Classifying and grouping sentences using machine learning
A device that includes an enterprise data indexing engine (EDIE) configured to receive a set of sentences and to compare the words in the sentences to a set of predefined keywords. The EDIE is further configured to identify one or more sentences that do not contain any of the keywords and to associate the identified sentences with a first classification type. The EDIE is further configured to identify a sentence that contains one or more keywords and to associate the sentence with a second classification type. The EDIE is further configured to link together the sentence that is associated with the second classification type and the sentences that are associated with the first classification type.
1. A device, comprising:
a memory operable to store a set of keywords, wherein each keyword is associated with an action; and
an enterprise data indexing engine implemented by a processor, configured to:
receive a plurality of sentences, wherein each sentence is linked with a location identifier that identifies a location in a document where a particular sentence is located;
compare words in each of the plurality of sentences to the set of keywords;
identify one or more sentences from the plurality of sentences that do not contain any of the keywords;
associate the one or more sentences that do not contain any of the keywords with a first classification type;
identify a sentence from the plurality of sentences that contains one or more keywords, wherein the sentence is identified after associating the one or more sentences with the first classification type;
associate the sentence that contains one or more keywords with a second classification type;
link the sentence associated with the second classification type with the one or more sentences associated with the first classification type; and
train a sentence classification neural network with the sentence associated with the second classification type and the one or more sentences associated with the first classification type that have been linked.
2. The device of claim 1 , wherein the plurality of sentences are ordered sequentially based on their location in the document.
3. The device of claim 1 , wherein:
the first classification type indicates that the sentence is an explanatory sentence that provides context information; and
the second classification type indicates that the sentence is an actionable sentence that provides instructions for performing an action.
4. The device of claim 1 , wherein the enterprise data indexing engine is further configured to:
input a set of sentences from the plurality of sentences that do not contain any of the keywords into a machine learning model;
receive a set of classifications for the set of sentences from the machine learning model;
determine one or more sentences from the set of sentences are associated with the first classification type based on the set of classifications;
determine a sentence from the set of sentences is associated with the second classification type based on the set of classifications; and
link the sentence from the set of sentences that is associated with the second classification type with the one or more sentences from the set of sentences that are associated with the first classification type.
5. The device of claim 1 , wherein a location identifier indicates a section within the document where a sentence is located.
6. The device of claim 1 , wherein a location identifier indicates a paragraph within the document where a sentence is located.
7. The device of claim 1 , wherein the enterprise data indexing engine is further configured to identify the one or more keywords contained within the sentence associated with the second classification type.
8. A method, comprising:
receiving a plurality of sentences, wherein each sentence is linked with a location identifier that identifies a location in a document where a particular sentence is located;
comparing words in each of the plurality of sentences to a set of predefined keywords, wherein each keyword is associated with an action;
identifying one or more sentences from the plurality of sentences that do not contain any of the keywords;
associating the one or more sentences that do not contain any of the keywords with a first classification type;
identifying a sentence from the plurality of sentences that contains one or more keywords, wherein the sentence is identified after associating the one or more sentences with the first classification type;
associating the sentence that contains one or more keywords with a second classification type;
linking the sentence associated with the second classification type with the one or more sentences associated with the first classification type; and
training a sentence classification neural network with the sentence associated with the second classification type and the one or more sentences associated with the first classification type that have been linked.
9. The method of claim 8 , wherein the plurality of sentences are ordered sequentially based on their location in the document.
10. The method of claim 8 , wherein:
the first classification type indicates that the sentence is an explanatory sentence that provides context information; and
the second classification type indicates that the sentence is an actionable sentence that provides instructions for performing an action.
11. The method of claim 8 , further comprising:
inputting a set of sentences from the plurality of sentences that do not contain any of the keywords into a machine learning model;
receiving a set of classifications for the set of sentences from the machine learning model;
determining one or more sentences from the set of sentences are associated with the first classification type based on the set of classifications;
determining a sentence from the set of sentences is associated with the second classification type based on the set of classifications; and
linking the sentence from the set of sentences that is associated with the second classification type with the one or more sentences from the set of sentences that are associated with the first classification type.
12. The method of claim 8 , wherein a location identifier indicates a section within the document where a sentence is located.
13. The method of claim 8 , wherein a location identifier indicates a paragraph within the document where a sentence is located.
14. The method of claim 8 , further comprising identifying the one or more keywords contained within the sentence associated with the second classification type.
15. A non-transitory computer readable medium storing instructions that when executed by a processor cause the processor to:
receive a plurality of sentences, wherein each sentence is linked with a location identifier that identifies a location in a document where a particular sentence is located;
compare words in each of the plurality of sentences to a set of predefined keywords, wherein each keyword is associated with an action;
identify one or more sentences from the plurality of sentences that do not contain any of the keywords;
associate the one or more sentences that do not contain any of the keywords with a first classification type;
identify a sentence from the plurality of sentences that contains one or more keywords, wherein the sentence is identified after associating the one or more sentences with the first classification type;
associate the sentence that contains one or more keywords with a second classification type;
link the sentence associated with the second classification type with the one or more sentences associated with the first classification type; and
train a sentence classification neural network with the sentence associated with the second classification type and the one or more sentences associated with the first classification type that have been linked.
16. The non-transitory computer readable medium of claim 15 , wherein the plurality of sentences are ordered sequentially based on their location in the document.
17. The non-transitory computer readable medium of claim 15 , wherein:
the first classification type indicates that the sentence is an explanatory sentence that provides context information; and
the second classification type indicates that the sentence is an actionable sentence that provides instructions for performing an action.
18. The non-transitory computer readable medium of claim 15 , further comprising instructions that when executed by the processor causes the processor to:
input a set of sentences from the plurality of sentences that do not contain any of the keywords into a machine learning model;
receive a set of classifications for the set of sentences from the machine learning model;
determine one or more sentences from the set of sentences are associated with the first classification type based on the set of classifications;
determine a sentence from the set of sentences is associated with the second classification type based on the set of classifications; and
link the sentence from the set of sentences that is associated with the second classification type with the one or more sentences from the set of sentences that are associated with the first classification type.
19. The non-transitory computer readable medium of claim 15 , wherein a location identifier indicates a section within the document where a sentence is located.
20. The non-transitory computer readable medium of claim 15 , wherein a location identifier indicates a paragraph within the document where a sentence is located.