The aim of this paper is to develop a principled and empirically motivated approach to robust, negotiative spoken dialogue with databases. Robustness is achieved by limiting the set of representable utterance types. Still, the vast majority of utterances that occur in practice can be handled.

This paper presents a robust parsing algorithm and semantic formalism for the interpretation of utterances in spoken negotiative dialogue with databases. The algorithm works in two passes: a domain-specific pattern-matching phase and a domain-independent semantic analysis phase. Robustness is achieved by limiting the set of representable utterance types to an empirically motivated subclass which is more expressive than propositional slot–value lists, but much less expressive than first-order logic. Our evaluation shows that in actual practice the vast majority of utterances that occur can be handled, and that the parsing algorithm is highly efficient and accurate.

This paper presents an algorithm for domain-dependent parsing of utterances in negotiative dialogue. To represent such utterances, the algorithm outputs semantic expressions that are more expressive than propositional slot-filler structures. It is very fast and robust, yet precise and capable of correctly combining information from different utterance fragments.

This paper describes an approach to contextual reasoning for interpretation ofspoken multimodal dialogue. The approach is based on combining recencybased search for antecedents with an object-oriented domain representation insuch a way that the search is highly constrained by the type information of theantecedents. By furthermore representingcandidate antecedents from the dialoguehistory and visual context in a uniformway, a single machinery (based on -reduction in lambda calculus) can be usedfor resolving many kinds of underspecified utterances. The approach has beenimplemented in two highly different domains.

This paper describes an approach to contextual reasoning for interpretation ofspoken multimodal dialogue. The approach is based on combining recencybased search for antecedents with an object-oriented domain representation insuch a way that the search is highly constrained by the type information of theantecedents. By furthermore representingcandidate antecedents from the dialoguehistory and visual context in a uniformway, a single machinery (based on -reduction in lambda calculus) can be usedfor resolving many kinds of underspecified utterances. The approach has beenimplemented in two highly different domains.

We describe an implemented spoken-language dialogue system for a travel-planning domain, which accesses a commercially available travel-information web-server and supports a flexible mixed-initiative dialogue strategy. We argue, based on data from initial Wizard-of-Oz experiments, that mixed-initiative strategies are appropriate for many types of user, but require more sophisticated architectures for processing of language and dialogue; we then use these observations to motivate an architecture which combines parallel deep and shallow natural language analysis engines and an agenda-driven dialogue manager. We outline the top-level processing strategy used by the dialogue manager, and also a novel formalism, which we call Flat Utterance Description, that allows us to reduce the output of the deep and shallow language-processing engines to a common representation.

Accurate classifiers for short texts are valuable assets in many applications. Especially in online communities, where users contribute to content in the form of posts and com- ments, an effective way of automatically categorising posts proves highly valuable. This paper investigates the use of N- grams as features for short text classification, and compares it to manual feature design techniques that have been popu- lar in this domain. We find that the N-gram representations greatly outperform manual feature extraction techniques.

The self-rationalising capabilities of LLMs are appealing because the generated explanations can give insights into the plausibility of the predictions. However, how faithful the explanations are to the predictions is questionable, raising the need to explore the patterns behind them further. To this end, we propose a hypothesis-driven statistical framework. We use a Bayesian network to implement a hypothesis about how a task (in our example, natural language inference) is solved, and its internal states are translated into natural language with templates. Those explanations are then compared to LLM-generated free-text explanations using automatic and human evaluations. This allows us to judge how similar the LLM’s and the Bayesian network’s decision processes are. We demonstrate the usage of our framework with an example hypothesis and two realisations in Bayesian networks. The resulting models do not exhibit a strong similarity to GPT-3.5. We discuss the implications of this as well as the framework’s potential to approximate LLM decisions better in future work.

Evaluation of machine translation (MT) is a difficult task, both for humans, and using automatic metrics. The main difficulty lies in the fact that there is not one single correct translation, but many alternative good translation options.MT systems are often evaluated using automatic metrics, which commonly rely on comparing a translation to only a single human reference translation. An alternative is different types of human evaluations, commonly ranking be-tween systems or estimations of adequacy and fluency on some scale, or error analyses.

We have explored four different evaluation methods on output from three different statistical MT systems. The main focus is on different types of human evaluation. We compare two conventional evaluation methods, human error analysis and automatic metrics, to two lesser used evaluation methods based on reading comprehension and eye-tracking. These two methods of evaluations are performed without the subjects seeing the source sentence. There have been few previous attempts of using reading comprehension and eye-tracking for MT evaluation.

