The identified bio-term associations are evaluated using Chem2Bio2RDF.
The higher the entropy is, the more diverse the bio-term is over topics.
The first bio-term, tuberculosis, is an infectious lung disease caused by various strains of mycobacterium.
In order to get a quantitative measurement of the goodness of our Bio-LDA model in discovering the bio-term associations, the bio-term pairs in Chem2Bio2Rdf are used as the gold standard.
As shown in table 8, only few bio-term pairs are identified using the co-occurrence method.
The last bio-term, cholesterol, is a waxy steroid metabolite found in the cell membranes and transported in the blood plasma of all animals.
As discussed in section 3.3.1, we can generate bio-term associations by combining the linked data resources (i.e. Chem2Bio2Rdf) and literature resources (Bio-LDA model).
The input query for the connectivity map is ((vi, t), t1, t2), where (vi, t) is the specified bio-term and its type (e.g. Alzheimer's Disease, Disease), t1 and t2 are the bio-term types that formed the connectivity maps (e.g., drug and gene).
In our Bio-LDA model, we can compute the bio-term entropies over topics as shown in equation 6, which indicates that bio-terms tend to address a single topic or cover multiple topics.
Table 10 shows the symmetric KL divergence for pairs of bio-terms in this use case, and n shows the number of co-occurrence of the given bio-term pair.
