The earlier semi-supervised models required experts to provide seed aspects.
Knowledge-base topic models are provided with domain specific knowledge rules instead of seed aspects, following a semi-supervised approach.
For the first proposed topic model, the Fine-grained Labeled LDA (FL-LDA), seeding aspects are applied to guide the model to discover words that are related to these seeding aspects.
More specifically, the proposed methodology first extracts seeding aspects and related terms from detailed product descriptions readily available on E-commerce websites.
In order to reduce the gap in accuracy of supervised and semi-supervised topic models, several knowledge-based models were proposed that improved results with manually provided knowledge rules instead of seed aspects.
For the second model, the Unified Fine-grained Labeled LDA (UFL-LDA), we incorporate unlabeled documents to extend the FL-LDA model so that words related to the seeding aspects or other high-frequency words in customer reviews are extracted.
Thus, in the case of metastable medium (DMEM), the "seeding" aspects of the boiled serum can be more readily revealed.
The flow cytometry of joint embryos and endosperms, seed aspects and microsatellite analysis together led to evidence that both male and female clones produced 2 n gametes that mated with x gametes to produce viable 3 x plants.
These changes may be one of the reasons for the poor LC of metastatic lung tumors, but both the "seed" aspect and the "soil" aspect may be important.
These derivatives include factor from density (FFD), volume of seeds (VOL), aspect ratio (ASPECT), horizontal and vertical deviation from optimal ellipses (HDFE and VDFE).
