Semantic Sparsity Research Articles

An Appearance-Semantic Descriptor with Coarse-to-Fine Matching for Robust VPR.

In recent years, semantic segmentation has made significant progress in visual place recognition (VPR) by using semantic information that is relatively invariant to appearance and viewpoint, demonstrating great potential. However, in some extreme scenarios, there may be semantic occlusion and semantic sparsity, which can lead to confusion when relying solely on semantic information for localization. Therefore, this paper proposes a novel VPR framework that employs a coarse-to-fine image matching strategy, combining semantic and appearance information to improve algorithm performance. First, we construct SemLook global descriptors using semantic contours, which can preliminarily screen images to enhance the accuracy and real-time performance of the algorithm. Based on this, we introduce SemLook local descriptors for fine screening, combining robust appearance information extracted by deep learning with semantic information. These local descriptors can address issues such as semantic overlap and sparsity in urban environments, further improving the accuracy of the algorithm. Through this refined screening process, we can effectively handle the challenges of complex image matching in urban environments and obtain more accurate results. The performance of SemLook descriptors is evaluated on three public datasets (Extended-CMU Season, Robot-Car Seasons v2, and SYNTHIA) and compared with six state-of-the-art VPR algorithms (HOG, CoHOG, AlexNet_VPR, Region VLAD, Patch-NetVLAD, Forest). In the experimental comparison, considering both real-time performance and evaluation metrics, the SemLook descriptors are found to outperform the other six algorithms. Evaluation metrics include the area under the curve (AUC) based on the precision-recall curve, Recall@100%Precision, and Precision@100%Recall. On the Extended-CMU Season dataset, SemLook descriptors achieve a 100% AUC value, and on the SYNTHIA dataset, they achieve a 99% AUC value, demonstrating outstanding performance. The experimental results indicate that introducing global descriptors for initial screening and utilizing local descriptors combining both semantic and appearance information for precise matching can effectively address the issue of location recognition in scenarios with semantic ambiguity or sparsity. This algorithm enhances descriptor performance, making it more accurate and robust in scenes with variations in appearance and viewpoint.

Improved Graph Contrastive Learning for Short Text Classification

Text classification occupies an important role in natural language processing and has many applications in real life. Short text classification, as one of its subtopics, has attracted increasing interest from researchers since it is more challenging due to its semantic sparsity and insufficient labeled data. Recent studies attempt to combine graph learning and contrastive learning to alleviate the above problems in short text classification. Despite their fruitful success, there are still several inherent limitations. First, the generation of augmented views may disrupt the semantic structure within the text and introduce negative effects due to noise permutation. Second, they ignore the clustering-friendly features in unlabeled data and fail to further utilize the prior information in few valuable labeled data. To this end, we propose a novel model that utilizes improved Graph contrastIve learning for short text classiFicaTion (GIFT). Specifically, we construct a heterogeneous graph containing several component graphs by mining from an internal corpus and introducing an external knowledge graph. Then, we use singular value decomposition to generate augmented views for graph contrastive learning. Moreover, we employ constrained kmeans on labeled texts to learn clustering-friendly features, which facilitate cluster-oriented contrastive learning and assist in obtaining better category boundaries. Extensive experimental results show that GIFT significantly outperforms previous state-of-the-art methods. Our code can be found in https://github.com/KEAML-JLU/GIFT.

A multi‐label social short text classification method based on contrastive learning and improved ml‐KNN

AbstractShort texts on social platforms often have the problems of diverse categories and semantic sparsity, making it challenging to identify the diverse intentions of users. To address this issue, this article proposes a multi‐label social short text classification method (IML‐CL) based on contrastive learning and improved ml‐KNN. First, a contrastive learning approach is employed to train a multi‐label text classification model. This approach improves semantic sparsity by leveraging the knowledge from the existing samples to enrich the feature representation of short texts. Simultaneously, an improved ml‐KNN algorithm is developed to enhance the accuracy of label prediction. This algorithm utilizes a two‐layer nearest neighbor rule and introduces a penalty function and weight optimization. Next, the model generates the feature representation for the test sample and predicts its label. Additionally, the improved ml‐KNN algorithm retrieves neighbors of the test sample and uses their label information for prediction. Finally, the two predictions are combined to obtain the final prediction, which accurately identifies the user's intention. The experimental results demonstrate that, on the dataset constructed in this article, the IML‐CL method effectively boosts the performance of the baseline model.

