State-of-the-art abstractive summarization models still suffer from the content contradiction between the summaries and the input text, which is referred to as the factual inconsistency problem. Recently, a large number of works have also been proposed to evaluate factual consistency or improve it by post-editing methods. However, these post-editing methods typically focus on replacing suspicious entities, failing to identify and modify incorrect content hidden in sentence structures. In this paper, we first verify that the correctable errors can be enriched by leveraging sentence structure pruning operation, and then we propose a post-editing method based on that. In the correction process, the pruning operation on possible errors is performed on the syntactic dependency tree with the guidance of multiple factual evaluation metrics. Experimenting on the FRANK dataset shows a great improvement in factual consistency compared with strong baselines and, when combined with them, can achieve even better performance. All the codes and data will be released on paper acceptance.

Financial reports are essential for the decision-making processes of various stakeholders, containing vast amounts of both quantitative and qualitative data. As the business world becomes increasingly intricate, stakeholders require a swift means to understand a company’s financial status. Text summarization is a useful tool in this regard, aiming to present long texts concisely without losing their essence. Given the complexity and the structured nature of financial documents, summarizing them poses a significant challenge. This paper suggests a method employing Positional Language Models (PLMs), a subset of non-neural language models that assess the sequence of tokens in input data, for financial report summarization. Our proposed method is unsupervised, eliminating the need for training and ensuring computational efficiency for lengthy documents.

Early diagnosis of skin diseases is crucial for effective treatment, preventing spread, minimizing long-term damage, managing chronic conditions, detecting underlying health issues, and promoting psychological well-being. It is often necessary to do difficult tests in order to obtain a diagnosis, which demands a lot of resources from patients as well as medical personnel. Motivated by this shortage, we introduce a user-friendly and private system for the automatic detection of skin disease from a picture. Our system employs a deep neural model for the accurate classification of a given picture to one of the predefined classes of diseases. We created a large dataset of six skin lesion types and tested different neural models. The best-performing model, which achieved 92% accuracy, was integrated into our system. The application and its web demo can be accessed at https://teledermatologis-ai.streamlit.app/.

Unfortunately, offensive language in social media is a common phenomenon nowadays. It harms many people and vulnerable groups. Therefore, automated detection of offensive language is in high demand and it is a serious challenge in multilingual domains. Various machine learning approaches combined with natural language techniques have been applied for this task lately. This paper contributes to this area from several aspects: (1) it introduces a new dataset of annotated Facebook comments in Hebrew; (2) it describes a case study with multiple supervised models and text representations for a task of offensive language detection in three languages, including two Semitic (Hebrew and Arabic) languages; (3) it reports evaluation results of cross-lingual and multilingual learning for detection of offensive content in Semitic languages; and (4) it discusses the limitations of these settings.

Automatic definition extraction from texts is an important task that has numerous applications in several natural language processing fields such as summarization, analysis of scientific texts, automatic taxonomy generation, ontology generation, concept identification, and question answering. For definitions that are contained within a single sentence, this problem can be viewed as a binary classification of sentences into definitions and non-definitions. Definitions in scientific literature can be generic (Wikipedia) or more formal (mathematical articles). In this paper, we focus on automatic detection of one-sentence definitions in mathematical texts, which are difficult to separate from surrounding text. We experiment with several data representations, which include sentence syntactic structure and word embeddings, and apply deep learning methods such as convolutional neural network (CNN) and recurrent neural network (RNN), in order to identify mathematical definitions. Our experiments demonstrate the superiority of CNN and its combination with RNN, applied on the syntactically-enriched input representation. We also present a new dataset for definition extraction from mathematical texts. We demonstrate that the use of this dataset for training learning models improves the quality of definition extraction when these models are then used for other definition datasets. Our experiments with different domains approve that mathematical definitions require special treatment, and that using cross-domain learning is inefficient.

Various Seq2Seq learning models designed for machine translation were applied for abstractive summarization task recently.Despite these models provide high ROUGE scores, they are limited to generate comprehensive summaries with a high level of abstraction due to its degenerated attention distribution.We introduce Diverse Convolutional Seq2Seq Model(DivCNN Seq2Seq) using Determinantal Point Processes methods(Micro DPPs and Macro DPPs) to produce attention distribution considering both quality and diversity.Without breaking the end to end architecture, Di-vCNN Seq2Seq achieves a higher level of comprehensiveness compared to vanilla models and strong baselines.All the reproducible codes and datasets are available online 1 .1 available at https://github.com/thinkwee/DPPCNN Sum marization Article: marseille , france the french prosecutor leading an investigation into the crash of germanwings flight 9525 insisted wednesday that he was not aware of any video footage from on board the plane .marseille prosecutor brice robin told cnn that so far no videos were used in the crash investigation ...... of a cell phone video showing the harrowing final seconds from on board germanwings flight 9525 as it crashed into the french alps .......paris match and bild reported that the video was recovered from a phone at the wreckage site ....... cnn 's frederik pleitgen , pamela boykoff , antonia mortensen , sandrine amiel and anna-maja rappard contributed to this report .CNN Seq2Seq: french prosecutor UNK robin says he was not aware of any video

Automatic headline generation is a sub-task of oneline summarization with many reported applications. Evaluation of systems generating headlines is a very challenging and undeveloped area. In this paper, we introduce a system that performs automatic evaluation of systems in terms of a quality of the generated headlines. The evaluation is performed using multiple metrics for comparing evaluated headlines with the gold standard ones or measuring their coverage of main document topics. Both types of metrics evaluate headline's content and informativeness, but not grammatical structure. The only input required by our system is a set of documents with gold standard and automatically generated headlines. The Headline Evaluation System (HEvaS) provides a user with a choice from multiple (10) metrics, then calculates the chosen metrics, performs statistical analysis of the evaluated systems and visualizes the results. Multiple headline generation systems can be evaluated at the same run. This paper describes all evaluation metrics and architecture, utilized by our system. As an evaluation of the HEvaS, we perform a case study with a couple of baseline systems and report the results. Although we tested the system on English content only, the multilingual content can also be supported.

