Papers by Alexander Spangher
Do LLMs Plan Like Human Writers? Comparing Journalist Coverage of Press Releases with LLMs (2024.emnlp-main)
Copied to clipboard
| Challenge: | Journalists engage in multiple steps in news writing that depend on human creativity, such as exploring different “angles” and selecting sources. |
| Approach: | They propose to use large language models to help journalists plan their news coverage . they find that LLMs recommend more creative angles and more informational sources . |
| Outcome: | The proposed models align better with humans when recommending angles, compared with informational sources. |
Multitask Semi-Supervised Learning for Class-Imbalanced Discourse Classification (2021.emnlp-main)
Copied to clipboard
| Challenge: | Discourse learning is a complex task, and schemas evolve across annotation efforts preventing compilation of smaller datasets into larger ones. |
| Approach: | They propose a multitask learning approach that can combine discourse datasets from similar and diverse domains to improve discourse classification. |
| Outcome: | The proposed approach improves on the NewsDiscourse dataset by 4.9% over current state-of-the-art benchmarks on one of the largest discourse datasets. |
LegalDiscourse: Interpreting When Laws Apply and To Whom (2024.naacl-long)
Copied to clipboard
| Challenge: | Recent advances in NLP and information retrieval have already enabled practical applications. |
| Approach: | They propose a 'discourse' taxonomy for span-and-relation parsing of legal texts . they use a dataset of 602 state-level law paragraphs with 3,715 discourse spans and 1,671 relations to investigate the increase in liquor licenses and decrease in applicable laws. |
| Outcome: | The proposed model performs poorly at span identification and relation classification, but lags far below human level. |
NewsInterview: a Dataset and a Playground to Evaluate LLMs’ Grounding Gap via Informational Interviews (2025.acl-long)
Copied to clipboard
Alexander Spangher, Michael Lu, Sriya Kalyan, Hyundong Justin Cho, Tenghao Huang, Weiyan Shi, Jonathan May
| Challenge: | Existing large datasets (1k-10k transcripts) are generated via crowdsourcing and are inherently unnatural. |
| Approach: | They curate a dataset of 40,000 two-person informational interviews from NPR and CNN . they find that LLMs are significantly less likely than human interviewers to use acknowledgements and pivot to higher-level questions. |
| Outcome: | The proposed model is based on 40,000 interviews with journalists and CNN . |
Are Large Language Models Capable of Generating Human-Level Narratives? (2024.emnlp-main)
Copied to clipboard
Yufei Tian, Tenghao Huang, Miri Liu, Derek Jiang, Alexander Spangher, Muhao Chen, Jonathan May, Nanyun Peng
| Challenge: | a recent HCI study has pointed to gaps in machine storytelling ability at the global level . authors show that LLMs have less suspense and less tension than human stories . |
| Approach: | They propose a computational framework to analyze narratives through three discourse-level aspects. |
| Outcome: | The proposed framework analyzes narratives through three discourse-level aspects . it shows that LLMs fall short of human abilities in discourse understanding . |
Explaining Mixtures of Sources in News Articles (2024.findings-emnlp)
Copied to clipboard
| Challenge: | a recent study shows that language models are essential for long-form article generation. |
| Approach: | They propose a generative process where a source-selection schema is first selected by a journalist, and then sources are chosen based on categories in that schema. |
| Outcome: | The proposed model can predict the most suitable schema given just the headline with reasonable accuracy. |
Identifying Informational Sources in News Articles (2023.emnlp-main)
Copied to clipboard
| Challenge: | Identifying sources of information in news articles is relevant to many tasks in NLP, including misinformation detection and argumentation. |
| Approach: | They propose a task to study compositionality of sources in news articles to understand how they are chosen to complement each other. |
| Outcome: | The proposed dataset can be used to train high-performing models for information detection and source attribution. |
Creative Planning with Language Models: Practice, Evaluation and Applications (2025.naacl-tutorial)
Copied to clipboard
| Challenge: | This tutorial explores how planning has been learned and deployed in creative workflows . many human creative tasks involve extensive planning, and actions need to be taken . |
| Approach: | This tutorial explores how planning has been learned and deployed in creative workflows . authors discuss forward and backward learning approaches for planning in LLMs - and evaluation metrics tailored to latent plans . |
| Outcome: | This tutorial examines how planning has been learned and deployed in creative workflows . it discusses forward and backward learning approaches for planning in LLMs - evaluation metrics tailored to latent plans . |
NewsEdits: A News Article Revision Dataset and a Novel Document-Level Reasoning Challenge (2022.naacl-main)
Copied to clipboard
| Challenge: | a large dataset of news article revision histories provides clues to narrative and factual evolution in news articles. |
| Approach: | They propose tasks to predict edit-actions performed during version updates . they define article-level edit actions: Addition, Deletion, Edit and Refactor . |
| Outcome: | The proposed dataset is large-scale and multilingual and spans 15 years . it shows that edit-actions are predictable and are likely to be based on factual evolution . |
Sequentially Controlled Text Generation (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Using GPT-2, long documents can ramble and do not follow human-like writing structure. |
| Approach: | They propose a controlled text generation task that generates documents with structure . they use a news article as a dataset to test different degrees of structural awareness . |
| Outcome: | The proposed task generates documents with a structure that is human-like, but long documents lack structure. |
Spatial Layouts in News Homepages Capture Human Preferences (2025.emnlp-main)
Copied to clipboard
| Challenge: | Homepage layouts are daily and manually curated by expert human news editors, and serve as a proxy for information prioritization. |
| Approach: | They propose a pairwise preference model to capture ranked preferences between news items . they use a dataset of 3,000 news homepages captured twice daily over a five-year period . |
| Outcome: | The proposed model infers editorial preference with over 0.7 F1 score (based on human trials) it also ranks local city council policies passed over ten years in San Francisco, evaluating their "newsworthiness". |
Learning Action Conditions from Instructional Manuals for Instruction Understanding (2023.acl-long)
Copied to clipboard
| Challenge: | a weakly supervised task is proposed to extract mentions of preconditions and postconditions of actions from instructional manuals. |
| Approach: | They propose a task dubbed action condition inference which extracts mentions of preconditions and postconditions of actions from instructional manuals. |
| Outcome: | The proposed approach improves on the existing models, but still far behind human performance. |
Tracking the Newsworthiness of Public Documents (2024.acl-long)
Copied to clipboard
| Challenge: | a new method to model news coverage of local government is needed . we show that newsworthiness predictions can be useful for journalists seeking to keep abreast of local governments. |
| Approach: | They propose a method that explicitly models when and why stories get press attention . they use an annotated corpus of news articles to build models that predict if a policy item will get covered . |
| Outcome: | The proposed model outperforms retrieval-based methods with limited annotated data and language use between corpora. |