Data-to-Text Generation
A classic problem in natural-language generation (NLG) involves taking structured data, such as a table, as input, and producing text that adequately and fluently describes this data as output. Unlike machine translation, which aims for complete transduction of the sentence to be translated, this form of NLG is usually taken to require addressing (at least) two separate challenges: what to say, the selection of an appropriate subset of the input data to discuss, and how to say it, the surface realization of a generation.
( Image credit: Data-to-Text Generation with Content Selection and Planning )
Papers
Showing 1–10 of 219 papers
All datasetsWebNLGE2E NLG ChallengeWebNLG FullCleaned E2E NLG ChallengeRotoWireRotoWire (Relation Generation)ToTToXAlignDARTMULTIWOZ 2.1RotoWire (Content Ordering)Rotowire (Content Selection)
Benchmark Results
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | S_1^R | BLEU | 68.6 | — | Unverified |
| 2 | EDA_CS | BLEU | 67.05 | — | Unverified |
| 3 | TrICy (trK = 0) | BLEU | 66.43 | — | Unverified |
| 4 | Slug | BLEU | 66.19 | — | Unverified |
| 5 | TGen | BLEU | 65.93 | — | Unverified |
| 6 | EDA_CS (TL) | BLEU | 65.8 | — | Unverified |
| 7 | Sys1-Primary | BLEU | 65.61 | — | Unverified |
| 8 | Zhang | BLEU | 65.45 | — | Unverified |
| 9 | Self-memory | BLEU | 65.11 | — | Unverified |
| 10 | Gong | BLEU | 64.22 | — | Unverified |