Differences
This shows you the differences between two versions of the page.
|
projs:clans:docs:entity_disambiguation_xu:testrelevancy [2014/01/21 09:33] xmill.zod created |
projs:clans:docs:entity_disambiguation_xu:testrelevancy [2014/01/21 09:33] (current) xmill.zod |
||
|---|---|---|---|
| Line 4: | Line 4: | ||
| ===== Parameters ===== | ===== Parameters ===== | ||
| - name: weiboid type: string description: the unique id of a weibo post | - name: weiboid type: string description: the unique id of a weibo post | ||
| - | ===== Returns ===== | + | ===== Return ===== |
| - A boolean value represents this post is related to the person or not | - A boolean value represents this post is related to the person or not | ||
| + | |||
| ===== Detail Information ===== | ===== Detail Information ===== | ||
| * First find out which filtering keywords that related to the person appeared in the post. | * First find out which filtering keywords that related to the person appeared in the post. | ||
| * Then, use the product of these keywords' weight as the weight of this post. | * Then, use the product of these keywords' weight as the weight of this post. | ||
| * Set the threshold as the weight product of the keywords that apeared in the stock name of the related company. If the weight of the post is bigger than the threshold, then we take it as a related one. Otherwise, it will be labbeled as an unrelated one. | * Set the threshold as the weight product of the keywords that apeared in the stock name of the related company. If the weight of the post is bigger than the threshold, then we take it as a related one. Otherwise, it will be labbeled as an unrelated one. | ||