A survey of topic models: From a whole-cycle perspective

With the rapid development of information science and social networks, the Internet has accumulated various data containing valuable information and topics. The topic model has become one of the primary semantic modeling and classification methods. It has been widely studied in academia and industry. However, most topic models only focus on long texts and often suffer from semantic sparsity problems. The sparse, short text content and irregular data have brought major challenges to the application of topic models in semantic modeling and topic discovery. To overcome these challenges, researchers have explored topic models and achieved excellent results. However, most of the current topic models are applicable to a specific model task. The majority of current reviews ignore the whole-cycle perspective and framework. It brings great challenges for novices to learn topic models. To deal with the above challenges, we investigate more than a hundred papers on topic models and summarize the research progress on the entire topic model process, including theory, method, datasets, and evaluation indicator. In addition, we also analyzed the statistical data results of the topic model through experiments and introduced its applications in different fields. The paper provides a whole-cycle learning path for novices. It encourages researchers to give more attention to the topic model algorithm and the theory itself without paying extra attention to understanding the relevant datasets, evaluation methods and latest progress.

News topic detection based on capsule semantic graph

Most news topic detection methods use word-based methods, which easily ignore the relationship among words and have semantic sparsity, resulting in low topic detection accuracy. In addition, the current mainstream probability methods and graph analysis methods for topic detection have high time complexity. For these reasons, we present a news topic detection model on the basis of capsule semantic graph (CSG). The keywords that appear in each text at the same time are modeled as a keyword graph, which is divided into multiple subgraphs through community detection. Each subgraph contains a group of closely related keywords. The graph is used as the vertex of CSG. The semantic relationship among the vertices is obtained by calculating the similarity of the average word vector of each vertex. At the same time, the news text is clustered using the incremental clustering method, where each text uses CSG; that is, the similarity among texts is calculated by the graph kernel. The relationship between vertices and edges is also considered when calculating the similarity. Experimental results on three standard datasets show that CSG can obtain higher precision, recall, and F1 values than several latest methods. Experimental results on large-scale news datasets reveal that the time complexity of CSG is lower than that of probabilistic methods and other graph analysis methods.

Research on user granularity-level personalized social text generation technology

With the introduction of large-scale pre-trained language models, breakthroughs have been made in text generation technology research. On this basis, in order to assist users to complete personalized creation, this paper proposes a user-level fine-grained control generation model. First, we design the Encoder-Decoder framework based on the GPT2 structure, and model and encode the user’s static personalized information on the Encoder side. Then add a bidirectional independent attention module to receive the personalized feature vector on the Decoder side. The attention module in the original GPT2 structure captures the dynamic personalized features in the user text, namely writing style, expression way, etc. Next, the scores of each attention module are weighted and fused to participate in the subsequent decoding to automatically generate social texts constrained by the user’s personalized feature attributes. However, the semantic sparsity of the user’s basic information will cause occasional conflicts between the generated text and some personalized features. Therefore, we use the Alignment module to perform the secondary enhancement and generation of consistent understanding between the output data of the Decoder and the user’s personalized features, and finally realize the personalized social text generation. Experiments show that compared with the GPT2 baseline model, the fluency of the model is improved by 0.3%-0.6%, and on the basis of no loss of language fluency, the social text generated by the model can have significant user personalization characteristics, among which personalization and consistency the two evaluation indicators of sexuality both increased significantly by 8.4% and 9%.

A spatial-temporal topic model with sparse prior and RNN prior for bursty topic discovering in social networks

With the rapid growth of social network users, the social network has accumulated massive social network topics. However, due to the randomness of content, it becomes sparse and noisy, accompanied by many daily chats and meaningless topics, which brings challenges to bursty topics discovery. To deal with these problems, this paper proposes the spatial-temporal topic model with sparse prior and recurrent neural networks (RNN) prior for bursty topic discovering (ST-SRTM). The semantic relationship of words is learned through RNN to alleviate the sparsity. The spatial-temporal areas information is introduced to focus on bursty topics for further weakening the semantic sparsity of social network context. Besides, we introduced the “Spike and Slab” prior to decouple the sparseness and smoothness. Simultaneously, we realized the automatic discovery of social network bursts by introducing the burstiness of words as the prior and binary switching variables. We constructed multiple sets of comparative experiments to verify the performance of ST-SRTM by leveraging different evaluation indicators on real Sina Weibo data sets. The experimental results confirm the superiority of our ST-SRTM.

