Helpful reviews are essential for e-commerce and review websites, as they can help customers make quick purchase decisions and merchants to increase profits. Due to a great number of online reviews with unknown helpfulness, it recently leads to promising research on building automatic mechanisms to assess review helpfulness. The mainstream methods generally extract various linguistic and embedding features solely from the text of a review as the evidence for helpfulness prediction. We, however, consider that the helpfulness of a review should be fully aware of the metadata (such as the title, the brand, the category, and the description) of its target product, besides the textual content of the review itself. Hence, in this paper we propose an end-to-end deep neural architecture directly fed by both the metadata of a product and the raw text of its reviews to acquire product-aware review representations for helpfulness prediction. The learned representations do not require tedious labor on feature engineering and are expected to be more informative as the target-aware evidence to assess the helpfulness of online reviews. We also construct two large-scale datasets which are a portion of the real-world web data in Amazon and Yelp, respectively, to train and test our approach. Experiments are conducted on two different tasks: helpfulness identification and regression of online reviews, and results demonstrate that our approach can achieve state-of-the-art performance with substantial improvements.
{"title":"Product-Aware Helpfulness Prediction of Online Reviews","authors":"M. Fan, Chao Feng, Lin Guo, Mingming Sun, Ping Li","doi":"10.1145/3308558.3313523","DOIUrl":"https://doi.org/10.1145/3308558.3313523","url":null,"abstract":"Helpful reviews are essential for e-commerce and review websites, as they can help customers make quick purchase decisions and merchants to increase profits. Due to a great number of online reviews with unknown helpfulness, it recently leads to promising research on building automatic mechanisms to assess review helpfulness. The mainstream methods generally extract various linguistic and embedding features solely from the text of a review as the evidence for helpfulness prediction. We, however, consider that the helpfulness of a review should be fully aware of the metadata (such as the title, the brand, the category, and the description) of its target product, besides the textual content of the review itself. Hence, in this paper we propose an end-to-end deep neural architecture directly fed by both the metadata of a product and the raw text of its reviews to acquire product-aware review representations for helpfulness prediction. The learned representations do not require tedious labor on feature engineering and are expected to be more informative as the target-aware evidence to assess the helpfulness of online reviews. We also construct two large-scale datasets which are a portion of the real-world web data in Amazon and Yelp, respectively, to train and test our approach. Experiments are conducted on two different tasks: helpfulness identification and regression of online reviews, and results demonstrate that our approach can achieve state-of-the-art performance with substantial improvements.","PeriodicalId":23013,"journal":{"name":"The World Wide Web Conference","volume":"357 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2019-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76328267","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hongyu Lu, Min Zhang, Weizhi Ma, Yunqiu Shao, Yiqun Liu, Shaoping Ma
User behaviors are widely used as implicit feedbacks of user preferences in personalized information systems. In previous works and online applications, the user's click signals are used as positive feedback for ranking, recommendation, evaluation, etc. However, when users click on a piece of low-quality news, they are more likely to have negative experiences and different reading behaviors. Hence, the ignorance of the quality effects of news may lead to the misinterpretation of user behaviors as well as consequence studies. To address these issues, we conducted an in-depth user study in mobile news streaming scenario to investigate whether and how the quality of news may affect user preferences and user behaviors. Firstly, we verify that quality does affect user preferences, and low-quality news results in a lower preference. We further find that this effect varies with both interaction phases and user's interest in the topic of the news. Secondly, we inspect how users interact with low-quality news. Surprisingly, we find that users are more likely to click on low-quality news because of its high title persuasion. Moreover, users will read less and slower with fewer revisits and examinations while reading the low-quality news. Based on these quality effects we have discovered, we propose the Preference Behavior Quality (PBQ) probability model which incorporates the quality into traditional behavior-only implicit feedback. The significant improvement demonstrates that incorporating quality can help build implicit feedback. Since the importance and difficulty in collecting news quality, we further investigate how to identify it automatically. Based on point-wise and pair-wise distinguishing experiments, we show that user behaviors, especially reading ratio and dwell time, have high ability to identify news quality. Our research has comprehensively analyzed the effects of quality on user preferences and behaviors, and raised the awareness of item quality in interpreting user behaviors and estimating user preferences.
