The local structures of the iron atoms for a series of strongly coupled Fe 2 (TIED)L 4 complexes (TIED = tetraiminethylenedimacrocycle, L = axail ligand) have been investigated by K-edge X-ray Absorption Spectroscopy (XAS). These complexes include not only the well characterized iso-valence CH 3 CN complex, mixed-valence CH 3 CN and Cl − complexes, and previously reported iso-valence CO complex but also the new isolated solids of iso-valence Fe 2 TIED complexes with Cl − , Br − , imidazole, pyridine, histidine, N,N-dimethyformamide (DMF), SCN − , and CN − as axial ligands and mixed-valence complexes with Br − and imidazole as axial ligands. The average Fe-N distances for the first coordination sphere of the iron atom obtained by EXAFS analysis are 1.94, 1.94, 1.95, 1.96, 1.94, 1.93, 1.96, 1.96, 1.96, and 1.96 Å for the iso- and mixed-valence CH 3 CN and imidazole complexes and iso-valence complexes with SCN − , CN − , CO, pyridine, histidine, and DMF as axial ligands, respectively. Two-shell fitting analyses of the complexes gave average iron to the four planar coordinated nitrogen distance of 1.90, 1.91, 1.91, 1.92, and 1.92 Å for the Fe 2 (TIED)L 4 with L = DMF, pyridine, Cl − , Br − , imidazole, and histidine complexes, respectively. The average distances from the center iron to: N(DMF), 2.05; N(pyridine), 2.05; Cl − , 2.33; Br − , 2.45; N(imidazole), 2.08; and N(histidine), 2.07 Å. They are all comparable to related bond distances in the literature. The above data indicate that there is no significant difference in the average Fe-N distances between each of the iso- and mixed-valence pairs. Also different axial ligands do not cause significant impact on the average Fe-N distances from the iron atom to the four coordinated N in the TIED ligand. The threshold edge positions shift about +1 eV from the iso-valence CH 3 CN, Cl − , and Br − complexes to their corresponding mixed-valence complexes. The relatively small shift compared with the normal +2 [special characters omitted] +3 eV edge shift from Fe 2+ to Fe 3+ reflects the oxidation state change of iron from Fe 2+ to Fe 2.5+ . The edge energy of the isovalence diiron complexes with different axial ligands increases in the order of the spectrochemical series of the axial Iigands from strong to weak field Iigands as follows [58, 59]: [special characters omitted] All the complexes studied here have a weak dipole-forbidden 1s → 3d pre-edge transition. The low intensity indicates only a small distortion of the octahedral coordination geometry of the central iron atom.

We, as humans, purposely alter our facial expression to convey information during our daily interactions. However, our facial expressions can also unconsciously change based on external stimuli. In the current thesis, we focus on visual stimuli and hypothesize that our facial expression is indicative of the perceived valence -namely the pleasantness- of the former. To evaluate our hypothesis, we experiment with different neural network architectures on a 3D facial mesh-valence dataset. At first, various VAE-based architectures, adapted accordingly to operate on 3D meshes, were employed to extract representation embeddings of facial meshes. Thereafter an LSTM head was utilized to address the different sequential downstream tasks, including the valence estimation. In addition, representation disentanglement approaches were considered, aiming at representing the facial shape and expression independently. Although our experiments suggest that the facial expression is not a reliable estimator of the perceived valence, we demonstrate that mesh VAE-based architectures can be employed to extract competent mesh representations and address less ambiguous downstream tasks such as expression classification. Additionally, we observed that representation disentanglement boosts the performance in both terms of facial expression classification and valence estimation. Finally, we highlight the mesh VAE capabilities in morphing between existing meshes as well as generating novel samples. / Vi, som människor, ändrar medvetet våra ansiktsuttryck för att förmedla information under våra dagliga interaktioner. Men våra ansiktsuttryck kan också omedvetet förändras baserat på yttre stimuli. I den aktuella avhandlingen fokuserar vi på visuella stimuli och antar att vårt ansiktsuttryck är ett tecken på den upplevda valensen -nämligen behagligheten- hos den förra. För att utvärdera vår hypotes experimenterar vi med olika neurala nätverksarkitekturer på en 3D-ansikts-mesh-valensdatauppsättning. Till en början användes olika VAE-baserade arkitekturer, anpassade för att fungera på 3D-nät, för att extrahera representationsinbäddningar av ansiktsnät. Därefter användes ett LSTM-huvud för att hantera de olika sekventiella nedströmsuppgifterna, inklusive valensuppskattningen. Dessutom övervägdes metoder för att distrahera representation, som syftade till att representera ansiktsformen och uttrycket oberoende. Även om våra experiment tyder på att ansiktsuttrycket inte är en tillförlitlig estimerare av den upplevda valensen, visar vi att mesh-VAE-baserade arkitekturer kan användas för att extrahera kompetenta mesh-representationer och adressera mindre tvetydiga nedströmsuppgifter såsom uttrycksklassificering. Dessutom observerade vi att representationsupplösning ökar prestandan både vad gäller klassificering av ansiktsuttryck och valensuppskattning. Slutligen lyfter vi fram mesh-VAE-förmågan vid morphing mellan befintliga maskor samt generering av nya prover.

