Apprentissage automatique de la morphologie : le cas des structures racine-schème

RÉSUMÉ Cette thèse porte sur le développement de méthodes algorithmiques pour découvrir automatiquement la structure morphologique des mots d'un corpus. On considère en particulier le cas des langues s'approchant du type introflexionnel, comme l'arabe ou l'hébreu. La tradition linguistique décrit la morphologie de ces langues en termes d'unités discontinues : les racines consonantiques et les schèmes vocaliques. Ce genre de structure constitue un défi pour les systèmes actuels d'apprentissage automatique, qui opèrent généralement avec des unités continues. La stratégie adoptée ici consiste à traiter le problème comme une séquence de deux sous-problèmes. Le premier est d'ordre phonologique : il s'agit de diviser les symboles (phonèmes, lettres) du corpus en deux groupes correspondant autant que possible aux consonnes et voyelles phonétiques. Le second est de nature morphologique et repose sur les résultats du premier : il s'agit d'établir l'inventaire des racines et schèmes du corpus et de déterminer leurs règles de combinaison. On examine la portée et les limites d'une approche basée sur deux hypothèses : (i) la distinction entre consonnes et voyelles peut être inférée sur la base de leur tendance à alterner dans la chaîne parlée; (ii) les racines et les schèmes peuvent être identifiés respectivement aux séquences de consonnes et voyelles découvertes précédemment. L'algorithme proposé utilise une méthode purement distributionnelle pour partitionner les symboles du corpus. Puis il applique des principes analogiques pour identifier un ensemble de candidats sérieux au titre de racine ou de schème, et pour élargir progressivement cet ensemble. Cette extension est soumise à une procédure d'évaluation basée sur le principe de la longueur de description minimale, dans- l'esprit de LINGUISTICA (Goldsmith, 2001). L'algorithme est implémenté sous la forme d'un programme informatique nommé ARABICA, et évalué sur un corpus de noms arabes, du point de vue de sa capacité à décrire le système du pluriel. Cette étude montre que des structures linguistiques complexes peuvent être découvertes en ne faisant qu'un minimum d'hypothèses a priori sur les phénomènes considérés. Elle illustre la synergie possible entre des mécanismes d'apprentissage portant sur des niveaux de description linguistique distincts, et cherche à déterminer quand et pourquoi cette coopération échoue. Elle conclut que la tension entre l'universalité de la distinction consonnes-voyelles et la spécificité de la structuration racine-schème est cruciale pour expliquer les forces et les faiblesses d'une telle approche. ABSTRACT This dissertation is concerned with the development of algorithmic methods for the unsupervised learning of natural language morphology, using a symbolically transcribed wordlist. It focuses on the case of languages approaching the introflectional type, such as Arabic or Hebrew. The morphology of such languages is traditionally described in terms of discontinuous units: consonantal roots and vocalic patterns. Inferring this kind of structure is a challenging task for current unsupervised learning systems, which generally operate with continuous units. In this study, the problem of learning root-and-pattern morphology is divided into a phonological and a morphological subproblem. The phonological component of the analysis seeks to partition the symbols of a corpus (phonemes, letters) into two subsets that correspond well with the phonetic definition of consonants and vowels; building around this result, the morphological component attempts to establish the list of roots and patterns in the corpus, and to infer the rules that govern their combinations. We assess the extent to which this can be done on the basis of two hypotheses: (i) the distinction between consonants and vowels can be learned by observing their tendency to alternate in speech; (ii) roots and patterns can be identified as sequences of the previously discovered consonants and vowels respectively. The proposed algorithm uses a purely distributional method for partitioning symbols. Then it applies analogical principles to identify a preliminary set of reliable roots and patterns, and gradually enlarge it. This extension process is guided by an evaluation procedure based on the minimum description length principle, in line with the approach to morphological learning embodied in LINGUISTICA (Goldsmith, 2001). The algorithm is implemented as a computer program named ARABICA; it is evaluated with regard to its ability to account for the system of plural formation in a corpus of Arabic nouns. This thesis shows that complex linguistic structures can be discovered without recourse to a rich set of a priori hypotheses about the phenomena under consideration. It illustrates the possible synergy between learning mechanisms operating at distinct levels of linguistic description, and attempts to determine where and why such a cooperation fails. It concludes that the tension between the universality of the consonant-vowel distinction and the specificity of root-and-pattern structure is crucial for understanding the advantages and weaknesses of this approach.

