Table 1.
Details of the texts used for topic modelling.
Fig 1.
Topic modelling framework for comparison of the Upanishads with the Bhagavad Gita.
Table 2.
Processed text after removing special characters and transforming archaic words into modern English.
Table 3.
Dataset statistics.
Fig 2.
Chapter wise word count for different texts in the dataset.
Fig 3.
Leading bigrams and trigrams for the Bhagavad Gita and Upanishads by Eknath Easwaran.
Fig 4.
Leading bigrams and trigrams for the Ten Principal Upanishads.
Fig 5.
Topics of the Ten Principal Upanishads and some of their relevant documents(Model: USE-HDBSCAN-UMAP).
Fig 6.
Visualisation of top 20 words, and top 10 bigrams and trigrams for the Bhagavad Gita.
Fig 7.
Topics of Bhagavad Gita and the most relevant documents(Model: USE-HDBSCAN-UMAP).
Table 4.
Value of topic coherence metric (TC-NPMI) for different corpus.
Fig 8.
Visualization of the semantic space of the Bhagavad Gita (Eknath Easwaran) and the Upanishads (Eknath Easwaran) with topic labels.
Fig 9.
Visualisation of different topics of 108 Upanishads.
Fig 10.
Visualization of the semantic space of different parts (based on 4 Vedas) of 108 Upanishads.
Fig 11.
Heatmap showing the similarity between different topics of Bhagavad Gita (Eknath Easwaran) and Upanishads (Eknath Easwaran) generated from a selected approach (SBERT-UMAP-HDBSCAN).
Table 5.
Topics of the Bhagavad Gita(Eknath Easwaran) with most similar topics from the Upanishads(Eknath Easwaran).
Table 6.
Topics of the Bhagavad Gita(Eknath Easwaran) with most similar topics from the Ten Principal Upanishads(Shri Purohit Swami & W.B. Yeats).
Fig 12.
Heatmap showing the similarity between different topics of Bhagavad Gita (Eknath Easwaran) and the Ten Principal Upanishads (Shri Purohit Swami & W.B. Yeats) generated from a selected approach (SBERT-UMAP-HDBSCAN).
Table 7.
Classification of 108 Upanishads based on the four key Vedas.
Note that the original Yajur Veda is divided into two parts (Krishna- Yajur-Veda and Sukla-Yajur-Veda).
Fig 13.
Comparison of dimensionality reduction and visualisation by PCA and UMAP for the combined semantic space of the Bhagavad Gita (Eknath Easwaran) and the Upanishads (Eknath Easwaran).
Note that the PCA has 4.4% explained variance ratio for dim = 1 and 3.7% for dim = 2, taken from total of 500 dimensions in original data.