Intelligent and autonomous robotic applications often require robots to have more information about their environment than provided by traditional occupancy maps. An example are semantic maps, which provide qualitative descriptions of the environment. While research in the area of semantic mapping has been performed, most robotic frameworks still offer only occupancy maps. In this thesis, a framework is developed to handle multi-layered 2D maps in ROS. The framework offers occupancy and semantic layers, but can be extended with new layer types in the future. Furthermore, an algorithm to automatically generate semantic maps from RGB-D images is presented. Software tests were performed to check if the framework fulfills all set requirements. It was shown that the requirements are accomplished. Furthermore, the semantic mapping algorithm was evaluated with different configurations in two test environments, a laboratory and a floor. While the object shapes of the generated semantic maps were not always accurate and some false detections occurred, most objects were successfully detected and placed on the semantic map. Possible ways to improve the accuracy of the mapping in the future are discussed.

The three structural possibilities marking a noun with an English article are a, the, and Ø (the absence of an article). Although these structural possibilities are simple, they encode a multitude of semantic and pragmatic functions, and it is these complex form-function interactions that this study explores and explains using a semantic map model. The semantic map that is proposed contains three dimensions, which I refer to as Grammatical Number, Referentiality, and Discourse Mode. Each of these dimensions contains a number of further semantic values or pragmatic functions - which I will label "attributes" - that are implicated in English article choice. Various semantic map versions are tested and compared with a methodological approach that uses data collected in a controlled protocol from an elicited conversational discourse. The version that performed best is used as a basis for proposing a comprehensive semantic map that includes the following dimensions and dimensional attributes: a Number dimension with 3 attributes (singular, plural, and uncountable); a Referentiality dimension with 11 attributes, including 7 referential attributes that describe kinds of identifiability (proper names, shared lexis, shared speech situation, frame, current discourse, identifiable to speaker only ["new reference"], and identifiable to neither speaker nor listener [non-specific]) as well as 4 non-referential attributes (categorization, general non-referential expressions, finite verb [verb-object] "noun incorporation", and idioms); and a Discourse Mode dimension with 4 attributes (headline, immediacy, normal, and reintroducing). This model of English articles contributes to the field of research on articles as well as to the field of English language instruction and learning. In addition, it is suggested that the methodological paradigm used to test the semantic map model may be useful as an experimental paradigm for testing semantic maps of other constructions and languages.

This thesis and appended papers present the process of tacking the problem of environment modeling for autonomous agent. More specifically, the focus of the work has been semantic mapping of warehouses. A semantic map for such purpose is expected to be layout-like and support semantics of both open spaces and infrastructure of the environment. The representation of the semantic map is required to be understandable by all involved agents (humans, AGVs and WMS.) And the process of semantic mapping is desired to lean toward full-autonomy, with minimum input requirement from human user. To that end, we studied the problem of semantic annotation over two kinds of spatial map from different modalities. We identified properties, structure, and challenges of the problem. And we have developed representations and accompanied methods, while meeting the set criteria. The overall objective of the work is “to develop and construct a layer of abstraction (models and/or decomposition) for structuring and facilitate access to salient information in the sensory data. This layer of abstraction connects high level concepts to low-level sensory pattern.” Relying on modeling and decomposition of sensory data, we present our work on abstract representation for two modalities (laser scanner and camera) in three appended papers. Feasibility and the performance of the proposed methods are evaluated over data from real warehouse. The thesis conclude with summarizing the presented technical details, and drawing the outline for future work. / Automatic Inventory and Mapping of Stock (AIMS)