{"title":"Quality Effects on User Preferences and Behaviorsin Mobile News Streaming","authors":"Hongyu Lu, Min Zhang, Weizhi Ma, Yunqiu Shao, Yiqun Liu, Shaoping Ma","doi":"10.1145/3308558.3313751","DOIUrl":"https://doi.org/10.1145/3308558.3313751","url":null,"abstract":"User behaviors are widely used as implicit feedbacks of user preferences in personalized information systems. In previous works and online applications, the user's click signals are used as positive feedback for ranking, recommendation, evaluation, etc. However, when users click on a piece of low-quality news, they are more likely to have negative experiences and different reading behaviors. Hence, the ignorance of the quality effects of news may lead to the misinterpretation of user behaviors as well as consequence studies. To address these issues, we conducted an in-depth user study in mobile news streaming scenario to investigate whether and how the quality of news may affect user preferences and user behaviors. Firstly, we verify that quality does affect user preferences, and low-quality news results in a lower preference. We further find that this effect varies with both interaction phases and user's interest in the topic of the news. Secondly, we inspect how users interact with low-quality news. Surprisingly, we find that users are more likely to click on low-quality news because of its high title persuasion. Moreover, users will read less and slower with fewer revisits and examinations while reading the low-quality news. Based on these quality effects we have discovered, we propose the Preference Behavior Quality (PBQ) probability model which incorporates the quality into traditional behavior-only implicit feedback. The significant improvement demonstrates that incorporating quality can help build implicit feedback. Since the importance and difficulty in collecting news quality, we further investigate how to identify it automatically. Based on point-wise and pair-wise distinguishing experiments, we show that user behaviors, especially reading ratio and dwell time, have high ability to identify news quality. Our research has comprehensively analyzed the effects of quality on user preferences and behaviors, and raised the awareness of item quality in interpreting user behaviors and estimating user preferences.","PeriodicalId":23013,"journal":{"name":"The World Wide Web Conference","volume":"60 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2019-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73341100","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The sequential recommendation, which models sequential behavioral patterns among users for the recommendation, plays a critical role in recommender systems. However, the state-of-the-art Recurrent Neural Networks (RNN) solutions rarely consider the non-linear feature interactions and non-monotone short-term sequential patterns, which are essential for user behavior modeling in sparse sequence data. In this paper, we propose a novel Recurrent Convolutional Neural Network model (RCNN). It not only utilizes the recurrent architecture of RNN to capture complex long-term dependencies, but also leverages the convolutional operation of Convolutional Neural Network (CNN) model to extract short-term sequential patterns among recurrent hidden states. Specifically, we first generate a hidden state at each time step with the recurrent layer. Then the recent hidden states are regarded as an “image”, and RCNN searches non-linear feature interactions and non-monotone local patterns via intra-step horizontal and inter-step vertical convolutional filters, respectively. Moreover, the output of convolutional filters and the hidden state are concatenated and fed into a fully-connected layer to generate the recommendation. Finally, we evaluate the proposed model using four real-world datasets from various application scenarios. The experimental results show that our model RCNN significantly outperforms the state-of-the-art approaches on sequential recommendation.
{"title":"Recurrent Convolutional Neural Network for Sequential Recommendation","authors":"Chengfeng Xu, Pengpeng Zhao, Yanchi Liu, Jiajie Xu, V. Sheng, Zhiming Cui, Xiaofang Zhou, Hui Xiong","doi":"10.1145/3308558.3313408","DOIUrl":"https://doi.org/10.1145/3308558.3313408","url":null,"abstract":"The sequential recommendation, which models sequential behavioral patterns among users for the recommendation, plays a critical role in recommender systems. However, the state-of-the-art Recurrent Neural Networks (RNN) solutions rarely consider the non-linear feature interactions and non-monotone short-term sequential patterns, which are essential for user behavior modeling in sparse sequence data. In this paper, we propose a novel Recurrent Convolutional Neural Network model (RCNN). It not only utilizes the recurrent architecture of RNN to capture complex long-term dependencies, but also leverages the convolutional operation of Convolutional Neural Network (CNN) model to extract short-term sequential patterns among recurrent hidden states. Specifically, we first generate a hidden state at each time step with the recurrent layer. Then the recent hidden states are regarded as an “image”, and RCNN searches non-linear feature interactions and non-monotone local patterns via intra-step horizontal and inter-step vertical convolutional filters, respectively. Moreover, the output of convolutional filters and the hidden state are concatenated and fed into a fully-connected layer to generate the recommendation. Finally, we evaluate the proposed model using four real-world datasets from various application scenarios. The experimental results show that our model RCNN significantly outperforms the state-of-the-art approaches on sequential recommendation.","PeriodicalId":23013,"journal":{"name":"The World Wide Web Conference","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2019-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73921176","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jyun-Yu Jiang, Mingyang Zhang, Cheng Li, Michael Bendersky, Nadav Golbandi, Marc Najork
