Bayesian Active Learning-Based Robot Tutor for Children’s Word-Reading Skills
Authors: Goren Gordon, Cynthia Breazeal
AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We developed a Bayesian active-learning algorithm that continuously and efficiently assesses a child s word-reading skills and implemented it in a social robot. We then developed an integrated experimental paradigm in which a child plays a novel story-creation tablet game with the robot. We show that our algorithm results in an accurate representation of the child s word-reading skills for a large age range, 4-8 year old children, and large initial reading skill range. We also show that employing childspecific assessment-based tutoring results in an ageand initial reading skill-independent learning, compared to random tutoring. |
| Researcher Affiliation | Academia | Goren Gordon and Cynthia Breazeal Personal Robots Group, MIT Media Lab, 20 Ames Street E15-468 Cambridge, MA 02139 {ggordon,cynthiab}@media.mit.edu |
| Pseudocode | No | The paper describes algorithms using mathematical equations and descriptive text, but no structured pseudocode block or explicitly labeled algorithm section was found. |
| Open Source Code | No | The paper mentions using 'an open-source natural language generation library' but does not state that the code for their own methodology is open-source or provide a link to it. |
| Open Datasets | No | The paper describes data collection through the 'TOWRE word assessment test' and an 'in-house developed app game,' but does not provide access information, citation to a public dataset, or indicate that the data used is publicly available. |
| Dataset Splits | No | The paper describes experimental phases (pre-test, story phase, post-test) and data collection within those phases but does not specify dataset splits (e.g., percentages, sample counts) for training, validation, or testing a model in a typical machine learning context. |
| Hardware Specification | No | The paper states, 'For the social robotic platform we used Dragon Bot (Setapen 2012), a squash-and-stretch Android smartphone based robot.' However, it does not provide specific CPU/GPU models, memory details, or other detailed hardware specifications for the smartphone or any other computing resources used to run the experiments. |
| Software Dependencies | No | The paper mentions 'a commercial child-like voice for the text-to-speech software' and 'an open-source natural language generation library,' but it does not provide specific version numbers for any software dependencies. |
| Experiment Setup | Yes | The paper provides extensive details on the experimental setup, including the selection of words ('chosen according to the active-learning method'), the criteria for distractors ('two words which are most similar...one word that the child should know...and one word that the child should not know'), the number of repetitions ('repeated ten times' for pre-test, 'total of ten words' for post-test), and the frequency of robot interaction ('In 50% of the sentences, the robot expresses a shy face and asks the child to show it a word'). |