JRC's Participation at TAC 2011: Guided and Multilingual Summarization Tasks
Bibliographic record
Abstract
The paper describes our participation in the Guided and Multilingual Summarization Tasks at the Text Analysis Conference 2011 (TAC’11). We participated in the Guided task with the system from the previous year which combines aspect identification by an event extraction system and automatically learned lexicons with LSA-based summarizer. This year we included temporal analysis to improve sentence ordering, detection of update information and dealing with the WHEN aspect. We made a first try to compress and paraphrase sentences with our second run. Multilingual summarization is our ultimate goal and thus all components of the system are either fully language independent or can be relatively easily adapted for other languages. The multilingual task provided a possibility to test the system on other languages then English. The sentence-extractive summarizer was ranked among the top systems in the case of readability and non-redundancy. Even if the content of its summaries was not ranked on the top for English in the main Guided task, it reached the top results in the Multilingual task. The generative run suffered from worse readability which affected also the content scores.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.012 | 0.019 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.002 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.003 | 0.001 |
| Scholarly communication | 0.003 | 0.003 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.004 | 0.004 |
| Insufficient payload (model declined to judge) | 0.012 | 0.009 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".