Automatic induction of lexical features

This thesis concerns artificially intelligent natural language processing systems that are capable of learning the properties of lexical items (properties like verbal valency or inflectional class membership) autonomously while they are fulfilling their tasks for which they have been deployed in the first place. Many of these tasks require a deep analysis of language input, which can be characterized as a mapping of utterances in a given input C to a set S of linguistically motivated structures with the help of linguistic information encoded in a grammar G and a lexicon L: G + L + C → S (1) The idea that underlies intelligent lexical acquisition systems is to modify this schematic formula in such a way that the system is able to exploit the information encoded in S to create a new, improved version of the lexicon: G + L + S → L' (2) Moreover, the thesis claims that a system can only be considered intelligent if it does not just make maximum usage of the learning opportunities in C, but if it is also able to revise falsely acquired lexical knowledge. So, one of the central elements in this work is the formulation of a couple of criteria for intelligent lexical acquisition systems subsumed under one paradigm: the Learn-Alpha design rule. The thesis describes the design and quality of a prototype for such a system, whose acquisition components have been developed from scratch and built on top of one of the state-of-the-art Head-driven Phrase Structure Grammar (HPSG) processing systems. The quality of this prototype is investigated in a series of experiments, in which the system is fed with extracts of a large English corpus. While the idea of using machine-readable language input to automatically acquire lexical knowledge is not new, we are not aware of a system that fulfills Learn-Alpha and is able to deal with large corpora. To instance four major challenges of constructing such a system, it should be mentioned that a) the high number of possible structural descriptions caused by highly underspeci ed lexical entries demands for a parser with a very effective ambiguity management system, b) the automatic construction of concise lexical entries out of a bulk of observed lexical facts requires a special technique of data alignment, c) the reliability of these entries depends on the system's decision on whether it has seen 'enough' input and d) general properties of language might render some lexical features indeterminable if the system tries to acquire them with a too high precision. The cornerstone of this dissertation is the motivation and development of a general theory of automatic lexical acquisition that is applicable to every language and independent of any particular theory of grammar or lexicon. This work is divided into five chapters. The introductory chapter first contrasts three different and mutually incompatible approaches to (artificial) lexical acquisition: cue-based queries, head-lexicalized probabilistic context free grammars and learning by unification. Then the postulation of the Learn-Alpha design rule is presented. The second chapter outlines the theory that underlies Learn-Alpha and exposes all the related notions and concepts required for a proper understanding of artificial lexical acquisition. Chapter 3 develops the prototyped acquisition method, called ANALYZE-LEARN-REDUCE, a framework which implements Learn-Alpha. The fourth chapter presents the design and results of a bootstrapping experiment conducted on this prototype: lexeme detection, learning of verbal valency, categorization into nominal count/mass classes, selection of prepositions and sentential complements, among others. The thesis concludes with a review of the conclusions and motivation for further improvements as well as proposals for future research on the automatic induction of lexical features.

