Table 1.
Comparison of previous studies on topic evolution analysis
References | Research area | Topic detection | Factors of measuring topic intensity | Factors of measuring topic status | Methods of path evolution |
---|---|---|---|---|---|
Wei et al., 2020 | Parliamentary text | Community detection | N | FC | DTS |
Jian et al., 2018 | Microblog | LDA | Probability distribution of blog-topic | N | DTS; CTS |
Liu et al., 2017 | Scientific literature | PLDA | ND | FC | DTS; CTS |
Miao et al., 2020 | Patents | LDA | CWT | N | DTS |
Chen et al., 2018 | Patents | LDA | ND | N | DTS; CTS; HMM |
Gao et al., 2020 | Short text | OCCTM | N | N | DTS; CTS |
Current study | Web text | LDA | ND; CWT | FC; CWT | Topic drifting probability |
PLDA parallel latent dirichlet allocation, ND number of documents, CWT contribution of words to a topic, FC frequency of co-occurrence, DTS division of time slices, CTS calculation of topic similarity
*The letter N represents ‘without the referred research point’