Semantic text matching is one of the most important research problems in many domains, including, but not limited to, information retrieval, question answering, and recommendation. Among the different types of semantic text matching, long-document-to-long-document text matching has many applications, but has rarely been studied. Most existing approaches for semantic text matching have limited success in this setting, due to their inability to capture and distill the main ideas and topics from long-form text. In this paper, we propose a novel Siamese multi-depth attention-based hierarchical recurrent neural network (SMASH RNN) that learns the long-form semantics, and enables long-form document based semantic text matching. In addition to word information, SMASH RNN is using the document structure to improve the representation of long-form documents. Specifically, SMASH RNN synthesizes information from different document structure levels, including paragraphs, sentences, and words. An attention-based hierarchical RNN derives a representation for each document structure level. Then, the representations learned from the different levels are aggregated to learn a more comprehensive semantic representation of the entire document. For semantic text matching, a Siamese structure couples the representations of a pair of documents, and infers a probabilistic score as their similarity. We conduct an extensive empirical evaluation of SMASH RNN with three practical applications, including email attachment suggestion, related article recommendation, and citation recommendation. Experimental results on public data sets demonstrate that SMASH RNN significantly outperforms competitive baseline methods across various classification and ranking scenarios in the context of semantic matching of long-form documents.
{"title":"Semantic Text Matching for Long-Form Documents","authors":"Jyun-Yu Jiang, Mingyang Zhang, Cheng Li, Michael Bendersky, Nadav Golbandi, Marc Najork","doi":"10.1145/3308558.3313707","DOIUrl":"https://doi.org/10.1145/3308558.3313707","url":null,"abstract":"Semantic text matching is one of the most important research problems in many domains, including, but not limited to, information retrieval, question answering, and recommendation. Among the different types of semantic text matching, long-document-to-long-document text matching has many applications, but has rarely been studied. Most existing approaches for semantic text matching have limited success in this setting, due to their inability to capture and distill the main ideas and topics from long-form text. In this paper, we propose a novel Siamese multi-depth attention-based hierarchical recurrent neural network (SMASH RNN) that learns the long-form semantics, and enables long-form document based semantic text matching. In addition to word information, SMASH RNN is using the document structure to improve the representation of long-form documents. Specifically, SMASH RNN synthesizes information from different document structure levels, including paragraphs, sentences, and words. An attention-based hierarchical RNN derives a representation for each document structure level. Then, the representations learned from the different levels are aggregated to learn a more comprehensive semantic representation of the entire document. For semantic text matching, a Siamese structure couples the representations of a pair of documents, and infers a probabilistic score as their similarity. We conduct an extensive empirical evaluation of SMASH RNN with three practical applications, including email attachment suggestion, related article recommendation, and citation recommendation. Experimental results on public data sets demonstrate that SMASH RNN significantly outperforms competitive baseline methods across various classification and ranking scenarios in the context of semantic matching of long-form documents.","PeriodicalId":23013,"journal":{"name":"The World Wide Web Conference","volume":"43 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2019-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84237312","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Automatically classifying extremely short texts, such as social media posts and web page titles, plays an important role in a wide range of content analysis applications. However, traditional classifiers based on bag-of-words (BoW) representations often fail in this task. The underlying reason is that the document similarity can not be accurately measured under BoW representations due to the extreme sparseness of short texts. This results in significant difficulty to capture the generality of short texts. To address this problem, we use a better regularized word mover's distance (RWMD), which can measure distances among short texts at the semantic level. We then propose a RWMD-based centroid classifier for short texts, named RWMD-CC. Basically, RWMD-CC computes a representative semantic centroid for each category under the RWMD measure, and predicts test documents by finding the closest semantic centroid. The testing is much more efficient than the prior art of K nearest neighbor classifier based on WMD. Experimental results indicate that our RWMD-CC can achieve very competitive classification performance on extremely short texts.