Behavioural analysis of mouse mutants with altered neuronal and glial genes

Im ersten Teil dieser Doktorarbeit beabsichtigte meine Arbeit, die funktionelle Beteiligung des CB1 Rezeptors, einer Hauptkomponente des neuronalen Endocannabinoid-Systems (ECS), an der Ausbildung von verschiedenen Verhaltensphänotypen mit Hilfe von konditionalen Mausmutanten, denen der CB1 Rezeptor auf verschiedenen neuronalen Unterpopulationen fehlt, aufzuschlüsseln und zu untersuchen. Verschiedene Verhaltensmodelle wurden hierzu getestet. Dabei lag der Fokus dieser Arbeit auf der CB1f/f;D1-Cre Mauslinie, welche der CB1 Rezeptor auf den D1 Rezeptor exprimierenden Neuronen des Striatums fehlt. Ich konnte zeigen, dass der Verlust des CB1 Rezeptors auf diesen Neuronen keinen Einfluss auf basale neurologische Funktionen, Gewicht, Bewegung, Exploration, Sozialverhalten, Angst und Stressbewältigung der Tiere hat, jedoch eine Beteiligung an der Entwicklung von Suchtverhalten gegeben ist. Bei Betrachtung des Kokain-induzierten Suchtverhaltens zeigten die konditionalen Mausmutanten eine reduzierte Suchtanfälligkeit sowohl im Vergleich zu Tieren mit einem totalen CB1 Rezeptor Verlust in allen Körperzellen, als auch zu genetisch unveränderten Kontrollmäusen beider Linien.rnDes Weiteren zeigen die Ergebnisse dieser Studie eine große, aber gegensätzliche Beteiligung des ECS bei der Regulation von Exploration in Abhängigkeit des Verlustes des CB1 Rezeptors auf GABAergen Neuronen des Vorderhirns und kortikalen glutamatergen Neuronen, jedoch nicht auf striatalen Neuronen alleine. Zusätzlich war ich in der Lage, die Wichtigkeit des genetischen Hintergrunds von Mauslinien nicht nur auf die Ausbildung von spezifischen Verhaltensphänotypen, sondern auch auf die Genexpression zu zeigen.rnIn dem zweiten Teil dieser Arbeit, in dem ich mich auf die Funktion von Gliazellen konzentrierte, wurden ebenfalls Mausmutanten in verschiedenen Verhaltensmodellen getestet. Ein genetisches Auslöschen des NG2 Glykoproteins in Gliazellen sorgt in den Knock-out Mäusen für ein schlechteres Hörvermögen und ein reduziertes Depressionsverhalten im Vergleich zu ihren Wildtyp-Kontrollmäusen. Interessanterweise zeigten diese Tiere auch eine reduzierte Empfänglichkeit bei chemisch induzierten epileptischen Krämpfen, was eine Rolle des NG2 Glykoproteins bei der Kontrolle der glutamatergen Homöostase vorschlägt, die wahrscheinlich durch Strukturänderungen der Neuron-Glia-Synapse verursacht wird. rn

Comparative functional analysis of factors controlling glial differentiation in Drosophila and mouse

The present study is a comparative functional analysis of three factors controlling glial differentiation in mouse (Fyn Src kinase, hnRNPF/H and NG2) and their homologues in Drosophila (Src42A and 64B, Glorund and Kon-tiki (Kon)). In Drosophila, mutations in any of these genes were not associated with major embryonic neurodevelopmental phenotypes. Src kinases and Glorund were shown to be ubiquitously expressed, whereas kon mRNA showed selective expression in muscles as well as in central and peripheral glia. Kon was also shown to be expressed in L3 larvae with high levels of protein accumulation at the neuromuscular junction (NMJ) and in muscles in the form of speckles. Knockdown of kon in glia resulted in NMJ phenotypes, mainly characterized by a significant increase in bouton number and a reduction in α-Konecto staining intensity at the NMJ. From the three glial layers ensheathing the peripheral nervous system, subperineurial glial showed to be the one contributing the most to kon knockdown dependent NMJ phenotypes, while perineurial glia only had a minor role. The knockdown of kon in glia also showed to affect Glutamate receptor subunit (α-GluRIIA) clustering in the postsynapse, same as microtubule arrangement in the presynapse, as seen by α-Futsch pattern interruptions and alterations. kon knockdown in glia also resulted in impaired axonal transport, as seen by the accumulation of Bruchpilot-positive vesicles along the nerves, abnormal formation of neuronal derived protrusions and swellings, filled with vacuole-like structures. Glia number along the peripheral nerves is also reduced as consequence of kon knockdown. Muscle derived Kon was shown to accumulate at the NMJ and play a role in bouton consolidation and to interfere with phagocytosis of ghost boutons. NMJ bouton and branch number was also significantly increased in Kon overexpression in glia. The overexpression of Kon in glia also resulted in a massive elongation of the ventral nerve cord, which served in a suppressor screen to identify intracellular interaction partners of Kon in glia. It was shown that Kon is processed in glia and preliminary results indicate that the metalloendopeptidase Kuzbanian (the fly homologue of ADAM10) may play a role in the shedding of Konecto. In the present work, Kon is shown as a multifunctional gene with various roles in glia-neuron and glia-neuron-muscle interaction.