xvi, 185 p. : ill. A print copy of this thesis is available through the UO Libraries. Search the library catalog for the location and call number. / This dissertation investigates constructional alternation among the English verb- at , verb- away-at , and verb- away constructions. The primary purpose is to lay a fundamental conceptual framework on the interrelation between how we perceive a situation in an external world and how we construe it as an event structure in a conceptualized world to encode it linguistically. This study suggests an Extended Semantic Map (hereafter ESM) model. It presents an in-depth analysis of the three constructions, derived from the BNC (British National Corpus), and resultative constructions in Korean and Japanese. I argue that language has conceptual bases rooted in perception and cognitive construal. Construal allows one to view the same situation in a number of alternative ways. Construal is closely related to distribution of attention, which has two main patterns: focus of attention and windowing of attention. Focus of attention is mainly based on perceptual prominence. It is placed on participants and is typically encoded in the selection and arrangement of nominals. Windowing of attention operates on cognitive prominence. It is a cognitive process to segment some relation(s) out of an event structure. It is typically encoded in predicate or adverbial expressions. I further argue that any mismatch between perceptual and cognitive prominence requires overt marking. For example, the English passive construction requires the overt marking of ' be/get + past participle,' which directs an addressee's primary focus of attention to a perceptually secondary but cognitively primary patient. It also places windowing of attention on the perceptually secondary but cognitively primary Change. Windowing and focus of attention will be used to define the X- and Y-axes of the ESM. The X-axis consists of five causal relations -- Volition, Activity, Force Transfer, Change, and State, on which attention is windowed. The Y-axis is composed of four types of configuration for the semantic roles of the participants -- Agent, Agent-Location, Agent-Theme, and Theme. The ESM visually maps relations among constructions within and across languages. It illustrates how event structures can be categorized typically as either [Activity]-windowing or [Change]-windowing. Finally, it also allows us to represent cross-linguistic differences in the available constructions for construing event structures. / Committee in charge: Eric Pederson, Chairperson, Linguistics; Scott DeLancey, Member, Linguistics; Doris Payne, Member, Linguistics; Kaori Idemaru, Outside Member, East Asian Languages & Literatures

Les ontologies offrent une modélisation des connaissances d'un domaine basée sur une hiérarchie des concepts clefs de ce domaine. Leur utilisation dans le cadre des Systèmes de Recherche d'Information (SRI), tant pour indexer les documents que pour exprimer une requête, permet notamment d'éviter les ambiguïtés du langage naturel qui pénalisent les SRI classiques. Les travaux de cette thèse portent essentiellement sur l'utilisation d'ontologies lors du processus d'appariement durant lequel les SRI ordonnent les documents d'une collection en fonction de leur pertinence par rapport à une requête utilisateur. Nous proposons de calculer cette pertinence à l'aide d'une stratégie d'agrégation de scores élémentaires entre chaque document et chaque concept de la requête. Cette agrégation, simple et intuitive, intègre un modèle de préférences dépendant de l'utilisateur et une mesure de similarité sémantique associée à l'ontologie. L'intérêt majeur de cette approche est qu'elle permet d'expliquer à l'utilisateur pourquoi notre SRI, OBIRS, estime que les documents qu'il a sélectionnés sont pertinents. Nous proposons de renforcer cette justification grâce à une visualisation originale où les résultats sont représentés par des pictogrammes, résumant leurs pertinences élémentaires, puis disposés sur une carte sémantique en fonction de leur pertinence globale. La Recherche d'Information étant un processus itératif, il est nécessaire de permettre à l'utilisateur d'interagir avec le SRI, de comprendre et d'évaluer les résultats et de le guider dans sa reformulation de requête. Nous proposons une stratégie de reformulation de requêtes conceptuelles basée sur la transposition d'une méthode éprouvée dans le cadre de SRI vectoriels. La reformulation devient alors un problème d'optimisation utilisant les retours faits par l'utilisateur sur les premiers résultats proposés comme base d'apprentissage. Nous avons développé une heuristique permettant de s'approcher d'une requête optimale en ne testant qu'un sous-espace des requêtes conceptuelles possibles. Nous montrons que l'identification efficace des concepts de ce sous-espace découle de deux propriétés qu'une grande partie des mesures de similarité sémantique vérifient, et qui suffisent à garantir la connexité du voisinage sémantique d'un concept.Les modèles que nous proposons sont validés tant sur la base de performances obtenues sur des jeux de tests standards, que sur la base de cas d'études impliquant des experts biologistes. / Domain ontologies provide a knowledge model where the main concepts of a domain are organized through hierarchical relationships. In conceptual Information Retrieval Systems (IRS), where they are used to index documents as well as to formulate a query, their use allows to overcome some ambiguities of classical IRSs based on natural language processes.One of the contributions of this study consists in the use of ontologies within IRSs, in particular to assess the relevance of documents with respect to a given query. For this matching process, a simple and intuitive aggregation approach is proposed, that incorporates user dependent preferences model on one hand, and semantic similarity measures attached to a domain ontology on the other hand. This matching strategy allows justifying the relevance of the results to the user. To complete this explanation, semantic maps are built, to help the user to grasp the results at a glance. Documents are displayed as icons that detail their elementary scores. They are organized so that their graphical distance on the map reflects their relevance to a query represented as a probe. As Information Retrieval is an iterative process, it is necessary to involve the users in the control loop of the results relevancy in order to better specify their information needs. Inspired by experienced strategies in vector models, we propose, in the context of conceptual IRS, to formalize ontology based relevance feedback. This strategy consists in searching a conceptual query that optimizes a tradeoff between relevant documents closeness and irrelevant documents remoteness, modeled through an objective function. From a set of concepts of interest, a heuristic is proposed that efficiently builds a near optimal query. This heuristic relies on two simple properties of semantic similarities that are proved to ensure semantic neighborhood connectivity. Hence, only an excerpt of the ontology dag structure is explored during query reformulation.These approaches have been implemented in OBIRS, our ontological based IRS and validated in two ways: automatic assessment based on standard collections of tests, and case studies involving experts from biomedical domain.

