PARLY Lithuania

PARLY Lithuania: What is said in Parliament?

How To Use PARLY

PARLY is available only in the official language.

Terms Mode: Search for topics and terms that interest you. The preview shows how frequent the term is used in debates overall.

Speaker Mode: Find out what your member of parliament is saying about the topics you care about.

Parties Mode: Find out which party is talking about what topics and terms more often.

Gender Mode: Find out what terms and topics are talked about by female as opposed to male MPs and vice-versa.

Two ways to display the results: Absolute Hits show the total number of hits within the selected year, while Relative Distribution shows the hits for the selected filter as a share of all hits for the searched term.

By clicking on the data points, you can click & read through the speeches in which the searched term was mentioned, and dig into the whole parliamentary sessions.

About The Data

Session & Transcript Collection: Debate transcripts of the Lithuanian Parliament (Seimas) were scraped from lrs.lt for the 12th (2016–2020) and 13th (2020–2024) Seimas on 7 January 2026. A total of 67 PDF collections were downloaded, covering sessions from November 2016 through November 2024.

Text Extraction: Text was extracted from all PDFs, speakers and party affiliations identified, and individual statements collapsed into single rows. This yielded 216,134 statements.

Translation of Session Metadata: Session names and dates were translated from Lithuanian to English using the Google Gemini AI model (gemini-3-flash-preview).

MP Data Extraction and Matching: MP data (name, party affiliation, mandate dates) was scraped from lrs.lt for the 12th–14th Seimas and matched to statements via abbreviated speaker name and session date. Duplicates from overlapping mandates were resolved by collapsing party affiliations; transcript-extracted party info served as fallback.

Gender Identification: Speaker gender was identified using Google Gemini (gemini-3-flash-preview) with a structured output schema (male/female/unknown).

---

Version 2.0 - February 2026

Contact: michael.jennewein@fes.de