Contextual Recommendation of Adages, Allusions, Anecdotes, Aphorisms, Jokes, Proverbs, Quotes, Lyrics, Poems, Stories, and Witticisms

Hello. I would like to share some ideas with the community with respect to a research and development project.

These ideas involve conversational search and recommender systems for wisdom materials including adages, allusions, anecdotes, aphorisms, jokes, proverbs, quotes, lyrics, poems, stories, and witticisms.

One implementational approach involves using vector databases for wisdom materials and using dialogue-context vectors for looking up the wisdom materials. As envisioned, when end-users engaged in dialogues with AI systems, e.g., narrating personal experiences or occurrences, AI systems could conversationally recommend ranked lists of wisdom materials to those end-users.

Any thoughts about or building upon these ideas? Thank you.

This is an interesting idea!

What you’ve described sounds like a Retrieval-Augmented-Generation approach. The assistants api may be one way to get started, though there’s quite abit of feedback about it from this forum.

One thing to consider is that the queries may not be in the same embedding space as the wisdom material. As a toy example, the individual may be seeking help about growing fruits (the literal situation), but the wisdom material mentioning fruits (as a metaphor) may not exactly be relevant to the situation.


Thank you. Those are good points about the applicability of retrieval-augmented generation and about there being potentially more than one embedding space involved.

Wisdom materials may have multiple interpretations and thus uses. For instance, the proverb “a rolling stone gathers no moss” has two popular interpretations: (1) people pay a price for being always on the move, in that they have no roots in a specific place (the original meaning), and (2) people who keep moving avoid picking up responsibilities and cares.

Without personalization, conversational search and recommender systems could utilize interactive dialogical processes to help end-users explicate their scenarios or queries to best sift and sort the space of wisdom materials.

With personalization, we might envision there being two input vectors. One would be a vector for a model of the particular end-user. The other would be a vector for their provided scenario or query. It should be possible to combine or to otherwise transform these two input vectors into a larger vector in a higher-dimensional embedding space.

Regarding modeling end-users with vectors, using the life2vec approach [1], or something like it, end-users could be modeled by means of using sequences of life-events.

We can also consider recommender systems for wisdom materials which present end-users with opportunities for providing feedback. After receiving their results, end-users could then click upon, or otherwise upvote, those wisdom materials which were most useful.


[1] Savcisens, Germans, Tina Eliassi-Rad, Lars Kai Hansen, Laust Hvas Mortensen, Lau Lilleholt, Anna Rogers, Ingo Zettler, and Sune Lehmann. “Using sequences of life-events to predict human lives.” Nature Computational Science (2023): 1-14. [PDF]

I wrote up a quick article on these interesting topics here: Artificial Intelligence and the Contextual Recommendation of Wit and Wisdom.