Tools
Tools
Heinig, S and Nanda, A (2021) Supervised Learning Algorithms to Extract Market Sentiment: An Application in the UK Commercial Real Estate Market. Real Estate Finance, 38 (2). pp. 147-160. ISSN 0748318X
|
Text
Supervised Learning Algorithms to Extract Market Sentiment An Application using Commercial Real Estate Market.pdf - Accepted Version Download (286kB) | Preview |
Abstract
Sentiment analysis has become a key area of research in economics and finance with methods evolving from traditional survey-based analysis to computational linguistic techniques. New developments in data handling and analysis have allowed extracting sentiment from vast amounts of written documents. However, these methods depend heavily on the existence of training and test data sets. The choice of training data is critical in such applications. We show a novel application from a unique market – commercial real estate. There are several unique attributes of the real estate market that makes such analysis critical for insightful market intelligence. In the absence of training data sets for the UK commercial real estate (CRE) market, we propose the use of Amazon book reviews for real estate related products. Our analysis has shown, that the use of more than 200,000 book reviews, can train different supervised learning algorithms, which in turn, can capture the sentiment and more importantly, it can help predict the direct commercial real estate market trends.
| Item Type: | Article |
|---|---|
| Additional Information: | Reprinted from Real Estate Finance, 38, 2, 1/10/2021, pp147-160, with permission of Kluwer Law International. |
| Subjects: | H Social Sciences > HF Commerce > HF5001 Business > HF5410 Marketing. Distribution of Products Q Science > QA Mathematics > QA75 Electronic computers. Computer science T Technology > T Technology (General) |
| Divisions: | Civil Engineering & Built Environment |
| Publisher: | Wolters Kluwer |
| Related URLs: | |
| Date Deposited: | 06 Sep 2021 12:12 |
| Last Modified: | 26 Aug 2022 10:15 |
| URI: | https://researchonline.ljmu.ac.uk/id/eprint/15127 |
 |
View Item |