One example of a reading comprehension study is Fuji (1999) who conducted an experiment to compare English-to-Japanese MT to several versions of manual corrections of the system output. He found significant differences be-tween texts with large differences on reading comprehension questions. Doherty and O’Brien (2009) is the only study we are aware of using eye-tracking for MT evaluation. They found that the average gaze time and fixation counts were significantly lower for sentences judged as excellent in an earlier evaluation, than for bad sentences.

Like previous research we find that both reading comprehension and eye-tracking can be useful for MT evaluation.

The results of these methods are consistent with the other methods for comparison between systems with a big quality difference. For systems with similar quality, however, the different evaluation methods often does not show any significant differences.

The Spoken Language Translator (SLT) is a multi-lingual speech-to-speech translation prototype supporting English, Swedish and French within the air traffic information system (ATIS) domain. The design of SLT is characterized by a strongly corpus-driven approach, which accentuates the need for cost-efficient collection procedures to obtain training data. This paper discusses various approaches to the data collection issue pursued within a speech translation framework. Original American English speech and language data have been collected using traditional Wizard-of-Oz (WOZ) techniques, a relatively costly procedure yielding high-quality results. The resulting corpus has been translated textually into Swedish by a large number of native speakers (427) and used as prompts for training the target language speech model. This ￜbudgetￜ collection method is compared to the accepted method, i.e., gathering data by means of a full-blown WOZ simulation. The results indicate that although translation in this case proved economical and produced considerable data, the method is not sensitive to certain features typical of spoken language, for which WOZ is superior

Privacy preservation of sensitive information is one of the main concerns in clinical text mining. Due to the inherent privacy-keeping problems that arise when handling clinical data, the clinical corpora used to create the clinical Named Entity Recognition (NER) models underlying clinical de-identification systems cannot be shared. This implies that clinical NER models are trained and tested on data coming from the same institution because it is rarely possible to evaluate them on data belonging to a different institution. Given this sharing restrictions, it is very to assess whether a clinical NER model has overfitted the data or if it is driven by undetected biases. In this paper we present the results of the first-ever cross-institution evaluation of a Swedish de-identification system on Swedish clinical data. Alongside the encouraging results, we present a discussion about differences and similarities across EHR naming conventions and NER tagsets.

Sharing data is an important part of the progress of science in many fields. In the largely deep learning dominated field of natural language processing, textual resources are in high demand. In certain domains, such as that of medical records, the sharing of data is limited by ethical and legal restrictions and therefore requires anonymization. The process of manual anonymization is tedious and expensive, thus automated anonymization is of great value. Since medical records consist of unstructured text, pieces of sensitive information have to be identified in order to be masked for anonymization. Named-entity recognition (NER) is the subtask of information extraction named entities, such as person names or locations, are identified and categorized. Recently, models that leverage unsupervised training on large quantities of unlabeled training data have performed impressively on the NER task, which shows promise in their usage for the problem of anonymization. In this study, a small set of medical records was annotated with named-entity tags. Because of the lack of any training data, a BERT model already fine-tuned for NER was then evaluated on the evaluation set. The aim was to find out how well the model would perform on NER on medical records, and to explore the possibility of using the model to anonymize medical records. The most positive result was that the model was able to identify all person names in the dataset. The average accuracy for identifying all entity types was however relatively low. It is discussed that the success of identifying person names shows promise in the model’s application for anonymization. However, because the overall accuracy is significantly worse than that of models fine-tuned on domain-specific data, it is suggested that there might be better methods for anonymization in the absence of relevant training data.

This thesis work adresses text classification in relation to genre identification using different feature sets, with a focus on syntactic based features. We built our models by means of traditional machine learning algorithms, i.e. Naive Bayes, K-nearest neighbour, Support Vector Machine and Random Forest in order to predict the literary genre of books. We trained our models using as feature sets bag-of-words (BOW), bigrams, syntactic-based bigrams and emotional features, as well as combinations of features. Results obtained using the best features, i.e. BOW combined with bigrams based on syntactic relations between words, on the test set showed an enhancement in performance by 2% in F1-score over the baseline using BOW features, which translates into a positive impact of using syntactic information in the task of text classification.

