Document understanding using data mining and machine learning techniques

With the explosive growth of the volume and complexity of document data (e.g., news, blogs, web pages), it has become a necessity to semantically understand documents and deliver meaningful information to users. Areas dealing with these problems are crossing data mining, information retrieval, and machine learning. For example, document clustering and summarization are two fundamental techniques for understanding document data and have attracted much attention in recent years. Given a collection of documents, document clustering aims to partition them into different groups to provide efficient document browsing and navigation mechanisms. One unrevealed area in document clustering is that how to generate meaningful interpretation for the each document cluster resulted from the clustering process. Document summarization is another effective technique for document understanding, which generates a summary by selecting sentences that deliver the major or topic-relevant information in the original documents. How to improve the automatic summarization performance and apply it to newly emerging problems are two valuable research directions. To assist people to capture the semantics of documents effectively and efficiently, the dissertation focuses on developing effective data mining and machine learning algorithms and systems for (1) integrating document clustering and summarization to obtain meaningful document clusters with summarized interpretation, (2) improving document summarization performance and building document understanding systems to solve real-world applications, and (3) summarizing the differences and evolution of multiple document sources.

Análise sintagmática aplicada ao processo de sumarização automática de documentos do português brasileiro

This research studies the application of syntagmatic analysis of written texts in the language of Brazilian Portuguese as a methodology for the automatic creation of extractive summaries. The automation of abstracts, while linked to the area of natural language processing (PLN) is studying ways the computer can autonomously construct summaries of texts. For this we use as presupposed the idea that switch to the computer the way a language is structured, in our case the Brazilian Portuguese, it will help in the discovery of the most relevant sentences, and consequently build extractive summaries with higher informativeness. In this study, we propose the definition of a summarization method that automatically perform the syntagmatic analysis of texts and through them, to build an automatic summary. The phrases that make up the syntactic structures are then used to analyze the sentences of the text, so the count of these elements determines whether or not a sentence will compose the summary to be generated

Testu arteko koherentziazko erlazio-egitura: lehen urratsak euskaraz

[EU]Testu bat koherente egiten duten arrazoiak ulertzea oso baliagarria da testuaren beraren ulermenerako, koherentzia eta koherentzia-erlazioak testu bat edo gehiago koherente diren ondorioztatzen laguntzen baitigu. Lan honetan gai bera duten testu ezberdinen arteko koherentziazko 3 Cross Document Structure Theory edo CST (Radev, 2000) erlazio aztertu eta sailkatu dira. Hori egin ahal izateko, euskaraz idatziriko gai berari buruzko testuak segmentatzeko eta beraien arteko erlazioak etiketatzeko gidalerroak proposatzen dira. 10 testuz osaturiko corpusa etiketatu da; horietako 3 cluster bi etiketatzailek aztertu dute. Etiketatzaileen arteko adostasunaren berri ematen dugu. Koherentzia-erlazioak garatzea oso garrantzitsua da Hizkuntzaren Prozesamenduko hainbat sistementzat, hala nola, informazioa erauzteko sistementzat, itzulpen automatikoarentzat, galde-erantzun sistementzat eta laburpen automatikoarentzat. Etorkizunean CSTko erlazio guztiak corpus esanguratsuan aztertuko balira, testuen arteko koherentzia- erlazioak euskarazko testuen prozesaketa automatikoa bideratzeko lehenengo pausua litzateke hemen egindakoa.

Towards automatic tweet generation: A comparative study from the text summarization perspective in the journalism genre

In recent years, Twitter has become one of the most important microblogging services of the Web 2.0. Among the possible uses it allows, it can be employed for communicating and broadcasting information in real time. The goal of this research is to analyze the task of automatic tweet generation from a text summarization perspective in the context of the journalism genre. To achieve this, different state-of-the-art summarizers are selected and employed for producing multi-lingual tweets in two languages (English and Spanish). A wide experimental framework is proposed, comprising the creation of a new corpus, the generation of the automatic tweets, and their assessment through a quantitative and a qualitative evaluation, where informativeness, indicativeness and interest are key criteria that should be ensured in the proposed context. From the results obtained, it was observed that although the original tweets were considered as model tweets with respect to their informativeness, they were not among the most interesting ones from a human viewpoint. Therefore, relying only on these tweets may not be the ideal way to communicate news through Twitter, especially if a more personalized and catchy way of reporting news wants to be performed. In contrast, we showed that recent text summarization techniques may be more appropriate, reflecting a balance between indicativeness and interest, even if their content was different from the tweets delivered by the news providers.

