Publicação
Natural Language Processing applied to Food Data: a smart food description mapping system
| dc.contributor.author | Tomé, Sidney | |
| dc.date.accessioned | 2020-07-14T11:01:43Z | |
| dc.date.available | 2020-07-14T11:01:43Z | |
| dc.date.issued | 2019-10-18 | |
| dc.description.abstract | Introduction: For the past years we have been working with EFSA on report of food data for the domains of chemical contaminants and food additives; Over the years, INSA has been the central point for gathering data from multiple national entities for processing and redirecting these information to EFSA; A National Data Management Systems was built from the start in order to facilitate such tasK; A consortium between 3 partners (INSA, ASAE & HAPIH) was created and they share the common interest of further developing their own official control National Data Management Systems (NDMS); INSA has developed PT.ON.DATA NDMS based on SSD and SSD2 data models in cooperation with ASAE and other national competent authorities. HAPIH has developed a NDMS with a similar approach; Both ASAE and HAPIH are interested in implementing real-time sample data collection based on preparatory digital forms already existing in PT.ON.DATA. The three partners are committed to investigate and implement an automatic approach to FoodEx2 classification of food samples using the knowledge and databases existing in HAPIH and INSA. Consortium aim: The consortium envisages improving the quality of raw occurrence data for risk assessment by reducing error, incrementing completeness and timeliness both in data fields and food classification, and simultaneously reducing human time and work and therefore releasing time of scientists for data analysis and for performing risk assessment; Together, the improvements will be reflected on the strengthening of food safety risk assessment capacity of the countries involved and contributing to a better assess on risks associated with the food chain by EFSA. Project IDRisk (Improving Data quality for RISK assessment) - This project emerged from two distinct problems regarding data report: Technicians/Data experts, whether they are working on-field sampling data or cleaning and preparing data for report, an absurd amount of manual work is done: too much time is spent; The more manual work we have, the more prone to errors the data will become. Main objectives: Improve/Restructure the dynamic sampling forms module; Develop the application that will run on the mobile devices; Plan and implement an automatic NDMS FoodEx2 classification system for sampling descriptions. | pt_PT |
| dc.description.version | N/A | pt_PT |
| dc.identifier.uri | http://hdl.handle.net/10400.18/7037 | |
| dc.language.iso | eng | pt_PT |
| dc.peerreviewed | no | pt_PT |
| dc.subject | Natural Language | pt_PT |
| dc.subject | Natural Language Processing | pt_PT |
| dc.subject | Food Data | pt_PT |
| dc.subject | Smart Food Description | pt_PT |
| dc.subject | Mapping System | pt_PT |
| dc.subject | IDRisk | pt_PT |
| dc.subject | Segurança Alimentar | pt_PT |
| dc.title | Natural Language Processing applied to Food Data: a smart food description mapping system | pt_PT |
| dc.type | conference object | |
| dspace.entity.type | Publication | |
| oaire.citation.conferencePlace | Lisboa, Portugal | pt_PT |
| oaire.citation.title | 13th International Food Data Conference (IFDC 2019), International Network of Food Data Systems, 5-18 outubro 2019 | pt_PT |
| rcaap.rights | restrictedAccess | pt_PT |
| rcaap.type | conferenceObject | pt_PT |