Characteristics Analysis of Customer Vocabulary from Vegetable-Product Reviews

To collect new entries in a custom dictionary for morphological analysis, we analyzed customer vocabulary used in vegetable-product reviews. Key words specific to vegetable-product reviews were chosen on the basis of scores determined by TermExtract software from nouns regarded as customer vocabular...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Agricultural Information Research 2016, Vol.25(3), pp.105-115
Hauptverfasser: Takezaki, Akane, Oura, Yuji, Kono, Yoshinobu, Kiura, Takuji, Hayashi, Takeshi
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:To collect new entries in a custom dictionary for morphological analysis, we analyzed customer vocabulary used in vegetable-product reviews. Key words specific to vegetable-product reviews were chosen on the basis of scores determined by TermExtract software from nouns regarded as customer vocabulary that were not already included in the morpheme basic dictionary. Sixty percent of selected key words represented the product name, including cultivar and brand name. Of customer concepts in which selected key words were represented, 66% were also included in researcher concepts used on the National Agricultural Research Organization’s website. The ratio of researcher concepts to customer concepts categorized by product name was low, whereas those categorized by food functionality and safety were high. These results suggest the importance of product names as new entries in the custom dictionary for morphological analysis; the importance of gleaning product names not only from researcher vocabulary, but also from diverse sources with uniquely identifiable information; and the possibility of efficiently gleaning terms related to food functionality and safety from researcher vocabulary.
ISSN:0916-9482
1881-5219
DOI:10.3173/air.25.105