Scientific data dissemination a data catalogue to assist research organizations

The constant scientific production in the universities and in the research centers makes these organizations produce and acquire a great amount of data in a short period of time. Due to the big quantity of data, the research organizations become potentially vulnerable to the impacts on information booms that may cause a chaos as far as information management is concerned. In this context, the development of data catalogues comes up as one possible solution to the problems such as (I) the organization and (II) the data management. In the scientific scope, the data catalogues are implemented with the standard for digital and geospatial metadata and are broadly utilized in the process of producing a catalogue of scientific information. The aim of this work is to present the characteristics of access and storage of metadata in databank systems in order to improve the description and dissemination of scientific data. Relevant aspects will be considered and they should be analyzed during the stage of planning, once they can determine the success of implementation. The use of data catalogues by research organizations may be a way to promote and facilitate the dissemination of scientific data, avoid the repetition of efforts while being executed, as well as incentivate the use of collected, processed an also stored.

Maintaining Iowa's Secondary Road Environment, HR-289, 1989

The Iowa counties have been successful in maintaining a good roadway environment on our 90,000 mile secondary road system. However, county highway personnel must remain vigilant in detecting, discovering and correcting potential problems if our roads are to remain so. This presentation was developed for those county personnel who work and travel on secondary roads. The presentation discusses things county personnel can look for during their daily operations which could possibly create a potential problem. If these situations are uncovered and corrected in a timely manner, our secondary road system will be maintained in an appropriate manner.

Experimental Use of Calcium Magnesium Acetate, HR-253, 1983

The Iowa Department of Transportation Materials Laboratory personnel announced in early 1982 a process to produce a road deicer consisting of sand grains coated with calcium magnesium acetate (CMA). From that point forward the Iowa DOT began searching for a means of economically producing CMA to their concept. During 1983 and 1984 the first attempts devised for commercially producing CMA were attempted by the W.G. Block Company, Davenport, Iowa, under Iowa Highway Research Board Project HR-253. This first attempt at commercially producing CMA was accomplished by the use of concrete transit mixer equipment. Even though this procedure proved successful in the batch mixing of CMA, the need for higher production rates to reduce the cost per ton still existed. During the fall of 1984, Cedarapids Inc, Cedar Rapids, Iowa, proposed to Iowa DOT personnel the application of their technology to a continuous mixing concept for CMA. Arrangements were made for the continuous test mixing of 60 to 100 tons of CMA/sand deicer. This report covers the production effort, description and results of procedures outlined in Cedarapids Inc's proposal of September 19, 1984. The objectives of this research were: 1. To produce the CMA/sand deicer concept on a continuous mixing basis to Iowa DOT CMA concentration levels. 2. To evaluate the results of preheating the carrying vehicle (sand) prior to CMA ingredient introduction. 3. To analyze the feasibility of production equipment and procedures necessary for portable and/or stationary applications of continuous mixing concepts.

Biennial thistles : bull, musk, tall, 1981

Includes the history, description and control measures of a variety of thistles in Iowa.

Global invasion history of the fire ant Solenopsis invicta.

The fire ant Solenopsis invicta is a significant pest that was inadvertently introduced into the southern United States almost a century ago and more recently into California and other regions of the world. An assessment of genetic variation at a diverse set of molecular markers in 2144 fire ant colonies from 75 geographic sites worldwide revealed that at least nine separate introductions of S. invicta have occurred into newly invaded areas and that the main southern U.S. population is probably the source of all but one of these introductions. The sole exception involves a putative serial invasion from the southern United States to California to Taiwan. These results illustrate in stark fashion a severe negative consequence of an increasingly massive and interconnected global trade and travel system.

La Geografia i l'estudi dels boscos a Espanya

This article makes a revision of the most eminent contributions in Geography in relation to the forest study, starting with principal Spanish geographical journals, congress transactions and other bibliography about forest questions. A selection of the most eminent bibliography has been made paying special attention to three main subjects: forest and forest science history, forest and landscape description, and forest management

Mitä katsoja kertoo elokuvasta : elokuvan sisäisten elementtien tunnistaminen ja sisällönkuvailu

Summary: What the viewer conveys of the movie - content description and identification of the key elements in movies

IAMG - Asociación Internacional para las Geociencias Matemáticas

We introduce the International Association for Mathematical Geosciences (IAMG) from its origins to the present and summarize the means by which is fulfilling its mission. The IAMG was set up in Prague in 1968 with the intention of “promoting throughout the world the advancement of mathematics, statistics and informatics in the Geosciences”. It is a non-profit organisation that includes among its members researchers, teachers, professionals and students dedicated to mathematical and statistical Applications in all branches of the earth sciences. The IAMG publishes three major journals and a newsletter and organises a yearly conference, whist at the same time supporting many others. It bestows 4 biannual prizes that recognise outstanding achievements in its field, organises a distinguished lecturer series, and encourages Young researchers by providing Student research and travel funds and promoting Student societies. The IAMG celebrated its 40tn anniversary in Oslo, Norway, in 2008 and continues to be a venue for a vast array of researchers in this broad field by providing a common driving force for their activities

