Search

Video-based trajectory analysis might be rather well discussed in sports, such as soccer or basketball, but in cycling, this is far less common. In this paper, a video processing pipeline to extract riding lines in cyclocross races is presented. The pipeline consists of a stepwise analysis process to extract riding behavior from a region (i.e., the fence) in a video camera feed. In the first step, the riders are identified by an Alphapose skeleton detector and tracked with a spatiotemporally aware pose tracker. Next, each detected pose is enriched with additional meta-information, such as rider modus (e.g., sitting on the saddle or standing on the pedals) and detected team (based on the worn jerseys). Finally, a post-processor brings all the information together and proposes ride lines with meta-information for the riders in the fence. The presented methodology can provide interesting insights, such as intra-athlete ride line clustering, anomaly detection, and detailed breakdowns of riding and running durations within the segment. Such detailed rider info can be very valuable for performance analysis, storytelling, and automatic summarization.

Water is a critically important element of human life. The best practice of effective water management comes from ancient civilizations that, despite their technologies and practices, were unable to prevent collapse from water scarcity. In the 21st century, in an era of climate change, pollution or population explosion, cities are looking for innovative ways to effectively manage scarce resources for future generations. Which elements should cities of the future follow to avoid water collapse? The following article aims to identify the key elements of effective management and to represent them graphically in the form of a recommended model, which will be verified in the future in Slovakia. The article uses case analysis of best past and current practices, comparison and summarization to identify the elements, creativity, and logic in the development of the model, including induction and deduction. The article serves as a basis for fellow researchers (analyses carried out) and strategic urban management (effective urban water management). The main finding of the article is that ecological change puts pressure on social elements and therefore it is necessary to focus on the area of strategic management. Cities should not only know how to manage resource abundance or short-term scarcity, but also long-term scarcity. They should use elements of trust, awareness and continuous improvement through modern monitoring technologies (UAVs, sensors) and prediction (machine learning). This is the only way to generate water sustainability in the urban concept of the future.

Text summarization remains a challenging task in the natural language processing field despite the plethora of applications in enterprises and daily life. One of the common use cases is the summarization of web pages which has the potential to provide an overview of web pages to devices with limited features. In fact, despite the increasing penetration rate of mobile devices in rural areas, the bulk of those devices offer limited features in addition to the fact that these areas are covered with limited connectivity such as the GSM network. Summarizing web pages into SMS becomes, therefore, an important task to provide information to limited devices. This work introduces WATS-SMS, a T5-based French Wikipedia Abstractive Text Summarizer for SMS. It is built through a transfer learning approach. The T5 English pre-trained model is used to generate a French text summarization model by retraining the model on 25,000 Wikipedia pages then compared with different approaches in the literature. The objective is twofold: (1) to check the assumption made in the literature that abstractive models provide better results compared to extractive ones, and (2) to evaluate the performance of our model compared to other existing abstractive models. A score based on ROUGE metrics gave us a value of 52% for articles with length up to 500 characters against 34.2% for transformer-ED and 12.7% for seq-2seq-attention, and a value of 77% for articles with larger size against 37% for transformers-DMCA. Moreover, an architecture including a software SMS-gateway has been developed to allow owners of mobile devices with limited features to send requests and to receive summaries through the GSM network.

Named entity recognition (NER) is a fundamental task in many natural language processing (NLP) applications, such as text summarization and semantic information retrieval. Recently, deep neural networks (NNs) with the attention mechanism yield excellent performance in NER by taking advantage of character-level and word-level representation learning. In this paper, we propose a deep context-aware bidirectional long short-term memory (CaBiLSTM) model for the Sindhi NER task. The model relies upon contextual representation learning (CRL), bidirectional encoder, self-attention, and sequential conditional random field (CRF). The CaBiLSTM model incorporates task-oriented CRL based on joint character-level and word-level representations. It takes character-level input to learn the character representations. Afterwards, the character representations are transformed into word features, and the bidirectional encoder learns the word representations. The output of the final encoder is fed into the self-attention through a hidden layer before decoding. Finally, we employ the CRF for the prediction of label sequences. The baselines and the proposed CaBiLSTM model are compared by exploiting pretrained Sindhi GloVe (SdGloVe), Sindhi fastText (SdfastText), task-oriented, and CRL-based word representations on the recently proposed SiNER dataset. Our proposed CaBiLSTM model achieved a high F1-score of 91.25% on the SiNER dataset with CRL without relying on additional handmade features, such as hand-crafted rules, gazetteers, or dictionaries.

Summaries of scientific publications enable readers to gain an overview of a large number of studies, but users’ preferences have not yet been explored. In this paper, we conduct two user studies (i.e., short- and long-term studies) where Japanese university students read summaries of English research articles that were either manually written or automatically generated using text summarization and/or machine translation. In the short-term experiment, subjects compared and evaluated the two types of summaries of the same article. We analyze the characteristics in the generated summaries that readers regard as important, such as content richness and simplicity. The experimental results show that subjects are mainly judged based on four criteria, including content richness, simplicity, fluency, and format. In the long-term experiment, subjects read 50 summaries and answered whether they would like to read the original papers after reading the summaries. We discuss the characteristics in the summaries that readers tend to use to determine whether to read the papers, such as topic, methods, and results. The comments from subjects indicate that specific components of scientific publications, including research topics and methods, are important to judge whether to read or not. Our study provides insights to enhance the effectiveness of automatic summarization of scientific publications.

