In this section
The catalogue of every piece of evidence on this site. Each entry has a stable internal identifier (E[unit].[seq]), a full Harvard Cite Them Right reference, an in-text citation form, and a direct anchor link. The full alphabetical bibliography lives on the References page.
Consolidated catalogue of every piece of evidence on this site. Each entry has an internal identifier (E[unit].[seq]), a full Harvard Cite Them Right reference for citation in the reflection, an in-text citation form, a direct anchor link, and a short description.
The Evidence Index is the navigational catalogue. The full alphabetical bibliography lives on the References page, where every entry below also appears as a Mella, A. (2026...) reference.
How to cite an evidence item from the reflection
In the body of the reflection or any unit page:
The team’s working agreement was the first artefact to operationalise the academic frame (Mella, 2026b).
In the consolidated references.md:
Mella, A. (2026b) Group D kick-off deck slide 3 - working agreement. [Online image]. Available at: https://protode908.github.io/eportfolio-uoe/units/unit-02/#evidence-unit2-deck-slide-3 (Accessed: 10 July 2026).
The Harvard letter suffix (2026a, 2026b, …) tracks alphabetical order in references.md. The E identifier (E2.3) tracks the item’s location on this site.
Identifier convention
E2.x- evidence on the Unit 2 pageE5.x- evidence on the Unit 5 page (and so on for each unit)E6.x- evidence on the Team Project page (Unit 6 hub)E11.x- evidence on the Individual Project page (Unit 11 hub)ECD1.x,ECD2.x- evidence on the Collaborative Discussion pagesEPD.x- evidence on the Professional Development pageEH.x- evidence anchored on the Home page (rare)
Index of evidence items
Unit 2 - Seminar Preparation, Team Formation, and Kick-off
E2.1 - Group D kick-off deck slide 1: title
Harvard reference (copy for citing):
Mella, A. (2026a) Group D kick-off deck slide 1 - title (Airbnb Business Analysis, Track 1, Group D). [Online image]. Available at: https://protode908.github.io/eportfolio-uoe/units/unit-02/#evidence-unit2-deck-slide-1 (Accessed: 10 July 2026).
In-text citation: (Mella, 2026a)
Direct link: Unit 2 - Slide 1
Description: Title slide of the kick-off deck. Establishes the project frame: Airbnb Business Analysis, Track 1 (Classical Machine Learning, Regression and Clustering), Group D. Evidences LO4.
E2.2 - Group D kick-off deck slide 2: team introduction template
Harvard reference:
Mella, A. (2026b) Group D kick-off deck slide 2 - team introduction template. [Online image]. Available at: https://protode908.github.io/eportfolio-uoe/units/unit-02/#evidence-unit2-deck-slide-2 (Accessed: 10 July 2026).
In-text citation: (Mella, 2026b)
Direct link: Unit 2 - Slide 2
Description: Team introduction template (background, prior group experience, skills, availability, target grade, constraints). Evidences LO4.
E2.3 - Group D kick-off deck slide 3: working agreement
Harvard reference:
Mella, A. (2026c) Group D kick-off deck slide 3 - working agreement and group etiquette. [Online image]. Available at: https://protode908.github.io/eportfolio-uoe/units/unit-02/#evidence-unit2-deck-slide-3 (Accessed: 10 July 2026).
In-text citation: (Mella, 2026c)
Direct link: Unit 2 - Slide 3
Description: Working agreement / group etiquette: nine explicit rules (on-time, finished-means-fully-usable, fair workload, disagree-and-commit, and others). Operationalises lessons from prior team-project reflective work. Evidences LO4 and LO1 (professional issues in team practice).
E2.4 - Group D kick-off deck slide 4: communication and tooling proposal (1 of 2)
Harvard reference:
Mella, A. (2026d) Group D kick-off deck slide 4 - communication and tooling proposal (1 of 2). [Online image]. Available at: https://protode908.github.io/eportfolio-uoe/units/unit-02/#evidence-unit2-deck-slide-4 (Accessed: 10 July 2026).
