Extracting brand names and product names from product review webpages: A case of mixed languages
Kanokwan Atchariyachanvanich, Sakda Jantararoungtong, Vichayanan Jirabovolvanit · 2016
Most of Thai product review websites display mixed contents by showing the product name and brand name in English language and reviewed contents in Thai language. However, the current product extraction application programming interfaces (API) could not extract product name written in English language from Thai product review websites. Therefore, this paper proposed the algorithm to extract brand names and product names from mixed Thai-English language product review websites. The experimental result revealed the proposed algorithm had a 97.48% of precision. The contribution of the proposed algorithm can be a part of location-based m-commerce. The algorithm extracts product name from the logs file of wireless access point in a shopping mall in order to recognize customers' interest. The sale promotion or advertising will be offered to the customers based on their interest via their mobile phone.