Linguists have been focused on a qualitative comparison of the semantics from different languages. Evaluation of the semantic interpretation among disparate language pairs like English and Tamil is an even more formidable task than for Slavic languages. The concept of word embedding in Natural Language Processing (NLP) has enabled a felicitous opportunity to quantify linguistic semantics. Multi-lingual tasks can be performed by projecting the word embeddings of one language onto the semantic space of the other. This research presents a suite of data-efficient deep learning approaches to deduce the transfer function from the embedding space of English to that of Tamil, deploying three popular embedding algorithms: Word2Vec, GloVe and FastText. A novel evaluation paradigm was devised for the generation of embeddings to assess their effectiveness, using the original embeddings as ground truths. Transferability across other target languages of the proposed model was assessed via pre-trained Word2Vec embeddings from Hindi and Chinese languages. We empirically prove that with a bilingual dictionary of a thousand words and a corresponding small monolingual target (Tamil) corpus, useful embeddings can be generated by transfer learning from a well-trained source (English) embedding. Furthermore, we demonstrate the usability of generated target embeddings in a few NLP use-case tasks, such as text summarization, part-of-speech (POS) tagging, and bilingual dictionary induction (BDI), bearing in mind that those are not the only possible applications.

While digitalization of cultural organizations is in full swing and growth, it is common knowledge that websites can be used as a beacon to expand the awareness and consideration of their services on the Web. Nevertheless, recent research results indicate the managerial difficulties in deploying strategies for expanding the discoverability, visibility, and accessibility of these websites. In this paper, a three-stage data-driven Search Engine Optimization schema is proposed to assess the performance of Libraries, Archives, and Museums websites (LAMs), thus helping administrators expand their discoverability, visibility, and accessibility within the Web realm. To do so, the authors examine the performance of 341 related websites from all over the world based on three different factors, Content Curation, Speed, and Security. In the first stage, a statistically reliable and consistent assessment schema for evaluating the SEO performance of LAMs websites through the integration of more than 30 variables is presented. Subsequently, the second stage involves a descriptive data summarization for initial performance estimations of the examined websites in each factor is taking place. In the third stage, predictive regression models are developed to understand and compare the SEO performance of three different Content Management Systems, namely the Drupal, WordPress, and custom approaches, that LAMs websites have adopted. The results of this study constitute a solid stepping-stone both for practitioners and researchers to adopt and improve such methods that focus on end-users and boost organizational structures and culture that relied on data-driven approaches for expanding the visibility of LAMs services.

Building Smart City management concepts is based on the implementation and use of advanced technologies. The primary impulse for writing the article was the ambition to identify the current advanced technologies of Smart City management. The aim of the article is to propose a general model for the implementation of advanced technologies for Smart City management, based on the knowledge gained from the analysis of literature and case studies. In order to fulfill the set goal, it is necessary to obtain answers to two research questions. The findings were obtained through a secondary analysis of the literature, i.e., relevant articles from the scientific databases Web of Science and Scopus analysis of case studies of the best Smart Cities practices. According to the Smart City Index 2020 and IESE Cities in Motion, the leaders among the Smart Cities are Singapore and London, followed by Helsinki. In addition to the analyses, the article also uses methods of summarization, comparison, creativity, logic, induction and deduction. Smart Cities use 12 identified advanced technologies in their practice. Strategic management in Singapore, London and Helsinki adapts technology to the needs and requirements of its citizens, thus connecting the technological aspect with the managerial and social aspects. The contributions of the work include results for fellow researchers and a model for strategic management of new Smart Cities. The results of the article provide fellow researchers with the findings of a secondary analysis of relevant articles, from which they can draw when writing their own publications without the need for time-consuming search of the articles about this topic in databases. The general model of implementation of advanced technologies serves as a basis for strategic management of new Smart Cities that want to implement a technological base and at the same time do not want to forget the managerial and social aspects. Testing the model in practice with a new Slovak Smart City is part of future research activities.

Energy disaggregation, known in the literature as Non-Intrusive Load Monitoring (NILM), is the task of inferring the power demand of the individual appliances given the aggregate power demand recorded by a single smart meter which monitors multiple appliances. In this paper, we propose a deep neural network that combines a regression subnetwork with a classification subnetwork for solving the NILM problem. Specifically, we improve the generalization capability of the overall architecture by including an encoder–decoder with a tailored attention mechanism in the regression subnetwork. The attention mechanism is inspired by the temporal attention that has been successfully applied in neural machine translation, text summarization, and speech recognition. The experiments conducted on two publicly available datasets—REDD and UK-DALE—show that our proposed deep neural network outperforms the state-of-the-art in all the considered experimental conditions. We also show that modeling attention translates into the network’s ability to correctly detect the turning on or off an appliance and to locate signal sections with high power consumption, which are of extreme interest in the field of energy disaggregation.

This article describes an AI-based solution to multiclass fire segmentation. The flame contours are divided into red, yellow, and orange areas. This separation is necessary to identify the hottest regions for flame suppression. Flame objects can have a wide variety of shapes (convex and non-convex). In that case, the segmentation task is more applicable than object detection because the center of the fire is much more accurate and reliable information than the center of the bounding box and, therefore, can be used by robotics systems for aiming. The UNet model is used as a baseline for the initial solution because it is the best open-source convolutional neural network. There is no available open dataset for multiclass fire segmentation. Hence, a custom dataset was developed and used in the current study, including 6250 samples from 36 videos. We compared the trained UNet models with several configurations of input data. The first comparison is shown between the calculation schemes of fitting the frame to one window and obtaining non-intersected areas of sliding window over the input image. Secondarily, we chose the best main metric of the loss function (soft Dice and Jaccard). We addressed the problem of detecting flame regions at the boundaries of non-intersected regions, and introduced new combinational methods of obtaining output signal based on weighted summarization and Gaussian mixtures of half-intersected areas as a solution. In the final section, we present UUNet-concatenative and wUUNet models that demonstrate significant improvements in accuracy and are considered to be state-of-the-art. All models use the original UNet-backbone at the encoder layers (i.e., VGG16) to demonstrate the superiority of the proposed architectures. The results can be applied to many robotic firefighting systems.