In-text citation: (Mella, 2026d)
Direct link: Unit 2 - Slide 4
Description: Communication and tooling proposal: meeting cadence, Zoom for synchronous meetings with recording, email for milestones, WhatsApp for daily coordination. Evidences LO4.
E2.5 - Group D kick-off deck slide 5: communication and tooling proposal (2 of 2)
Harvard reference:
Mella, A. (2026e) Group D kick-off deck slide 5 - communication and tooling proposal (2 of 2). [Online image]. Available at: https://protode908.github.io/eportfolio-uoe/units/unit-02/#evidence-unit2-deck-slide-5 (Accessed: 10 July 2026).
In-text citation: (Mella, 2026e)
Direct link: Unit 2 - Slide 5
Description: Communication and tooling proposal continued: GitHub for code, Google or OneDrive for non-code, Word with track changes, Excel, R or Python, PDF for final deliverable. Evidences LO4.
E2.6 - Group D kick-off deck slide 6: submission scope
Harvard reference:
Mella, A. (2026f) Group D kick-off deck slide 6 - submission scope. [Online image]. Available at: https://protode908.github.io/eportfolio-uoe/units/unit-02/#evidence-unit2-deck-slide-6 (Accessed: 10 July 2026).
In-text citation: (Mella, 2026f)
Direct link: Unit 2 - Slide 6
Description: Submission scope: 1,000-word executive analytical report, supporting Python notebooks, methodologically sound and decision-driven. Evidences LO3 and LO4.
Team Project (Unit 6) - Track 1 Airbnb Business Analysis
E6.1 - Track 1 decision register (D-001 to D-018)
Harvard reference:
Mella, A. (2026g) Track 1 decision register: D-001 to D-018, audit trail of methodological decisions for the Airbnb business analysis. [Project document]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-decision-register (Accessed: 10 July 2026).
In-text citation: (Mella, 2026g)
Direct link: Team Project - Decision register
Description: Chronological audit trail of every methodological choice for Track 1. Each entry captures decision, alternatives considered, tier of evidence (T1 / T2 / T3), citation, in-report visibility flag. Evidences LO3 (critical appraisal of ML techniques) and LO4 (effective team-member documentation discipline).
E6.2 - Track 1 design document
Harvard reference:
Mella, A. (2026h) Track 1 design document: handover-ready synthesis of the CRISP-DM five-script pipeline, regression specification, clustering specification, and figure ownership. [Project document]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-design-document (Accessed: 10 July 2026).
In-text citation: (Mella, 2026h)
Direct link: Team Project - Design document
Description: Synthesised by topic (project goal, methodology backbone, dataset specification, ML design constraints, cleaning decisions, feature engineering, regression handover, clustering handover, pipeline contract). A different team member could pick up the work using only this document plus the cleaned CSV. Evidences LO2 and LO4.
E6.3 - Briefing pack to colleague (data-cleaning lead)
Harvard reference:
Mella, A. (2026i) Track 1 briefing pack for the data-cleaning lead, version 0.6: cleaning rules with academic citations, three-tier framework, deliverable specification and quality checklist. [Project document]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-briefing-pack (Accessed: 10 July 2026).
In-text citation: (Mella, 2026i)
Direct link: Team Project - Briefing pack
Description: Self-contained briefing pack that pre-loaded every cleaning rule with its academic citation, the three-tier framework, the deliverable specification and the quality checklist - so the colleague could defend her work without re-deriving it. Evidences LO4 (virtual team leadership).
E6.4 - Three-tier justification framework
Harvard reference:
Mella, A. (2026j) Three-tier justification framework: project-internal rule tagging every methodological decision under Tier 1 (peer-reviewed literature), Tier 2 (recognised library documentation), or Tier 3 (pragmatic with future-work framing). [Project document]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-three-tier-framework (Accessed: 10 July 2026).
