Keith Trnka

User Interaction with Word Prediction: The Effects of Prediction Quality

Word prediction systems can reduce the number of keystrokes required to form a message in a letter-based AAC system. It has been questioned, however, whether such savings translate into an enhanced communication rate due to the additional overhead (e.g., shifting of focus and repeated scanning of a prediction list) required in using such a system. Our hypothesis is that word prediction has high potential for enhancing AAC communication rate, but the amount is dependent in a complex way on the accuracy of the predictions. Due to significant user interface variations in AAC systems and the potential bias of prior word prediction experience on existing devices, this hypothesis is difficult to verify. We present a study of two different word prediction methods compared against letter-by-letter entry at simulated AAC communication rates. We find that word prediction systems can in fact speed communication rate (an advanced system gave a 58.6% improvement), and that a more accurate word prediction system can raise the communication rate higher than is explained by the additional accuracy of the system alone due to better utilization (93.6% utilization for advanced versus 78.2% for basic).

Topic modeling in fringe word prediction for AAC

Word prediction can be used for enhancing the communication ability of persons with speech and language impairments. In this work, we explore two methods of adapting a language model to the topic of conversation, and apply these methods to the prediction of fringe words.

Exploring and Exploiting the Limited Utility of Captions in Recognizing Intention in Information Graphics

This paper presents a corpus study that explores the extent to which captions contribute to recognizing the intended message of an information graphic. It then presents an implemented graphic interpretation system that takes into account a variety of communicative signals, and an evaluation study showing that evidence obtained from shallow processing of the graphics caption has a significant impact on the systems success. This work is part of a larger project whose goal is to provide sight-impaired users with effective access to information graphics.

Corpus studies in word prediction

Word prediction can be used to enhance the communication rate of people with disabilities who use Augmentative and Alternative Communication (AAC) devices. We use statistical methods in a word prediction system, which are trained on a corpus, and then measure the efficacy of the resulting system by calculating the theoretical keystroke savings on some held out data. Ideally training and testing should be done on a large corpus of AAC text covering a variety of topics, but no such corpus exists. We discuss training and testing on a wide variety of corpora meant to approximate text from AAC users. We show that training on a combination of in-domain data with out-of-domain data is often more beneficial than either data set alone and that advanced language modeling such as topic modeling is portable even when applied to very different text.

Evaluating Word Prediction: Framing Keystroke Savings

Researchers typically evaluate word prediction using keystroke savings, however, this measure is not straightforward. We present several complications in computing keystroke savings which may affect interpretation and comparison of results. We address this problem by developing two gold standards as a frame for interpretation. These gold standards measure the maximum keystroke savings under two different approximations of an ideal language model. The gold standards additionally narrow the scope of deficiencies in a word prediction system.

Adaptive Language Modeling for Word Prediction

We present the development and tuning of a topic-adapted language model for word prediction, which improves keystroke savings over a comparable baseline. We outline our plans to develop and integrate style adaptations, building on our experience in topic modeling to dynamically tune the model to both topically and stylistically relevant texts.

Adapting word prediction to subject matter without topic-labeled data

Word prediction helps to increase communication rate when using Augmentative and Alternative Communication devices. Basic prediction systems offer topically inappropriate predictions for the context, thus we adapt the predictions to the topic of discourse. However, previous work has relied on texts that are grouped into topics by humans. In contrast, we avoid this restriction by treating each document as a topic. The results are comparable to human-labeled topics and also the method is applicable to unlabeled text.