In this paper, a deep stacked auto-encoder (SAE) scheme followed by a hierarchical Sparse Modeling for Representative Selection (SMRS) algorithm is proposed to summarize dance video sequences, recorded using the VICON Motion capturing system. SAE&rsquo, s main task is to reduce the redundant information embedding in the raw data and, thus, to improve summarization performance. This becomes apparent when two dancers are performing simultaneously and severe errors are encountered in the humans&rsquo, point joints, due to dancers&rsquo, occlusions in the 3D space. Four summarization algorithms are applied to extract the key frames, density based, Kennard Stone, conventional SMRS and its hierarchical scheme called H-SMRS. Experimental results have been carried out on real-life dance sequences of Greek traditional dances while the results have been compared against ground truth data selected by dance experts. The results indicate that H-SMRS being applied after the SAE information reduction module extracts key frames which are deviated in time less than 0.3 s to the ones selected by the experts and with a standard deviation of 0.18 s. Thus, the proposed scheme can effectively represent the content of the dance sequence.

One of the main motives for creating this article was to explore the importance of cultural aspects in building smart city approaches. The aim of this article was to obtain answers to three research questions, the answers to which made it possible to identify the elements of multiculturalism that affect the development of smart cities, to find out how multiculturalism affects smart cities and how to manage diversity. The ambition was to create and organize the most important findings into a comprehensive framework. To achieve this goal, secondary analysis methods were used by examining the literature and case studies of best practices from Europe, North America, and Asia. The choice of case studies was conditioned by the placement of smart cities in four global indices (smart city index, Arcadis, IESE and global power index), the existence of a multicultural strategy and elements of successful diversity management, including positive effects and possible limitations. In addition, methods of analysis, comparison and summarization were used. Effective diversity management acts as an accelerator of the sustainable development of smart cities. The results of the analysis of the case studies serve as a basis for recommendations and the creation of a proposed general model, whose task is to simplify the adoption of intelligent concepts, which creates space for the specification of local or cultural conditions of the country. Testing the model in practice is the subject of the following research activities of the authors.

The development of the economy and the transition to industry 4.0 creates new challenges for artificial intelligence methods. Such challenges include the processing of large volumes of data, the analysis of various dynamic indicators, the discovery of complex dependencies in the accumulated data, and the forecasting of the state of processes. The main point of this study is the development of a set of analytical and prognostic methods. The methods described in this article based on fuzzy logic, statistic, and time series data mining, because data extracted from dynamic systems are initially incomplete and have a high degree of uncertainty. The ultimate goal of the study is to improve the quality of data analysis in industrial and economic systems. The advantages of the proposed methods are flexibility and orientation to the high interpretability of dynamic data. The high level of the interpretability and interoperability of dynamic data is achieved due to a combination of time series data mining and knowledge base engineering methods. The merging of a set of rules extracted from the time series and knowledge base rules allow for making a forecast in case of insufficiency of the length and nature of the time series. The proposed methods are also based on the summarization of the results of processes modeling for diagnosing technical systems, forecasting of the economic condition of enterprises, and approaches to the technological preparation of production in a multi-productive production program with the application of type 2 fuzzy sets for time series modeling. Intelligent systems based on the proposed methods demonstrate an increase in the quality and stability of their functioning. This article contains a set of experiments to approve this statement.

Under the constraint of memory capacity of the neural network and the document length, it is difficult to generate summaries with adequate salient information. In this work, the self-matching mechanism is incorporated into the extractive summarization system at the encoder side, which allows the encoder to optimize the encoding information at the global level and effectively improves the memory capacity of conventional LSTM. Inspired by human coarse-to-fine understanding mode, localness is modeled by Gaussian bias to improve contextualization for each sentence, and merged into the self-matching energy. The refined self-matching mechanism not only establishes global document attention but perceives association with neighboring signals. At the decoder side, the pointer network is utilized to perform a two-hop attention on context and extraction state. Evaluations on the CNN/Daily Mail dataset verify that the proposed model outperforms the strong baseline models and statistical significantly.

This paper proposes a novel method for sports video scene classification with the particular intention of video summarization. Creating and publishing a shorter version of the video is more interesting than a full version due to instant entertainment. Generating shorter summaries of the videos is a tedious task that requires significant labor hours and unnecessary machine occupation. Due to the growing demand for video summarization in marketing, advertising agencies, awareness videos, documentaries, and other interest groups, researchers are continuously proposing automation frameworks and novel schemes. Since the scene classification is a fundamental component of video summarization and video analysis, the quality of scene classification is particularly important. This article focuses on various practical implementation gaps over the existing techniques and presents a method to achieve high-quality of scene classification. We consider cricket as a case study and classify five scene categories, i.e., batting, bowling, boundary, crowd and close-up. We employ our model using pre-trained AlexNet Convolutional Neural Network (CNN) for scene classification. The proposed method employs new, fully connected layers in an encoder fashion. We employ data augmentation to achieve a high accuracy of 99.26% over a smaller dataset. We conduct a performance comparison against baseline approaches to prove the superiority of the method as well as state-of-the-art models. We evaluate our performance results on cricket videos and compare various deep-learning models, i.e., Inception V3, Visual Geometry Group (VGGNet16, VGGNet19), Residual Network (ResNet50), and AlexNet. Our experiments demonstrate that our method with AlexNet CNN produces better results than existing proposals.

Deep reinforcement learning (DRL) has achieved significant results in many machine learning (ML) benchmarks. In this short survey, we provide an overview of DRL applied to trading on financial markets with the purpose of unravelling common structures used in the trading community using DRL, as well as discovering common issues and limitations of such approaches. We include also a short corpus summarization using Google Scholar. Moreover, we discuss how one can use hierarchy for dividing the problem space, as well as using model-based RL to learn a world model of the trading environment which can be used for prediction. In addition, multiple risk measures are defined and discussed, which not only provide a way of quantifying the performance of various algorithms, but they can also act as (dense) reward-shaping mechanisms for the agent. We discuss in detail the various state representations used for financial markets, which we consider critical for the success and efficiency of such DRL agents. The market in focus for this survey is the cryptocurrency market; the results of this survey are two-fold: firstly, to find the most promising directions for further research and secondly, to show how a lack of consistency in the community can significantly impede research and the development of DRL agents for trading.

