NetMiner

10/08/2026

Map attractions and villages as a co-visitation network, and a hidden mutual-benefit structure emerges.

Zhou et al. (2020) combined social media and survey data to map Shenzhen in China. Villages depend on spillover, yet sustain attractions through lodging and meals.

👉 Read the full case study: https://netminer.medium.com/1f7d7a5bc8b7

26/07/2026

Community power over construction projects grows as they near completion, not fades.

Researchers analyzed 89 responses from Hong Kong construction practitioners using NetMiner's two-mode social network analysis.
Contrary to conventional wisdom, district councils and NGOs gained power toward the closing stage.

- Governments, developers, and main contractors form the top power tier
- End users have the lowest average power, but grow relatively stronger at the start and end

See how stakeholder power shifts across each project stage.
👉 Read the full case study: https://medium.com/p/e44ade570f02

23/07/2026

Important links aren't just about frequency,
they can be identified through statistical significance.

Apply Backbone filtering to a keyword co-occurrence network,
and you can keep only the core connections, no arbitrary threshold needed.

Here's how to extract statistically significant links
with a no-code approach.

👉 Read the tutorial
https://medium.com//becfca8e9c68

No-Code Text Analysis

21/07/2026

Technical support for NetMiner 4 ends on December 31, 2026.

Your license still works after that date, so NetMiner 4 keeps running. New features and improvements will roll out through NetMiner 5 from here on.

👉 See the benefits: https://stib.ee/dbZM

20/07/2026

Left alone, AI analyzes graphs with just 36% accuracy.

Split the work by strength and the story changes. AI handles unstructured data interpretation and automation, SNA handles structural analysis and verification.

- GNNs extend SNA into predictive models, and GraphRAG gives AI SNA style context.
- SNA makes GNN's reasoning explainable, and LLMs help with SNA's hardest step, building the network itself.

See what Graph Intelligence changes, in the link below.

👉 Read the full piece: https://netminer.medium.com/ef5d264f861f

14/07/2026

AI can describe data it never actually queried, or present a number as an analysis result when no analysis ran.

Pairing it with analytics changes that.

- Analytics builds an analysis plan for the question first
- Verifies the data scope, then runs the calculation
- Grounds the interpretation in evidence

AI lowers the barrier to analysis. Analytics makes AI's answers verifiable. Five ways AI and analytics change each other, in the link.

👉 Read the full article: https://medium.com//4c96d8205cf5

Key Takeaways - On its own, AI tends to stay a ‘plausible-sounding generation and automation tool’ — it turns into real results only when…

12/07/2026

Map 6,400+ frailty papers as a network and 40 years of research shifts become visible.

Kim and Jang (2018) analyzed author keywords from 6,424 frailty-related papers (1981–2016) using NetMiner and text network analysis — revealing the knowledge structure that systematic reviews alone cannot capture.

- "Malnutrition" ranked first in degree centrality in the 2010s — the strongest signal yet that nutritional-status research has moved back to the center of frailty science.
- The field moved from mortality and disability in the 1980s, to home care and sarcopenia in the 2000s, to malnutrition and geriatric oncology in the 2010s — all mapped in a single network diagram.

See 40 years of knowledge structure change for yourself.

👉 Read more: https://medium.com//b4f10750cd8e

08/07/2026

AI use in academic research isn't a plagiarism question. It's a judgement question.

A recent framework redefines responsible AI use around four properties, ethical, valid, reproducible, transparent, and sorts AI use into four modes, search, co-author, validator, tutor, each with its own rule.

Only 6 of 38 leading universities surveyed have a policy covering this.

👉 See the five-dimension framework: https://medium.com//f9cae9122d35

06/07/2026

Regression gives you one number per variable.
SHAP gives you a breakdown for each case.

In some regions, urbanization drives crime rate predictions. In others, education level could play a bigger role.
SHAP shows exactly which variables mattered — and how much — for each individual case.

Follow the steps and you'll see what regression alone doesn't show.
No coding required.
👉 https://netminer.medium.com/20d74e8503a1

Regression analysis is a powerful tool in social science research.

02/07/2026

23,614 customer posts point to the same answer: some features need fixing before others.

Jeong et al. (2017) analyzed Samsung Galaxy Note 5 Reddit posts using NetMiner, LDA topic modeling, and deep learning-based sentiment analysis.
Priorities that word-frequency analysis cannot detect became measurable for the first time.

- Samsung Pay (score: 12.18): Most discussed, least satisfied. Missing NFC support drove the highest opportunity score across all 65 product topics.
- Camera (satisfaction: 8.14): Low opportunity score despite high visibility. Customer expectations already exceeded — further investment is unlikely to pay off.

See which features to improve first, and the data behind the decision.
👉 Read more: https://netminer.medium.com/5e83c7102f00