A Connectionist Corpus-Based Approach to the Building of Word Representations (1994)
Rupert L. Tang
A system for building distributed representations for words using a sliding window is described here which takes a corpora of sentences from either artificial data or natural data. The system is able to develop representations that capture the regularity of word occurrences underlying a given body of text. The paradigm presented here is based on a FGREP module with some variations on the architecture and the training methodology. Experiments are done to show that the system is capable of capturing the surface regularity of words.
View:
PDF
Citation:
Technical Report HR-94-01, Department of Computer Science, The University of Texas at Austin, 1994.
Bibtex:

Lappoon R. Tang Undergraduate Alumni ltang [at] utb edu
FGREPNET The FGREPNET package contains the C-code and data for training and testing an FGREP network in developing distributed re... 1994