Finite mixture models are widely used for modeling and clustering data. When they are used for clustering, they are often interpreted by regarding each component as one cluster. However, this assumption may be invalid when the components overlap. It leads to the issue of analyzing such overlaps to correctly understand the models. The primary purpose of this paper is to establish a theoretical framework for interpreting the overlapping mixture models by estimating how they overlap, using measures of information such as entropy and mutual information. This is achieved by merging components to regard multiple components as one cluster and summarizing the merging results. First, we propose three conditions that any merging criterion should satisfy. Then, we investigate whether several existing merging criteria satisfy the conditions and modify them to fulfill more conditions. Second, we propose a novel concept named clustering summarization to evaluate the merging results. In it, we can quantify how overlapped and biased the clusters are, using mutual information-based criteria. Using artificial and real datasets, we empirically demonstrate that our methods of modifying criteria and summarizing results are effective for understanding the cluster structures. We therefore give a new view of interpretability/explainability for model-based clustering.

The paper describes a convex optimization formulation of the extractive text summarization problem and a simple and scalable algorithm to solve it. The optimization program is constructed as a convex relaxation of an intuitive but computationally hard integer programming problem. The objective function is highly symmetric, being invariant under unitary transformations of the text representations. Another key idea is to replace the constraint on the number of sentences in the summary with a convex surrogate. For solving the program we have designed a specific projected gradient descent algorithm and analyzed its performance in terms of execution time and quality of the approximation. Using the datasets DUC 2005 and Cornell Newsroom Summarization Dataset, we have shown empirically that the algorithm can provide competitive results for single document summarization and multi-document query-based summarization. On the Cornell Newsroom Summarization Dataset, it ranked second among the unsupervised methods tested. For the more challenging task of multi-document query-based summarization, the method was tested on the DUC 2005 Dataset. Our algorithm surpassed the other reported methods with respect to the ROUGE-SU4 metric, and it was at less than 0.01 from the top performing algorithms with respect to ROUGE-1 and ROUGE-2 metrics.

In recent years, the explainable artificial intelligence (XAI) paradigm is gaining wide research interest. The natural language processing (NLP) community is also approaching the shift of paradigm: building a suite of models that provide an explanation of the decision on some main task, without affecting the performances. It is not an easy job for sure, especially when very poorly interpretable models are involved, like the almost ubiquitous (at least in the NLP literature of the last years) transformers. Here, we propose two different transformer-based methodologies exploiting the inner hierarchy of the documents to perform a sentiment analysis task while extracting the most important (with regards to the model decision) sentences to build a summary as the explanation of the output. For the first architecture, we placed two transformers in cascade and leveraged the attention weights of the second one to build the summary. For the other architecture, we employed a single transformer to classify the single sentences in the document and then combine the probability scores of each to perform the classification and then build the summary. We compared the two methodologies by using the IMDB dataset, both in terms of classification and explainability performances. To assess the explainability part, we propose two kinds of metrics, based on benchmarking the models’ summaries with human annotations. We recruited four independent operators to annotate few documents retrieved from the original dataset. Furthermore, we conducted an ablation study to highlight how implementing some strategies leads to important improvements on the explainability performance of the cascade transformers model.

The exponential growth of user-generated content has increased the need for efficient video summarization schemes. However, most approaches underestimate the power of aural features, while they are designed to work mainly on commercial/professional videos. In this work, we present an approach that uses both aural and visual features in order to create video summaries from user-generated videos. Our approach produces dynamic video summaries, that is, comprising the most “important” parts of the original video, which are arranged so as to preserve their temporal order. We use supervised knowledge from both the aforementioned modalities and train a binary classifier, which learns to recognize the important parts of videos. Moreover, we present a novel user-generated dataset which contains videos from several categories. Every 1 s part of each video from our dataset has been annotated by more than three annotators as being important or not. We evaluate our approach using several classification strategies based on audio, video and fused features. Our experimental results illustrate the potential of our approach.

Currently, reviews on the Internet contain abundant information about users and products, and this information is of great value to recommendation systems. As a result, review-based recommendations have begun to show their effectiveness and research value. Due to the accumulation of a large number of reviews, it has become very important to extract useful information from reviews. Automatic summarization can capture important information from a set of documents and present it in the form of a brief summary. Therefore, integrating automatic summarization into recommendation systems is a potential approach for solving this problem. Based on this idea, we propose a joint summarization and pre-trained recommendation model for review-based rate prediction. Through automatic summarization and a pre-trained language model, the overall recommendation model learns a fine-grained summary representation of the key content as well as the relationships between words and sentences in each review. The review summary representations of users and items are finally incorporated into a neural collaborative filtering (CF) framework with interactive attention mechanisms to predict the rating scores. We perform experiments on the Amazon dataset and compare our method with several competitive baselines. Experimental results show that the performance of the proposed model is obviously better than that of the baselines. Relative to the current best results, the average improvements obtained on four sub-datasets randomly selected from the Amazon dataset are approximately 3.29%.

Temporal salience considers how visual attention varies over time. Although visual salience has been widely studied from a spatial perspective, its temporal dimension has been mostly ignored, despite arguably being of utmost importance to understand the temporal evolution of attention on dynamic contents. To address this gap, we proposed Glimpse, a novel measure to compute temporal salience based on the observer-spatio-temporal consistency of raw gaze data. The measure is conceptually simple, training free, and provides a semantically meaningful quantification of visual attention over time. As an extension, we explored scoring algorithms to estimate temporal salience from spatial salience maps predicted with existing computational models. However, these approaches generally fall short when compared with our proposed gaze-based measure. Glimpse could serve as the basis for several downstream tasks such as segmentation or summarization of videos. Glimpse’s software and data are publicly available.