{"title":"Classifying Extremely Short Texts by Exploiting Semantic Centroids in Word Mover's Distance Space","authors":"C. Li, Jihong Ouyang, Ximing Li","doi":"10.1145/3308558.3313397","DOIUrl":"https://doi.org/10.1145/3308558.3313397","url":null,"abstract":"Automatically classifying extremely short texts, such as social media posts and web page titles, plays an important role in a wide range of content analysis applications. However, traditional classifiers based on bag-of-words (BoW) representations often fail in this task. The underlying reason is that the document similarity can not be accurately measured under BoW representations due to the extreme sparseness of short texts. This results in significant difficulty to capture the generality of short texts. To address this problem, we use a better regularized word mover's distance (RWMD), which can measure distances among short texts at the semantic level. We then propose a RWMD-based centroid classifier for short texts, named RWMD-CC. Basically, RWMD-CC computes a representative semantic centroid for each category under the RWMD measure, and predicts test documents by finding the closest semantic centroid. The testing is much more efficient than the prior art of K nearest neighbor classifier based on WMD. Experimental results indicate that our RWMD-CC can achieve very competitive classification performance on extremely short texts.","PeriodicalId":23013,"journal":{"name":"The World Wide Web Conference","volume":"76 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2019-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85514769","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper introduces RecBoard, a unified web-based platform that facilitates researchers and practitioners to train, test, deploy, and monitor recommendation systems. RecBoard streamlines the end-to-end process of building recommendation systems by providing a collaborative user interface that automates repetitive tasks related to dataset management, model training, visualization, deployments, and monitoring. Our demo prototype demonstrates how RecBoard can empower common tasks in research and development. RecBoard will be open-sourced and publicly available upon publication.
{"title":"RecBoard: A Web-based Platform for Recommendation System Research and Development","authors":"M. Chawla, Kriti Singh, Longqi Yang, D. Estrin","doi":"10.1145/3308558.3314133","DOIUrl":"https://doi.org/10.1145/3308558.3314133","url":null,"abstract":"This paper introduces RecBoard, a unified web-based platform that facilitates researchers and practitioners to train, test, deploy, and monitor recommendation systems. RecBoard streamlines the end-to-end process of building recommendation systems by providing a collaborative user interface that automates repetitive tasks related to dataset management, model training, visualization, deployments, and monitoring. Our demo prototype demonstrates how RecBoard can empower common tasks in research and development. RecBoard will be open-sourced and publicly available upon publication.","PeriodicalId":23013,"journal":{"name":"The World Wide Web Conference","volume":"46 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2019-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79752667","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper we present a web-based open source tool and a method for generating insight from any text or discourse using text network analysis. The tool (InfraNodus) can be used by researchers and writers to organize and to better understand their notes, to measure the level of bias in discourse, and to identify the parts of the discourse where there is a potential for insight and new ideas. The method is based on text network analysis algorithm, which represents any text as a network and identifies the most influential words in a discourse based on the terms' co-occurrence. Graph community detection algorithm is then applied in order to identify the different topical clusters, which represent the main topics in the text as well as the relations between them. The community structure is used in conjunction with other measures to identify the level of bias or cognitive diversity of the discourse. Finally, the structural gaps in the graph can indicate the parts of the discourse where the connections are lacking, therefore highlighting the areas where there's a potential for new ideas. The tool can be used as stand-alone software by end users as well as implemented via an API into other tools. Another interesting application is in the field of recommendation systems: structural gaps could indicate potentially interesting non-trivial connections to any connected datasets.