When assessing the readability of a text it is helpful to consider all its interacting elements. This includes its syntactic complexity, but other aspects, such as that of cohesion, are no less important. The thesis explores how these are reflected in each other and in the readability of books in a dataset provided by the publisher Nypon och Vilja, which consists of easy-to-read books divided into six levels of readability. To provide additional nuance, the interrelated concepts of epistemic stance and narrativity are introduced for the purpose of deepening the analysis of the statistical findings. They also prove useful in further discussion surrounding complexity and cohesion as they relate to reading skill and knowledge asymmetries. Principal component analysis (PCA) is employed to uncover these statistical relationships on a broader scale, though more specific in-depth analysis are performed relating to certain metrics. While the findings have some support in literature, re-affirming the importance of narrativity for contextualizing cohesion, the clear link between higher complexity and less narrative text was not expected. Furthermore, the PCA indicates a more nuanced picture of referential cohesion and the use of its constituent metrics, depending both on narrativity and complexity.

Given the explosion in the size of social media, the amount of hate speech is also growing. To efficiently combat this issue we need reliable and scalable machine learning models. Current solutions rely on crowdsourced datasets that are limited in size, or using training data from self-identified hateful communities, that lacks specificity. In this thesis we introduce a novel semi-supervised modelling strategy. It is first trained on the freely available data from the hateful communities and then fine-tuned to classify hateful tweets from crowdsourced annotated datasets. We show that our model reach state of the art performance with minimal hyper-parameter tuning.

This paper introduces a new research project Simulating Intonational Varieties of Swedish (SIMULEKT). The basic goal of the project is to produce more precise and thorough knowledge about some major intonational varieties of Swedish. In this research effort the Swedish prosody model plays a prominent role. A fundamental idea is to take advantage of speech synthesis in different forms. In our analysis and synthesis work we will focus on some major intonational types: South, Göta, Svea, Gotland, Dala, North, and Finland Swedish. The significance of our project work will be within basic research as well as in speech technology applications.

In order for game characters to be believable, they must appear to possess qualities such as emotions, the ability to learn and adapt as well as being able to communicate in natural language. With this paper we aim to contribute to the development of believable non-player characters (NPCs) in games, by presenting a method for managing NPC dialogues. We have selected the trade scenario as an example setting since it offers a well-known and limited domain common in games that support ownership, such as role-playing games. We have developed a dialogue manager in State Chart XML, a newly introduced W3C standard, as part of DEAL -- a research platform for exploring the challenges and potential benefits of combining elements from computer games, dialogue systems and language learning.

For the vast majority of languages, Natural Language Processing (NLP) tools are either absent entirely, or leave much to be desired in their final performance. Despite having nearly 4 million speakers, one such low-resource language is Galician. In an effort to expand available NLP resources, this project sought to construct a dataset for Semantic Role Labeling (SRL) and produce a baseline for future research to use in comparisons. SRL is a task which has shown success in amplifying the final output for various NLP systems, including Machine Translation and other interactive language models. This project was successful in that fact and produced 24 SRL models and two SRL datasets; one Galician and one Spanish. mBERT and XLM-R were chosen as the baseline architectures; additional models were first pre-trained on the SRL task in a language other than the target to measure the effects of transfer-learning. Scores are reported on a scale of 0.0-1.0. The best performing Galician SRL model achieved an f1 score of 0.74, introducing a baseline for future Galician SRL systems. The best performing Spanish SRL model achieved an f1 score of 0.83, outperforming the baseline set by the 2009 CoNLL Shared Task by 0.025. A pre-processing method, verbal indexing, was also introduced which allowed for increased performance in the SRL parsing of highly complex sentences; effects were amplified in scenarios where the model was both pre-trained and fine-tuned on datasets utilizing the method, but still visible even when only used during fine-tuning.

In this paper, we leverage existing corpora, WordNet, and dependency parsing to build the first Galician dataset for training semantic role labeling systems in an effort to expand available NLP resources. Additionally, we introduce verb indexing, a new pre-processing method, which helps increase the performance when semantically parsing highly-complex sentences. We use transfer-learning to test both the resource and the verb indexing method. Our results show that the effects of verb indexing were amplified in scenarios where the model was both pre-trained and fine-tuned on datasets utilizing the method, but improvements are also noticeable when only used during fine-tuning. The best-performing Galician SRL model achieved an f1 score of 0.74, introducing a baseline for future Galician SRL systems. We also tested our method on Spanish where we achieved an f1 score of 0.83, outperforming the baseline set by the 2009 CoNLL Shared Task by 0.025 showing the merits of our verb indexing method for pre-processing.

