https://dataprd.com/tags/llm/ Recent content in Llm on Dataprd.Com Hugo en-us Sat, 15 Aug 2026 09:00:00 +0000 https://dataprd.com/posts/llm-price-list-extraction/ Sat, 15 Aug 2026 09:00:00 +0000 https://dataprd.com/posts/llm-price-list-extraction/ <p>Every accommodation booking system eventually runs into the same wall. Owners describe their prices in prose, and software needs rows in a table.</p> <p>What arrives looks like this, give or take a language. The colours show which sentence becomes which kind of row:</p> <div class="owner-text"> <div class="owner-text-label">Owner's text</div> <div class="owner-line"><span class="hl hl-season">High season from 15 June to 22 August, 35,000 per night for up to 6 guests, extra person 5,000 per night.</span></div> <div class="owner-line"><span class="hl hl-season">Cleaning 15,000, free from 7 nights.</span></div> <div class="owner-line"><span class="hl hl-holiday">Christmas package 24 to 28 December, 4 nights 180,000.</span></div> <div class="owner-line"><span class="hl hl-pet">Dogs 2,000 per night, maximum two, up to 15 kg.</span></div> <div class="owner-line"><span class="hl hl-discount">Booking of 7 nights gets 1 night free.</span></div> <div class="owner-line"><span class="hl hl-rule">Saturday to Saturday in high season.</span></div> </div> <table> <thead> <tr> <th>Row</th> <th>Captures</th> </tr> </thead> <tbody> <tr> <td><span class="tag tag-season">season</span></td> <td>Date range, nightly rate, guests included, extra guest fee, cleaning fee and when it is waived</td> </tr> <tr> <td><span class="tag tag-holiday">holiday</span></td> <td>Fixed night count and a package total</td> </tr> <tr> <td><span class="tag tag-pet">pet fee</span></td> <td>Per-animal rate, maximum animals, maximum weight</td> </tr> <tr> <td><span class="tag tag-discount">discount</span></td> <td>Type <code>night_bonus</code>: 7 nights required, 1 night free</td> </tr> <tr> <td><span class="tag tag-rule">rule</span></td> <td>Arrival and departure day, minimum stay</td> </tr> </tbody> </table> <p>Six sentences become five rows, with every ambiguity resolved on the way. The model returns this:</p>