Table 1 Examples of labeling functions.
From: Materials information extraction via automatically generated corpus
Labeling function | Description |
|---|---|
LF_temperature_words | If the sentence contains words related to temperature, such as solvus, γ’, we label True. |
LF_for | If the sentence contains “ ‘γ’ solvus temperature’ for ‘superalloy’”, we label True. |
LF_oneMatch | If there is only one superalloy and one γ’ solvus temperature in a sentence, we label True. |
LF_temperature_left | If the γ’ solvus temperature is included in the parentheses after the superalloy, we label True. |
LF_alloy_twoExpress | If a superalloy has two expressions, we label True. |
LF_in | If there is a sentence of “ ‘γ’ solvus temperature’ in ‘superalloy’”, we label True. |
LF_equal | If there is a keyword “equal” near the superalloy and γ’ solvus temperature, we label True. |
LF_hasTem | If there is a sentence pattern “has a temperature of”, we label True. |
LF_between_and | If the γ’ solvus temperature is within a certain interval, we label True. |
LF_be | If it is clear what the γ’ solvus temperature of the superalloy is, we label True. |
