Language Modeling with Sentence-Level Mixtures
نویسندگان
چکیده
This paperintroduces a simple mixtare language model that attempts to capture long distance conslraints in a sentence or paragraph. The model is an m-component mixture of Irigram models. The models were constructed using a 5K vocabulary and trained using a 76 million word Wail Street Journal text corpus. Using the BU recognition system, experiments show a 7% improvement in recognition accuracy with the mixture trigram models as compared to using a Irigram model. 1. I N T R O D U C T I O N The overall performance of a large vocabulary continuous speech recognizer is greatly impacted by the constraints imposed by a language model, or the effective constraints of a stochastic language model that provides the a priori probability estimates of the word sequence P ( w z , . . . , w r ) . The most commonly used statistical language model assumes that the word sequence can be described as a high order Marker process, typically referred to as an n-gram model, where the probability of a word sequence is given by:
منابع مشابه
The Role of Emotioncy in Cognitive Load and Sentence Comprehension of Language Learners
Emotion and cognition are both considered influential factors in language learning. In this study, the role of "emotioncy" (which is a combination of emotion and frequency) in the cognitive load and sentence comprehension of a group of language learners was examined. Emotioncy includes emotions that are evoked by the senses. To this aim, 200 English as a foreign language (EFL) learners were ask...
متن کاملThe Effect of Sentence-Writing Practice on Iranian low-intermediate EFL Learners’ L2 Grammatical Accuracy
This study aimed to investigate the effect of sentence writing practice on male and female low-intermediate students’ English grammatical accuracy. The question this study tried to answer does English grammatical accuracy can be affected by sentence writing practice. To find the answer to the question, 15 low intermediate level students from Kish away institute were selected. They were both mal...
متن کاملIranian EFL Learners’ Lexical Inferencing Strategies at Both Text and Sentence levels
Lexical inferencing is one of the most important strategies in vocabulary learning and it plays an important role in dealing with unknown words in a text. In this regard, the aim of this study was to determine the lexical inferencing strategies used by Iranian EFL learners when they encounter unknown words at both text and sentence levels. To this end, forty lower intermediate students were div...
متن کاملEvaluating the Success of the Visual Learners in Vocabulary Learning through Word List versus Sentence Making Approaches
Thisstudy sought to evaluate the learners' achievements with the visual learning style when exposed to the sentence making and word list approaches. On that account, 45 basic level participants who studied at the Iran Language Institute (ILI), Bushehr, took part in this research study. At the outset, the learners were given Barsch learning style inventory (1991) to determine the learners' learn...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 1994