ارسل ملاحظاتك

ارسل ملاحظاتك لنا







Automatic Annotation and Text Categorization of Marketing Document Corpus

المصدر: مجلة جامعة الزيتونة
الناشر: جامعة الزيتونة
المؤلف الرئيسي: Casteleiro, Mercedes Arguello (Author)
مؤلفين آخرين: Abusa, Mukhtar (Co-Author)
المجلد/العدد: ع10
محكمة: نعم
الدولة: ليبيا
التاريخ الميلادي: 2014
الصفحات: 34 - 68
DOI: 10.35778/1742-000-010-025
ISSN: 2523-1006
رقم MD: 840043
نوع المحتوى: بحوث ومقالات
قواعد المعلومات: EduSearch, EcoLink, IslamicInfo, HumanIndex
مواضيع:
رابط المحتوى:
صورة الغلاف QR قانون

عدد مرات التحميل

3

حفظ في:
المستخلص: The ever-evolving field of marketing possesses an ever-growing lexicon, which aspires to capture the new trends and processes in the field. Neologisms constantly emerge and they can be widespread across marketing publications before proper terminological resources can have the time to acknowledge and add them. Although automated text processing is highly desirable; it can be easily hampered, as it is possible to find sets of words with different form that have the same meaning. Therefore, detecting the marketing topics mentioned in a document and annotating them accordingly, it is a burdensome task performed by marketing experts and trained terminologists that cannot be easily delegated to computers. This paper exemplifies the use of Web services for automatic annotation and text categorisation of highly specialised documents in the field of marketing. The approach presented outlines how Semantic Web technologies can be crucial to obtain a Web-based application that easily deals with terminological background knowledge. The annotation application developed and presented in this paper is based on Web services and considers three types of annotations: Dublin Core annotation, thesaurus annotation, and ontology-based annotation. To validate the proposal, the research makes use of a marketing document corpus that contains documents in different formats. The paper provides details of the evaluation performed for this corpus.

ISSN: 2523-1006

عناصر مشابهة