Translation of unstructured text into semantic models
Semantic model usage includes receiving, via a graphical user interface, input of unstructured text from a user. While receiving the input, the input is compared to semantic models in a semantic library, and from the semantic model, a semantic model satisfying a threshold is selected. A suggestion is generated from the semantic model, and presented, in the graphical user interface, to the user. An acceptance of the suggestion may be received. The input is augmented with the suggestion and stored.
1. A method for semantic model usage, the method comprising:
receiving, via a graphical user interface on a computer system, input of unstructured text from a user;
while receiving the input:
providing the input to a plurality of semantic models in a semantic library,
obtaining a plurality of similarities, wherein each of the plurality of similarities is calculated using the input and one of the plurality of semantic models, wherein a similarity of the plurality of similarities is a cosine similarity,
selecting, using at least one of the plurality of similarities, a semantic model from the plurality of semantic models, wherein the semantic model is selected based on having the cosine similarity of the semantic model being less than a threshold,
generating a suggestion from the semantic model, wherein the suggestion comprises at least a sub-sentence from the semantic model, wherein the semantic model comprises a plurality of sub-sentences, wherein at least one of the plurality of sub-sentences is associated with a parameter, and wherein the sub-sentence is one of the plurality of sub-sentences,
presenting, in the graphical user interface, the suggestion to the user,
receiving, from the user, an acceptance of the suggestion, and
augmenting the input with the suggestion to create an augmented input; and
storing the augmented input.
2. The method of claim 1 , further comprising:
extracting content from the augmented input; and
processing the extracted content.
3. The method of claim 1 , wherein the augmented input is a revision of the unstructured text to match the suggestion.
4. The method of claim 1 , wherein calculating the cosine similarity is performed using vectorized skipgrams.
5. The method of claim 1 , further comprising:
processing a plurality of unstructured text documents to obtain the plurality of semantic models.
6. The method of claim 1 , wherein the plurality of semantic models are specific to a professional domain.
7. A system for semantic model usage, the system comprising:
a computer processor;
a graphical user interface that:
receives, via the graphical user interface, input of unstructured text from a user, and
while receiving the input:
presents, in the graphical user interface, a suggestion to the user,
receives, from the user, an acceptance of the suggestion, and
a text analyzer that:
while receiving the input:
provides the input to a plurality of semantic models in a semantic library,
obtains a plurality of similarities, wherein each of the plurality of similarities is calculated using the input and one of the plurality of semantic models, wherein a similarity of the plurality of similarities is a cosine similarity,
selects, using at least one of the plurality of similarities, a semantic model from the plurality of semantic models, wherein the semantic model is selected based on having the cosine similarity of the semantic model being less than the threshold
generates the suggestion from the semantic model, wherein the suggestion comprises at least a sub-sentence from the semantic model, wherein the semantic model comprises a plurality of sub-sentences, wherein at least one of the plurality of sub-sentences is associated with a parameter, and wherein the sub-sentence is one of the plurality of sub-sentences,
augments the input with the suggestion to create augmented input; and
stores the augmented input.
8. The system of claim 7 , further comprising:
a content extractor that:
extracts content from the augmented input; and
processes the extracted content.
9. The system of claim 7 , wherein the augmented input is a revision of the unstructured text to match the suggestion.
10. The system of claim 7 , wherein calculating the cosine similarity is performed using vectorized skipgrams.
11. The system of claim 7 , further comprising:
a semantic modeling tool that:
processes a plurality of unstructured text documents to obtain the plurality of semantic models.
12. The system of claim 7 , wherein the plurality of semantic models are specific to a professional domain.
13. A non-transitory computer readable medium for semantic model usage, the non-transitory computer readable medium comprising computer readable program code for:
receiving, via a graphical user interface, input of unstructured text from a user;
while receiving the input:
providing the input to a plurality of semantic models in a semantic library,
obtaining a plurality of similarities, wherein each of the plurality of similarities is calculated using the input and one of the plurality of semantic models, wherein a similarity of the plurality of similarities is a cosine similarity,
selecting, using at least one of the plurality of similarities, from the plurality of semantic models, a semantic model from the plurality of semantic models, wherein the semantic model is selected based on having the cosine similarity of the semantic model being less than the threshold,
generating a suggestion from the semantic model, wherein the suggestion comprises at least a sub-sentence from the semantic model, wherein the semantic model comprises a plurality of sub-sentences, wherein at least one of the plurality of sub-sentences is associated with a parameter, and wherein the sub-sentence is one of the plurality of sub-sentences,
presenting, in the graphical user interface, the suggestion to the user,
receiving, from the user, an acceptance of the suggestion, and
augmenting the input with the suggestion to create augmented input; and
storing the augmented input.
14. The non-transitory computer readable medium of claim 13 , further comprising computer readable program code for:
extracting content from the augmented input; and
processing the extracted content.
15. The non-transitory computer readable medium of claim 13 , wherein the augmented input is a revision of the unstructured text to match the suggestion.
16. The non-transitory computer readable medium of claim 13 , wherein calculating the cosine similarity is performed using vectorized skipgrams.
17. The non-transitory computer readable medium of claim 13 , wherein the plurality of semantic models are specific to a professional domain.