In-text citation: (Mella, 2026j)
Direct link: Team Project - Three-tier framework
Description: Original methodological apparatus designed to interrupt “just-because” reasoning. Every decision tagged T1 (preferred), T2 (when T1 absent) or T3 (last resort, flagged for future work). Evidences LO3 (systematic reasoning).
E6.5 - Triple-source evaluation map
Harvard reference:
Mella, A. (2026k) Triple-source evaluation map: report sections matched against the assignment brief, the tutor’s structure guidance, and the PG Grading rubric. [Project document]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-triple-source-map (Accessed: 10 July 2026).
In-text citation: (Mella, 2026k)
Direct link: Team Project - Triple-source map
Description: Pre-empts the “we hit the brief but missed the rubric” trap by mapping every report section to all three concurrent evaluation sources. Evidences LO3 (disciplined preparation).
E6.6 - CRISP-DM five-script Python pipeline
Harvard reference:
Mella, A. (2026l) Track 1 CRISP-DM five-script Python pipeline: 01_data_profiling, 02_data_preparation, 03_feature_analysis, 04_regression, 05_clustering. [Source code]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-pipeline (Accessed: 10 July 2026).
In-text citation: (Mella, 2026l)
Direct link: Team Project - Pipeline
Description: Five-script pipeline operationalising CRISP-DM (Chapman et al., 2000) with explicit phase ownership and feedback loops. Evidences LO2 and LO3.
E6.7 - Demand-mapping pivot (D-017)
Harvard reference:
Mella, A. (2026m) Demand-mapping pivot decision (D-017): re-anchoring the analytical narrative from a Smart-Pricing prescriptive framing to a demand-mapping plus competitive-map informational framing after the regression’s predictive ceiling. [Decision register entry]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-demand-mapping-pivot (Accessed: 10 July 2026).
In-text citation: (Mella, 2026m)
Direct link: Team Project - Demand-mapping pivot
Description: A Phase 5 to Phase 1 CRISP-DM iteration in action. The regression’s modest R-squared made the original Smart-Pricing framing untenable; the pivot was logged with academic justification rather than as a face-saving reframing. Evidences LO3 (theory-and-practice critical thinking).
E6.8 - Will-NOT-claim list (Section 5 of the report)
Harvard reference:
Mella, A. (2026n) Will-NOT-claim list: defensive criticality framing in the report, naming the specific recommendations the analysis cannot support. [Report extract]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-will-not-claim (Accessed: 10 July 2026).
In-text citation: (Mella, 2026n)
Direct link: Team Project - Will-NOT-claim list
Description: Explicit list of recommendations the analysis cannot support: no amenity recommendations, no NLP, no seasonal pricing, no per-host commands, no booking-conversion forecasts, and explicitly not a Smart-Pricing product. Pre-empts over-claim risk. Evidences LO1 (ethics) and LO3 (critical evaluation).
E6.9 - Committed ML design constraints upfront (D-011)
Harvard reference:
Mella, A. (2026o) Committed ML design constraints upfront (D-011): the six pre-cleaning constraints that pushed downstream ML choices upstream into data preparation. [Decision register entry]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-ml-design-constraints (Accessed: 10 July 2026).
In-text citation: (Mella, 2026o)
Direct link: Team Project - ML design constraints
Description: Six constraints committed before profiling: regression target equals log1p(price), numeric encoding, zero-NaN modelling columns, outlier philosophy, scaling stage, provisional feature list. Demonstrates methodological foresight. Evidences LO2 (datasets) and LO3 (technique appraisal).
E6.10 - Pre-run design document for EDA and cleaning
Harvard reference:
Mella, A. (2026p) Pre-run design document for the EDA and data-cleaning workstream: output specification, seven cleaning tasks with literature citations, exploration step, findings-document spec, folder-structure spec. [Project document]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-eda-pre-run (Accessed: 10 July 2026).