[EN] In this paper, we present an approach to Spanish talk shows summarization. Our approach is based on the use of Siamese Neural Networks on the transcription of the show audios. Specifically, we propose to use Hierarchical Attention Networks to select the most relevant sentences for each speaker about a given topic in the show, in order to summarize his opinion about the topic. We train these networks in a siamese way to determine whether a summary is appropriate or not. Previous evaluation of this approach on summarization task of English newspapers achieved performances similar to other state-of-the-art systems. In the absence of enough transcribed or recognized speech data to train our system for talk show summarization in Spanish, we acquire a large corpus of document-summary pairs from Spanish newspapers and we use it to train our system. We choose this newspapers domain due to its high similarity with the topics addressed in talk shows. A preliminary evaluation of our summarization system on Spanish TV programs shows the adequacy of the proposal., This work has been partially supported by the Spanish MINECO and FEDER founds under project AMIC (TIN2017-85854-C4-2-R). Work of Jose-Angel Gonzalez is financed by Universitat Politecnica de Valencia under grant PAID-01-17.

This article describes techniques employed in the production of a synthetic dataset of driver telematics emulated from a similar real insurance dataset. The synthetic dataset generated has 100,000 policies that included observations about driver's claims experience together with associated classical risk variables and telematics-related variables. This work is aimed to produce a resource that can be used to advance models to assess risks for usage-based insurance. It follows a three-stage process using machine learning algorithms. The first stage is simulating values for the number of claims as multiple binary classifications applying feedforward neural networks. The second stage is simulating values for aggregated amount of claims as regression using feedforward neural networks, with number of claims included in the set of feature variables. In the final stage, a synthetic portfolio of the space of feature variables is generated applying an extended $\texttt{SMOTE}$ algorithm. The resulting dataset is evaluated by comparing the synthetic and real datasets when Poisson and gamma regression models are fitted to the respective data. Other visualization and data summarization produce remarkable similar statistics between the two datasets. We hope that researchers interested in obtaining telematics datasets to calibrate models or learning algorithms will find our work valuable., Comment: 24 pages, 11 figures, 6 tables

Automatic extractive text summarization retrieves a subset of data that represents most notable sentences in the entire document. In the era of digital explosion, which is mostly unstructured textual data, there is a demand for users to understand the huge amount of text in a short time, this demands the need for an automatic text summarizer. From summaries, the users get the idea of the entire content of the document and can decide whether to read the entire document or not. This work mainly focuses on generating a summary from multiple news documents. In this case, the summary helps to reduce the redundant news from the different newspapers. A multi-document summary is more challenging than a single-document summary since it has to solve the problem of overlapping information among sentences from different documents. Extractive text summarization yields the sensitive part of the document by neglecting the irrelevant and redundant sentences. In this paper, we propose a framework for extracting a summary from multiple documents in the Malayalam Language. Also, since the multi-document summarization data set is sparse, methods based on deep learning are difficult to apply. The proposed work discusses the performance of existing standard algorithms in multi-document summarization of the Malayalam Language. We propose a sentence extraction algorithm that selects the top ranked sentences with maximum diversity. The system is found to perform well in terms of precision, recall, and F-measure on multiple input documents.

Detecting key frames in videos is a common problem in many applications such as video classification, action recognition and video summarization. These tasks can be performed more efficiently using only a handful of key frames rather than the full video. Existing key frame detection approaches are mostly designed for supervised learning and require manual labelling of key frames in a large corpus of training data to train the models. Labelling requires human annotators from different backgrounds to annotate key frames in videos which is not only expensive and time consuming but also prone to subjective errors and inconsistencies between the labelers. To overcome these problems, we propose an automatic self-supervised method for detecting key frames in a video. Our method comprises a two-stream ConvNet and a novel automatic annotation architecture able to reliably annotate key frames in a video for self-supervised learning of the ConvNet. The proposed ConvNet learns deep appearance and motion features to detect frames that are unique. The trained network is then able to detect key frames in test videos. Extensive experiments on UCF101 human action and video summarization VSUMM datasets demonstrates the effectiveness of our proposed method.

Opinion mining and summarization of the increasing user-generated content on different digital platforms (e.g., news platforms) are playing significant roles in the success of government programs and initiatives in digital governance, from extracting and analyzing citizen&rsquo, s sentiments for decision-making. Opinion mining provides the sentiment from contents, whereas summarization aims to condense the most relevant information. However, most of the reported opinion summarization methods are conceived to obtain generic summaries, and the context that originates the opinions (e.g., the news) has not usually been considered. In this paper, we present a context-aware opinion summarization model for monitoring the generated opinions from news. In this approach, the topic modeling and the news content are combined to determine the &ldquo, importance&rdquo, of opinionated sentences. The effectiveness of different developed settings of our model was evaluated through several experiments carried out over Spanish news and opinions collected from a real news platform. The obtained results show that our model can generate opinion summaries focused on essential aspects of the news, as well as cover the main topics in the opinionated texts well. The integration of term clustering, word embeddings, and the similarity-based sentence-to-news scoring turned out the more promising and effective setting of our model.

Explosive growth in geospatial and temporal data as well as the emergence of new technologies emphasize the need for automated discovery of spatiotemporal knowledge. Spatiotemporal data mining studies the process of discovering interesting and previously unknown, but potentially useful patterns from large spatiotemporal databases. It has broad application domains including ecology and environmental management, public safety, transportation, earth science, epidemiology, and climatology. The complexity of spatiotemporal data and intrinsic relationships limits the usefulness of conventional data science techniques for extracting spatiotemporal patterns. In this survey, we review recent computational techniques and tools in spatiotemporal data mining, focusing on several major pattern families: spatiotemporal outlier, spatiotemporal coupling and tele-coupling, spatiotemporal prediction, spatiotemporal partitioning and summarization, spatiotemporal hotspots, and change detection. Compared with other surveys in the literature, this paper emphasizes the statistical foundations of spatiotemporal data mining and provides comprehensive coverage of computational approaches for various pattern families. ISPRS Int. J. Geo-Inf. 2015, 4 2307 We also list popular software tools for spatiotemporal data analysis. The survey concludes with a look at future research needs.