MVGAN: Multi-View Graph Attention Network for Social Event Detection

Social networks are critical sources for event detection thanks to the characteristics of publicity and dissemination. Unfortunately, the randomness and semantic sparsity of the social network text bring significant challenges to the event detection task. In addition to text, time is another vital element in reflecting events since events are often followed for a while. Therefore, in this article, we propose a novel method named Multi-View Graph Attention Network (MVGAN) for event detection in social networks. It enriches event semantics through both neighbor aggregation and multi-view fusion in a heterogeneous social event graph. Specifically, we first construct a heterogeneous graph by adding the hashtag to associate the isolated short texts and describe events comprehensively. Then, we learn view-specific representations of events through graph convolutional networks from the perspectives of text semantics and time distribution, respectively. Finally, we design a hashtag-based multi-view graph attention mechanism to capture the intrinsic interaction across different views and integrate the feature representations to discover events. Extensive experiments on public benchmark datasets demonstrate that MVGAN performs favorably against many state-of-the-art social network event detection algorithms. It also proves that more meaningful signals can contribute to improving the event detection effect in social networks, such as published time and hashtags.

HGAT: Heterogeneous Graph Attention Networks for Semi-supervised Short Text Classification

Short text classification has been widely explored in news tagging to provide more efficient search strategies and more effective search results for information retrieval. However, most existing studies, concentrating on long text classification, deliver unsatisfactory performance on short texts due to the sparsity issue and the insufficiency of labeled data. In this article, we propose a novel heterogeneous graph neural network-based method for semi-supervised short text classification, leveraging full advantage of limited labeled data and large unlabeled data through information propagation along the graph. Specifically, we first present a flexible heterogeneous information network (HIN) framework for modeling short texts, which can integrate any type of additional information and meanwhile capture their relations to address the semantic sparsity. Then, we propose Heterogeneous Graph Attention networks (HGAT) to embed the HIN for short text classification based on a dual-level attention mechanism, including node-level and type-level attentions. To efficiently classify new coming texts that do not previously exist in the HIN, we extend our model HGAT for inductive learning, avoiding re-training the model on the evolving HIN. Extensive experiments on single-/multi-label classification demonstrates that our proposed model HGAT significantly outperforms state-of-the-art methods across the benchmark datasets under both transductive and inductive learning.

Dual‐Level Attention Based on a Heterogeneous Graph Convolution Network for Aspect‐Based Sentiment Classification

With the development of 5G, the advancement of basic infrastructure has led to considerable development in related research and technology. It also promotes the development of various smart devices and social platforms. More and more people are now using smart devices to post their reviews right after something happens. In order to keep pace with this trend, we propose a method to analyze users’ sentiment by using their text data. When analyzing users’ text data, it is noted that a user’s review may contain many aspects. Traditional text classification methods used by smart devices, however, usually ignore the importance of multiple aspects of a review. Additionally, most algorithms usually ignore the network structure information between the words in a sentence and the sentence itself. To address these issues, we propose a novel dual‐level attention‐based heterogeneous graph convolutional network for aspect‐based sentiment classification which minds more context information through information propagation along with graphs. Particularly, we first propose a flexible HIN (heterogeneous information network) framework to model the user‐generated reviews. This framework can integrate various types of additional information and capture their relationships to alleviate semantic sparsity of some labeled data. This framework can also leverage the full advantage of the hidden network structure information through information propagation along with graphs. Then, we propose a dual‐level attention‐based heterogeneous graph convolutional network (DAHGCN), which includes node‐level and type‐level attentions. The attention mechanisms can analyze the importance of different adjacent nodes and the importance of different types of nodes for the current node. The experimental results on three real‐world datasets demonstrated the effectiveness and reliability of our model.

Leverage Label and Word Embedding for Semantic Sparse Web Service Discovery

Information retrieval-based Web service discovery approach suffers from the semantic sparsity problem caused by lacking of statistical information when the Web services are described in short texts. To handle this problem, external information is often utilized to improve the discovery performance. Inspired by this, we propose a novel Web service discovery approach based on a neural topic model and leveraging Web service labels. More specifically, words in Web services are mapped into continuous embeddings, and labels are integrated by a neural topic model simultaneously for embodying external semantics of the Web service description. Based on the topic model, the services are interpreted into hierarchical models for building a service querying and ranking model. Extensive experiments on several datasets demonstrated that the proposed approach achieves improved performance in terms of F-measure. The results also suggest that leveraging external information is useful for semantic sparse Web service discovery.

