Integrasi Data dan Visualisasi Graf pada Aplikasi Halal Nutrition Food
Bibliographic record
Abstract
Halal Nutrition Food merupakan sebuah aplikasi pencarian produk halal berbasis web yang menampilkan berbagai informasi produk halal secara rinci yang didapatkan dari gabungan dataset yang diintegrasikan dalam bentuk Linked Open Data. Namun, saat ini, Halal Nutrition Food masih memiliki sedikit jumlah produk beserta informasinya dan juga masih belum memiliki visualisasi sehingga pengguna tidak dapat memiliki wawasan mengenai data dengan mudah. Open Food Fact merupakan open data produk makanan beserta informasinya yang berasal dari seluruh dunia. Untuk itu, perlu adanya integrasi data antara Open Food Fact dan Halal Nutrition Food. Kami hanya mengambil data produk yang memiliki bahan makanan dalam bahasa inggris dan perancis. Data tersebut dibersihkan melalui data cleansing yang di dalamnya terdapat pengukuran kesamaan bahan makanan menggunakan Levenshtein dan Jaccard distance untuk memperbaki kesalahan penulisan. Hasil data cleansing diintegrasikan ke Halal Nutrition Food. Untuk mengetahui kelompok makanan apa saja yang ada, data yang sudah dalam bentuk graf produk dan bahan makanan, dikelompokkan menggunakan METIS graph partitioning. Hasil data cleansing menunjukkan bahwa untuk memperbaiki kesalahan penulisan, Levenshtein distance lebih baik daripada Jaccard distance. Pengukuran kesamaan menggunakan Jaccard distance menunjukkan bahwa ada terlalu banyak bahan makanan yang mirip. Padahal, sebenarnya tidak ada kesalahan penulisan dan bahan makanan tersebut memiliki perbedaan arti. Produk dan bahan makanan dikelompokkan menjadi 20 kelompok, sesuai pengelompokan produk yang dilakukan oleh MUI. Hasil graph partitioning divisualisikan bersama dengan visualisasi graf hubungan produk dan bahan makanan, visualisasi graf produk makanan yang berstatus haram dan visualisasi graf produk makanan yang mengandung MSG. =================== Halal Nutrition Food is a web-based halal product search application featuring a variety of halal product information in detail obtained from a combined dataset that is integrated into the form of Linked Open Data. However, currently, Halal Nutrition Food still has a small number of products along with its information and also still has no visualization so users can not have insight into data easily. Open Food Fact is an open data of food products along with information from all over the world. So, there is a need for data integration between Open Food Fact and Halal Nutrition Food. We only collect food products in English and French from Open Food Fact. Those data are cleaned through data cleansing with similarity measurement using Levenshtein and Jaccard distance to correct typos. The data cleansing results are integrated into Halal Nutrition Food. To reveal what food product/ingredient clusters are, the data that already in the product and food graphs, are clustered using METIS graph partitioning. The data cleansing results show that to correct typos, the Levenshtein distance is better than the Jaccard distance. Jaccard distance similarity measurement shows that there are too many similar ingredients. In fact, there are no typos and those have the different meaning. Food products and ingredients are clustered into 20 clusters, according to the MUI product grouping. The results of graph partitioning are visualized along with graph visualization of food products and ingredients relationships, haram food products and food products that contain MSG.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.003 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; both teacher heads agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".