Towards a unified framework for opinion retrieval, mining and summarization

The exponential increase of subjective, user-generated content since the birth of the Social Web, has led to the necessity of developing automatic text processing systems able to extract, process and present relevant knowledge. In this paper, we tackle the Opinion Retrieval, Mining and Summarization task, by proposing a unified framework, composed of three crucial components (information retrieval, opinion mining and text summarization) that allow the retrieval, classification and summarization of subjective information. An extensive analysis is conducted, where different configurations of the framework are suggested and analyzed, in order to determine which is the best one, and under which conditions. The evaluation carried out and the results obtained show the appropriateness of the individual components, as well as the framework as a whole. By achieving an improvement over 10% compared to the state-of-the-art approaches in the context of blogs, we can conclude that subjective text can be efficiently dealt with by means of our proposed framework.

Extractive text summarization: can we use the same techniques for any text?

In this paper we address two issues. The first one analyzes whether the performance of a text summarization method depends on the topic of a document. The second one is concerned with how certain linguistic properties of a text may affect the performance of a number of automatic text summarization methods. For this we consider semantic analysis methods, such as textual entailment and anaphora resolution, and we study how they are related to proper noun, pronoun and noun ratios calculated over original documents that are grouped into related topics. Given the obtained results, we can conclude that although our first hypothesis is not supported, since it has been found no evident relationship between the topic of a document and the performance of the methods employed, adapting summarization systems to the linguistic properties of input documents benefits the process of summarization.

COMPENDIUM: A text summarization system for generating abstracts of research papers

This article analyzes the appropriateness of a text summarization system, COMPENDIUM, for generating abstracts of biomedical papers. Two approaches are suggested: an extractive (COMPENDIUM E), which only selects and extracts the most relevant sentences of the documents, and an abstractive-oriented one (COMPENDIUM E–A), thus facing also the challenge of abstractive summarization. This novel strategy combines extractive information, with some pieces of information of the article that have been previously compressed or fused. Specifically, in this article, we want to study: i) whether COMPENDIUM produces good summaries in the biomedical domain; ii) which summarization approach is more suitable; and iii) the opinion of real users towards automatic summaries. Therefore, two types of evaluation were performed: quantitative and qualitative, for evaluating both the information contained in the summaries, as well as the user satisfaction. Results show that extractive and abstractive-oriented summaries perform similarly as far as the information they contain, so both approaches are able to keep the relevant information of the source documents, but the latter is more appropriate from a human perspective, when a user satisfaction assessment is carried out. This also confirms the suitability of our suggested approach for generating summaries following an abstractive-oriented paradigm.

Application of Text Summarization techniques to the Geographical Information Retrieval task

Automatic Text Summarization has been shown to be useful for Natural Language Processing tasks such as Question Answering or Text Classification and other related fields of computer science such as Information Retrieval. Since Geographical Information Retrieval can be considered as an extension of the Information Retrieval field, the generation of summaries could be integrated into these systems by acting as an intermediate stage, with the purpose of reducing the document length. In this manner, the access time for information searching will be improved, while at the same time relevant documents will be also retrieved. Therefore, in this paper we propose the generation of two types of summaries (generic and geographical) applying several compression rates in order to evaluate their effectiveness in the Geographical Information Retrieval task. The evaluation has been carried out using GeoCLEF as evaluation framework and following an Information Retrieval perspective without considering the geo-reranking phase commonly used in these systems. Although single-document summarization has not performed well in general, the slight improvements obtained for some types of the proposed summaries, particularly for those based on geographical information, made us believe that the integration of Text Summarization with Geographical Information Retrieval may be beneficial, and consequently, the experimental set-up developed in this research work serves as a basis for further investigations in this field.

