[INDOLOGY] Introduction of MITRA Search, Deep Research, and DharmaNexus

[Sebastian Nehrdich]

Dear Colleagues,

We are happy to announce the launch of several new, powerful, and
freely-available digital tools from the Dharmamitra <http://dharmamitra.rg>
project, designed to aid in the study of Sanskrit and other classical Asian
languages. While many of you may be familiar with our translation and
grammar tools, we have now released a suite of services focused on semantic
search and intertextuality exploration.


*1. MITRA Search*MITRA Search is a semantic search engine that understands
the meaning behind a query, not just keywords. This allows for nuanced
discovery within large textual corpora. For Sanskritists, this means you
can:
Search for specific philosophical concepts (e.g., the definition of cetanā)
and find relevant passages even if they do not use the exact wording or are
even written in the same language.
Find instances of a particular metaphor or analogy across different texts
and genres.
Use a Sanskrit passage as a query to find conceptual parallels elsewhere in
the corpus, or in its Tibetan and Chinese translations.
In addition to semantic search, we also provide an 'exact search' that uses
the word segmentation system developed by Oliver Hellwig and me in
combination with BM25 retrieval to provide a more focussed search
experience. Since large scale search across the various Sanskrit datasets
is a significant need in our field, we hope that this can be useful to
people on this mailing list.
You can learn more here:
https://dharmamitra.github.io/dharmamitra-guides/mitra_tools/search/

*2. DharmaNexus <http://dharmanexus.org>*
As the actively maintained successor to the BuddhaNexus platform,
DharmaNexus is a powerful tool for exploring intertextual relationships. It
allows researchers to see how passages from Sanskrit texts are cited,
adapted, or paralleled in other works. It also features comprehensive
parallel data aligned on sentence level between Buddhist Sanskrit works and
their counterparts in Tibetan and Chinese, which can be especially helpful
for people working on material that has translations into these languages.
Explore DharmaNexus:
https://dharmamitra.github.io/dharmamitra-guides/dharmanexus/

*3. MITRA Deep Research Translation Mode*
While not many on this list are longing for machine translation, we now
provide a system that combines our powerful semantic search with a broader
analytical framework, integrating references into the actual Sanskrit and
related literature to provide a contextually informed translation with
references into the texts stored in DharmaNexus. This system is also able
of locating relevant passages in other languages like Tibetan and Chinese,
which can have positive impact on translation quality and provide further
important references.
Explore MITRA Deep Research:
https://dharmamitra.github.io/dharmamitra-guides/mitra_tools/deep_research/


*Sanskrit Data Sources*These tools are powered by a growing collection of
digitized Sanskrit texts. The current core of our Sanskrit data is sourced
from GRETIL, the recently releeased Muktabodha collection, and the  Digital
Sanskrit Buddhist Canon (DSBC) as well as a number of other individual
texts that the MITRA team has collected and digitized. We are extremely
grateful to these institutions and proejcts to make their data available
under permissive licenses, which enables the creation of meta resources
like this.
We are also continuously working to expand our corpus. There is currently a
number of texts held back for technical reasons, but we will update the
system very soon to patch the gaps. Going forward, Dharmamitra will aim for
quarterly data updates. You can find more details on the Sanskrit data
here: https://github.com/dharmamitra/dharmanexus-sanskrit/

If you are interested in collaboration, of course feel free to reach out to
us. We are happy to provide API access and share data resources
wherever possible.

All the best,

The Dharmamitra Team
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://list.indology.info/pipermail/indology/attachments/20250727/cdee2c3a/attachment.htm>