Research meetings: Spring 2026

SpråkbankenText's research meetings during Spring 2026, will take place on Thursdays, from 10:15 to 11.30 (occasionally to 12:00), typically in room F314, unless otherwise announced. The meetings are normally open for all members of SpråkbankenText. On some occasions, we invite a broader audience, these public seminars are announced in SpråkbankenText's calendar.

Please update the list of conference deadlines. Moreover, to access previous meeting periods, which remain in this document as commented sections, you will need to log in and navigate through it.

The schedule for the Spring of 2026 is as follows (cells highlighted in orange indicate related events that occur outside the meeting’s usual date and/or time):

Date	Topic	Comments
January
8	Discussion of grant proposals – drafts are welcome and may be circulated in advance of the meeting.	Basically for: RJ or VR
15	Discussion of grant proposals (RJ; VR) – drafts are welcome and may be circulated in advance of the meeting - and Planning meeting for Spring 2026	Be prepared with suggestions and relevants events we should note in the calender - for even greater convenience, feel free to edit it yourself before the meeting!
15	Prof Beáta Megyesi: Cracking Hidden Codes and Undeciphered Languages	Time: 15:00-17:00, Place: C350, Lisebergssalen
22	Presentation av VR-projekt - Moved later in Feb	Irene Elmerot
27	RJ Projects and Programmes 2026	Calls close today: 27 January at 03:00 PM
29	Dry-runs (i) Arianna: for presentation of the SweLL treebank (in Swedish) at the CLARIN consortium meeting; (ii) Any LREC-workshops' drafts at this stage?	Arianna: "Trädbanken SweLL: när andraspråkstalare annoterar andraspråkstexter" and possibly LREC-workshops' drafts Dimitris & Herb: "Disfluencies and ASR Performance on Swedish Spontaneous Speech from the ‘Trip to Stockholm’ Discourse Narrative Task" - we can have the 19th or the 26/2 for LREC as well - the Notification of acceptance for the main conference is on the 13th
February
4	Cheap(er) human-in-the-loop labelling strategy for better datasets? - Speaker: Dylan Pashley, Lund University	Time: 12.00 to 13.00 CET; Where: Online - link by registration HERE
4	Comprehensively Evaluating Language in Language Models - Leonie Weissweiler, Uppsala University	Time: 13:15-15:00 CET; Where: In-person (room J336) and Zoom Abstract: Abstract: As Large Language Models (LLMs) are being increasingly used in high-stakes situations, it is vital that we accurately assess their strengths, but also their limitations. To this end, I ask: how can we ensure that we neither over- nor underestimate Language Models’ linguistic capabilities? For this, evaluations must consider the full breadth of human language. In my talk, I will demonstrate how progress can be made towards this goal in two aspects: multilingual evaluation, and evaluation for the long tail of language. For multilingual evaluation, I will show how agreement evaluation can be scaled to over 100 languages. For the long tail of language, I will report results from two investigations of language models’ understanding of the so-that construction, with which even state-of-the-art models struggle, even though rich distributional information is available in their training data. I will further demonstrate how LLMs themselves can be leveraged to annotate corpora for long-tail constructions. This will further stretch the boundaries of what we are able to test. All evaluations together paint a nuanced picture of the linguistic capabilities of large language models, showing achievements as well as remaining deficits.
5	Canceled — several coworkers are away on a trip 🚆🧳🏨	---
12	Presentation av VR-projektet De förslavades röster, där vi skapar en korpus av afrikansk-amerikansk engelska.	Irene Elmerot and Leif-Jöran Olsson
19	LREC main and workshop draft papers presentations	Please feel free to share your draft paper contributions with everyone before the meeting. LREC-2026 Main Conference: Maria Irena Szawerna & Jacob Lee Suchardt; Fill-in-the-Blanks: Generating Pseudonyms for English and Swedish Texts with RoBERTa and Qwen Niklas Deworetzki & Arianna Masciolini; Syntactic Sugar for Syntactic Queries: Sequential Representations for Dependency Queries LREC-2026 Workshops: CALD-pseudo: Maria Irena Szawerna & Simon Dobnik; Birds of a Feather: Do Embedding Representations of Personal Information Stick Together? 9th UDW workshop: Christina Klironomou, [et al.] & Arianna Masciolini; Towards Universal Dependencies for L2 Learners of Modern Greek: Annotation and Challenges 6th RaPID: Dimitrios Kokkinakis, Herbert Lange and Ricardo Muñoz Sanchez; Disfluencies and ASR Performance on Swedish Spontaneous Speech from the ‘Trip to Stockholm’ Discourse Narrative Task
19	HumAI	Speaker: Olle Häggström, Time: 15-17, Location:C350, Humanisten and online
25	Explainable AI (XAI) – Concepts, Tools and Applications	Time: 12.00 to 13.00; Where: Online - link by registration; Speaker: Sule Tekkesinoglu; Spoken language: English. Abstract: Abstract: As AI continues to advance, the need to understand how these systems make decisions is becoming increasingly critical. This seminar introduces the growing field of Explainable Artificial Intelligence (XAI), which aims to make AI systems more transparent, interpretable, and accountable. We will explore why explainability matters not only for developers and researchers but also for non-technical users. Core XAI concepts, including local and global explanations and model-specific and model-agnostic methods, will be introduced, along with some application examples. We will also discuss how explanations are presented and evaluated, as well as the challenges that remain in this evolving field.
26	No meeting	Cancelled.
March
