Social Media Bots, Recommender Systems and Wiki Technology

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Social Media Bots, Recommender Systems and Wiki Technology

Adam Sobieski
Introduction

A recent research article of mine, Artificial Wisdom [1], pertains to search engines and recommender systems for: anecdotes, proverbs, quotations, lyrics, poetry, narratives (e.g. parables, allegories), and humor. I refer to these materials as wisdom materials.

A variety of social media bot is described with an integrated recommender system for wisdom materials. Wiki technology is described as a reasonable option to consider for producing and maintaining dynamic corpora of wisdom materials to be utilized by such systems.

Social Media Bots with Integrated Recommender Systems

Wisdom materials could be provided to users daily, quote-of-the-day services, or on an as-needed basis, whenever a social media bot determines that a wisdom material could provide value to a user. Recommended wisdom materials could be sent to users via a number of possible channels (e.g. social media websites, instant messenger applications, e-mail).

To ensure that recommended wisdom materials provide value to users, recommender systems could utilize users’ social media posts and other content to obtain context data. Systems could process users’ posts and content including to estimate users’ affect or complex mental states. Should a user indicate that they are sad, utilizing a status update or by otherwise sharing that information in a post with their friends and a social media bot, the social media bot could send one or more wisdom materials to uplift the user.

Recommender systems for wisdom materials, operating at scale, could learn and improve over the course of time. Users could be provided with a number of options for providing feedback with respect to wisdom material recommendations. Systems could also process users’ social media posts and content before, during and after encounters with recommended wisdom materials. Did a recommended wisdom material provide value to, e.g. uplift, a user in a context?

Utilizing technology such as Solid, social media bots with integrated recommender systems can provide value to users while simultaneously protecting users’ data privacy.

Wiki Technology

Wiki technology is a reasonable option to consider for producing and maintaining dynamic corpora of wisdom materials. Such Wiki systems could be utilized by organizations stewarding social media bots, by larger numbers of users, or by the general public.

Some varieties of wisdom materials tend to make use of figurative language. Accordingly, the indexing of such materials shouldn’t necessarily utilize the literal text of the items. For instance, the proverb “a rolling stone gathers no moss” is not best indexed and searched for by its lexemes. One desires to search, somehow, for the contents of the interpretations of the items rather than for the literal text of the items. Contextual searching is also desirable. Varieties of wisdom materials should be modeled; a wisdom material, for instance, could have multiple interpretations and each interpretation could have text content for indexing, keywords, categories, and so forth.

The modeling of wisdom materials, Wiki-based technology for producing and maintaining dynamic corpora of wisdom materials, and social media bots with integrated recommender systems for wisdom materials are indicated to be contemporary research topics.


Best regards,
Adam Sobieski

[1] http://www.phoster.com/artificial-wisdom/

_______________________________________________
Wiki-research-l mailing list
[hidden email]
https://lists.wikimedia.org/mailman/listinfo/wiki-research-l