Is a Wizard-of-Oz Required for Robot-Led Conversation Practice in a Second Language?

The large majority of previous work on human-robot conversations in a second language has been performed with a human wizard-of-Oz. The reasons are that automatic speech recognition of non-native conversational speech is considered to be unreliable and that the dialogue management task of selecting robot utterances that are adequate at a given turn is complex in social conversations. This study therefore investigates if robot-led conversation practice in a second language with pairs of adult learners could potentially be managed by an autonomous robot. We first investigate how correct and understandable transcriptions of second language learner utterances are when made by a state-of-the-art speech recogniser. We find both a relatively high word error rate (41%) and that a substantial share (42%) of the utterances are judged to be incomprehensible or only partially understandable by a human reader. We then evaluate how adequate the robot utterance selection is, when performed manually based on the speech recognition transcriptions or autonomously using (a) predefined sequences of robot utterances, (b) a general state-of-the-art language model that selects utterances based on learner input or the preceding robot utterance, or (c) a custom-made statistical method that is trained on observations of the wizard’s choices in previous conversations. It is shown that adequate or at least acceptable robot utterances are selected by the human wizard in most cases (96%), even though the ASR transcriptions have a high word error rate. Further, the custom-made statistical method performs as well as manual selection of robot utterances based on ASR transcriptions. It was also found that the interaction strategy that the robot employed, which differed regarding how much the robot maintained the initiative in the conversation and if the focus of the conversation was on the robot or the learners, had marginal effects on the word error rate and understandability of the transcriptions but larger effects on the adequateness of the utterance selection. Autonomous robot-led conversations may hence work better with some robot interaction strategies.

Novel Approach for Statistical Interpretation: A Case Study from Long-Term Crop Production Experiments (Hungary)

In recent decades, the agricultural sector has witnessed rapid technological interventions from field to the production stage. Thus, the importance of these technological interventions must be strictly evaluated. The traditional statistical method often deems low statistical differences as a significant one, which cannot be considered effective from different perspectives. In this sense, the aim of this research was to develop a new statistical method for evaluating agricultural experiments based on different criteria; hence, the significant importance of the technological interventions can be clearly determined. Data were collected from of a long-term (13-year) crop production experiment (Central Europe, Hungary), which involved five different fertilization levels, along with non-fertilized treatment (control), two irrigation treatments (irrigated and non-irrigated), and 15–20 genotypes of maize. The output of this research showed that the classic statistical approach for testing the significant differences among treatments should be accompanied with our new suggested approach (i.e., professional test), which reflect whether treatments were professionally effective or not. Also, results showed that good statistical background is not enough for interoperating the analysis of agricultural experiments. This research suggested that erroneous conclusions can be avoided by merging classical and professional statistical tests, and correct recommendations could be provided to decision makers and farmers based on their financial resources.