A novel concept-level approach for ultra-concise opinion summarization

The Web 2.0 has resulted in a shift as to how users consume and interact with the information, and has introduced a wide range of new textual genres, such as reviews or microblogs, through which users communicate, exchange, and share opinions. The exploitation of all this user-generated content is of great value both for users and companies, in order to assist them in their decision-making processes. Given this context, the analysis and development of automatic methods that can help manage online information in a quicker manner are needed. Therefore, this article proposes and evaluates a novel concept-level approach for ultra-concise opinion abstractive summarization. Our approach is characterized by the integration of syntactic sentence simplification, sentence regeneration and internal concept representation into the summarization process, thus being able to generate abstractive summaries, which is one the most challenging issues for this task. In order to be able to analyze different settings for our approach, the use of the sentence regeneration module was made optional, leading to two different versions of the system (one with sentence regeneration and one without). For testing them, a corpus of 400 English texts, gathered from reviews and tweets belonging to two different domains, was used. Although both versions were shown to be reliable methods for generating this type of summaries, the results obtained indicate that the version without sentence regeneration yielded to better results, improving the results of a number of state-of-the-art systems by 9%, whereas the version with sentence regeneration proved to be more robust to noisy data.

Abstractive Long Document Summarization: Studio e Sperimentazione di Modelli Generativi Retrieval-Augmented

In questa tesi si trattano lo studio e la sperimentazione di un modello generativo retrieval-augmented, basato su Transformers, per il task di Abstractive Summarization su lunghe sentenze legali. La sintesi automatica del testo (Automatic Text Summarization) è diventata un task di Natural Language Processing (NLP) molto importante oggigiorno, visto il grandissimo numero di dati provenienti dal web e banche dati. Inoltre, essa permette di automatizzare un processo molto oneroso per gli esperti, specialmente nel settore legale, in cui i documenti sono lunghi e complicati, per cui difficili e dispendiosi da riassumere. I modelli allo stato dell’arte dell’Automatic Text Summarization sono basati su soluzioni di Deep Learning, in particolare sui Transformers, che rappresentano l’architettura più consolidata per task di NLP. Il modello proposto in questa tesi rappresenta una soluzione per la Long Document Summarization, ossia per generare riassunti di lunghe sequenze testuali. In particolare, l’architettura si basa sul modello RAG (Retrieval-Augmented Generation), recentemente introdotto dal team di ricerca Facebook AI per il task di Question Answering. L’obiettivo consiste nel modificare l’architettura RAG al fine di renderla adatta al task di Abstractive Long Document Summarization. In dettaglio, si vuole sfruttare e testare la memoria non parametrica del modello, con lo scopo di arricchire la rappresentazione del testo di input da riassumere. A tal fine, sono state sperimentate diverse configurazioni del modello su diverse tipologie di esperimenti e sono stati valutati i riassunti generati con diverse metriche automatiche.

Evidence for divided automatic attention

A long-standing debate in the literature is whether attention can form two or more independent spatial foci in addition to the well-known unique spatial focus. There is evidence that voluntary visual attention divides in space. The possibility that this also occurs for automatic visual attention was investigated here. Thirty-six female volunteers were tested. In each trial, a prime stimulus was presented in the left or right visual hemifield. This stimulus was characterized by the blinking of a superior, middle or inferior ring, the blinking of all these rings, or the blinking of the superior and inferior rings. A target stimulus to which the volunteer should respond with the same side hand or a target stimulus to which she should not respond was presented 100 ms later in a primed location, a location between two primed locations or a location in the contralateral hemifield. Reaction time to the positive target stimulus in a primed location was consistently shorter than reaction time in the horizontally corresponding contralateral location. This attentional effect was significantly smaller or absent when the positive target stimulus appeared in the middle location after the double prime stimulus. These results suggest that automatic visual attention can focus on two separate locations simultaneously, to some extent sparing the region in between.