Semantic Sparse Service Discovery Using Word Embedding and Gaussian LDA

Nowadays, a growing number of web services are offered in API marketplaces browsed by service developers or third-party registries. Under this situation, API marketplaces' users greatly rely on a search engine to find suitable web services. However, due to the fact that functional attributes of web services are usually described in short texts, the search engine-based discovery approach suffers from the semantic sparsity problem, which hinders the effect of service discovery. To address this issue, we propose a novel web service discovery approach using word embedding and Gaussian latent Dirichlet allocation (Gaussian LDA). Unlike most existing service discovery approaches, our approach first uses context information generated by word embedding to enrich the semantics of service descriptions and users' queries. Then, the enriched service description is loaded into the Gaussian LDA model to acquire service description representation. Finally, the services are ranked by considering the relevance between the extended user's query and service description representation. The experiments conducted on a real-world web service dataset and the results demonstrate that the proposed approach achieves superior effectiveness on web service discovery.

A multi-feature probabilistic graphical model for social network semantic search

With the rapid development of social network platforms, more and more people are using them to search for material related to their interests. As the texts of social media messages are usually so short, when traditional existing document modeling methods are used in social network search tasks, the problem of semantic sparsity arises, leading to low-quality semantic representation and low-precision social network search results. Fortunately, besides of short text, social media data also has other features, such as timestamps, locations, and its user information. In light of this, to realize precise social network search, we propose a multi-feature probabilistic graphical model (MFPGM), which can generate high-quality semantic representation. To deal with the problem of semantic sparsity, we exploit two strategies in MFPGM. First, we propose a concept named special region and utilize location information to aggregate short text into long text. Second, we introduce the biterm pattern that can generate dense semantic space by supposing that a biterm occurring in the same context has the same topic. In order to generate high-quality semantic representations, we simultaneously model multiple features (i.e., biterm, user, location and timestamp) of social network data to enhance the semantic learning process of MFPGM. We conduct a lot of experiments on real-word datasets, and the comparisons with several state-of-art baseline methods have demonstrated the superiority of our MFPGM on topic quality and search performance. Additionally, with the help of the generated semantic representations, MFPGM allows people to analyze the relationships between time and the popularities of topics.

Extended search method based on a semantic hashtag graph combining social and conceptual information

Searching for microblog short text by their meaning is a challenging task because of the semantic sparsity of the information in social networks. The extended search approaches are commonly accepted which facilitate short text understanding and search by enriching the short text. However, they only analyze the literal semantics of short text, and the unique social characteristics of social network which also contain semantic information are not utilized well. To better capture the rich semantics in microblog short text, we propose a new microblog short text extended search method based on a semantic hashtag graph by combining social and conceptual information, which enriches each short text by concepts and associated hashtags to represent whole semantic features. Considering the microblog context, we introduce concepts through Wikipedia, as well as semantic consistency of hashtags. Specifically, for conceptual semantics, we propose a conceptual analysis method which merges explicit and implicit information in Wikipedia. For social semantics in hashtags, a semantic hashtag graph which combines social and conceptual information is put forward to generate semantic associated hashtags. We conduct experiments and the results show that our method is obviously better than the other existing state-of-the-art approaches in semantic understanding and search of short text.

Tagging augmented neural topic model for semantic sparse Web service discovery

SummarySearch engine based Web service discovery model suffers from the semantic sparsity problem due to the fact that Web services are described in short texts, which in turn leads to poor recall. To address this issue, external information that enriches the semantics of the Web service and improves discovery performance has been highly concerned. In light of this, we propose a novel Web service discovery approach that uses the neural topic model, which seamlessly integrates tagging information and word embedding for semantic sparsity problem. More specifically, instead of clustering Web services as done in most existing service discovery approaches, we use word embedding to map the words as continuous embeddings to embody external semantics of the service description. We also leverage the neural topic model in service discovery, which takes continuous word distribution as the input and interprets the Web service description as a hierarchical model. Based on the neural topic model and word embedding, we propose an efficient Web service query and ranking approach. Experiments conducted on a real‐world Web service dataset demonstrate the effectiveness of the proposed approach.