Combining structured knowledge and neural language models to tackle natural language processing tasks is a recent research trend that catalyzes community attention. This integration holds a lot of potential in document summarization, especially in the biomedical domain, where the jargon and the complex facts make the overarching information truly hard to interpret. In this context, graph construction via semantic parsing plays a crucial role in unambiguously capturing the most relevant parts of a document. However, current works are limited to extracting open-domain triples, failing to model real-world n-ary and nested biomedical interactions accurately. To alleviate this issue, we present EASumm, the first framework for biomedical abstractive summarization enhanced by event graph extraction (i.e., graphical representations of medical evidence learned from scientific text), relying on dual text-graph encoders. Extensive evaluations on the CDSR dataset corroborate the importance of explicit event structures, with better or comparable performance than previous state-of-the-art systems. Finally, we offer some hints to guide future research in the field.

Though the strong evolution of knowledge learning models has characterized the last few years, the explanation of a phenomenon from text documents, called descriptive text mining, is still a difficult and poorly addressed problem. The need to work with unlabeled data, explainable approaches, unsupervised and domain independent solutions further increases the complexity of this task. Currently, existing techniques only partially solve the problem and have several limitations. In this paper, we propose a novel methodology of descriptive text mining, capable of offering accurate explanations in unsupervised settings and of quantifying the results based on their statistical significance. Considering the strong growth of patient communities on social platforms such as Facebook, we demonstrate the effectiveness of the contribution by taking the short social posts related to Esophageal Achalasia as a typical case study. Specifically, the methodology produces useful explanations about the experiences of patients and caregivers. Starting directly from the unlabeled patient’s posts, we derive correct scientific correlations among symptoms, drugs, treatments, foods and so on.

Localisation algorithms based on the estimation of the time-of-arrival of the received signal are particularly interesting when ultra-wide band (UWB) signaling is adopted for high-definition location aware applications. In this context non-line-of-sight (NLOS) propagation condition may drastically degrade the localisation accuracy if not properly recognised. We propose a new NLOS identification technique based on the analysis of UWB signals through supervised and unsupervised machine learning algorithms, which are typically adopted to extract knowledge from data according to the data mining approach. Thanks to these algorithms we can automatically generate a very reliable model that recognises if an UWB received signal has crossed obstacles (NLOS situation). The main advantage of this solution is that it extracts the model for NLOS identification directly from example waveforms gathered in the environment and does not rely on empirical tuning of parameters as required by other NLOS identification algorithms. Moreover experiments show that accurate NLOS classifiers can be extracted from measured signals either pre-classified or unclassified and even from samples algorithmically-generated from statistical models, allowing the application of the method in real scenarios without training it on real data.

Though recurrent neural networks (RNN) outperform traditional machine learning algorithms in the detection of long-term dependencies among the training instances, such as in term sequences in sentences or among values in time series, surprisingly few studies so far have deployed concrete solutions with RNNs for the stock market trading. Presumably the current difficulties of training RNNs have contributed to discourage their wide adoption.This work presents a simple but effective solution, based on a deep RNN, whose gains in trading with Dow Jones Industrial Average (DJIA) outperform the state-of-the-art, moreover the gain is 50% higher than that produced by similar feed forward deep neural networks. The trading actions are driven by the predictions of the price movements of DJIA, using simply its publicly available historical series. To improve the reliability of results with respect to the literature, we have experimented the approach on a long consecutive period of 18 years of hist orical DJIA series, from 2000 to 2017. In 8 years of trading in the test set period from 2009 to 2017, the solution has quintupled the initial capital, moreover since DJIA has on average an increasing trend, we also tested the approach with a decreasing averagely trend by simply inverting the same historical series of DJIA. In this extreme case, in which hardly any investor would risk money, the approach has more than doubled the initial capital.

Efficiently detecting conversation threads from a pool of messages, such as social network chats, emails, comments to posts, news etc., is relevant for various applications, including Web Marketing, Information Retrieval and Digital Forensics. Existing approaches focus on text similarity using keywords as features that are strongly dependent on the dataset. Therefore, dealing with new corpora requires further costly analyses conducted by experts to find out new relevant features. This paper introduces a novel method to detect threads from any type of conversational texts overcoming the issue of previously determining specific features for each dataset. To automatically determine the relevant features of messages we map each message into a three dimensional representation based on its semantic content, the social interactions in terms of sender/recipients and its timestamp; then clustering is used to detect conversation threads. In addition, we propose a supervised approach to detect conversation threads that builds a classification model which combines the above extracted features for predicting whether a pair of messages belongs to the same thread or not. Our model harnesses the distance measure of a message to a cluster representing a thread to capture the probability that a message is part of that same thread. We present our experimental results on seven datasets, pertaining to different types of messages, and demonstrate the effectiveness of our method in the detection of conversation threads, clearly outperforming the state of the art and yielding an improvement of up to a 19%.

Within text categorization and other data mining tasks, the use of suitable methods for term weighting can bring a substantial boost in effectiveness. Several term weighting methods have been presented throughout literature, based on assumptions commonly derived from observation of distribution of words in documents. For example, the idf assumption states that words appearing in many documents are usually not as important as less frequent ones. Contrarily to tf.idf and other weighting methods derived from information retrieval, schemes proposed more recently are supervised, i.e. based on knownledge of membership of training documents to categories. We propose here a supervised variant of the tf.idf scheme, based on computing the usual idf factor without considering documents of the category to be recognized, so that importance of terms frequently appearing only within it is not underestimated. A further proposed variant is additionally based on relevance frequency, considering occurrences of words within the category itself. In extensive experiments on two recurring text collections with several unsupervised and supervised weighting schemes, we show that the ones we propose generally perform better than or comparably to other ones in terms of accuracy, using two different learning methods.