ABSTRACT In this study we show empirically on a Turkish corpus that Zipf's power laws rel... more ABSTRACT In this study we show empirically on a Turkish corpus that Zipf's power laws relating frequencies of words to rank, vocabulary and number of meanings can also be derived from term x document analysis. This supports the idea put forward by Ding that Zipfian distributions and SVD carry similar hidden (latent) semantics.
In this study we show empirically on a Turkish corpus that Zipf's power laws relating frequencies... more In this study we show empirically on a Turkish corpus that Zipf's power laws relating frequencies of words to rank, vocabulary and number of meanings can also be derived from term x document analysis. This supports the idea put forward by Ding that Zipfian distributions and SVD carry similar hidden (latent) semantics.
ABSTRACT In this study we show empirically on a Turkish corpus that Zipf's power laws rel... more ABSTRACT In this study we show empirically on a Turkish corpus that Zipf's power laws relating frequencies of words to rank, vocabulary and number of meanings can also be derived from term x document analysis. This supports the idea put forward by Ding that Zipfian distributions and SVD carry similar hidden (latent) semantics.
In this study we show empirically on a Turkish corpus that Zipf's power laws relating frequencies... more In this study we show empirically on a Turkish corpus that Zipf's power laws relating frequencies of words to rank, vocabulary and number of meanings can also be derived from term x document analysis. This supports the idea put forward by Ding that Zipfian distributions and SVD carry similar hidden (latent) semantics.
Uploads
Papers