SEQUENTIAL DECISIONS AND PREDICTIONS IN NATURAL LANGUAGE PROCESSING

Natural language processing has achieved great success in a wide range of ap- plications, producing both commercial language services and open-source language tools. However, most methods take a static or batch approach, assuming that the model has all information it needs and makes a one-time prediction. In this disser- tation, we study dynamic problems where the input comes in a sequence instead of all at once, and the output must be produced while the input is arriving. In these problems, predictions are often made based only on partial information. We see this dynamic setting in many real-time, interactive applications. These problems usually involve a trade-off between the amount of input received (cost) and the quality of the output prediction (accuracy). Therefore, the evaluation considers both objectives (e.g., plotting a Pareto curve). Our goal is to develop a formal understanding of sequential prediction and decision-making problems in natural language processing and to propose efficient solutions. Toward this end, we present meta-algorithms that take an existent batch model and produce a dynamic model to handle sequential inputs and outputs. Webuild our framework upon theories of Markov Decision Process (MDP), which allows learning to trade off competing objectives in a principled way. The main machine learning techniques we use are from imitation learning and reinforcement learning, and we advance current techniques to tackle problems arising in our settings. We evaluate our algorithm on a variety of applications, including dependency parsing, machine translation, and question answering. We show that our approach achieves a better cost-accuracy trade-off than the batch approach and heuristic-based decision- making approaches. We first propose a general framework for cost-sensitive prediction, where dif- ferent parts of the input come at different costs. We formulate a decision-making process that selects pieces of the input sequentially, and the selection is adaptive to each instance. Our approach is evaluated on both standard classification tasks and a structured prediction task (dependency parsing). We show that it achieves similar prediction quality to methods that use all input, while inducing a much smaller cost. Next, we extend the framework to problems where the input is revealed incremen- tally in a fixed order. We study two applications: simultaneous machine translation and quiz bowl (incremental text classification). We discuss challenges in this set- ting and show that adding domain knowledge eases the decision-making problem. A central theme throughout the chapters is an MDP formulation of a challenging problem with sequential input/output and trade-off decisions, accompanied by a learning algorithm that solves the MDP.

THE ROLE OF RULES, EXAMPLES AND INDIVIDUAL DIFFERENCES IN THE ACQUISITION OF DECLARATIVE AND PROCEDURAL SECOND LANGUAGE KNOWLEDGE

The relevance of explicit instruction has been well documented in SLA research. Despite numerous positive findings, however, the issue continues to engage scholars worldwide. One issue that was largely neglected in previous empirical studies - and one that may be crucial for the effectiveness of explicit instruction - is the timing and integration of rules and practice. The present study investigated the extent to which grammar explanation (GE) before practice, grammar explanation during practice, and individual differences impact the acquisition of L2 declarative and procedural knowledge of two grammatical structures in Spanish. In this experiment, 128 English-speaking learners of Spanish were randomly assigned to four experimental treatments and completed comprehension-based task-essential practice for interpreting object-verb (OV) and ser/estar (SER) sentences in Spanish. Results confirmed the predicted importance of timing of GE: participants who received GE during practice were more likely to develop and retain their knowledge successfully. Results further revealed that the various combinations of rules and practice posed differential task demands on the learners and consequently drew on language aptitude and WM to a different extent. Since these correlations between individual differences and learning outcomes were the least observed in the conditions that received GE during practice, we argue that the suitable integration of rules and practice ameliorated task demands, reducing the burden on the learner, and accordingly mitigated the role of participants’ individual differences. Finally, some evidence also showed that the comprehension practice that participants received for the two structures was not sufficient for the formation of solid productive knowledge, but was more effective for the OV than for the SER construction.

Robotic Sensing and Manipulation of Deformable Linear Objects with Learning-based methods

Nowadays robotic applications are widespread and most of the manipulation tasks are efficiently solved. However, Deformable-Objects (DOs) still represent a huge limitation for robots. The main difficulty in DOs manipulation is dealing with the shape and dynamics uncertainties, which prevents the use of model-based approaches (since they are excessively computationally complex) and makes sensory data difficult to interpret. This thesis reports the research activities aimed to address some applications in robotic manipulation and sensing of Deformable-Linear-Objects (DLOs), with particular focus to electric wires. In all the works, a significant effort was made in the study of an effective strategy for analyzing sensory signals with various machine learning algorithms. In the former part of the document, the main focus concerns the wire terminals, i.e. detection, grasping, and insertion. First, a pipeline that integrates vision and tactile sensing is developed, then further improvements are proposed for each module. A novel procedure is proposed to gather and label massive amounts of training images for object detection with minimal human intervention. Together with this strategy, we extend a generic object detector based on Convolutional-Neural-Networks for orientation prediction. The insertion task is also extended by developing a closed-loop control capable to guide the insertion of a longer and curved segment of wire through a hole, where the contact forces are estimated by means of a Recurrent-Neural-Network. In the latter part of the thesis, the interest shifts to the DLO shape. Robotic reshaping of a DLO is addressed by means of a sequence of pick-and-place primitives, while a decision making process driven by visual data learns the optimal grasping locations exploiting Deep Q-learning and finds the best releasing point. The success of the solution leverages on a reliable interpretation of the DLO shape. For this reason, further developments are made on the visual segmentation.

