System for creating a dictionary
AssigneeMicrosoft Corporation
InventorJoseph E. Pentheroudakis
Life of this patent
Estimating the term…
Abstract
A computer readable medium has computer executable components that include a morphological analyzer capable of using a corpus of words to automatically form a dictionary containing words associated with respective lemmas and respective parts of speech. The computer executable components also include a dictionary analyzer capable of automatically improving such a dictionary.
Claims22
A method for creating a dictionary of words for a language, each entry in the dictionary indicating a part of speech for the word and a lemma for the word, the method comprising:
- selecting a corpus of words;
- analyzing the corpus of words with a morphological analyzer to assign a part of speech and a lemma to the words of the corpus to generate a dictionary entry;
- and
- storing the dictionary entry in the dictionary.
A computer readable medium having computer executable components comprising:
- a morphological analyzer capable of using a corpus of words to form a dictionary containing words associated with a lemma and a part of speech;
- and
- a dictionary analyzer capable of automatically improving the dictionary.
Dependencies are inferred from the claim text.