langproc - natural language processing solution
Not working yet. Using: Just run the tests.py, it holds a setup with english dictionary and sample sentence. The sentence gets scrambled in a random order and is fed to the retrieval function. The retrieval function spits out a variety of strings that have complete words and are arranged as closely to the given sentence as possible.
Tags: dictionary search, semantic and linguisstic analysis
Task: The input data is divided into scrambled columns of three characters. The objective is to retrieve the original text from this scrambled input.
Languages: Lithuanian(only latin alphabet), English.
a) Use a dictionary that is freely available or personally constructed. b) Don't use a dictionary
Guidelines:
Dictionary part can be solved by just iteratinv over all permutations of given input and finding the one with the best match. Of course, some words(especially the Proper noungs) may not be in the dictionary. Part without dictionary will have to use a model of a generalized language, that would be trained using dictionaries of various languages, by derviation of tasks or in any other way.