The wood-processing industry currently does not sufficiently use modern technologies, unlike the automotive sector. The primary motive for writing this article was in cooperation with a Slovak wood processing company, which wanted to improve its logistics processes and increase competitiveness in the wood processing sector through the implementation of new technologies. The aim of this article was to identify the positives and limitations of the implementation of Internet of Things (IoT) technology into the wood processing industry, based on a secondary analysis of case studies and the best practice of American wood processing companies such as West Fraser Timber in Canada, and Weyerhaeuser in the USA. The selection of case studies was conditional on criteria of time relevance, size of the sawmills, and production volume in m3. These conditional criteria reflected the conditions for the introduction of similar concepts for wood-processing enterprises in Slovakia. The implementation of the IoT can reduce operating costs by up to 20%, increase added value for customers, and collect real-time data that can serve as the basis for support of management and decision-making at the operational, tactical, and strategic levels. In addition to the secondary analysis, methods of comparison of global wood processing companies, synthesis of knowledge, and summarization of positives and limitations of IoT implementation or deduction were used to reach our conclusions. The results were used as the basis for the design of a general model for the implementation of IoT technology for Slovak wood processing enterprises. This model may represent best practice for the selected locality and industry. The implications and verification of the designed model in practice will form part of other research activities, already underway in the form of a primary survey.

Supervised neural network models have achieved outstanding performance in the document summarization task in recent years. However, it is hard to get enough labeled training data with a high quality for these models to generate different types of summaries in reality. In this work, we mainly focus on improving the performance of the popular unsupervised Textrank algorithm that requires no labeled training data for extractive summarization. We first modify the original edge weight of Textrank to take the relative position of sentences into account, and then combine the output of the improved Textrank with K-means clustering to improve the diversity of generated summaries. To further improve the performance of our model, we innovatively incorporate external knowledge from open-source knowledge graphs into our model by entity linking. We use the knowledge graph sentence embedding and the tf-idf embedding as the input of our improved Textrank, and get the final score for each sentence by linear combination. Evaluations on the New York Times data set show the effectiveness of our knowledge-enhanced approach. The proposed model outperforms other popular unsupervised models significantly.

The advent of the sequence-to-sequence model and the attention mechanism has increased the comprehension and readability of automatically generated summaries. However, most previous studies on text summarization have focused on generating or extracting sentences only from an original text, even though every text has a latent topic category. That is, even if a topic category helps improve the summarization quality, there have been no efforts to utilize such information in text summarization. Therefore, this paper proposes a novel topical category-aware neural text summarizer which is differentiated from legacy neural summarizers in that it reflects the topic category of an original text into generating a summary. The proposed summarizer adopts the class activation map (CAM) as topical influence of the words in the original text. Since the CAM excerpts the words relevant to a specific category from the text, it allows the attention mechanism to be influenced by the topic category. As a result, the proposed neural summarizer reflects the topical information of a text as well as the content information into a summary by combining the attention mechanism and CAM. The experiments on The New York Times Annotated Corpus show that the proposed model outperforms the legacy attention-based sequence-to-sequence model, which proves that it is effective at reflecting a topic category into automatic summarization.

In an age overflowing with information, the task of converting unstructured data into structured data are a vital task of great need. Currently, most relation extraction modules are more focused on the extraction of local mention-level relations&mdash, usually from short volumes of text. However, in most cases, the most vital and important relations are those that are described in length and detail. In this research, we propose GREG: A Global level Relation Extractor model using knowledge graph embeddings for document-level inputs. The model uses vector representations of mention-level &lsquo, local&rsquo, relation&rsquo, s to construct knowledge graphs that can represent the input document. The knowledge graph is then used to predict global level relations from documents or large bodies of text. The proposed model is largely divided into two modules which are synchronized during their training. Thus, each of the model&rsquo, s modules is designed to deal with local relations and global relations separately. This allows the model to avoid the problem of struggling against loss of information due to too much information crunched into smaller sized representations when attempting global level relation extraction. Through evaluation, we have shown that the proposed model yields high performances in both predicting global level relations and local level relations consistently.

The main goal of this Special Issue is to present the latest advances in research and novel applications of speech and language technologies based on the works presented at the IberSPEECH edition held in Barcelona in 2018, paying special attention to those focused on Iberian languages. IberSPEECH is the international conference of the Special Interest Group on Iberian Languages (SIG-IL) of the International Speech Communication Association (ISCA) and of the Spanish Thematic Network on Speech Technologies (Red Temática en Tecnologías del Habla, or RTTH for short). Several researchers were invited to extend their contributions presented at IberSPEECH2018 due to their interest and quality. As a result, this Special Issue is composed of 13 papers that cover different topics of investigation related to perception, speech analysis and enhancement, speaker verification and identification, speech production and synthesis, natural language processing, together with several applications and evaluation challenges.

Multimedia information requires large repositories of audio-video data. Retrieval and delivery of video content is a very time-consuming process and is a great challenge for researchers. An efficient approach for faster browsing of large video collections and more efficient content indexing and access is video summarization. Compression of data through extraction of keyframes is a solution to these challenges. A keyframe is a representative frame of the salient features of the video. The output frames must represent the original video in temporal order. The proposed research presents a method of keyframe extraction using the mean of consecutive k frames of video data. A sliding window of size k / 2 is employed to select the frame that matches the median entropy value of the sliding window. This is called the Median of Entropy of Mean Frames (MME) method. MME is mean-based keyframes selection using the median of the entropy of the sliding window. The method was tested for more than 500 videos of sign language gestures and showed satisfactory results.