This study investigates expressions akin to ‘again’, which inhabit the semantic domain of repetition and restitution, from a cross-linguistic perspective. Using massively parallel corpora as the primary source of data the aim of this study is to investigate whether the encoding of repetitive and restitutive meaning is a cross-linguistically valid difference and if there are any patterns in the language specific variation of the repetitive and restitutive domain. By using Multi-Dimensional Scaling and Partitioning Around Medoids to investigate how the expressions ‘third time’, ‘second time’, ‘again’, ‘back’ and ‘return’ make up the semantic space of the domain, it was determined that the domain in question forms a continuum of meanings. This scale, named the TURN-hierarchy, is comprised of repetitive expressions like ‘third time’ to the far left, ambiguous expressions like ‘again’ in the intermediate section and restitutive expressions such as ‘return, back’ to the far right. Furthermore, the results show that repetitive and restitutive meaning is encoded differently in a majority of the sample languages, and that there is asymmetry in the encoding of repetition and restitution where repetitive meaning is privileged. Thus,it is proposed that all languages have at least one exclusively repetitive expression.

Dans le cadre de notre thèse, nous avons proposé une approche générique d’extraction automatique des événements et de leur exploitation. L’approche est organisée en quatre composantes indépendantes et réutilisables. Une première composante de prétraitement, où les textes sont nettoyés et segmentés. Au cours de la seconde étape, les événements sont extraits en sebasant sur notre algorithme AnnotEC qui dispose d’une complexité polynomiale et qui est associé à des cartes sémantiques et des ressources linguistiques dédiées. Nous avons mis en place deux nouvelles mesures de similarité SimCatégoreille et SimEvent pour regrouper les événementssimilaires dans le cadre de la troisième composante de clustering. Les annotations, ajoutées tout au long des trois premières étapes, sont exploitées au niveau de la dernière composante par le bais des fichiers de synthèse paramétrables par l’utilisateur.L’approche a été évaluée sur un corpus issu du Web 2.0. Nous avons comparé les résultats avec des méthodes d’apprentissage automatique et des méthodes linguistiques par compilation et nous avons obtenu de meilleurs résultats. / In the framework of our thesis, we proposed a generic approach for the automatic extraction of events and their exploitation. This approach is divided into four independent and reusable components. The first component of pretreatment, in which texts are cleaned and segmented. During the second stage, events are extracted based on our algorithm AnnotEC which has polynomial complexity. AnnotEC is associated with semantic maps and dedicated linguistic resources. We have proposed two new similarity measures SimCatégoreille and SimEvent to group similar events using clustering algorithms.Annotations, added throughout the first three steps, are used at the last component by summarizing files configurable by users. The approach was evaluated on a corpus of Web 2.0, we compared the obtained results with machine learning methods and linguistic compiling methods and we got good results.

Denna uppsats undersöker kontinuativer i elva germanska språk genom en kvantitiv och kvalitativ analys av korpusdata. Syftet är att undersöka om den lexikala förändring som beskrivs hos kontinuativer i svenska av Britse (2022) kan observeras i andra germanska språk. Studien ämnar även undersöka polysemin hos kontinuativer och föreslå en semantisk karta för dessa i germanska språk. För att ge ett diakront perspektiv har det numera utdöda språket gotiska inkluderats. Studien baseras till stor del på beskrivningar av faspolaritet, i synnerhet kontinuitet, i van der Auwera (2021) och Van Baar (1997), men även Panova (under revision). Resultaten visar att den lexikala förändring som förut beskrivits observeras i svenska, norskt bokmål, nynorsk och danska. I resten av de undersökta germanska språken finns svaga tendenser eller inga tendenser alls. Studien visar även på rika polysemiska nätverk, vilket följer beskrivning- arna i Van Baar (1997) och Panova (under revision). Gotiska följer tvärspråkliga mönster och presenterar även möjliga källor till komparativa partiklar vars användning överlappar kontinuativa uttryck. / This thesis examines continuatives in eleven Germanic languages in a quantitative and qualitative analysis of corpus data. The purpose is to investigate if the lexical change described in continuatives in Swedish by Britse (2022) can be observed in other Germanic languages. The study also aims to examine polysemy in continuous expressions och suggest a semantic map for these in Germanic languages. The now extinct language Gothic has been included for a diachronic perspective. The study is largely based on descriptions of phasal polarity, par- ticularly continuity, in van der Auwera (2021) and Van Baar (1997), but also Panova (under revision). The results show that the lexical change is observed in Swedish, Norwegian Bokmål, Nynorsk and Danish. In the rest of the analysed Germanic languages, it is not observed. The study shows rich polysemous networks, that follow descriptions in Van Baar (1997) and Panova (under revision). Gothic follows cross-linguistic patterns and also presents possible sources of comparative particles who’s usage overlaps that of continuatives.