{"title":"InfraNodus: Generating Insight Using Text Network Analysis","authors":"Dmitry Paranyushkin","doi":"10.1145/3308558.3314123","DOIUrl":"https://doi.org/10.1145/3308558.3314123","url":null,"abstract":"In this paper we present a web-based open source tool and a method for generating insight from any text or discourse using text network analysis. The tool (InfraNodus) can be used by researchers and writers to organize and to better understand their notes, to measure the level of bias in discourse, and to identify the parts of the discourse where there is a potential for insight and new ideas. The method is based on text network analysis algorithm, which represents any text as a network and identifies the most influential words in a discourse based on the terms' co-occurrence. Graph community detection algorithm is then applied in order to identify the different topical clusters, which represent the main topics in the text as well as the relations between them. The community structure is used in conjunction with other measures to identify the level of bias or cognitive diversity of the discourse. Finally, the structural gaps in the graph can indicate the parts of the discourse where the connections are lacking, therefore highlighting the areas where there's a potential for new ideas. The tool can be used as stand-alone software by end users as well as implemented via an API into other tools. Another interesting application is in the field of recommendation systems: structural gaps could indicate potentially interesting non-trivial connections to any connected datasets.","PeriodicalId":23013,"journal":{"name":"The World Wide Web Conference","volume":"AES-10 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2019-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84515742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Qiang Zhang, Shangsong Liang, Aldo Lipani, Z. Ren, Emine Yilmaz
Stance detection has gained increasing interest from the research community due to its importance for fake news detection. The goal of stance detection is to categorize an overall position of a subject towards an object into one of the four classes: agree, disagree, discuss, and unrelated. One of the major problems faced by current machine learning models used for stance detection is caused by a severe class imbalance among these classes. Hence, most models fail to correctly classify instances that fall into minority classes. In this paper, we address this problem by proposing a hierarchical representation of these classes, which combines the agree, disagree, and discuss classes under a new related class. Further, we propose a two-layer neural network that learns from this hierarchical representation and controls the error propagation between the two layers using the Maximum Mean Discrepancy regularizer. Compared with conventional four-way classifiers, this model has two advantages: (1) the hierarchical architecture mitigates the class imbalance problem; (2) the regularization makes the model to better discern between the related and unrelated stances. An extensive experimentation demonstrates state-of-the-art accuracy performance of the proposed model for stance detection.
{"title":"From Stances' Imbalance to Their HierarchicalRepresentation and Detection","authors":"Qiang Zhang, Shangsong Liang, Aldo Lipani, Z. Ren, Emine Yilmaz","doi":"10.1145/3308558.3313724","DOIUrl":"https://doi.org/10.1145/3308558.3313724","url":null,"abstract":"Stance detection has gained increasing interest from the research community due to its importance for fake news detection. The goal of stance detection is to categorize an overall position of a subject towards an object into one of the four classes: agree, disagree, discuss, and unrelated. One of the major problems faced by current machine learning models used for stance detection is caused by a severe class imbalance among these classes. Hence, most models fail to correctly classify instances that fall into minority classes. In this paper, we address this problem by proposing a hierarchical representation of these classes, which combines the agree, disagree, and discuss classes under a new related class. Further, we propose a two-layer neural network that learns from this hierarchical representation and controls the error propagation between the two layers using the Maximum Mean Discrepancy regularizer. Compared with conventional four-way classifiers, this model has two advantages: (1) the hierarchical architecture mitigates the class imbalance problem; (2) the regularization makes the model to better discern between the related and unrelated stances. An extensive experimentation demonstrates state-of-the-art accuracy performance of the proposed model for stance detection.","PeriodicalId":23013,"journal":{"name":"The World Wide Web Conference","volume":"70 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2019-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84096458","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In traditional machine learning, classifiers training is typically undertaken in the setting of single-task learning, so the trained classifier can discriminate between different classes. However, this must be based on the assumption that different classes are mutually exclusive. In real applications, the above assumption does not always hold. For example, the same book may belong to multiple subjects. From this point of view, researchers were motivated to formulate multi-label learning problems. In this context, each instance can be assigned multiple labels but the classifiers training is still typically undertaken in the setting of single-task learning. When probabilistic approaches are adopted for classifiers training, multi-task learning can be enabled through transformation of a multi-labelled data set into several binary data sets. The above data transformation could usually result in the class imbalance issue. Without the above data transformation, multi-labelling of data results in an exponential increase of the number of classes, leading to fewer instances for each class and a higher difficulty for identifying each class. In addition, multi-labelling of data is very time consuming and expensive in some application areas, such as hate speech detection. In this paper, we introduce a novel formulation of the hate speech type identification problem in the setting of multi-task learning through our proposed fuzzy ensemble approach. In this setting, single-labelled data can be used for semi-supervised multi-label learning and two new metrics (detection rate and irrelevance rate) are thus proposed to measure more effectively the performance for this kind of learning tasks. We report an experimental study on identification of four types of hate speech, namely: religion, race, disability and sexual orientation. The experimental results show that our proposed fuzzy ensemble approach outperforms other popular probabilistic approaches, with an overall detection rate of 0.93.
