Automatic Generation of Product Association Networks Using Latent Dirichlet Allocation
| Autor: | Johannes Putzke, Javier Sanchez-Monzon, Kai Fischbach |
|---|---|
| Rok vydání: | 2011 |
| Předmět: |
Topic model
social networks LDA Computer science Association (object-oriented programming) webmining Machine learning computer.software_genre collaborative innovation networks Latent Dirichlet allocation Semantic network symbols.namesake General Materials Science COINs Product (category theory) Information retrieval business.industry Concept map Probabilistic logic data mining brand concept maps symbols Thematic structure Artificial intelligence business computer |
| Zdroj: | Procedia - Social and Behavioral Sciences. 26:63-75 |
| ISSN: | 1877-0428 |
| DOI: | 10.1016/j.sbspro.2011.10.563 |
| Popis: | We present a method for extracting semantic networks of words that consumers associate with products and brands, and illustrate the method using reviews of McDonald's products from the opinion platform www.ciao.de as examples. We model the generation of each product review with the probabilistic topic model Latent Dirichlet Allocation (LDA), which enables us to discover the hidden thematic structure of all the reviews in our text collection. We conduct an association analysis of all the words used, revealing the semantic networks of words. Our approach may be highly relevant for marketing managers, for example, as they analyze brand concept maps or seek to optimize ad campaigns with the best words. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|