[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-135654-105":59,"doc-detail-135654-en":130},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":123,"head_meta":125,"extra_data":127,"updated_unix":129},105,"en","topic-modeling-with-wasserstein-autoencoders","Topic Modeling with Wasserstein Autoencoders","","A novel neural topic model is proposed within the Wasserstein autoencoders (WAE) framework. Instead of variational autoencoder methods, the approach directly enforces a Dirichlet prior on latent document-topic vectors and performs distribution matching via a kernel-based Maximum Mean Discrepancy (MMD). Results show MMD better matches high-dimensional Dirichlet distributions than GANs, and adding randomness to encoder outputs improves topic coherence. A topic uniqueness metric complements NPMI for a holistic evaluation, with experiments on multiple real datasets yielding significantly better topics than existing models.",{"@graph":69,"@context":122},[70,84,105],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":30,"@type":76,"position":81},"https://docshare.wps.com/document/technology/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/topic-modeling-with-wasserstein-autoencoders/135654/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/topic-modeling-with-wasserstein-autoencoders/135654.png","ImageObject",300,407,{"name":92,"@type":93},"Margaret","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-08","2026-08-21",true,{"@type":102,"interactionType":103,"userInteractionCount":24},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118],{"name":109,"@type":110,"acceptedAnswer":111},"How does the proposed W-LDA model differ from VAE-based topic models?","Question",{"text":112,"@type":113},"W-LDA directly enforces a Dirichlet prior on latent document-topic vectors through distribution matching, avoiding Gaussian approximations used to handle Dirichlet in VAE settings.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"Why is Maximum Mean Discrepancy (MMD) emphasized in the model?",{"text":117,"@type":113},"MMD is used with an appropriate kernel to match distributions, and the work reports that MMD performs much better than GAN-based matching for high-dimensional Dirichlet distributions.",{"name":119,"@type":110,"acceptedAnswer":120},"How is topic quality evaluated beyond coherence?",{"text":121,"@type":113},"Topic coherence is measured using NPMI, and a new Topic Uniqueness metric is introduced to quantify diversity so topics are not repetitive.","https://schema.org",{"og:url":83,"og:type":124,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":126,"canonical":83},"index,follow",{"doc_id":128,"site_id":62},135654,1790706963,{"code":4,"msg":5,"data":131},{"doc_id":128,"user_id":132,"nickname":92,"user_avatar":133,"doc_module":4,"category_id":29,"category_name":30,"doc_title":65,"doc_description":67,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":24,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":139,"language":140,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":67,"update_tm":144,"read_time":145},137451207643,"https://ap-avatar.wpscdn.com/davatar_3d24733baf745e90a7e4bdd5f77d97b2","Topic Modeling with Wasserstein Autoencoders  \nFeng Nany, Ran Ding 􀀃 z, Ramesh Nallapatiy, Bing Xiangy  \nAmazon Web Servicesy , Compass Inc.z  \nfnanfen, rnallapa, bxiangg@amazon .comy , ran .ding@compass .comz  \nAbstract  \nWe propose a novel neural topic model in the Wasserstein autoencoders (WAE) framework.  \nUnlike existing variational autoencoder based models, we directly enforce Dirichlet prior on the latent document-topic vectors. We exploit the structure of the latent space and apply a suitable kernel in minimizing the Maximum Mean Discrepancy (MMD) to perform distribution matching. We discover that MMD performs much better than the Generative Adversarial Network (GAN) in matching high dimensional Dirichlet distribution. We further discover that incorporating randomness in the encoder output during training leads to significantly more coherent topics. To measure the diversity of the produced topics, we propose a simple topic uniqueness metric. Together with the widely used coherence measure NPMI, we offer a more wholistic evaluation of topic quality. Experiments on several real datasets show that our model produces significantly better topics than existing topic models.  \n1 Introduction  \nProbabilistic topic models (Hoffman et al., 2010) have been widely used to explore large collections of documents in an unsupervised manner. They can discover the underlying themes and organize the documents accordingly. The most popular probabilistic topic model is the Latent Dirichlet Allocation (LDA) (Blei et al., 2003), where the authors developed a variational Bayesian (VB) algorithm to perform approximate inference; subsequently (Griffiths and Steyvers, 2004) proposed an alternative inference method using collapsed Gibbs sampling.  \nMore recently, deep neural networks have been successfully used for such probabilistic models with the emergence of variational autoencoders  \n􀀃 This work was done when the author was with Amazon.  \n(VAE) (Kingma and Welling, 2013) . The key advantage of such neural network based models is that inference can be carried out easily via a forward pass of the recognition network, without the need for expensive iterative inference scheme per example as in VB and collapsed Gibbs sampling. Topic models that fall in this framework include NVDM (Miao et al., 2016), ProdLDA (Srivastava and Sutton, 2017) and NTM-R (Ding et al., 2018) . At a high level, these models consist of an encoder network that maps the Bag-of-Words (BoW) input to a latent document-topic vector and a decoder network that maps the document-topic vector to a discrete distribution over the words in the vocabulary. They are autoencoders in the sense that the output of the decoder aims to reconstruct the word distribution of the input BoW representation. Besides the reconstruction loss, VAE-based methods also minimize a KL-divergence term between the prior and posterior of the latent vector distributions. Despite their popularity, these VAEbased topic models suffer from several conceptual and practical challenges. First, the Auto-Encoding Variational Bayes (Kingma and Welling, 2013) framework of VAE relies on a reparameterization trick that only works with the “location-scale”family of distributions. Unfortunately, the Dirichlet distribution, which largely accounted for the modeling success of LDA, does not belong to this family. The Dirichlet prior on the latent documenttopic vector nicely captures the intuition that a document typically belongs to a sparse subset of topics. The VAE-based topic models have to resort to various Gaussian approximations toward this effect. For example, NVDM and NTM-R simply use Gaussian instead of Dirichlet prior; ProdLDA uses Laplace approximation of the Dirichlet distribution in the softmax basis as prior. Second, the KL divergence term in the VAE objective forces posterior distributions for all examples to match  \n6345  \nProceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages","cbCaiq04TASdyOkF","https://ap.wps.com/l/cbCaiq04TASdyOkF","pdf",1015248,37,"English","# Abstract\n# 1 Introduction","[{\"question\":\"How does the proposed W-LDA model differ from VAE-based topic models?\",\"answer\":\"W-LDA directly enforces a Dirichlet prior on latent document-topic vectors through distribution matching, avoiding Gaussian approximations used to handle Dirichlet in VAE settings.\"},{\"question\":\"Why is Maximum Mean Discrepancy (MMD) emphasized in the model?\",\"answer\":\"MMD is used with an appropriate kernel to match distributions, and the work reports that MMD performs much better than GAN-based matching for high-dimensional Dirichlet distributions.\"},{\"question\":\"How is topic quality evaluated beyond coherence?\",\"answer\":\"Topic coherence is measured using NPMI, and a new Topic Uniqueness metric is introduced to quantify diversity so topics are not repetitive.\"}]","Topic Modeling with Wasserstein Autoencoders | PDF",1787316728,93]