Computerized method of training a computer executed model for recognizing numerical quantities
View Patent ↗A computerized method for training a computer executed model for recognizing numerical quantities is provided. An input, atleast one unit expression, is received by an input module. The input module may then search for numeric values and the unit expression in a text corpus, wherein, the text corpus comprises sets of words and frequency of occurrence of each of the sets. The input module may identify identified sets, wherein the identified sets may comprise a combination of a numeric value and the unit expression. A synthetic text generation module may then generate sentences from the text corpus by applying the identified sets as input. A training dataset may be generated by a labeling module by auto labelling features in the generated sentences based on the numeric value and the unit expression and further a training module may train the training model by providing input based on the training dataset.
1. A computerized method of training a computer executed model for recognizing numerical quantities, the method carried out by one of more processors, the method comprising:
receiving, as input, atleast one unit expression;
searching for numeric values and the unit expression in a text corpus, the text corpus comprising sets of words and frequency of occurrence of each of the sets, the search resulting in identification of sets that comprise a combination of a numeric value and the unit expression;
generating sentences from the text corpus by applying the identified sets as input;
generating a training dataset by auto labelling in the generated sentences based on the numeric value and the unit expression; and
training the model by providing input based on the training dataset.
2. The method of claim 1 , further comprising:
evaluating performance of the model using a validation dataset;
obtaining sentence generation feedback based on the evaluation; and
applying the sentence generation feedback for tuning generation of the sentences from the text corpus.
3. The method of claim 1 , further comprising:
evaluating performance of the model using a validation dataset;
obtaining training module feedback based on the evaluation; and
applying the training module feedback for tuning training of the model.
4. The method of claim 1 , wherein the sets of words are sets of bigrams.
5. The method of claim 1 , wherein generating sentences from the text corpus comprises identifying, based on the text corpus, words that appear before and after each of the words in the identified sets.
6. The method of claim 5 , wherein each of the identified sets comprises two words, wherein generating the sentences comprises identifying, words that appear before a first of the two words and words that appear after a second of the two words.
7. The method of claim 5 , wherein words in the identified set are adjacent to each other in the sentence formed based on the identified set.
8. The method of claim 1 , comprises receiving a plurality of unit expressions as input, wherein each of the identified sets comprise a combination of a numeric value and any one of the unit expressions.
9. A computerized system for training a computer executed model for recognizing numerical quantities, the system comprising one or more processors configured to:
receive, as input, atleast one unit expression;
search for numeric values and the unit expression in a text corpus, the text corpus comprising sets of words and frequency of occurrence of each of the sets, the search resulting in identification of sets that comprise a combination of a numeric value and the unit expression;
generate sentences from the text corpus by applying the identified sets as input;
generate a training dataset by auto labelling in the generated sentences based on the numeric value and the unit expression; and
train the model by providing input based on the training dataset.
10. The system of claim 9 , wherein the one or more processors are further configured to:
evaluate performance of the model using a validation dataset;
obtain sentence generation feedback based on the evaluation; and
apply the sentence generation feedback for tuning generation of the sentences from the text corpus.
11. The system of claim 9 , wherein the sets of words are sets of bigrams.
12. The system of claim 9 , wherein the one or more processors are configured to generate sentences from the text corpus by identifying, based on the text corpus, words that appear before and after each of the words in the identified sets.
13. The system of claim 12 , wherein words in the identified set are adjacent to each other in the sentence formed based on the identified set.
14. The system of claim 9 , wherein the one or more processors are configured to receive a plurality of unit expressions as input, wherein each of the identified sets comprise a combination of a numeric value and any one of the unit expressions.