Stars-894

| Risk | Impact | Mitigation | |------|--------|------------| | NLP model returns irrelevant terms (low precision) | Poor author trust, extra manual work | Start with a conservative confidence threshold (e.g., 0.75); adjust based on analytics. | | Increased latency on slow connections | Frustrating editing experience | Debounce, cache recent suggestions, fallback to “unavailable” banner after 1 s timeout. | | Taxonomy drift (new terms not in taxonomy) | Suggestions cannot be mapped → no UI display | Show raw term in tooltip with “Add to taxonomy” link (future ticket). | | Over‑reliance on suggestions leading to tag homogenization | Loss of niche tags | Encourage authors to add custom tags; keep the UI for free‑form tag entry. | | Service outage of NLP micro‑service | Feature unavailable | Graceful fallback; continue publishing without suggestions. |

| ID | As a… | I want to… | So that… | |----|-------|------------|----------| | STARS‑894‑US1 | Content author | See a list of suggested tags while I type | I can quickly add the most relevant tags without searching | | STARS‑894‑US2 | Content author | Accept or reject each suggestion with a single click or keyboard shortcut | I retain full control over the final tag set | | STARS‑894‑US3 | Content author | View why a tag was suggested (highlighted snippet) | I can trust the recommendation and understand its relevance | | STARS‑894‑US4 | Product analyst | Export acceptance/rejection data to the analytics dashboard | I can measure the impact of the feature and spot gaps in the taxonomy |

| Phase | Timeframe | Major Milestones | |-------|-----------|-------------------| | A – Concept | Jan 2022 – Jun 2022 | Mission Concept Review (MCR) completed. | | B – Preliminary Design | Jul 2022 – Dec 2023 | Preliminary Design Review (PDR). | | C – Detailed Design | Jan 2024 – Dec 2025 | Critical Design Review (CDR); Procurement contracts signed. | | D – Fabrication & I&T | Jan 2026 – Dec 2027 | Subsystem qualification; Environmental tests (vibe, thermal vacuum). | | E – Launch & Early Ops | Q1 2028 | Launch; 30‑day commissioning; First Light. | | F – Full Science Operations | Q2 2028 – Q4 2033 | Routine observations, data release, end‑of‑mission de‑orbit. | STARS-894

Key Critical Path: Detector module assembly → Integration → Environmental test → Launch.

Schedule Buffer: 6 months allocated for launch‑vehicle integration and early‑orbit checkout. | ID | As a… | I want

| Entity | Fields | |--------|--------| | TagSuggestion | articleId (uuid), term (string), taxonomyId (uuid), confidence (float), suggestedAt (timestamp) | | TagSuggestionEvent | eventId (uuid), articleId, taxonomyId, action (enum: ACCEPT/DISMISS), eventTime |

| Sub‑system | Description | |------------|-------------| | Primary Detector | Cadmium‑Zinc‑Telluride (CZT) pixel array, 64 × 64 pixels, 5 mm pitch; energy range 0.1–150 keV; effective area 1200 cm². | | Secondary Scintillator | LaBr₃(Ce) crystal coupled to SiPM array, 150 keV–10 MeV; timing resolution 200 ps. | | Anti‑Coincidence Shield | Plastic scintillator + photomultiplier tubes to suppress charged‑particle background (< 2 % residual). | | On‑board Processing | FPGA‑based trigger engine; real‑time burst localization via coded‑mask deconvolution; ≤ 5 ms latency. | | Calibration | On‑board ^⁵⁷Co and ^⁶⁰Co sources; periodic celestial calibrators (Crab Nebula). | | Phase | Timeframe | Major Milestones |

The Smart Tag Recommendations feature will automatically suggest relevant tags as authors create or edit articles in the STARS content editor. By leveraging natural‑language processing (NLP) and the existing taxonomy, the system will surface context‑aware tags, improving discoverability, SEO, and editorial consistency across the platform.

| Criterion | Test | |-----------|------| | Real‑time suggestions – When the author types ≥ 5 characters in title/abstract/body, the system returns up to 7 tag suggestions. | Unit test of the suggestion API mock; integration test verifying UI updates within 500 ms of keystroke. | | Relevance ranking – Suggestions are ordered by confidence score (high → low). | Verify that confidence scores are decreasing; manual spot‑check on a set of sample articles. | | Accept/reject UI – Each suggestion has an “Add” button and a “Dismiss” (X) button; keyboard shortcuts Enter (accept) and Esc (dismiss) work. | End‑to‑end UI test using Cypress/Playwright. | | Snippet preview – Hovering (or pressing ?) on a suggestion shows a short snippet of the article where the term appears. | Visual regression test confirming tooltip content. | | No duplicate tags – Already‑assigned tags do not appear in the suggestion list. | Test with article pre‑populated with #science; ensure science is not suggested again. | | Graceful fallback – If the NLP service is unavailable, the UI shows a non‑intrusive “Tag suggestions unavailable” banner and does not block publishing. | Simulated service outage; verify UI behavior and that publishing proceeds. | | Analytics logging – Each accept/reject event fires a POST to /api/analytics/tag‑suggestion with articleId, tag, action, and timestamp. | Mock server intercept; verify payload structure. | | Performance – End‑to‑end latency from keystroke to visible suggestions ≤ 800 ms on a typical 3G connection. | Lighthouse/Performance test suite. | | Accessibility – All suggestion controls are keyboard‑navigable, ARIA‑labelled, and pass WCAG 2.1 AA contrast checks. | Axe automated audit + manual screen‑reader test. |