Automated text summarization extracts essential information from original text and presents it in a predefined number of words. In this paper, we introduce an unsupervised extractive summarization approach that takes its roots from the SLIM dataset compression algorithm [1] based on the Minimum Description Length (MDL) principle [2], [3]. Our approach represents text as a transactional dataset, where sentences are transactions and normalized words are items. We use the SLIM algorithm (SLIM is not an abbreviation, it is Dutch word for 'smart') to solve the main bottleneck of the MDL computation, which is the generation of all frequent itemsets as a first step of the model construction. Additionally, we add a diversity constraint to the model in order to decrease appearance of repeated information in a summary. We introduce DRIM (Diversed SLIM) algorithm that performs unsupervised summarization, both generic and query-based, and does not require parameter tuning. We evaluate our summarizer on texts in English, but it can be easily extended to other languages.

Automated text summarization is aimed at extracting essential information from original text and presenting it in a minimal, often predefined, number of words.In this paper, we introduce a new approach for unsupervised extractive summarization, based on the Minimum Description Length (MDL) principle, using the Krimp dataset compression algorithm (Vreeken et al., 2011).Our approach represents a text as a transactional dataset, with sentences as transactions, and then describes it by itemsets that stand for frequent sequences of words.The summary is then compiled from sentences that compress (and as such, best describe) the document.The problem of summarization is reduced to the maximal coverage, following the assumption that a summary that best describes the original text, should cover most of the word sequences describing the document.We solve it by a greedy algorithm and present the evaluation results.

The problem of extractive text summarization for a collection of documents is defined as the problem of selecting a small subset of sentences so that the contents and meaning of the original document set are preserved in the best possible way.In this paper we describe the linear programming-based global optimization model to rank and extract the most relevant sentences to a summary.We introduce three different objective functions being optimized.These functions define a relevance of a sentence that is being maximized, in different manners, such as: coverage of meaningful words of a document, coverage of its bigrams, or coverage of frequent sequences of words.We supply here an overview of our system's participation in the MultiLing contest of SIGDial 2015.

MOTIVATION: Structural knowledge, extracted from the Protein Data Bank (PDB), underlies numerous potential functions and prediction methods. The PDB, however, is highly biased: many proteins have more than one entry, while entire protein families are represented by a single structure, or even not at all. The standard solution to this problem is to limit the studies to non-redundant subsets of the PDB. While alleviating biases, this solution hides the many-to-many relations between sequences and structures. That is, non-redundant datasets conceal the diversity of sequences that share the same fold and the existence of multiple conformations for the same protein. A particularly disturbing aspect of non-redundant subsets is that they hardly benefit from the rapid pace of protein structure determination, as most newly solved structures fall within existing families. RESULTS: In this study we explore the concept of redundancy-weighted datasets, originally suggested by Miyazawa and Jernigan. Redundancy-weighted datasets include all available structures and associate them (or features thereof) with weights that are inversely proportional to the number of their homologs. Here, we provide the first systematic comparison of redundancy-weighted datasets with non-redundant ones. We test three weighting schemes and show that the distributions of structural features that they produce are smoother (having higher entropy) compared with the distributions inferred from non-redundant datasets. We further show that these smoothed distributions are both more robust and more correct than their non-redundant counterparts. We suggest that the better distributions, inferred using redundancy-weighting, may improve the accuracy of knowledge-based potentials and increase the power of protein structure prediction methods. Consequently, they may enhance model-driven molecular biology.

Whereas data mining in structured data focuses on frequent data values, in semistructured and graph data mining, the issue is frequent labels and common specific topologies. The structure of the data is just as important as its content. We study the problem of discovering typical patterns of graph data, a task made difficult because of the complexity of required subtasks, especially subgraph isomorphism. In this paper, we propose a new apriori-based algorithm for mining graph data, where the basic building blocks are relatively large, disjoint paths. The algorithm is proven to be sound and complete. Empirical evidence shows practical advantages of our approach for certain categories of graphs

Whereas data mining in structured data focuses on frequent data values, in semistructured and graph data the emphasis is on frequent labels and common topologies. Here, the structure of the data is just as important as its content. When data contains large amount of different labels, both fully labeled and partially labeled data may be useful. More informative patterns can be found in the database if some of the pattern nodes can be regarded as 'unlabeled'. We study the problem of discovering typical fully and partially labeled patterns of graph data. Discovered patterns are useful in many applications, including: compact representation of source information and a road-map for browsing and querying information sources.

Whereas data mining in structured data focuses on frequent data values, in semistructured and graph data the emphasis is on frequent labels and common topologies. Here, the structure of the data is just as important as its content. We study the problem of discovering typical patterns of graph data. The discovered patterns can be useful for many applications, including: compact representation of source information and a road-map for browsing and querying information sources. Difficulties arise in the discovery task from the complexity of some of the required sub-tasks, such as sub-graph isomorphism. This paper proposes a new algorithm for mining graph data, based on a novel definition of support. Empirical evidence shows practical, as well as theoretical, advantages of our approach.