2c:null 2d:["$","title",null,{"children":"Paper: An Intelligent Bootstrap-Based Framework for Negative Sentiment Classification of E-Commerce Reviews | anon author | Observable"}] 32:T1e24,const deck = (() => { const names = {oob_acc: "OOB accuracy", accuracy_632: ".632 accuracy", train_acc: "Train accuracy"}; const colors = {oob_acc: "#6f7f91", accuracy_632: "#b59b78", train_acc: "#9a9a9a"}; const textStyle = {fontSize: "14px", color: "#555555"}; const pct = x => (x * 100).toFixed(2) + "%"; const statsFor = (key, n = data.length) => { const v = data.slice(0, n).map(d => d[key]).sort(d3.ascending); return { mean: d3.mean(v), sd: d3.deviation(v) ?? 0, lo: d3.quantile(v, 0.025), hi: d3.quantile(v, 0.975), min: v[0], max: v.at(-1) }; }; // ---------- Slide 1: introduction ---------- const slide1 = html`

An Intelligent Bootstrap-Based Framework for Negative Sentiment Classification of E-Commerce Reviews

Interactive Visualisation of the Bootstrap Results

The model was evaluated with ${data.length} bootstrap iterations. In each iteration, the model is trained on a bootstrap sample and tested on the examples left out of that sample (out-of-bag, OOB).

OOB accuracy: accuracy on the out-of-bag examples.
.632 accuracy: weighted combination of OOB accuracy and training accuracy, which corrects the pessimism of OOB.
Train accuracy: always 1.0, so it carries no information and is not shown.

Use the arrow keys or the buttons below to move through the results.

`; // ---------- Slide 2: distribution with controls ---------- const metricInput = Inputs.select( new Map([[names.oob_acc, "oob_acc"], [names.accuracy_632, "accuracy_632"]]), {label: "Metric", value: "oob_acc"} ); const nInput = Inputs.range([1, data.length], {label: "Bootstrap iterations", step: 1, value: data.length}); const histBox = html`

`; const statsBox = html`

`; const drawHist = () => { const key = metricInput.value; const n = nInput.value; const subset = data.slice(0, n); const s = statsFor(key, n); const labelOnLeft = s.mean > (s.min + s.max) / 2; statsBox.innerHTML = `Mean ${pct(s.mean)} SD ${pct(s.sd)} 95% percentile interval ${pct(s.lo)} to ${pct(s.hi)}`; histBox.replaceChildren(Plot.plot({ height: 300, marginTop: 25, style: {...textStyle, background: "white"}, x: {label: "Accuracy"}, y: {label: "Density"}, marks: [ Plot.gridX({stroke: "#000000", strokeOpacity: 0.08, strokeWidth: 0.5}), Plot.gridY({stroke: "#000000", strokeOpacity: 0.08, strokeWidth: 0.5}), Plot.rectY(subset, Plot.binX( {y: (v, extent) => v.length / (subset.length * (extent.x2 - extent.x1))}, {x: key, thresholds: 20, fill: colors[key]} )), Plot.ruleX([s.mean], {stroke: "#555555", strokeWidth: 2}), Plot.text([s.mean], { x: d => d, frameAnchor: "top", dy: -12, dx: labelOnLeft ? -10 : 10, textAnchor: labelOnLeft ? "end" : "start", text: d => `Mean: ${pct(d)}`, fill: "#555555", fontWeight: "bold" }), Plot.ruleY([0]) ] })); }; metricInput.addEventListener("input", drawHist); nInput.addEventListener("input", drawHist); drawHist(); const slide2 = html`

Distribution of the bootstrap estimates

Density histogram of the first N iterations. The vertical line is the mean. Move the slider: with few iterations the shape is irregular and the mean shifts; with more iterations both stabilise.

${metricInput}${nInput}

${statsBox} ${histBox}

`; // ---------- Slide 3: convergence ---------- const long = [["oob_acc"], ["accuracy_632"]].flatMap(([key]) => { let sum = 0; return data.map((d, i) => ({ iteration: d.iteration, metric: names[key], value: d[key], mean: (sum += d[key]) / (i + 1) })); }); const lineChart = Plot.plot({ height: 400, marginLeft: 60, marginBottom: 50, grid: true, style: {...textStyle, background: "white"}, x: {label: "Bootstrap iteration"}, y: {label: "Accuracy (cumulative mean)", domain: [0.8, 0.95]}, color: {legend: true, domain: [names.oob_acc, names.accuracy_632], range: [colors.oob_acc, colors.accuracy_632]}, marks: [ Plot.lineY(long, {x: "iteration", y: "mean", stroke: "metric", strokeWidth: 2.5}), Plot.tip(long, Plot.pointerX({ x: "iteration", y: "mean", stroke: "metric", fontSize: 12, title: d => `${d.metric}\nIteration: ${d.iteration}\nCumulative mean: ${d.mean.toFixed(4)}\nSingle value: ${d.value.toFixed(4)}` })) ] }); const slide3 = html`

The estimates stabilise as iterations are added

Each point is the mean accuracy over all iterations up to that point. The curves fluctuate at first and then flatten. Hover to read the exact values. The y axis is zoomed to 0.80-0.95.