For software maintenance, bug reports provide useful information to developers because they can be used for various tasks such as debugging and understanding previous changes. However, as they are typically written in the form of conversations among developers, bug reports tend to be unnecessarily long and verbose, with the consequence that developers often have difficulties reading or understanding bug reports. To mitigate this problem, methods that automatically generate a summary of bug reports have been proposed, and various related studies have been conducted. However, existing bug report summarization methods have not fully exploited the inherent characteristics of bug reports. In this paper, we propose a bug report summarization method that uses the weighted-PageRank algorithm and exploits the 'duplicates&rsquo, &lsquo, blocks&rsquo, and &lsquo, depends-on&rsquo, relationships between bug reports. The experimental results show that our method outperforms the state-of-the-art method in terms of both the quality of the summary and the number of applicable bug reports.

Current popular abstractive summarization is based on an attentional encoder-decoder framework. Based on the architecture, the decoder generates a summary according to the full text that often results in the decoder being interfered by some irrelevant information, thereby causing the generated summaries to suffer from low saliency. Besides, we have observed the process of people writing summaries and find that they write a summary based on the necessary information rather than the full text. Thus, in order to enhance the saliency of the abstractive summarization, we propose an attentive information extraction model. It consists of a multi-layer perceptron (MLP) gated unit that pays more attention to the important information of the source text and a similarity module to encourage high similarity between the reference summary and the important information. Before the summary decoder, the MLP and the similarity module work together to extract the important information for the decoder, thus obtaining the skeleton of the source text. This effectively reduces the interference of irrelevant information to the decoder, therefore improving the saliency of the summary. Our proposed model was tested on CNN/Daily Mail and DUC-2004 datasets, and achieved a 42.01 ROUGE-1 f-score and 33.94 ROUGE-1, recall respectively. The result outperforms the state-of-the-art abstractive model on the same dataset. In addition, by subjective human evaluation, the saliency of the generated summaries was further enhanced.

In general, humans and animals often interact within the same environment at the same time. Human activities may disturb or affect some bird activities. Therefore, it is important to monitor and study the relationships between human and animal activities. This paper proposed a system able not only to automatically classify human and bird activities using bioacoustic data, but also to automatically summarize patterns of events over time. To perform automatic summarization of acoustic events, a frequency&ndash, duration graph (FDG) framework was proposed to summarize the patterns of human and bird activities. This system first performs data pre-processing work on raw bioacoustic data and then applies a support vector machine (SVM) model and a multi-layer perceptron (MLP) model to classify human and bird chirping activities before using the FDG framework to summarize results. The SVM model achieved 98% accuracy on average and the MLP model achieved 98% accuracy on average across several day-long recordings. Three case studies with real data show that the FDG framework correctly determined the patterns of human and bird activities over time and provided both statistical and graphical insight into the relationships between these two events.

Wireless capsule endoscopy (WCE) has great advantages over traditional endoscopy because it is portable and easy to use, especially in remote monitoring health-services. However, during the WCE process, the large amount of captured video data demands a significant deal of computation to analyze and retrieve informative video frames. In order to facilitate efficient WCE data collection and browsing task, we present a resource- and bandwidth-aware WCE video summarization framework that extracts the representative keyframes of the WCE video contents by removing redundant and non-informative frames. For redundancy elimination, we use Jeffrey-divergence between color histograms and inter-frame Boolean series-based correlation of color channels. To remove non-informative frames, multi-fractal texture features are extracted to assist the classification using an ensemble-based classifier. Owing to the limited WCE resources, it is impossible for the WCE system to perform computationally intensive video summarization tasks. To resolve computational challenges, mobile-cloud architecture is incorporated, which provides resizable computing capacities by adaptively offloading video summarization tasks between the client and the cloud server. The qualitative and quantitative results are encouraging and show that the proposed framework saves information transmission cost and bandwidth, as well as the valuable time of data analysts in browsing remote sensing data.

text summarization aims to offer a highly condensed and valuable information that expresses the main ideas of the text. Most previous researches focus on extractive models. In this work, we put forward a new generative model based on convolutional seq2seq architecture. A hierarchical CNN framework is much more efficient than the conventional RNN seq2seq models. We also equip our model with a copying mechanism to deal with the rare or unseen words. Additionally, we incorporate a hierarchical attention mechanism to model the keywords and key sentences simultaneously. Finally we verify our model on two real-life datasets, GigaWord and DUC corpus. The experiment results verify the effectiveness of our model as it outperforms state-of-the-art alternatives consistently and statistical significantly.

User reviews, blogs, and social media data are widely used for various types of decision-making. In this connection, Machine Learning and Natural Language Processing techniques are employed to automate the process of opinion extraction and summarization. We have studied different techniques of opinion mining and found that the extraction of opinion target and opinion words and the relation identification between them are the main tasks of state-of-the-art techniques. Furthermore, domain-independent features extraction is still a challenging task, since it is costly to manually create an extensive list of features for every domain. In this study, we tested different syntactic patterns and semantic rules for the identification of evaluative expressions containing relevant target features and opinion. We have proposed a domain-independent framework that consists of two phases. First, we extract Best Fit Examples (BFE) consisting of short sentences and candidate phrases and in the second phase, pruning is employed to filter the candidate opinion targets and opinion words. The results of the proposed model are significant.

Inexpensive remote temperature data loggers have allowed for a dramatic increase of data describing water temperature regimes. This data is used in understanding the ecological functioning of natural riverine systems and in quantifying changes in these systems. However, an increase in the quantity of yearly temperature data necessitates complex data management, efficient summarization, and an effective data-cleaning regimen. This note focuses on identifying events where data loggers failed to record correct temperatures using data from the Sauk River in Northwest Washington State as an example. By augmenting automated checks with visual comparisons against air temperature, related sites, multiple years, and available flow data, dewatering events can be more accurately and efficiently identified.

