Automatic clustering of tokens from a corpus for grammar acquisition
View Patent ↗A system for recognizing patterns is disclosed. Grammar learning from a corpus includes, for the other non-context words, generating frequency vectors for each non-context token in a corpus based upon counted occurrences of a predetermined relationship of the non-context tokens to identified context tokens. Clusters are grown from the frequency vectors according to a lexical correlation or a cluster tree among the non-context tokens. The cluster tree is used for pattern recognition.
1. A system that recognizes patterns, the system comprising:
a first module configured to control a processor to generate frequency vectors for each non-context token in a corpus based upon counted occurrences of a predetermined relationship of the non-context tokens to context tokens;
a second module configured to control the processor to cluster the non-context tokens into a cluster tree based upon the frequency vectors according to a lexical correlation among the non-context tokens; and
a third module configured to control the processor to use the cluster tree for pattern recognition.
2. A system that performs grammar learning from a corpus, the system comprising:
a first module configured to control a processor to generate frequency vectors for each non-context token in a corpus based upon counted occurrences of a predetermined relationship of the non-context tokens to context tokens;
a second module configured to control the processor to cluster the non-context tokens based upon the frequency vectors according to a lexical correlation among the non-context tokens; and
a third module configured to control the processor to use a cluster tree for pattern recognition.
3. The system of claim 2 , wherein the second module configured to control the processor to cluster is further configured to control the processor to cluster the non-context tokens into a cluster tree.
4. The system of claim 3 , wherein the cluster tree represents a grammatical relationship among the non-context tokens.
5. The system of claim 3 , further comprising a fourth module configured to control the processor to cut the cluster tree along a cutting line to separate large clusters from small clusters.
6. The system of claim 2 , wherein small clusters are ranked according to a compactness value.
7. The system of claim 2 , wherein the predetermined relationship is a measure of adjacency.
8. The system of claim 2 , wherein the second module configured to control the processor to cluster performs the clustering based on Euclidean distances between the frequency vectors.
9. The system of claim 2 , wherein the second module configured to control the processor to cluster performs the clustering based on Manhattan distances between the frequency vectors.
10. The system of claim 2 , wherein the second module configured to cluster performs the clustering based on maximum distance metrics between the frequency vectors.
11. The system of claim 2 , further comprising a fourth module configured to control the processor to normalize the frequency vectors based upon a number of occurrences of the non-context token in the corpus.
12. The system of claim 2 , wherein the frequency vectors are multi-dimensional vectors, the number of dimensions being determined by the number of context tokens and a number of predetermined relationships of non-context tokens to the context token being counted.