Detecting social patterns within 20th century documentary photos: a deep learning based approach

The job of a historian is to understand what happened in the past, resorting in many cases to written documents as a firsthand source of information. Text, however, does not amount to the only source of knowledge. Pictorial representations, in fact, have also accompanied the main events of the historical timeline. In particular, the opportunity of visually representing circumstances has bloomed since the invention of photography, with the possibility of capturing in real-time the occurrence of a specific events. Thanks to the widespread use of digital technologies (e.g. smartphones and digital cameras), networking capabilities and consequent availability of multimedia content, the academic and industrial research communities have developed artificial intelligence (AI) paradigms with the aim of inferring, transferring and creating new layers of information from images, videos, etc. Now, while AI communities are devoting much of their attention to analyze digital images, from an historical research standpoint more interesting results may be obtained analyzing analog images representing the pre-digital era. Within the aforementioned scenario, the aim of this work is to analyze a collection of analog documentary photographs, building upon state-of-the-art deep learning techniques. In particular, the analysis carried out in this thesis aims at producing two following results: (a) produce the date of an image, and, (b) recognizing its background socio-cultural context,as defined by a group of historical-sociological researchers. Given these premises, the contribution of this work amounts to: (i) the introduction of an historical dataset including images of “Family Album” among all the twentieth century, (ii) the introduction of a new classification task regarding the identification of the socio-cultural context of an image, (iii) the exploitation of different deep learning architectures to perform the image dating and the image socio-cultural context classification.

Machine learning as a service for high energy physics (MLaaS4HEP): a service for ML-based data analyses

With the CERN LHC program underway, there has been an acceleration of data growth in the High Energy Physics (HEP) field and the usage of Machine Learning (ML) in HEP will be critical during the HL-LHC program when the data that will be produced will reach the exascale. ML techniques have been successfully used in many areas of HEP nevertheless, the development of a ML project and its implementation for production use is a highly time-consuming task and requires specific skills. Complicating this scenario is the fact that HEP data is stored in ROOT data format, which is mostly unknown outside of the HEP community. The work presented in this thesis is focused on the development of a ML as a Service (MLaaS) solution for HEP, aiming to provide a cloud service that allows HEP users to run ML pipelines via HTTP calls. These pipelines are executed by using the MLaaS4HEP framework, which allows reading data, processing data, and training ML models directly using ROOT files of arbitrary size from local or distributed data sources. Such a solution provides HEP users non-expert in ML with a tool that allows them to apply ML techniques in their analyses in a streamlined manner. Over the years the MLaaS4HEP framework has been developed, validated, and tested and new features have been added. A first MLaaS solution has been developed by automatizing the deployment of a platform equipped with the MLaaS4HEP framework. Then, a service with APIs has been developed, so that a user after being authenticated and authorized can submit MLaaS4HEP workflows producing trained ML models ready for the inference phase. A working prototype of this service is currently running on a virtual machine of INFN-Cloud and is compliant to be added to the INFN Cloud portfolio of services.

Intermediate linguistic task fine-tuning on Multi-lingual models

State-of-the-art NLP systems are generally based on the assumption that the underlying models are provided with vast datasets to train on. However, especially when working in multi-lingual contexts, datasets are often scarce, thus more research should be carried out in this field. This thesis investigates the benefits of introducing an additional training step when fine-tuning NLP models, named Intermediate Training, which could be exploited to augment the data used for the training phase. The Intermediate Training step is applied by training models on NLP tasks that are not strictly related to the target task, aiming to verify if the models are able to leverage the learned knowledge of such tasks. Furthermore, in order to better analyze the synergies between different categories of NLP tasks, experimentations have been extended also to Multi-Task Training, in which the model is trained on multiple tasks at the same time.