Dans le cadre de notre thèse, nous avons proposé une approche générique multilingue d'extraction automatique de connaissances. Nous avons validé l‟approche sur l'extraction des événements de variations des cours pétroliers et l‟extraction des expressions temporelles liées à des référentiels. Notre approche est basée sur la constitution de plusieurs cartes sémantiques par analyse des données non structurées afin de formaliser les traces linguistiques textuelles exprimées par des catégories d'un point de vue de fouille. Nous avons mis en place un système expert permettant d‟annoter la présence des catégories en utilisant des groupes de règles. Deux algorithmes d'annotation AnnotEV et AnnotEC ont été appliqués, dans la plateforme SemanTAS. Le rappel et précision de notre système d'annotation est autour de 80%. Nous avons présenté les résultats aussi sous forme des fiches de synthèses. Nous avons validé l'aspect Multilingue de l'approche sur la langue française et arabe, et l'aspect généricité et scalabilité en testant sur plusieurs corpus de taille confédérale. / In this thesis, we proposed a multilingual generic approach for the automatic information extraction. Particularly, events extraction of price variation and temporal information extraction linked to temporal referential. Our approach is based on the constitution of several semantic maps by textual analysis in order to formalize the linguistic traces expressed by categories. We created a database for an expert system to identify and annotate information (categories and their characteristics) based on the contextual rule groups. Two algorithms AnnotEC and AnnotEV have been applied in the SemanTAS platform to validate our assumptions. We have obtained a satisfactory result; Accuracy and recall are around 80%. We presented extracted knowledge by a summary file. In order to approve the multilingual aspect of our approach, we have carried out experiments on French and Arabic. We confirmed the scalability level by the annotation of large corpus.

This study investigates the cross-linguistic encoding of bad and evil expressions. Using parallel data from the Bible corpus consisting of translations of the New Testament into 30 languages, probabilistic semantic maps have been created using Multi-Dimensional scaling. Special attention has been paid to the presence of morphological and syntactic negation withinthe domain. The results show that languages either have one broader expression that is used within the entire domain, or they have at least two expressions of which one is broader, i.e. expresses a bad state, action or character flaw, and the other one narrower, i.e. is restricted to themost evil actions or characters which require a moral agent. Languages with several expressions vary largely in how broad or restricted the expressions are within the domain. Therefore, a scalar view of the domain has been proposed, rather than dividing the domain into discrete semantic categories. In the languages where negation marking was present within the domain, it only occurred in the broader expressions. / I denna studie undersöks den tvärspråkliga kodningen av uttryck med dålig och ond. Probabilistiska semantiska kartor har skapats med hjälp av Multi-Dimensional scaling genom att använda parallel data från Bibelkorpusen som består av 30 översättningar av Nya Testamentet. Förekomsten av eventuell morfologisk och syntaktisk negation inom domänen har tillägnats särskild uppmärksamhet. Resultaten visar att de flesta språken antingen har ett bredare uttryck som används inom hela domänen, eller har minst två uttryck varav ett är bredare, dvs används för dåliga tillstånd, handlingar eller karaktärsdrag, och det andra är mer begränsad, dvs används endast för de mest onda handlingar och karaktärer som kräver en moralisk agent. Språk med flera uttryck varierar mycktet i hur breda eller begränsade uttrycken är. En representation av den semantiska domänen som en skala föreslås därför, snarare än att dela uppdomänen i diskreta semantiska kategorier. I de språken där negation förekom inom domänen fanns det endast i de bredare uttrycken.