We discuss methodological choices in diagnostic evaluation and error analysis in meaning representation parsing (MRP), i.e. mapping from natural language utterances to graph-based encodings of semantic structure. We expand on a pilot quantitative study in contrastive diagnostic evaluation, inspired by earlier work in syntactic dependency parsing, and propose a novel methodology for qualitative error analysis. This two-pronged study is performed using a selection of submissions, data, and evaluation tools featured in the 2019 shared task on MRP. Our aim is to devise methods for identifying strengths and weaknesses in different broad families of parsing techniques, as well as investigating the relations between specific parsing approaches, different meaning representation frameworks, and individual linguistic phenomena—by identifying and comparing common error patterns. Our preliminary empirical results suggest that the proposed methodologies can be meaningfully applied to parsing into graph-structured target representations, as a side-effect uncovering hitherto unknown properties of the different systems that can inform future development and cross-fertilization across approaches.

We discuss methodological choices in contrastive and diagnostic evaluation in meaning representation parsing, i.e. mapping from natural language utterances to graph-based encodings of semantic structure. Drawing inspiration from earlier work in syntactic dependency parsing, we transfer and refine several quantitative diagnosis techniques for use in the context of the 2019 shared task on Meaning Representation Parsing (MRP). As in parsing proper, moving evaluation from simple rooted trees to general graphs brings along its own range of challenges. Specifically, we seek to begin to shed light on relative strenghts and weaknesses in different broad families of parsing techniques. In addition to these theoretical reflections, we conduct a pilot experiment on a selection of top-performing MRP systems and two of the five meaning representation frameworks in the shared task. Empirical results suggest that the proposed methodology can be meaningfully applied to parsing into graph-structured target representations, uncovering hitherto unknown properties of the different systems that can inform future development and cross-fertilization across approaches.

We discuss methodological choices in diagnostic evaluation and error analysis in meaning representation parsing (MRP), i.e. mapping from natural language utterances to graph-based encodings of semantic structure. We expand on a pilot quantitative study in contrastive diagnostic evaluation, inspired by earlier work in syntactic dependency parsing, and propose a novel methodology for qualitative error analysis. This two-pronged study is performed using a selection of submissions, data, and evaluation tools featured in the 2019 shared task on MRP. Our aim is to devise methods for identifying strengths and weaknesses in different broad families of parsing techniques, as well as investigating the relations between specific parsing approaches, different meaning representation frameworks, and individual linguistic phenomena—by identifying and comparing common error patterns. Our preliminary empirical results suggest that the proposed methodologies can be meaningfully applied to parsing into graph-structured target representations, as a side-effect uncovering hitherto unknown properties of the different systems that can inform future development and cross-fertilization across approaches.

Native language identification (NLI) is a classification task which identifies the mother tongue of a language learner based on spoken or written material. The task gained popularity when it was featured in the 2017 BEA-12-workshop and since then many applications have been successfully found for NLI - ranging from language learning to authorship identification and forensic science. While a considerable amount of research has already been done in this area, we introduce a novel approach of incorporating syntactic information into the implementation of a BERT-based NLI model. In addition, we train separate models to test whether erroneous input sequences perform better than corrected sequences. To answer these questions we carry out both a quantitative and qualitative analysis. In addition, we test our idea of implementing a BERT-based GEC model to supply more training data to our NLI model without the need for manual annotation. Our results suggest that our models do not outperform the SVM baseline, but we attribute this result to the lack of training data in our dataset, as transformer-based architectures like BERT need huge amounts of data to be successfully fine-tuned. In turn, simple linear models like SVM perform well on small amounts of data. We also find that erroneous structures in data come useful when combined with syntactic information but neither boosts the performance of NLI model separately. Furthermore, our implemented GEC system performs well enough to produce more data for our NLI models, as their scores increase after implementing the additional data, resulting from our second experiment. We believe that our proposed architecture is potentially suitable for the NLI task if we incorporate extensions which we suggest in the conclusion section.

We present an overview of speech research applications of gaze tracking technology, where gaze behaviours are exploited as a tool for analysis rather than as a primary object of study. The methods presented are all in their infancy, but can greatly assist the analysis of digital audio and video as well as unlock the relationship between writing and other encodings on the one hand, and natural language, such as speech, on the other. We discuss three directions in this type of gaze tracking application: modelling of text that is read aloud, evaluation and annotation with naïve informants, and evaluation and annotation with expert annotators. In each of these areas, we use gaze tracking information to gauge the behaviour of people when working with speech and conversation, rather than when reading text aloud or partaking in conversations, in order to learn something about how the speech may be ana-lysed from a human perspective.