Extractive summarization aims to produce a concise version of a document by extracting information-rich sentences from the original texts. The graph-based model is an effective and efficient approach to rank sentences since it is simple and easy to use. However, its performance depends heavily on good text representation. In this paper, an integrated graph model (iGraph) for extractive text summarization is proposed. An enhanced embedding model is used to detect the inherent semantic properties at the word level, bigram level and trigram level. Words with part-of-speech (POS) tags, bigrams and trigrams were extracted to train the embedding models. Based on the enhanced embedding vectors, the similarity values between the sentences were calculated in three perspectives. The sentences in the document were treated as vertexes and the similarity between them as edges. As a result, three different types of semantic graphs were obtained for every document, with the same nodes and different edges. These three graphs were integrated into one enriched semantic graph in a naive Bayesian fashion. After that, TextRank, which is a graph-based ranking algorithm, was applied to rank the sentences, before the top scored sentences were selected for the summary according to the compression rate. Evaluated on the DUC 2002 and DUC 2004 datasets, our proposed method shows competitive performance compared to the state-of-the-art methods.

Made available in DSpace on 2018-12-11T16:52:21Z (GMT). No. of bitstreams: 0 Previous issue date: 2018-03-07 With the advent of the Internet of Things, billions of objects or devices are inserted into the global computer network, generating and processing data at a volume never imagined before. This paper proposes a way to collect and process local data through a data fusion technology called summarization. The main feature of the proposal is the local data fusion, through parameters provided by the application, ensuring the quality of data collected by the sensor node. In the evaluation, the sensor node was compared when performing the data summary with another that performed a continuous recording of the collected data. Two sets of nodes were created, one with a sensor node that analyzed the luminosity of the room, which in this case obtained a reduction of 97% in the volume of data generated, and another set that analyzed the temperature of the room, obtaining a reduction of 80% in the data volume. Through these tests, it has been proven that the local data fusion at the node can be used to reduce the volume of data generated, consequently decreasing the volume of messages generated by IoT environments. Department of Statistics Applied Mathematics and Computing Institute of Geosciences and Exact Sciences UNESP—University Estadual Paulista Department of Informatics and Statistics Federal University of Santa Catarina Computer and Networking Federal Institute of São Paulo (IFSP) Department of Statistics Applied Mathematics and Computing Institute of Geosciences and Exact Sciences UNESP—University Estadual Paulista

Many text mining tasks such as text retrieval, text summarization, and text comparisons depend on the extraction of representative keywords from the main text. Most existing keyword extraction algorithms are based on discrete bag-of-words type of word representation of the text. In this paper, we propose a patent keyword extraction algorithm (PKEA) based on the distributed Skip-gram model for patent classification. We also develop a set of quantitative performance measures for keyword extraction evaluation based on information gain and cross-validation, based on Support Vector Machine (SVM) classification, which are valuable when human-annotated keywords are not available. We used a standard benchmark dataset and a homemade patent dataset to evaluate the performance of PKEA. Our patent dataset includes 2500 patents from five distinct technological fields related to autonomous cars (GPS systems, lidar systems, object recognition systems, radar systems, and vehicle control systems). We compared our method with Frequency, Term Frequency-Inverse Document Frequency (TF-IDF), TextRank and Rapid Automatic Keyword Extraction (RAKE). The experimental results show that our proposed algorithm provides a promising way to extract keywords from patent texts for patent classification.

Text summarization namely, automatically generating a short summary of a given document, is a difficult task in natural language processing. Nowadays, deep learning as a new technique has gradually been deployed for text summarization, but there is still a lack of large-scale high quality datasets for this technique. In this paper, we proposed a novel deep learning method to identify high quality document–summary pairs for building a large-scale pairs dataset. Concretely, a long short-term memory (LSTM)-based model was designed to measure the quality of document–summary pairs. In order to leverage information across all parts of each document, we further proposed an improved LSTM-based model by removing the forget gate in the LSTM unit. Experiments conducted on the training set and the test set built upon Sina Weibo (a Chinese microblog website similar to Twitter) showed that the LSTM-based models significantly outperformed baseline models with regard to the area under receiver operating characteristic curve (AUC) value.

"Public legal information from all countries and international institutions is part of the common heritage of humanity. Maximizing access to this information promotes justice and the rule of law". In accordance with the aforementioned declaration on Free Access to Law by Legal information institutes of the world, a plethora of legal information is available through the Internet, while the provision of legal information has never before been easier. Given that law is accessed by a much wider group of people, the majority of whom are not legally trained or qualified, diversification techniques, should be employed in the context of legal information retrieval, as to increase user satisfaction. We address diversification of results in legal search by adopting several state of the art methods from the web search, network analysis and text summarization domains. We provide an exhaustive evaluation of the methods, using a standard data set from the Common Law domain that we subjectively annotated with relevance judgments for this purpose. Our results i) reveal that users receive broader insights across the results they get from a legal information retrieval system, ii) demonstrate that web search diversification techniques outperform other approaches (e.g., summarization-based, graph-based methods) in the context of legal diversification and iii) offer balance boundaries between reinforcing relevant documents or sampling the information space around the legal query. Data Set: https://github.com/mkoniari/LegalDivEval

Hyperspectral imaging technologies in the food and agricultural area have been evolving rapidly over the past 15 years owing to tremendous interest from both academic and industrial fields. Line-scan hyperspectral imaging is a major method that has been intensively researched and developed using different physical principles (e.g., reflectance, transmittance, fluorescence, Raman, and spatially resolved spectroscopy) and wavelength regions (e.g., visible (VIS), near infrared (NIR), and short-wavelength infrared (SWIR)). Line-scan hyperspectral imaging systems are mainly developed and used for surface inspection of food and agricultural products using area or line light sources. Some of these systems can also be configured to conduct spatially resolved spectroscopy measurements for internal or subsurface food inspection using point light sources. This paper reviews line-scan hyperspectral imaging techniques, with introduction, demonstration, and summarization of existing and emerging techniques for food and agricultural applications. The main topics include related spectroscopy techniques, line-scan measurement methods, hardware components and systems, system calibration methods, and spectral and image analysis techniques. Applications in food safety and quality are also presented to reveal current practices and future trends of line-scan hyperspectral imaging techniques.