Analysis of human posterior EEG alpha frequency as a function of implicit learning of probabilistic lateralized stimuli in a visual detection task

Alpha oscillatory activity has long been associated with perceptual and cognitive processes related to attention control. The aim of this study is to explore the task-dependent role of alpha frequency in a lateralized visuo-spatial detection task. Specifically, the thesis focuses on consolidating the scientific literature's knowledge about the role of alpha frequency in perceptual accuracy, and deepening the understanding of what determines trial-by-trial fluctuations of alpha parameters and how these fluctuations influence overall task performance. The hypotheses, confirmed empirically, were that different implicit strategies are put in place based on the task context, in order to maximize performance with optimal resource distribution (namely alpha frequency, associated positively with performance): “Lateralization” of the attentive resources towards one hemifield should be associated with higher alpha frequency difference between contralateral and ipsilateral hemisphere; “Distribution” of the attentive resources across hemifields should be associated with lower alpha frequency difference between hemispheres; These strategies, used by the participants according to their brain capabilities, have proven themselves adaptive or maladaptive depending on the different tasks to which they have been set: "Distribution" of the attentive resources seemed to be the best strategy when the distribution probability between hemifields was balanced: i.e. the neutral condition task. "Lateralization" of the attentive resources seemed to be more effective when the distribution probability between hemifields was biased towards one hemifield: i.e., the biased condition task.

Graph-based Vision-Language Model for Multi-Modal Retrieval of Fashion Products

Artificial Intelligence is reshaping the field of fashion industry in different ways. E-commerce retailers exploit their data through AI to enhance their search engines, make outfit suggestions and forecast the success of a specific fashion product. However, it is a challenging endeavour as the data they possess is huge, complex and multi-modal. The most common way to search for fashion products online is by matching keywords with phrases in the product's description which are often cluttered, inadequate and differ across collections and sellers. A customer may also browse an online store's taxonomy, although this is time-consuming and doesn't guarantee relevant items. With the advent of Deep Learning architectures, particularly Vision-Language models, ad-hoc solutions have been proposed to model both the product image and description to solve this problems. However, the suggested solutions do not exploit effectively the semantic or syntactic information of these modalities, and the unique qualities and relations of clothing items. In this work of thesis, a novel approach is proposed to address this issues, which aims to model and process images and text descriptions as graphs in order to exploit the relations inside and between each modality and employs specific techniques to extract syntactic and semantic information. The results obtained show promising performances on different tasks when compared to the present state-of-the-art deep learning architectures.

Sviluppo di Retrieval-based Chatbot per l'Italiano con Transformer

L’Intelligenza Artificiale negli ultimi anni sta plasmando il futuro dell’umanità in quasi tutti i settori. È già il motore principale di diverse tecnologie emergenti come i big data, la robotica e l’IoT e continuerà ad agire come innovatore tecnologico nel futuro prossimo. Le recenti scoperte e migliorie sia nel campo dell’hardware che in quello matematico hanno migliorato l’efficienza e ridotto i tempi di esecuzione dei software. È in questo contesto che sta evolvendo anche il Natural Language Processing (NLP), un ramo dell’Intelligenza Artificiale che studia il modo in cui fornire ai computer l'abilità di comprendere un testo scritto o parlato allo stesso modo in cui lo farebbe un essere umano. Le ambiguità che distinguono la lingua naturale dalle altre rendono ardui gli studi in questo settore. Molti dei recenti sviluppi algoritmici su NLP si basano su tecnologie inventate decenni fa. La ricerca in questo settore è quindi in continua evoluzione. Questa tesi si pone l'obiettivo di sviluppare la logica di una chatbot help-desk per un'azienda privata. Lo scopo è, sottoposta una domanda da parte di un utente, restituire la risposta associata presente in una collezione domande-risposte. Il problema che questa tesi affronta è sviluppare un modello di NLP in grado di comprendere il significato semantico delle domande in input, poiché esse possono essere formulate in molteplici modi, preservando il contenuto semantico a discapito della sintassi. A causa delle ridotte dimensioni del dataset italiano proprietario su cui testare il modello chatbot, sono state eseguite molteplici sperimentazioni su un ulteriore dataset italiano con task affine. Attraverso diversi approcci di addestramento, tra cui apprendimento metrico, sono state raggiunte alte accuratezze sulle più comuni metriche di valutazione, confermando le capacità del modello proposto e sviluppato.

Ecological Literacy And Beyond: Problem-based Learning For Future Professionals.

Ecological science contributes to solving a broad range of environmental problems. However, lack of ecological literacy in practice often limits application of this knowledge. In this paper, we highlight a critical but often overlooked demand on ecological literacy: to enable professionals of various careers to apply scientific knowledge when faced with environmental problems. Current university courses on ecology often fail to persuade students that ecological science provides important tools for environmental problem solving. We propose problem-based learning to improve the understanding of ecological science and its usefulness for real-world environmental issues that professionals in careers as diverse as engineering, public health, architecture, social sciences, or management will address. Courses should set clear learning objectives for cognitive skills they expect students to acquire. Thus, professionals in different fields will be enabled to improve environmental decision-making processes and to participate effectively in multidisciplinary work groups charged with tackling environmental issues.

