challenges exist in handling these tasks. (II) How to precisely estimate the intent similarity between user queries which only consist of a few words. (2) How to properly rank intents in terms of multiple factors, e.g. relevance, diversity, intent drift and so on. For the first challenge, we first investigate two interesting phenomena by analyzing query logs and document datasets, namely ' Same-Intent-Co-Click' ( SICC) and ' Same-Intent-Similar-Rank' ( SISR). SICC means that when users issue different queries, these queries represent the same intent if they click on the same URL. SISR means that if two queries denote the same intent, we should get similar search results when issuing them to a search engine. Then, we propose similarity functions for QIM based on the two phenomena. For the second challenge, we propose a novel intent ranking model which considers multiple factors as a whole. We perform extensive experiments and an interesting case study on the Chinese dataset of NTCIR-11 IMine task. Experimental results demonstrate the effectiveness of our proposed approaches in terms of both QIM and QIR. [ABSTRACT FROM AUTHOR]

This study uses precipitation oxygen isotopes (δ18Op) to examine key dynamics that deliver moisture to the southern slope of central Himalayas over different seasons. Results show that the majority of pre‐monsoon δ18Op values are relatively high and controlled by the westerlies and local moisture. However, some abnormally low δ18Op values coincide with higher precipitation amounts during the pre‐monsoon season due to moisture driven northwards from the Bay of Bengal and Arabian Sea to central Himalayas by anomalous circulations (quasi‐anticyclone, anticyclone, or/and westerlies trough). The size and location of the quasi‐anticyclone also influences the magnitude of the δ18Op decrease. In comparison, the monsoon δ18Op values are lower due to the combined effects of the Indian summer monsoon and convection. Our findings indicate that researchers need to consider the signals of abnormally low δ18Op values during the pre‐monsoon season when attempting to interpret ice core and tree‐ring records from central Himalayas. Plain Language Summary: How moisture is transported to the southern slope of central Himalayas remains unclear, especially for the frequent heavy precipitation events that occur during the pre‐monsoon season. Here, we address this issue using δ18Op measurements from the Asang station on the southern slope of central Himalayas during 2018–2019. We find that some abnormally low δ18Op values coincide with heavy precipitation during the pre‐monsoon season. These abnormally low δ18Op values are caused by the development of anomalous circulations that drives the Indian Ocean moisture to the Asang station. During the monsoon season, the δ18Op values are much lower than other seasons. Such low values are the product of the combined effects of the Indian summer monsoon and convection. We propose that the abnormally low δ18Op values during the pre‐monsoon season need to be considered in paleoclimate reconstructions using ice core and tree‐ring records in the region. The abnormally low δ18Op values during the pre‐monsoon season are closely correlated to anomalous circulations. This finding implies that δ18Op records from ice core and tree ring archives may have potential to reconstruct the frequency and intensity of such anomalous circulations during the pre‐monsoon season. Key Points: Abnormally low δ18Op values during the pre‐monsoon season coincide with heavy precipitation eventsOccurrences of anomalous circulations lead to the abnormally low δ18Op values during the pre‐monsoon seasonCombined effects of the Indian summer monsoon and convection cause lower δ18Op values during the monsoon season [ABSTRACT FROM AUTHOR]

Factorization machines are a generic supervised method for a wide range of tasks in the field of artificial intelligence, such as prediction, inference, etc., which can effectively model feature interactions. However, handling combinations of features is expensive due to the exponential growth of feature interactions with the order. In nature, not all feature interactions are equally useful for prediction. Recently, a large number of methods that perform feature interaction selection have attracted great attention because of their effectiveness at filtering out useless feature interactions. Current feature interaction selection methods suffered from the following limitations: (1) they assume that all users share the same feature interactions; and (2) they select pairwise feature interactions only. In this paper, we propose novel Bayesian variable selection methods, targeting feature interaction selection for factorization machines, which effectively reduce the number of interactions. We study personalized feature interaction selection to account for individual preferences, and further extend the model to investigate higher-order feature interaction selection on higher-order factorization machines. We provide empirical evidence for the advantages of the proposed Bayesian feature interaction selection methods using different prediction tasks. [ABSTRACT FROM AUTHOR]

Many parallel bridges have been recently designed and built around the world to accommodate the ever-increasing volumes of vehicle and rail traffic. Accordingly, the aerodynamic interference around the parallel girders becomes important to investigate for a more accurate estimation of wind-induced effects on bridges. In this study, a wide range of gap-width ratios (0–15) for parallel box girders are selected to investigate the aerodynamic interference using three-dimensional large eddy simulation at zero wind attack angle. Specific aerodynamic characteristics, including the flow structures, pressure distributions, mean values and standard deviations of the force coefficients, Strouhal number as well as spanwise correlations of the aerodynamic coefficients are examined with various gap-width ratios. The detailed comparison work demonstrates a good agreement with the experimental work. The simulation results reveal that all the aerodynamic characteristics are significantly influenced by the gap-width ratio. The three regions are bounded by critical gap-width ratios of G/B = 0.25 and G/B = 5. There are large changes in behavior at G/B = 0.25, whereas, for G/B > 5, interference effects are not evident. The interference effects on both mean and fluctuating aerodynamic coefficients of parallel box girders are also quantified and summarized. • A wide range of gap-width ratios(0–15) for parallel box girders is selected to investigate the aerodynamic interference using three-dimensional large eddy simulation. • The three regions are bounded by critical gap-width ratios of G/B = 0.25 and G/B = 5. There are large changes in behavior at G/B = 0.25, whereas for G/B > 5, interference effects are not evident.. • The aerodynamic interference factors for root mean square of fluctuating force coefficients are proposed to express the impact of aerodynamic interference effect. • The spanwise correlations of force coefficients suddenly increase at G/B = 0.25. [ABSTRACT FROM AUTHOR]

Ambient temperature changes may cause considerable and complicated global deformation in long-span bridges. This study investigates the mechanisms of temperature-induced mid-span vertical displacement of symmetrical twin-pylon cable-stayed bridges through geometric models. General formulas are presented to separately quantify the sensitivity of the concerned displacement with respect to the cable, girder, and pylon temperature variations. The applicability of the proposed formulas is verified using numerical models of 15 bridges. The subsequent parametric analysis reveals the relationships among the temperature sensitivity, material property (linear expansion coefficients), and structural geometries (main span L 0 , side-to-main span ratio γ 0 , pylon height-to-span ratio ξ , and pylon height ratio λ 0). Furthermore, the relative impacts of the temperature changes of different structural components are compared. This study focuses on the calculation of thermal deformation of a general category of cable-stayed bridges rather than a specific one. The formulas provided show the advantages of conceptual clarity, calculation simplicity, and general applicability. They also have far-reaching implications on the design of structural health monitoring systems and the establishment of the load–response baseline models. [ABSTRACT FROM AUTHOR]