We present an experiment designed for extracting construction candidates for a Swedish constructicon from text corpora. We have explored the use of hybrid n-grams with the practical goal to discover previously undescribed partially schematic constructions. The experiment was successful, in that quite a few new constructions were discovered. The precision is low, but as a push-button tool for construction discovery, it has proven a valuable tool for the work on a Swedish constructicon.

In this paper, we describe a method for mapping the phonological feature location of Swedish Sign Language (SSL) signs to the meanings in the Swedish semantic dictionary SALDO. By doing so, we observe clear differences in the distribution of meanings associated with different locations on the body. The prominence of certain locations for specific meanings clearly point to iconic mappings between form and meaning in the lexicon of SSL, which pinpoints modalityspecific properties of the visual modality.

This thesis applies the word embedding mapping approach to make a lexical comparison from academic word usage perspective. We aim to demonstrate the differences in academic word usage between a corpus of student writings and a corpus of academic English, as well as a corpus of student writings and social media texts. The Vecmap mapping algorithm, commonly used in solving cross-language mapping problems, was used to map academic English vector space and social media text vector space into the common student writing vector space to facilitate the comparison of word representations from different corpora and to visualize the comparison results. The average distance was defined as a measure of word usage differences of 420 typical academic words between each two corpora, and principal component analysis was applied to visualize the differences. A rank-biased overlap approach was adopted to evaluate the results of the proposed approach. The experimental results show that the usage of academic words of student writings corpus is more similar to the academic English corpus than to the social media text corpus. 

CoNLL 2015 featured a shared task on shallow discourse parsing. In 2016, the efforts continued with an increasing focus on sense classification. In the case of implicit sense classification, there was an interesting mix of traditional and modern machine learning classifiers using word representation models. In this thesis, we explore the performance of a number of these models, and investigate how they perform using a variety of word representation models. We show that there are large performance differences between word representation models for certain machine learning classifiers, while others are more robust to the choice of word representation model. We also show that with the right choice of word representation model, simple and traditional machine learning classifiers can reach competitive scores even when compared with modern neural network approaches. 

For some language pairs, pronoun translation is a discourse-driven task which requires information that lies beyond its local context. This motivates the task of predicting the correct pronoun given a source sentence and a target translation, where the translated pronouns have been replaced with placeholders. For cross-lingual pronoun prediction, we suggest a neural network-based model using preceding nouns and determiners as features for suggesting antecedent candidates. Our model scores on par with similar models while having a simpler architecture.

In this paper, we study the use of goodness of pronunciation (GOP) on child speech. We first compare the distributions of GOP scores on several open datasets representing various dimensions of speech variability. We show that the GOP distribution over CMU Kids, corresponding to young age, has larger spread than those on datasets representing other dimensions, i.e., accent, dialect, spontaneity and environmental conditions. We hypothesize that the increased variability of pronunciation in young age may impair the use of traditional mispronunciation detection methods for children. To support this hypothesis, we perform simulated mispronunciation experiments both for children and adults using different variants of the GOP algorithm. We also compare the results to real-case mispronunciations for native children showing that GOP is less effective for child speech than for adult speech.

Despite many years of research on Swedish language technology, there is still no well-documented standard for Swedish word processing covering the whole spectrum from low-level tokenization to morphological analysis and disambiguation. SWORD is a new initiative within the SWE-CLARIN consortium aiming to develop documented standards for Swedish word processing. In this paper, we report on a pilot study of Swedish tokenization, where we compare the output of six different tokenizers on four different text types. For one text type (Wikipedia articles), we also compare to the tokenization produced by six manual annotators.

Many approaches to solving tasks in the field of Natural Language Processing (NLP) use syntactic dependency trees (SDTs) as a feature to represent the latent nonlinear structure within sentences. Recently, work in parsing sentences to graph-based structures which encode semantic relationships between words—called semantic dependency graphs (SDGs)—has gained interest. This thesis seeks to explore the use of SDGs in place of and alongside SDTs within a relation classification system based on long short-term memory (LSTM) neural networks. Two methods for handling the information in these graphs are presented and compared between two SDG formalisms. Three new relation extraction system architectures have been created based on these methods and are compared to a recent state-of-the-art LSTM-based system, showing comparable results when semantic dependencies are used to enhance syntactic dependencies, but with significantly fewer training parameters.

In this position statement, we wish to contribute to the discussion about how to assess quality and coverage of a model.