Biologically Plausible Connectionist Prediction of Natural Language Thematic Relations

In Natural Language Processing (NLP) symbolic systems, several linguistic phenomena, for instance, the thematic role relationships between sentence constituents, such as AGENT, PATIENT, and LOCATION, can be accounted for by the employment of a rule-based grammar. Another approach to NLP concerns the use of the connectionist model, which has the benefits of learning, generalization and fault tolerance, among others. A third option merges the two previous approaches into a hybrid one: a symbolic thematic theory is used to supply the connectionist network with initial knowledge. Inspired on neuroscience, it is proposed a symbolic-connectionist hybrid system called BIO theta PRED (BIOlogically plausible thematic (theta) symbolic-connectionist PREDictor), designed to reveal the thematic grid assigned to a sentence. Its connectionist architecture comprises, as input, a featural representation of the words (based on the verb/noun WordNet classification and on the classical semantic microfeature representation), and, as output, the thematic grid assigned to the sentence. BIO theta PRED is designed to ""predict"" thematic (semantic) roles assigned to words in a sentence context, employing biologically inspired training algorithm and architecture, and adopting a psycholinguistic view of thematic theory.

Use of Medical Subject Headings (MeSH) in Portuguese for categorizing web-based healthcare content

Introduction: Internet users are increasingly using the worldwide web to search for information relating to their health. This situation makes it necessary to create specialized tools capable of supporting users in their searches. Objective: To apply and compare strategies that were developed to investigate the use of the Portuguese version of Medical Subject Headings (MeSH) for constructing an automated classifier for Brazilian Portuguese-language web-based content within or outside of the field of healthcare, focusing on the lay public. Methods: 3658 Brazilian web pages were used to train the classifier and 606 Brazilian web pages were used to validate it. The strategies proposed were constructed using content-based vector methods for text classification, such that Naive Bayes was used for the task of classifying vector patterns with characteristics obtained through the proposed strategies. Results: A strategy named InDeCS was developed specifically to adapt MeSH for the problem that was put forward. This approach achieved better accuracy for this pattern classification task (0.94 sensitivity, specificity and area under the ROC curve). Conclusions: Because of the significant results achieved by InDeCS, this tool has been successfully applied to the Brazilian healthcare search portal known as Busca Saude. Furthermore, it could be shown that MeSH presents important results when used for the task of classifying web-based content focusing on the lay public. It was also possible to show from this study that MeSH was able to map out mutable non-deterministic characteristics of the web. (c) 2010 Elsevier Inc. All rights reserved.

Effect of bandwidth knowledge of results on the learning of a grip force control task

An experiment was conducted to investigate the persistence of the effect of ""bandwidth knowledge of results (KR)"" manipulated during the learning phase of performing a manual force-control task. The experiment consisted of two phases, an acquisition phase with the goal of maintaining 60% maximum force in 30 trials, and a second phase with the objective of maintaining 40% of maximum force in 20 further trials. There were four bandwidths of KR: when performance error exceeded 5, 10, or 15% of the target, and a control group (0% bandwidth). Analysis showed that 5, 10, and 15% bandwidth led to better performance than 0% bandwidth KR at the beginning of the second phase and persisted during the extended trials.

Learning Profile Identification Based on the Analysis of the User Context of Interaction

One of the e-learning environment goal is to attend the individual needs of students during the learning process. The adaptation of contents, activities and tools into different visualization or in a variety of content types is an important feature of this environment, bringing to the user the sensation that there are suitable workplaces to his profile in the same system. Nevertheless, it is important the investigation of student behaviour aspects, considering the context where the interaction happens, to achieve an efficient personalization process. The paper goal is to present an approach to identify the student learning profile analyzing the context of interaction. Besides this, the learning profile could be analyzed in different dimensions allows the system to deal with the different focus of the learning.

Mummy! Ball! Fish!: Why English-learning children produce nouns earlier than verbs

When English-learning children begin using words the majority of their early utterances (around 80%) are nouns. Compared to nouns, there is a paucity of verbs or non-verb relational words, such as 'up' meaning 'pick me up'. The primary explanations to account for these differences in use either argue in support of a 'cognitive account', which claims that verbs entail more cognitive complexity than nouns, or they provide evidence challenging this account. In this paper I propose an additional explanation for children's noun/verb asymmetry. Presenting a 'multi-modal account' of word-learning based on children's gesture and word combinations, I show that at the one-word stage English-learning children use gestures to express verb-like elements which leaves their words free to express noun-like elements.