Unknown words

These long sentences are further complicated by the fact that they often include difficult vocabulary. Seeing words you don t know may send your anxiety level soaring, and nobody does his or her best work when anxious. With practice, though, you can learn to take those long sentences and unknown words in stride. Try following these seven steps ... 

Substitute words or sounds of your choosing in place of unknown words as you read. 

Archaic Gold Au 79 known to the ancients Unknown Word "aurum" meaning "gold" in Latin. 

Many times you ll see words in a story you don t know. Sometimes you can figure them out from their context, but if you find unknown words that might indeed have a critical bearing on your understanding of a character, for example, look these up now. 

Context. The words and ideas surrounding a vocabulary word are its context. If the context is clear, you should be able to substitute a nonsense word for the unknown word and still make the right answer choice. For example ... 

Synonyms or antonyms. Search for nearby clues that may be synonyms (same meaning) or antonyms (opposite meaning) of the unknown word ... 

Easy-to-understand phonetic pronunciation guides, in cases where it may be tough to figure out how to say an unknown word. 

Unknown words There will always be words that are not in the lexicon and when we find these we have to deal with them somehow. 

This is elearly a problem of data sparsity in that we ean t eofleet or design models for such a large number of types. When we consider that the required voeabulary for a real system may in fact be even larger, and that unknown words will have to be handled somehow, we see that in reality the situation may be even worse. 

Grapheme-to-phoneme conversion by rule is perhaps the classic application of traditional knowledge based rules in TTS. Indeed, in the past it was thought that this was how humans performed the task when presented with an unknown word they would apply a set of rules to find the pronunciation. The most common approach is to process the character sequence left-to-right, and for each character apply one or more rules in order to generate a phoneme. It should be clear that these rules can t operate in isolation, otherwise a t character would always result in a /t/ phoneme. Rather the character context of the character is used and in this was we can generate /th/ when t is followed by h. 

We can deal with this in a number of ways, but the most common approach is to store all roots and common inflections and derivations in the lexicon. When other inflections or derivations are encountered, these are treated as unknown words and passed to the grapheme-to-phoneme convertor. This then can attempt a basic morphological decomposition which attempts to find... 

Introduces exercises carefully. The teacher should not use unknown words in exercises. Write short sentences and take care not to formulate more than one question within one exercise. 

---