${lineChart}

`; // ---------- Deck shell ---------- const slides = [slide1, slide2, slide3]; let index = 0; const root = html`

`; const stage = root.querySelector(".stage"); const count = root.querySelector(".count"); const show = i => { index = Math.max(0, Math.min(slides.length - 1, i)); stage.replaceChildren(slides[index]); count.textContent = `${index + 1} / ${slides.length}`; root.querySelector(".prev").disabled = index === 0; root.querySelector(".next").disabled = index === slides.length - 1; }; root.querySelector(".prev").onclick = () => show(index - 1); root.querySelector(".next").onclick = () => show(index + 1); root.querySelector(".fs").onclick = () => root.requestFullscreen?.(); const onKey = e => { if (["INPUT", "SELECT", "TEXTAREA"].includes(e.target.tagName)) return; if (e.key === "ArrowRight" || e.key === " ") { e.preventDefault(); show(index + 1); } if (e.key === "ArrowLeft") { e.preventDefault(); show(index - 1); } }; document.addEventListener("keydown", onKey); if (typeof invalidation !== "undefined") invalidation.then(() => document.removeEventListener("keydown", onKey)); show(0); return root; })(); display(deck);33:[{"id":"4e27aa813a94a3df4ab1e3a24eb156a818074f941b56c36afcc6f78e3122e80d9bc2a65cc07c3f2031646d9ddfde9eb171205d270377a33e74f055a9108603fd","href":"https://static.observableusercontent.com/files/4e27aa813a94a3df4ab1e3a24eb156a818074f941b56c36afcc6f78e3122e80d9bc2a65cc07c3f2031646d9ddfde9eb171205d270377a33e74f055a9108603fd","name":"bootstrap_results.csv","type":"text/csv","lastModified":1791270060005,"size":10790}] 30:{"id":"50eec6c59a56b526","slug":"bootstrap-results","title":"Paper: An Intelligent Bootstrap-Based Framework for Negative Sentiment Classification of E-Commerce Reviews","thumbnail":"832b3537383d3903e053abfe3516ae73f29d98b922134726c8ca3f173251d9ee","access_level":"public","create_time":"$D2026-10-06T07:53:40.197Z","update_time":"$D2026-10-06T08:51:13.358Z","view_time":null,"first_public_time":"$D2026-10-06T08:04:13.544Z","has_unpublished_edits":null,"fork":false,"stars":1,"starred":false,"owner":{"type":"team","id":"3978fa3cff65daec","login":"anon-reviewer","name":"anon author","avatar_url":"https://avatars-next.observableusercontent.com/9cda5a7414b2af7f0be8ea1ea53fb9db6e586c2b1fb2d7cb8f4fef96d014554e","tier":"free_2026","bio":""},"version":27,"publish_version":27,"role":null,"authored":false,"authors":[{"type":"individual","id":"9d915d1e2421bbb2","login":"anon-reviewer","name":"anon author","avatar_url":"https://avatars-next.observableusercontent.com/9cda5a7414b2af7f0be8ea1ea53fb9db6e586c2b1fb2d7cb8f4fef96d014554e","tier":null,"bio":""}],"members":[{"type":"team","id":"3978fa3cff65daec","login":"anon-reviewer","name":"anon author","avatar_url":"https://avatars-next.observableusercontent.com/9cda5a7414b2af7f0be8ea1ea53fb9db6e586c2b1fb2d7cb8f4fef96d014554e","tier":"free_2026","role":"editor"}],"comments":[],"custom_thumbnail":true,"trashed":false,"lapsed":null,"subscribed":null,"forks":1,"collection":null,"type":"notebook","body":{"titling":"auto","title":"Paper: An Intelligent Bootstrap-Based Framework for Negative Sentiment Classification of E-Commerce Reviews","cells":[{"id":1,"value":"# Paper: An Intelligent Bootstrap-Based Framework for Negative Sentiment Classification of E-Commerce Reviews\n## Interactive Visualisation of the Bootstrap Results","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":3,"value":"const data = await FileAttachment(\"bootstrap_results.csv\").csv({typed: true});","pinned":true,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":2,"value":"$32","pinned":true,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W33","stdlib":"2"}} 31:"$30:body"