We believe that BERT's prominence as a single-step pipeline for contextualization and classification highlights the need for benchmarks to evolve concurrently with models. Much recent work has touted BERT's raw power for solving natural language tasks, so we used a 12-layer uncased BERT pipeline with a linear classifier as a quick-and-dirty model to score well on the SemEval 2010 Task 8 dataset for relation classification between nominals. We initially expected there to be significant enough bias from BERT's training to influence downstream tasks, since it is well-known that biased training corpora can lead to biased language models (LMs). Gender bias is the most common example, where gender roles are codified within language models. To handle such training data bias, we took inspiration from work in the field of computer vision. Tang et al. (2020) mitigate human reporting bias over the labels of a scene graph generation task using a form of causal reasoning based on counterfactual analysis. They extract the total direct effect of the context image on the prediction task by "blanking out" detected objects, intuitively asking "What if these objects were not here?" If the system still predicts the same label, then the original prediction is likely caused by bias in some form. Our goal was to remove any effects from biases learned during BERT's pre-training, so we analyzed total effect (TE) instead. However, across several experimental configurations we found no noticeable effects from using TE analysis. One disappointing possibility was that BERT might be resistant to causal analysis due to its complexity. Another was that BERT is so powerful (or blunt?) that it can find unanticipated trends in its input, rendering any human-generated causal analysis of its predictions useless. We nearly concluded that what we expected to be delicate experimentation was more akin to trying to carve a masterpiece sculpture with a self-driven sledgehammer. We then found related work where BERT fooled humans by exploiting unexpected characteristics of a benchmark. When we used BERT to predict a relation for random words in the benchmark sentences, it guessed the same label as it would have for the corresponding marked entities roughly half of the time. Since the task had nineteen roughly-balanced labels, we expected much less consistency. This finding repeated across all pipeline configurations; BERT was treating the benchmark as a sequence classification task! Our final conclusion was that the benchmark is inadequate: all sentences appeared exactly once with exactly one pair of entities, so the task was equivalent to simply labeling each sentence. We passionately claim from our experience that the current trend of using larger and more complex LMs must include concurrent evolution of benchmarks. We as researchers need to be diligent in keeping our tools for measuring as sophisticated as the models being measured, as any scientific domain does.

We and others have found it fruitful to assume that users, when interacting with spoken dialogue systems, perceive the systems and their actions metaphorically. Common metaphors include the human metaphor and the interface metaphor (cf. Edlund, Heldner, & Gustafson, 2006). In the interface metaphor, the spoken dialogue system is perceived as a machine interface – often but not always a computer interface. Speech is used to accomplish what would have otherwise been accomplished by some other means of input, such as a keyboard or a mouse. In the human metaphor, on the other hand, the computer is perceived as a creature (or even a person) with humanlike conversational abilities, and speech is not a substitute or one of many alternatives, but rather the primary means of communicating with this creature. We are aware that more “natural ” or human-like behaviour does not automatically make a spoken dialogue system “better ” (i.e. more efficient or more well-liked by its users). Indeed, we are quite convinced that the advantage (or disadvantage) of humanlike behaviour will be highly dependent on the application. However, a dialogue system that is coherent with a human metaphor may profit from a number of characteristics.

This paper reports on a study that explores to what extent listeners are able to judge where a particular utterance fragment is located in a speaker's pitch range. The research consists of a perception study that makes use of 100 stimuli, selected from 50 different speakers whose speech was originally collected for a multi-speaker database of Swedish speech materials. The fragments are presented to subjects whom are asked to estimate whether the fragment is located in the lower or higher part of that speaker's range. Results reveal that listeners' judgments are dependent on the gender of the speaker, but that within a gender they tend to hear differences in range.

In this chapter, we review some of the issues in rule-based synthesis and specifically discuss formant synthesis. Formant synthesis and the theory behind have played an important role in both the scientific progress in understanding how humans talk and also the development of the first speech technology applications. Its flexibility and small footprint makes the approach still of interest and a valuable complement to the current dominant methods based on concatenative data-driven synthesis. As already mentioned in the overview by Schroeter (Chap. 19) we also see a new trend to combine the rule-based and data-driven approaches. Formant features from a database that can be used both to optimize a rule-based formant synthesis system and to optimize the search for good units in a concatenative system.

The goal of the project is to model the prosodic structuring of speech in terms of boundaries and groupings. The modeling will include different communicative situations and be based on existing as well as new speech corpora. Production and perception studies will be used in parallel with automatic methods developed for analysis, modeling and prediction of prosody. The model will be perceptually evaluated using synthetic speech.