Characterizing chaotic melodies in automatic music composition

In this paper, we initially present an algorithm for automatic composition of melodies using chaotic dynamical systems. Afterward, we characterize chaotic music in a comprehensive way as comprising three perspectives: musical discrimination, dynamical influence on musical features, and musical perception. With respect to the first perspective, the coherence between generated chaotic melodies (continuous as well as discrete chaotic melodies) and a set of classical reference melodies is characterized by statistical descriptors and melodic measures. The significant differences among the three types of melodies are determined by discriminant analysis. Regarding the second perspective, the influence of dynamical features of chaotic attractors, e.g., Lyapunov exponent, Hurst coefficient, and correlation dimension, on melodic features is determined by canonical correlation analysis. The last perspective is related to perception of originality, complexity, and degree of melodiousness (Euler's gradus suavitatis) of chaotic and classical melodies by nonparametric statistical tests. (c) 2010 American Institute of Physics. [doi: 10.1063/1.3487516]

Automatic Network Fingerprinting through Single-Node Motifs

Complex networks have been characterised by their specific connectivity patterns (network motifs), but their building blocks can also be identified and described by node-motifs-a combination of local network features. One technique to identify single node-motifs has been presented by Costa et al. (L. D. F. Costa, F. A. Rodrigues, C. C. Hilgetag, and M. Kaiser, Europhys. Lett., 87, 1, 2009). Here, we first suggest improvements to the method including how its parameters can be determined automatically. Such automatic routines make high-throughput studies of many networks feasible. Second, the new routines are validated in different network-series. Third, we provide an example of how the method can be used to analyse network time-series. In conclusion, we provide a robust method for systematically discovering and classifying characteristic nodes of a network. In contrast to classical motif analysis, our approach can identify individual components (here: nodes) that are specific to a network. Such special nodes, as hubs before, might be found to play critical roles in real-world networks.

An algorithm for computerized automatic tuning of power system stabilizers

The main objective of this paper is to relieve the power system engineers from the burden of the complex and time-consuming process of power system stabilizer (PSS) tuning. To achieve this goal, the paper proposes an automatic process for computerized tuning of PSSs, which is based on an iterative process that uses a linear matrix inequality (LMI) solver to find the PSS parameters. It is shown in the paper that PSS tuning can be written as a search problem over a non-convex feasible set. The proposed algorithm solves this feasibility problem using an iterative LMI approach and a suitable initial condition, corresponding to a PSS designed for nominal operating conditions only (which is a quite simple task, since the required phase compensation is uniquely defined). Some knowledge about the PSS tuning is also incorporated in the algorithm through the specification of bounds defining the allowable PSS parameters. The application of the proposed algorithm to a benchmark test system and the nonlinear simulation of the resulting closed-loop models demonstrate the efficiency of this algorithm. (C) 2009 Elsevier Ltd. All rights reserved.

Automatic tuning method for the design of supplementary damping controllers for flexible alternating current transmission system devices

The design of supplementary damping controllers to mitigate the effects of electromechanical oscillations in power systems is a highly complex and time-consuming process, which requires a significant amount of knowledge from the part of the designer. In this study, the authors propose an automatic technique that takes the burden of tuning the controller parameters away from the power engineer and places it on the computer. Unlike other approaches that do the same based on robust control theories or evolutionary computing techniques, our proposed procedure uses an optimisation algorithm that works over a formulation of the classical tuning problem in terms of bilinear matrix inequalities. Using this formulation, it is possible to apply linear matrix inequality solvers to find a solution to the tuning problem via an iterative process, with the advantage that these solvers are widely available and have well-known convergence properties. The proposed algorithm is applied to tune the parameters of supplementary controllers for thyristor controlled series capacitors placed in the New England/New York benchmark test system, aiming at the improvement of the damping factor of inter-area modes, under several different operating conditions. The results of the linear analysis are validated by non-linear simulation and demonstrate the effectiveness of the proposed procedure.