Les dictionnaires de valence sont utiles dans plusieurs tâches en traitement automatique des langues. Or, les dictionnaires de qualité de ce type sont créés au moins en partie manuellement; ils nécessitent donc beaucoup de ressources et sont difficiles à mettre à jour. De plus, plusieurs de ces ressources ne prennent pas en compte les différents sens des lemmes, qui sont pourtant importants puisque les arguments sélectionnés ont tendance à varier selon le sens du verbe. Dans ce mémoire, nous créons automatiquement un dictionnaire de valence des verbes du français qui tient compte de la polysémie. Nous extrayons 20 000 exemples de phrases pour chacun des 2 000 verbes les plus fréquents du franc¸ais. Nous obtenons ensuite les plongements lexicaux de ces verbes en contexte à l’aide d’un modèle de langue monolingue et de deux modèles de langue multilingues. Puis, nous utilisons des algorithmes de regroupement pour induire les différents sens de ces verbes. Enfin, nous analysons automatiquement les phrases à l’aide de différents analyseurs syntaxiques afin de trouver leurs arguments. Nous déterminons que la combinaison du modèle de langue français CamemBERT et d’un algorithme de regroupement agglomératif offre les meilleurs résultats dans la tâche d’induction de sens (58,19% de F1 B3), et que pour l’analyse syntaxique, Stanza est l’outil qui a les meilleures performances (83,29% de F1). En filtrant les cadres syntaxiques obtenus à l’aide d’une estimation de la vraisemblance maximale, une méthode statistique très simple qui permet de trouver les paramètres les plus vraisemblables d’un modèle de probabilité qui explique nos données, nous construisons un dictionnaire de valence qui se passe presque complètement d’intervention humaine. Notre procédé est ici utilisé pour le français, mais peut être utilisé pour n’importe quelle autre langue pour laquelle il existe suffisamment de données écrites. / Valency dictionaries are useful for many tasks in automatic language processing. However, quality dictionaries of this type are created at least in part manually; they are therefore resource-intensive and difficult to update. In addition, many of these resources do not take into account the different meanings of lemmas, which are important because the arguments selected tend to vary according to the meaning of the verb. In this thesis, we automatically create a French verb valency dictionary that takes polysemy into account. We extract 20 000 example sentences for each of the 2 000 most frequent French verbs. We then obtain the lexical embeddings of these verbs in context using a monolingual and two multilingual language models. Then, we use clustering algorithms to induce the different meanings of these verbs. Finally, we automatically parse the sentences using different parsers to find their arguments. We determine that the combination of the French language model CamemBERT and an agglomerative clustering algorithm offers the best results in the sense induction task (58.19% of F1 B3), and that for syntactic parsing, Stanza is the tool with the best performance (83.29% of F1). By filtering the syntactic frames obtained using maximum likelihood estimation, a very simple statistical method for finding the most likely parameters of a probability model that explains our data, we build a valency dictionary that almost completely dispenses with human intervention. Our procedure is used here for French, but can be used for any other language for which sufficient written data exists.

In this thesis, I investigate idiom processing from two angles through three different studies. First, I approached idiom processing from a constraint-based perspective. According to this view, not all idioms are alike: they can differ regarding lexical, and linguistic characteristics, such as their level of familiarity. In this first study, I investigated the underlying processes during the comprehension of idioms with different characteristics. I used the Event Related Potential (ERP) technique, which has high temporal resolution, to investigate this issue. I provided evidence that idioms' characteristics impact their processing. More specifically, idioms which are more familiar to language users (i.e., the ones that are encountered more frequently) showed processing facilitation compared to less familiar idioms. Also, idioms with plausible literal interpretation showed processing advantages over idioms which are less likely to be interpreted literally. The second aim of the current thesis was to investigate idiom processing from an embodied account of language processing. According to this view, various sources of information (including linguistic, affective, and sensory-motor) are available and used during the comprehension of language. While, this view has become popular in many language processing studies, studies of idiom processing are still at the beginning of this journey. To be able to investigate idiom processing while considering the role of affective and sensory-motor factors, we require access to norming data. In the second (descriptive) study, I conducted a large-scale survey and collected measures of valence, arousal, concreteness, and imageability for a set of English idioms, by both native speakers of English and proficient second language speakers. In the last study, I explored how the emotional status of idioms and their concreteness contributes to their processing, and whether this contribution is modulated by idiom familiarity. We found that the impact of non-linguistic sources of information (affective and sensory-motor) is determined by idiom familiarity, such that low familiar and high familiar idioms show different behaviour where these factors are concerned. For highly familiar idioms, behaviour aligns with the findings on word processing: for example, idioms with more positive valence showed facilitative processing. Unlike highly familiar idioms, valence had an inhibitory impact on idioms with low familiarity level, such that greater valence increased the reading time.

The Internet is becoming the main source for various tasks, from learning, to working, and shopping. There are many websites one can use to shop. Almost all stores have a website from which you can order anything you might want. As online shopping becomes more prominent, it is important to understand the effects of the Internet and its product reviewers and, specific to this study, consumer decision making. This study seeks to understand the effect of star ratings and valence on review perception between the different cognitive levels of individuals. Recognition review perception, and intent to purchase were being measured. Results showed that need for cognition did have an effect on accuracy of recognition and perceived valence. Need for cognition and congruency as well as actual valence had an effect on perceived valence. Need for cognition, actual valence, and congruency all had an effect on purchase intention. This research is important because it is relevant to a growing trend around the world. Technology is already integrated into nearly everyone’s lives and it is only going to more so as we continue to evolve. Just as it is becoming more common for people to receive education from online institutions, and for employers to use more Internet based applications, it is only natural consumers will continue the trend of purchasing items online. Learning the social and cognitive influences of online reviews on perception and purchasing intentions is something everyone needs to be aware of.

No description available.

No description available.

No description available.

No description available.

No description available.