It is worthy to note that in most of opinion mining tasks, the opinions do not have the validity period. However, since the financial market changes all the time, the investor’s opinions do have a validity period, even the opinions of professional stock analysts are the same. Most of the analysis reports of professional analysts set the validity period within one year even shorter.

Validity Time Period of Opinions

Fine-grained financial opinions can be represented in the following 9-tuple form: 
$$(e, s, h, t^p, t^v, a, d, C, M^e_{t^p})$$ 

where $e$ denotes the target entity, $s$ denotes the sentiment, $h$ denotes the opinion holder, $t^p$ denotes the publishing time, $t^v$ denotes the validity period, $a$ denotes the analysis aspect, $d$ denotes the degree of sentiment, $C$ denotes the set of investor’s claims, and $M^e_{t^p}$ denotes the market information set of target entity before publishing time.

Fine-Grained Financial Opinion Information Representation

Fine-grained financial opinions can be seen as an extension of coarse-grained financial opinion, potentially in the following aspects:
- Analysis aspect
- Degree of sentiment
- Investor's claims
- Market information
- Validity time period

University of Michigan - Ann Arbor

Financial opinion can be separated into two main aspects:
- Investor's opinion: the opinion about financial instruments
- Customer's opinion: the opinion about the financial product or financial service

Based on the two opinion types, we can also define the opinions by both coarse-grained and fine-grained viewpoints.

Financial Opinion Mining

Chung-Chi Chen, Hen-Hsen Huang, Hsin-Hsi Chen. 2020. Fine-grained Financial Opinion Mining: A Survey and Research Agenda. arXiv:2005.01897 [cs.CL]. https://arxiv.org/pdf/2005.01897.pdf.

Fine-grained Financial Opinion Mining: A Survey and Research Agenda

Coarse-grained financial opinions can be separated into two (bullish/bearish or positive/negative) or three (bullish/bearish/neutral or positive/negative/neutral) classes, and each opinion related to one target entity.

Learn Before

Related

Learn After