In-text citation: (Mella, 2026p)
Direct link: Team Project - EDA pre-run
Description: Handover document produced for the colleague leading the EDA and cleaning workstream. Defines output specification for the cleaned CSV, seven cleaning tasks (Rules 2.1 to 2.7) with Python code and literature citation each, logging task 2.8, standard EDA commands plus useful one-liners, findings-document specification, folder-structure spec. Treats CRISP-DM Phase 4 to Phase 3 feedback loop as a design constraint on Phase 3 rather than a loop discovered reactively. Evidences LO2 (dataset preparation) and LO4 (handover-ready artefact for virtual team).
E6.11 - Cleaning script (02_data_preparation.py)
Harvard reference:
Mella, A. (2026q) Cleaning script
02_data_preparation.py: imports, raw CSV load, Tasks 2.1 to 2.8 in order, produces cleaned CSV and step-2 cleaning log. [Source code]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-cleaning-script (Accessed: 10 July 2026).
In-text citation: (Mella, 2026q)
Direct link: Team Project - Cleaning script
Description: Single-file Python script alongside the Word brief. Defensive artefact: provided as a .py file because Word’s auto-correction of straight single quotes to curly quotes (and occasional indentation loss on copy-paste) produces Python syntax errors that look like brief errors but are tooling artefacts. Demonstrates risk-aware artefact design. Evidences LO2, LO3 and LO4.
E6.12 - Group D Google Drive workspace
Harvard reference:
Mella, A. (2026r) Group D Google Drive workspace: workstream-driven folder structure with role-appropriate access permissions and explicit sharing policy. [Project workspace]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-drive-workspace (Accessed: 10 July 2026).
In-text citation: (Mella, 2026r)
Direct link: Team Project - Drive workspace
Description: Workstream-driven folder structure under a Group D root. Differentiated permissions: Coordinator has editor access across; workstream leads have owner access on their sub-folders and editor elsewhere; others have read access by default with edit access on their workstream when active. External-sharing policy explicit in the meeting-summary email. Evidences LO4 (operational discipline).
E6.13 - Group D project plan and timeline
Harvard reference:
Mella, A. (2026s) Group D project plan and timeline: locked, owner-named milestones with input and output columns through to submission on 6 June 2026. [Project document]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-timeline (Accessed: 10 July 2026).
In-text citation: (Mella, 2026s)
Direct link: Team Project - Timeline
Description: Locked, owner-named milestones from kick-off through to submission. Each row corresponds to a CRISP-DM phase or to a downstream integration step. Makes the dependency chain visible: EDA and cleaning is on the critical path for modelling, modelling for the report draft, the report draft for QA. Evidences LO4 (project management).
E6.14 - 9 May 2026 meeting notes and summary email (redacted)
Harvard reference:
Mella, A. (2026t) Group D meeting on 9 May 2026: meeting notes and summary email circulating the locked business question, timeline, task allocation, and Coordinator role to the full distribution list, 9 May 2026. [Email]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/email-2026-05-09/ (Accessed: 10 July 2026).
In-text citation: (Mella, 2026t)
Direct link: Email of record - 9 May 2026 (redacted)
Description: Email-of-record for the 9 May decisions meeting (3 of 6 attended). Locked business question, timeline, task allocation and Coordinator role; security note on link sharing; explicit invitation to absent members for asynchronous review or counter-proposal. The published version on the e-portfolio has peer email addresses redacted; the original PDF with full headers is held in the internal record. Evidences LO4 (transparent decision-making with documented async-review opportunity).
E6.15 - Decision register entry D-019
Harvard reference:
Mella, A. (2026u) Decision register entry D-019: pre-run design document for EDA and cleaning issued; Coordinator role made explicit; timeline locked through to 6 June; partial-attendance decision-making rule applied at the 9 May meeting. [Decision register entry]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/#evidence-team-d019 (Accessed: 10 July 2026).
