Loading video...

Video Failed to Load

Go Home

“Can we get a new text analysis tool?” “No—we have Topic Model at home” Topic Model at home: outputs vague keywords; needs constant parameter fiddling🫠 Is there a better way? We introduce LLooM, a concept induction tool to explore text data in terms of interpretable concepts🧵

38,102 views • 2 years ago •via X (Twitter)

11 Comments

Michelle Lam's profile picture
Michelle Lam2 years ago

Analysts have questions like “How are women in power described?” Vague topics like “women, power, female” aren’t what they’re after—they want to understand data with nuanced concepts like “Criticism of traditional gender roles,” which are central to theory-driven data analysis.

Michelle Lam's profile picture
Michelle Lam2 years ago

LLooM fills this gap with concept induction, extracting high-level concepts defined by natural language descriptions & explicit inclusion criteria (e.g., “Dismissal of women’s concerns: Does the text dismiss or invalidate women’s concerns or experiences?”)

Michelle Lam's profile picture
Michelle Lam2 years ago

Our algorithm draws on the ability of large language models (LLMs) to generalize from examples. LLooM samples extracted text and iteratively synthesizes proposed concepts of increasing generality. Once concepts are induced, LLooM can classify the entire dataset.

Michelle Lam's profile picture
Michelle Lam2 years ago

We instantiate our algorithm in the LLooM Workbench, an open-source text analysis tool for computational notebooks. Users can explore text data in terms of high-level concepts—from a dataset overview to concept details to document-level scores, highlights, and rationale.

Michelle Lam's profile picture
Michelle Lam2 years ago

In evals, LLooM exceeds baselines to recover 70-90% of human-annotated, generic topics, and LLooM is significantly better at surfacing specific, nuanced concepts. From content moderation to AI ethics statements, LLooM concepts help us make sense of data:

Michelle Lam's profile picture
Michelle Lam2 years ago

Going further, expert data analysts used LLooM to uncover novel insights even on familiar datasets. Analysts were particularly excited that LLooM facilitated theory-driven analysis: they could read out patterns and write hypotheses through the language of LLooM concepts.

Michelle Lam's profile picture
Michelle Lam2 years ago

This work would not be possible without my amazing collaborators and advisors—Janice Teoh, @landay, @jeffrey_heer, and @msbernst! Start using LLooM to explore text data via concepts ( or see our #CHI2024 paper to learn more (

dana calacci 🦋 @dana.witchy.business's profile picture
dana calacci 🦋 @dana.witchy.business2 years ago

hell yeah. we've been using a similar method that's far less general than this for analyzing unstructured reddit data—so seeing some of the same ideas here, but generalized and tested rigorously is awesome

Michelle Lam's profile picture
Michelle Lam2 years ago

that's great! yes, there's a lot of open ground for text analysis tools to better align with the way researchers want to think about their data—with this release we'd really love to learn how LLooM works for a broader range of domains and research questions!

Sheikh Shafayat's profile picture
Sheikh Shafayat2 years ago

This looks a pretty generalizable tool! Thanks for sharing, I was looking for something like this for my class project 👌

Michelle Lam's profile picture
Michelle Lam2 years ago

thank you! awesome—hope LLooM can be helpful :)

Related Videos