{"title":"Fuzzy Multi-task Learning for Hate Speech Type Identification","authors":"Han Liu, P. Burnap, Wafa Alorainy, M. Williams","doi":"10.1145/3308558.3313546","DOIUrl":"https://doi.org/10.1145/3308558.3313546","url":null,"abstract":"In traditional machine learning, classifiers training is typically undertaken in the setting of single-task learning, so the trained classifier can discriminate between different classes. However, this must be based on the assumption that different classes are mutually exclusive. In real applications, the above assumption does not always hold. For example, the same book may belong to multiple subjects. From this point of view, researchers were motivated to formulate multi-label learning problems. In this context, each instance can be assigned multiple labels but the classifiers training is still typically undertaken in the setting of single-task learning. When probabilistic approaches are adopted for classifiers training, multi-task learning can be enabled through transformation of a multi-labelled data set into several binary data sets. The above data transformation could usually result in the class imbalance issue. Without the above data transformation, multi-labelling of data results in an exponential increase of the number of classes, leading to fewer instances for each class and a higher difficulty for identifying each class. In addition, multi-labelling of data is very time consuming and expensive in some application areas, such as hate speech detection. In this paper, we introduce a novel formulation of the hate speech type identification problem in the setting of multi-task learning through our proposed fuzzy ensemble approach. In this setting, single-labelled data can be used for semi-supervised multi-label learning and two new metrics (detection rate and irrelevance rate) are thus proposed to measure more effectively the performance for this kind of learning tasks. We report an experimental study on identification of four types of hate speech, namely: religion, race, disability and sexual orientation. The experimental results show that our proposed fuzzy ensemble approach outperforms other popular probabilistic approaches, with an overall detection rate of 0.93.","PeriodicalId":23013,"journal":{"name":"The World Wide Web Conference","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2019-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80358298","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Huichu Zhang, Siyuan Feng, Chang Liu, Yaoyao Ding, Yichen Zhu, Zihan Zhou, Weinan Zhang, Yong Yu, Haiming Jin, Z. Li
Traffic signal control is an emerging application scenario for reinforcement learning. Besides being as an important problem that affects people's daily life in commuting, traffic signal control poses its unique challenges for reinforcement learning in terms of adapting to dynamic traffic environment and coordinating thousands of agents including vehicles and pedestrians. A key factor in the success of modern reinforcement learning relies on a good simulator to generate a large number of data samples for learning. The most commonly used open-source traffic simulator SUMO is, however, not scalable to large road network and large traffic flow, which hinders the study of reinforcement learning on traffic scenarios. This motivates us to create a new traffic simulator CityFlow with fundamentally optimized data structures and efficient algorithms. CityFlow can support flexible definitions for road network and traffic flow based on synthetic and real-world data. It also provides user-friendly interface for reinforcement learning. Most importantly, CityFlow is more than twenty times faster than SUMO and is capable of supporting city-wide traffic simulation with an interactive render for monitoring. Besides traffic signal control, CityFlow could serve as the base for other transportation studies and can create new possibilities to test machine learning methods in the intelligent transportation domain.
{"title":"CityFlow: A Multi-Agent Reinforcement Learning Environment for Large Scale City Traffic Scenario","authors":"Huichu Zhang, Siyuan Feng, Chang Liu, Yaoyao Ding, Yichen Zhu, Zihan Zhou, Weinan Zhang, Yong Yu, Haiming Jin, Z. Li","doi":"10.1145/3308558.3314139","DOIUrl":"https://doi.org/10.1145/3308558.3314139","url":null,"abstract":"Traffic signal control is an emerging application scenario for reinforcement learning. Besides being as an important problem that affects people's daily life in commuting, traffic signal control poses its unique challenges for reinforcement learning in terms of adapting to dynamic traffic environment and coordinating thousands of agents including vehicles and pedestrians. A key factor in the success of modern reinforcement learning relies on a good simulator to generate a large number of data samples for learning. The most commonly used open-source traffic simulator SUMO is, however, not scalable to large road network and large traffic flow, which hinders the study of reinforcement learning on traffic scenarios. This motivates us to create a new traffic simulator CityFlow with fundamentally optimized data structures and efficient algorithms. CityFlow can support flexible definitions for road network and traffic flow based on synthetic and real-world data. It also provides user-friendly interface for reinforcement learning. Most importantly, CityFlow is more than twenty times faster than SUMO and is capable of supporting city-wide traffic simulation with an interactive render for monitoring. Besides traffic signal control, CityFlow could serve as the base for other transportation studies and can create new possibilities to test machine learning methods in the intelligent transportation domain.","PeriodicalId":23013,"journal":{"name":"The World Wide Web Conference","volume":"41 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2019-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80408086","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：