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.

Radiochemical aspects of production and processing of radiometals for preparation of metalloradiopharmaceuticals

Radiometals play an important role in nuclear medicine as involved in diagnostic or therapeutic agents. In the present work the radiochemical aspects of production and processing of very promising radiometals of the third group of the periodic table, namely radiogallium and radiolanthanides are investigated. The 68Ge/68Ga generator (68Ge, T½ = 270.8 d) provides a cyclotron-independent source of positron-emitting 68Ga (T½ = 68 min), which can be used for coordinative labelling. However, for labelling of biomolecules via bifunctional chelators, particularly if legal aspects of production of radiopharmaceuticals are considered, 68Ga(III) as eluted initially needs to be pre-concentrated and purified. The first experimental chapter describes a system for simple and efficient handling of the 68Ge/68Ga generator eluates with a cation-exchange micro-chromatography column as the main component. Chemical purification and volume concentration of 68Ga(III) are carried out in hydrochloric acid – acetone media. Finally, generator produced 68Ga(III) is obtained with an excellent radiochemical and chemical purity in a minimised volume in a form applicable directly for the synthesis of 68Ga-labelled radiopharmaceuticals. For labelling with 68Ga(III), somatostatin analogue DOTA-octreotides (DOTATOC, DOTANOC) are used. 68Ga-DOTATOC and 68Ga-DOTANOC were successfully used to diagnose human somatostatin receptor-expressing tumours with PET/CT. Additionally, the proposed method was adapted for purification and medical utilisation of the cyclotron produced SPECT gallium radionuclide 67Ga(III). Second experimental chapter discusses a diagnostic radiolanthanide 140Nd, produced by irradiation of macro amounts of natural CeO2 and Pr2O3 in natCe(3He,xn)140Nd and 141Pr(p,2n)140Nd nuclear reactions, respectively. With this produced and processed 140Nd an efficient 140Nd/140Pr radionuclide generator system has been developed and evaluated. The principle of radiochemical separation of the mother and daughter radiolanthanides is based on physical-chemical transitions (hot-atom effects) of 140Pr following the electron capture process of 140Nd. The mother radionuclide 140Nd(III) is quantitatively absorbed on a solid phase matrix in the chemical form of 140Nd-DOTA-conjugated complexes, while daughter nuclide 140Pr is generated in an ionic species. With a very high elution yield and satisfactory chemical and radiolytical stability the system could able to provide the short-lived positron-emitting radiolanthanide 140Pr for PET investigations. In the third experimental chapter, analogously to physical-chemical transitions after the radioactive decay of 140Nd in 140Pr-DOTA, the rapture of the chemical bond between a radiolanthanide and the DOTA ligand, after the thermal neutron capture reaction (Szilard-Chalmers effect) was evaluated for production of the relevant radiolanthanides with high specific activity at TRIGA II Mainz nuclear reactor. The physical-chemical model was developed and first quantitative data are presented. As an example, 166Ho could be produced with a specific activity higher than its limiting value for TRIGA II Mainz, namely about 2 GBq/mg versus 0.9 GBq/mg. While free 166Ho(III) is produced in situ, it is not forming a 166Ho-DOTA complex and therefore can be separated from the inactive 165Ho-DOTA material. The analysis of the experimental data shows that radionuclides with half-life T½ < 64 h can be produced on TRIGA II Mainz nuclear reactor, with specific activity higher than any available at irradiation of simple targets e.g. oxides.

Topologically correct intersection curves of tori and natural quadrics

This thesis provides efficient and robust algorithms for the computation of the intersection curve between a torus and a simple surface (e.g. a plane, a natural quadric or another torus), based on algebraic and numeric methods. The algebraic part includes the classification of the topological type of the intersection curve and the detection of degenerate situations like embedded conic sections and singularities. Moreover, reference points for each connected intersection curve component are determined. The required computations are realised efficiently by solving quartic polynomials at most and exactly by using exact arithmetic. The numeric part includes algorithms for the tracing of each intersection curve component, starting from the previously computed reference points. Using interval arithmetic, accidental incorrectness like jumping between branches or the skipping of parts are prevented. Furthermore, the environments of singularities are correctly treated. Our algorithms are complete in the sense that any kind of input can be handled including degenerate and singular configurations. They are verified, since the results are topologically correct and approximate the real intersection curve up to any arbitrary given error bound. The algorithms are robust, since no human intervention is required and they are efficient in the way that the treatment of algebraic equations of high degree is avoided.