Enhancing food composition databases : predicting missing values via knowledge graph embeddings

Možina, Marko (author), Žitnik, Slavko (mentor), Eftimov, Tome (comentor)

Food composition databases (FCDBs) have presented an integral part of food and nutritional research, dietary assessment, and related (e.g., health, environmental) fields. However, as with other scientific disciplines, the domain of nutrition and food composition is no exception to the problem of missing data. This can significantly reduce the accuracy and reliability of analyses based on food composition, as it introduces an element of ambiguity and can, therefore, limit their usage. To address this issue, researchers have explored various methods for imputing missing data. The easiest and most common approach to this problem is to calculate the mean or median from available data in the same FCDB or to borrow values from other FCDBs. However, such simple methods may produce notable errors. In this thesis, we investigate the use of knowledge graph embedding models for imputing missing values in FCDB. We used the ComplEx model from the Ampligraph library and results are very promising. By employing the approach described in our paper, the model can capture the underlying structure and relationships in the data, providing accurate imputations of missing values. Ultimately, the use of the proposed technique could lead to more accurate and reliable analyses in the field of nutritional research and dietary monitoring.

• Language: English
• Keywords: food composition database, nutrient values, missing data, graph machine learning, knowledge graph embeddings, data exploration, missing value imputation, ampligraph
• Work type: Bachelor thesis/paper
• Typology: 2.11 - Undergraduate Thesis
• Organization: FRI - Faculty of Computer and Information Science ; FMF - Faculty of Mathematics and Physics
• Year: 2023
• PID: 20.500.12556/RUL-149148
• COBISS.SI-ID: 164598019
• Publication date in RUL: 04.09.2023

MOŽINA, Marko, 2023, Enhancing food composition databases : predicting missing values via knowledge graph embeddings [online]. Bachelor’s thesis. [Accessed 8 April 2025]. Retrieved from: https://repozitorij.uni-lj.si/IzpisGradiva.php?lang=eng&id=149148

Secondary language

• Language: Slovenian
• Title: Izboljševanje podatkovnih baz sestave živil : dopolnjevanje manjkajočih vrednosti preko vložitev grafa znanja

Abstract

Podatkovne baze sestave živil (PBSŽ) so temeljno orodje pri raziskavah na področju prehrane, ocenah hranilnih vrednosti in sorodnih domenah (npr. zdravje, okolje). Vendar se, tako kot druge znanstvene discipline, tudi področje prehrane in sestave živil spopada s težavo manjkajočih podatkov. To lahko znatno zmanjša natančnost in zanesljivost analiz, ki temeljijo na strukturi živil, saj vpeljuje element dvoumnosti in s tem omejuje njihovo uporabo. Za rešitev tega problema so bile predlagane različne metode za dopolnjevanje manjkajočih podatkov. Najlažji in najpogostejši pristop je izračun povprečja oziroma mediane iz razpoložljivih podatkov v isti bazi ali pa izposoja vrednosti iz drugih. Vendar pa lahko takšne preproste metode povzročijo znatne napake. V tem diplomskem delu se raziskuje uporaba modela ComplEx iz knjižnice Ampligraph, ki temelji na vektorskih vložitvah grafa znanja za dopolnjevanje manjkajočih vrednosti v PBSŽ. S pristopom opisanim v tem delu lahko model zajame temeljno strukturo in odnose med podatki, kar omogoča natančno dopolnjevanje manjkajočih vrednosti. To dodatno potrjujejo rezultati tega dela, saj so primerljivi s tistimi najsodobnejših modelov. Uporaba predlagane metode bi lahko v prihodnje privedla do natančnejših in zanesljivejših analiz na področju prehranskih raziskav.

• Keywords: podatkovna baza sestave živil, hranilne vrednosti, manjkajoči podatki, strojno učenje na grafih, vložitev grafa znanja, analiza in raziskovanje podatkov, dopolnjevanje manjkajočih vrednosti, ampligraph