In-text citation: (Mella, 2026u)
Direct link: Team Project - D-019
Description: Decision-register entry capturing the four-fold v0.0.11 update: pre-run design document issued; Coordinator role made explicit; timeline locked; partial-attendance rule applied. Evidences LO3 (decision audit) and LO4 (project management).
E6.16 - Business question slide
Harvard reference:
Mella, A. (2026v) Business question slide for Group D Track 1: regression for price prediction, clustering segmentation, business value supporting hosts, with rationale and business outcome. [Slide]. Available at: https://protode908.github.io/eportfolio-uoe/assets/evidence/unit-6/2026-05-09_business-question-slide.pdf (Accessed: 10 July 2026).
In-text citation: (Mella, 2026v)
Direct link: Download business-question slide (PDF)
Description: Single slide capturing the locked Track 1 business question with the visual mapping of regression-for-price-prediction and clustering-segmentation onto business value supporting hosts. Includes assignment-brief Task 1 / Task 2 verbatim text. Evidences LO2 (dataset framing) and LO3 (technique selection rationale).
E6.17 - 9 May 2026 group meeting presentation (12 slides)
Harvard reference:
Mella, A. (2026w) Group D 9 May 2026 meeting presentation: 12-slide deck covering team status, business question, activities, roles, dependency map, CRISP-DM operationalisation, timeline and wrap-up. [Slide deck]. Available at: https://protode908.github.io/eportfolio-uoe/assets/evidence/unit-6/2026-05-09_group-meeting-presentation-12-slides.pdf (Accessed: 10 July 2026).
In-text citation: (Mella, 2026w)
Direct link: Download 12-slide deck (PDF); slides 3, 7, 8 anchored on the Team Project page.
Description: Full 12-slide presentation deck used at the 9 May meeting. Three slides carry direct evidentiary weight: Slide 3 documents the partial-attendance handling and the asynchronous consent obtained from Tamim plus the proactive QA/reviewer role-offer; Slide 7 is the colour-coded dependency map with hard and soft dependencies between workstreams; Slide 8 is the CRISP-DM-like methodology overlay with explicit feedback loops. Evidences LO1 (ethics role explicit on slide 6), LO2, LO3, LO4 (especially Slide 3).
E6.18 - 10 May 2026 EDA pre-run handover email
Harvard reference:
Mella, A. (2026x) EDA and Data cleaning design pre-run: handover email to the workstream lead delivering the folder structure, the pre-run design document and the cleaning script as a three-artefact pack, 10 May 2026. [Email]. Available at: https://protode908.github.io/eportfolio-uoe/projects/team-project/email-2026-05-10/ (Accessed: 10 July 2026).
In-text citation: (Mella, 2026x)
Direct link: Email of record - 10 May 2026 (redacted)
Description: Concrete email delivering the three-artefact pre-run pack to the data-cleaning workstream lead. Named recipient, dated subject line, structured numbered items, target completion date (16 May), open offer for ad-hoc clarification. Sister evidence to E6.10 (the design document) and E6.11 (the cleaning script). Evidences LO4 (transparent written cadence) and LO3 (rationale for the pre-run as a CRISP-DM Phase 4 to Phase 3 anticipation).
Collaborative Discussion 1 (Units 1 to 3)
ECD1.1 - Initial post (automotive: BMW and Jaguar Land Rover)
Harvard reference:
Mella, A. (2026y) Collaborative Discussion 1 initial post: Industry 4.0 and 5.0 in the automotive sector, BMW AI-enabled production and the Jaguar Land Rover 2025 cyber incident, 3 May 2026. [Forum post]. Available at: https://protode908.github.io/eportfolio-uoe/discussions/cd1/#my-initial-post (Accessed: 10 July 2026).
In-text citation: (Mella, 2026y)
Direct link: CD1 - Initial post
Description: Initial CD1 contribution exploring Industry 4.0 benefits (BMW AI-enabled predictive maintenance and quality control) and Industry 4.0 vulnerabilities (Jaguar Land Rover 2025 six-week cyber shutdown, £196M cost). Frames the discussion around Metcalf (2024) on Industry 5.0 resilience and human-centricity, with the OECD and Korinek-Stiglitz angle on AI ethics and labour. Evidences LO1 primarily; LO2 secondarily.