The Role of Third-Party Service Provider in B2B Integration

Sähköisen kaupankäynnin kasvun myötä, itsenäisten yritysten tietojärjestelmien integraation tarve on moninkertaistunut viime vuosien aikana. Yritykset ovat huomanneet, että tilaus-toimitusketjun automatisointiin tähtäävällä kokonaisvaltaisella integraatio-ratkaisulla on mahdollista päästä kattaviin kustannussäästöihin sekä tulojen kasvuun. Pääsääntöisesti yritykset kuitenkin etenevät hitaammin, integroimalla aluksi pienempiä liiketoiminnan tietojärjestelmien toimintoja. Positiivisten kokemusten perusteella yritykset ovat valmiitalaajentamaan sähköisen kaupankäynnin automatisointia myös muissa toiminnoissa. Tässä työssä keskitytään tarkastelemaan eri lähestymistapojayritystenvälisen integraation toteuttamiseen, sekä analysoimaan eri keinojen liiketoiminnallisia ja teknisiä vaikutuksia. Työ on tehty yhteistyössä UPM-KymmeneWood Oy:n kanssa, jonka tavoitteena oli saada perusteelliset tiedot yrityksenvälisestä integraatiosta ja syventää tietoja sekä integraatio-palveluita tarjoavien kolmansien osapuolten toimintatavoista että heidän tarjoamista palveluista ja niiden käyttökelpoisuudesta puutuoteteollisuudessa toimivassa yrityksessä. Käytännön osuudessa on tarkemmin esitelty integraatio-palveluita tarjoavien operaattoreiden kanssa käytyjen palaverien sekä heidän toimittamien materiaalien perusteella tehdyn tutkimustyön tuloksia, sisältäen yksityiskohtaiset kuvaukset yritystenvälisen integraation mahdollistavista palveluista.

Johdon palkkiojärjestelmätsuomalaisten pörssiyhtiöiden sijoittajaviestinnässä

Tutkimuksen tavoitteena on selvittää, mitä vaatimuksia suomalainen lainsäädäntö, IFRS-standardit, corporate governance - suosituksetja pörssin suositukset asettavat johdon palkkiosta tiedottamiselle ja kuinka laajasti tietoa tosiasiassa annetaan. Pörssiyhtiöiden tiedotuksen avoimuutta lähestytään tutkimalla, millaisia palkkioita ja kannustimia sekä päätöksentekotapoja on olemassa. Yhtiöiden omistuksen ja johtamisen eriytyessä syntyy päämies-agentti-ongelma. Johdon ja omistajien intressejä lähennetään palkkiojärjestelmien avulla. Palkitsemisen tarkoitus on kannustaa yrityksen johtoa toimimaan omistajan edun mukaisesti. Tutkimus on kvalitatiivinen ja tutkimusmenetelmä on deskriptiivinen: kuvailun ja luokittelun kautta pyritään analyysiin yritysten avoimuudesta palkkioiden suhteen. Tutkimus toteutettiinvertaamalla yhtiöiden Internet-sivuja ja vuosikertomuksia vuodelta 2004 määritettyihin kriteereihin. Tutkimuksen aineistona toimivat Helsingin arvopaperipörssin päälistan yhtiöt. Tutkimuksessa havaittiin, että palkkauksen muotoa tärkeämpää on palkitsemisen kokonaisuus ja siihen liittyvä päätöksenteko. Suomalaiset pörssiyhtiöt tiedottavat johdon palkkioista kohtuullisen hyvin, mutta erityisesti palkkioiden perusteista voisi kertoa tarkemminkin. Internetin tarjoamia mahdollisuuksia hyödynnetään kiitettävästi. Liike-vaihdoltaan suurimmatyritykset tiedottavat kautta linjan pienempiä tarkemmin.

La enseñanza de estrategias de resolución de problemas matemáticos en la ESO: un ejemplo concreto

Within this article is included a description and justification of a didactical sequence to improve the learning of problem solving strategies of Secondary Education students. The design of the didactical sequence is guided by four principles: a) to context the activities and problems to solve in students’ every day practices; b) to use teaching methods that show the problem solving processes; c) to design didactical materials in order to guide strategies such as selection, organisation and control; d) to facilitate the co-operative learning of problem solving strategies.

Regioselectividad en fullerenos, una visión computacional

Fullerenes are wonderful molecules that exhibit a wealth of interesting properties, which promises a major role for this family in the future in the (bio)medicine and nanotechnology fields. However, there are still many unknowns such as for instance the reactivity of these molecules and how this is affected by metals. These aspects are difficult to apprehend by experiments alone, for which reason there is a widespread interest in the description and simulation of them by theoretical chemistry. Here we briefly describe our results from the past five years and their future impact