5	Céline Leuzinger: Automatic Feedback Generation - Overview of the field and initial ideas	Overview: Automatic Feedback Generation (AFG) for second language learners is a relatively new field: early works started in 2022 and focused on fine-tuning LLMs on datasets of human-crafted feedback, achieving low to medium results (0.3-0.6 F1). But LLMs’ performances have largely improved over the last few years, offering new possibilities in AFG: studies now mainly use prompting to generate feedback, and employ human annotators to review and evaluate the output. This comes with a range of complex (yet exciting) challenges: to begin with, how should we evaluate the system's output? In other words, what is "good" feedback? Besides, which linguistic features (error classification, atomic edit, etc.) should we feed into the model to further improve on the performances? Existing studies disagree on this question, but have adopted radically different evaluation methods, making comparison of results impractical. The presentation has four parts: I) What is feedback? - SLA perspectives, II) Previous approaches in AFG, III) RQ 1: how can we evaluate feedback?, IV) RQ 2: which linguistic features improve performances? The goal of this presentation is to introduce my field, and hear everyone’s thoughts, ideas and critiques on the work carried out so far.
10	Maria Thorson from the Research and Innovation Services will give a research seminar and go over what applies to research data and what support we can get.	Time: 13:15-15 ; Place: C350, Lisebergssalen
11	Amy Han Qiu: "I vote to not kill the researcher 😣/ 😂 : An experimental study of congruent and incongruent emojis in spontaneous text-based conversation"	Time: 13:15-15 ; Place: J335 or via Zoom, https://gu-se.zoom.us/j/67063108947?pwd=kPpjvMLCekxNTBVzq4uYP5gFZ6Y6vd.1 - Abstract: Emotive symbols such as emojis and emoticons play a crucial role in text-based dialogues, allowing interlocutors to express affective, pragmatic, and interpersonal nuances that are difficult to convey through text alone. While prior research has primarily examined the perceived effects of emotive symbols in non-interactive settings, their role in spontaneous, interactive conversations remains underexplored. This study investigates how exposure to emotive symbols expressing contrasting affective stances influences the unfolding of text-based chats. We conducted psycholinguistic experiments with 48 participants, who engaged in 16 triadic chats to make decisions about an ethical dilemma. Emojis with either positive or negative emotional valence were systematically inserted into participants’ messages, creating either congruent or incongruent combinations with the semantic content of the message that can be interpreted as signalling opposite affective or pragmatic stances, such as empathy or irony. In this talk, I will present our findings about emoji use and response patterns and plans for future papers.
12	Draft discussion: the SweLL treebank (Arianna, Sasha, Maria, Elena and Caroline from the MLT programme)	This is a long paper that we are planning to submit to the Language Resources and Evaluation journal. We will share the draft in advance to make it possible to read it beforehand.
17	Bill Noble (MIT); Title: Words that work: Ad-hoc lexical meaning in facilitated civic dialogue	Time: 13:15-15:00; Location: Room J406 or via Zoom. Abstract: One of the things that makes human language so incredibly flexible is the ability to use common words and phrases with meanings that are specific to a particular interaction. This phenomenon, which I will refer to as ad hoc lexical meaning, touches on a number of subfields in linguistics, including lexical and compositional semantics, pragmatics, dialogue theory, and psycholinguistics. With accounts in prominent frameworks such as relevance theory, generative lexicon and construction grammar, there is no shortage of theoretical work on the topic. The work-in-progress presented in this talk takes inspiration from theoretical accounts to develop a computational approach for studying the role of ad hoc lexical meaning in the furtherance of constructive communication. Can a better understanding ad hoc lexical meaning aid in the development of facilitation strategies for advancing constructive civic dialogue? We use the Fora Corpus, a collection of 262 facilitated multi-party dialogues gathered in partnership with civic organizations across the US, as a starting point for an empirical investigation of this question.
18	LLMs and Responsible AI	Time: 12-17. Location: Room EC, Chalmers. Information and registration here.
19	Presentation of two newly launched projects NordForsk and VR	Dana (NordForsk): Culture-Sensitive Assessment and Adjustment of Large Language Models – Adaptation to the Nordic-Baltic Societies (3 yrs) and Dimitris (VR): AI-driven language biomarkers for early detection and progression of cognitive decline (4 yrs)
19	HumAI	Speaker: Evie Coussé, Time: 15-17, Location: J222 and online Read more here
25	Ricardos final seminar	Time: 13.15-15:00; Room C350, Opponent är prof. Beáta Megyesi.
26	Can be claimed	---
April
02	Can be claimed	---
09	Can be claimed	---
16	Visit from Språkbanken CLARIN (Sara and Eva)	Samordnare: Sasha - Location: J411
20	Marias mid-term seminar	Time: 13.15–15.00, Place: J444
23	HumAI	Speaker: Anton Tönberg, Time: 15-17, Location: C350, Humanisten and online(DK out of office)
30	Can be claimed	---
May
07	Two ERASMUS+ (Mobility for Traineeship) students from the University of Parma present their background and current work at the department	Giuseppe CHIRIATTI: Whisper Adaptation for Atypical Speech (preliminary) and Davide SAPONARA: computational models for meaning and meaning change (preliminary)
7	Word Rains and Word Clouds (hybrid) Workshop	Time: 13:00-16:00. Place: C360 - Note! Reply to Mats Fridlund before the 23rd of April
14	LREC 2026 - Main conference and Workshops	11-16 May 2026 - Palau de Congressos de Palma
21	HumAI	Speaker: Ashely Green, Time: 15-17, Location: C350, Humanisten and online
28	Utpal Pandey from UT Austin	---
June
04	Learner corpus research metadata	Herbert Lange
09	SBX retreat - details TBA	Marstrands Havshotell
11	Can be claimed	---
18	HumAI	Speaker: Ted Underwood, Time: 15-17, Location: C350, Humanisten and online
25	Can be claimed	---

---