ECD1.2 - Peer response on HSE Ireland ransomware and healthcare resilience
Harvard reference:
Mella, A. (2026z) Collaborative Discussion 1 peer response to a peer’s post on the HSE Ireland 2021 ransomware attack, with Office of the Controller and Auditor General findings on weak IT infrastructure and the NIST Recover function, 13 May 2026. [Forum post]. Available at: https://protode908.github.io/eportfolio-uoe/discussions/cd1/#my-first-peer-response (Accessed: 10 July 2026).
In-text citation: (Mella, 2026z)
Direct link: CD1 - First peer response
Description: Peer response anchoring the HSE Ireland 2021 cyber attack in the Office of the Controller and Auditor General (2022) findings on weak IT infrastructure (legacy systems, flat network, no standardised backup), connecting Industry 5.0’s resilience pillar (European Commission, 2021) to practical recovery testing, citing Moore et al. (2023) on paper-based workarounds and the NIST CSF 2.0 Recover function. Closes with a discussion question on conditional approval of digital-transformation programmes. Evidences LO1.
ECD1.3 - Peer response on pharmaceutical Industry 4.0 and Zero Trust
Harvard reference:
Mella, A. (2026aa) Collaborative Discussion 1 peer response to a peer’s post on Industry 4.0 in the pharmaceutical sector, with Zero Trust Architecture, NIST CSF Recover, and Merck NotPetya and West Pharmaceutical Services cyberattack examples, 13 May 2026. [Forum post]. Available at: https://protode908.github.io/eportfolio-uoe/discussions/cd1/#my-second-peer-response (Accessed: 10 July 2026).
In-text citation: (Mella, 2026aa)
Direct link: CD1 - Second peer response
Description: Peer response anchoring the pharmaceutical-sector cyber risk in Accenture (2025) and World Economic Forum (2025a) data on the cybersecurity investment lag (only 28% of organisations embed security into transformation initiatives). Introduces Zero Trust Architecture (Rose et al., 2020) and the NIST CSF 2.0 six-function model, plus Merck NotPetya (Crosignani et al., 2020) and West Pharmaceutical Services (2026) as real-world supply-chain cascade examples. Evidences LO1.
ECD1.4 - Summary post: cross-sector implementation-gap synthesis
Harvard reference:
Mella, A. (2026ab) Collaborative Discussion 1 summary post: cross-sector synthesis (automotive, healthcare, pharmaceutical, education) exposing the implementation gap between Industry 5.0 aspirations and organisational practice, 13 May 2026. [Forum post]. Available at: https://protode908.github.io/eportfolio-uoe/discussions/cd1/#my-summary-post (Accessed: 10 July 2026).
In-text citation: (Mella, 2026ab)
Direct link: CD1 - Summary post
Description: Synthesising summary across the four sectors discussed in the thread (automotive, healthcare, pharmaceutical, education). Introduces the Janardhan (2021) Meta outage as cross-organisation cascade evidence. Surfaces the implementation gap between Industry 5.0’s human-centric / resilient framing (European Commission, 2021; Metcalf, 2024) and operational reality (Accenture, 2025; World Economic Forum, 2025a). Documents the personal frame-change “from seeing it as an efficiency opportunity to seeing it as a socio-technical dependency problem”. Recommends measurement (recovery drills, manual fallback plans, supplier-risk controls, reskilling budgets, labour impact assessments) per OECD (2023) and World Economic Forum (2025b). Evidences LO1 primarily; LO2 secondarily.
Other units
(This index grows as evidence is added across the module. The pattern above is the canonical entry shape for every new evidence item: identifier, full Harvard reference, in-text citation, direct link, and short description with LO mapping.)