Diet Recipe Recommendation for Chronic Patients by Social Network Analysis

• Main Authors: Chia-Tzu Chang, 張嘉慈
• Other Authors: Chi-Bin Cheng
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/g6frd6

碩士 === 淡江大學 === 資訊管理學系碩士在職專班 === 106 === Finding a suitable recipe from Internet is not an easy task for chronic patients. Most of healthy recipe recommendation techniques are deigned only on daily intake of energy without considering personal dietary restrictions. The purpose of this thesis is to help chronic patients find proper recipes from Internet. This study first randomly selects recipes from Internet as a sample set. The associations among ingredients are measured by the pointwise mutual information algorithm based on the sampled recipes. An ingredient social network is then constructed according to the obtained ingredient associations. By using social network analysis to explore the structure of the ingredient network and the Taiwanese Food Nutrient Database and Dietary Guideline to compute healthy scores of ingredients, key ingredients are identified and used as suggesting keywords for Internet recipe search. Referring to the ranks of ingredients’ healthy scores, users could choose their preferred ingredients as searching keywords. Our experiments reveal that the searched recipes are efficient and effective by using the proposed approach.