Skip to content

Grant, team, licences, data version

About

This site is a demonstration of how a biomedical text mining model maps correlations between palliative drugs and their side effects in advanced cancer.

Not a clinical decision aid

This site accepts no patient data and gives no recommendations for practice. An uploaded file is deleted the moment processing ends and its contents are never logged; what does persist is a cache keyed by sentence hash holding entity spans, in memory, until the service restarts — so the same document is not processed twice. Its output is material for consideration when drawing up a care plan, not a substitute for clinical judgement. The candidate tier has not been reviewed by an expert and part of it is clinically inverted.

Android app

The app wraps this same site rather than copying it. It loads pallicaremap.id directly every time it opens, so there is no second version to fall behind — anything fixed here is in the app at once, with no update to install.

Download APK · version 1.0.0

Android 5.0 and above · requires an internet connection

Installed directly, not through the Play Store. Android asks once for permission to “install unknown apps”, and Play Protect may show a warning. Both are ordinary for a direct install, and the file is signed with a key whose fingerprint is published at /.well-known/assetlinks.json.

The research

Document number
PP-25070125481
Scheme
Beginner Lecturer Research Grant, Basic Research
Institution
Universitas Harapan Bangsa
Discipline
Engineering · Electrical Engineering and Informatics · Biomedical Engineering
Focus area
Health · Applying AI and health big data
Final TRL target
3 — analytical and experimental proof of concept
Duration
1 year (2026)
Required output
An article in a nationally reputable journal indexed in SINTA 1–4, and IP registration for the biomedical text mining model
Target journal
Jurnal Berita Ilmu Keperawatan, Universitas Muhammadiyah Surakarta
Corpus
405 palliative oncology articles, 263 passed the relevance filter
Gold annotations
498 of 500 pairs, judged by a clinical pharmacist

What “model” means here

There are no self-trained weight files. What this study develops is the pipeline: the normalisation rules, the term-mapping guards, the sentence-meaning filters, the token-distance threshold, the association scoring, and the translation into nursing output. Entity tagging and dependency parsing use third-party pretrained weights.

A trained relation classifier is scheduled for the next stage of the roadmap (2027, TRL 3–4); the 498-annotation gold standard built this year becomes its test set.

Third-party pretrained models and their licences.
ModelLicenceUsed for
alvaroalon2/biobert_chemical_nerApache-2.0Drug entity tagging
alvaroalon2/biobert_diseases_nerApache-2.0Side-effect entity tagging
spaCy en_core_web_sm 3.8.0MITDependency parsing and negation detection

Team

The research team with their identification numbers and assigned duties.
NameRoleStudy programmeDutiesSinta ID
Martyarini Budi SetyawatiNIDN 0618038401Principal investigatorNursingLeads the study, designs the model, supervises the experiments, and writes the article and the progress and final reports.6198530
Etika Dewi CahyaningrumNIDN 0601048602Co-investigatorNursingManages the data, designs the model, and writes the Methods section of the article.6005849
Iis Setiawan Mangku NegaraNIDN 0616027601Co-investigatorInformation TechnologyBuilds and runs the model experiments, and writes the Results section of the article.6198371
Rizka KhumaidaNIM 240112015Student researcherInformation SystemsAssists with corpus data processing, term normalisation, and research administration.—
Agriby Diandra ChaniagoNIM 240111017Student researcherInformaticsAssists with the drug–side-effect association analysis, the Sankey diagram visualisations, and the drug–side-effect matrix for the nursing output.—

National lecturer registration numbers and Sinta IDs are listed so that a reviewer can verify the team against the Sinta database.

Citation

This site is a demonstration, not a publication. Until the article appears, cite the research proposal:

Setyawati, M. B., Cahyaningrum, E. D., Mangku Negara, I. S., Khumaida, R., & Chaniago, A. D. (2026). Pemetaan Korelasi Obat Paliatif dan Efek Samping untuk Perencanaan Asuhan Keperawatan Paliatif Kanker Stadium Lanjut melalui Biomedical Text Mining (Proposal Penelitian Dosen Pemula No. PP-25070125481). Universitas Harapan Bangsa. pallicaremap.id, data v1.0.

The data version is part of the citation because the numbers will change once expert validation is complete.

Data version

Version
v1.0
Built
11 August 2026 at 15:36
Notebook hash
fdbe541174ca
Numeric regime
hasil penelitian = GPU fp16

This site will be updated once expert validation is complete; its version rises with its data.