The vocabulary transfer is the proportion of entities (without repetition) that appear both in the training/development set as well as in the test corpus.
A widespread baseline approach for NER methods is the vocabulary transfer, defined as the proportion of entities (without repetition) that are present both in the training/development data as well as in the test corpus.
Therefore, a large proportion of entities present in one cluster under optimal nitrogen may be separated into a significantly different cluster when nitrogen is limiting.
At the industry level, there is a statistically significant interaction effect between party membership and proportion of state-owned entities in an industry.
Especially, as the proportion of state-owned entities in an industry increases, the income gaps among various social groups widen substantially.
As the proportion of state-owned entities increases, the income advantage of working at state/collective units becomes even larger, suggesting that people enjoy cumulative benefits for working at monopolized sectors in monopolized industries.
Finally, model 6 also shows that the income gap between different educational groups widens as the industry becomes more monopolized, i.e., the income return from education increases as the proportion of state-owned entities in an industry increases.
At the region level, we include GDP per capita and the average years of schooling; at the industry level, we include variables indicating industry type, industry age, industry scale, the sexual and educational composition of industry employees, average revenue per capita, as well as proportion of state-owned entities representing the level of state monopoly for an industry.
Similar to what we found for place of residence and party membership, among all the region and industry factors considered here, the only significant cross-level interaction effect is for industrial monopoly (i.e., the proportion of state-owned entities in an industry).
Similarly, the high average income for high technology industries (with more employees with higher education) and monopolized industries (with a higher proportion of state-owned entities) is also partly a reflection that these industries attract more individuals having better human and political capital (Zhang 2004).
PPInterFinder was evaluated at the sentence level where a true positive represents the number or proportion of interacting protein entities that were correctly identified by the system.
