dataset_info:
features:
- name: id
dtype: int64
- name: org_index
dtype: int64
- name: data_source
dtype: string
- name: industry
dtype: string
- name: text
dtype: string
- name: labels
sequence:
sequence: string
- name: label_codes
dtype: string
splits:
- name: train
num_bytes: 2599501.8469831664
num_examples: 7930
- name: validation
num_bytes: 346490.977586533
num_examples: 1057
- name: test
num_bytes: 520228.17543030076
num_examples: 1587
download_size: 1010316
dataset_size: 3466221
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*
However, the development of such manually curated
datasets is a labour-intensive process and therefore existing ABSA datasets cover only a few domains and they are limited in size. In response, we present FABSA (Feedback ABSA), a new large-scale and multi-domain ABSA dataset of feedback reviews. FABSA consists of approximately 10,500 reviews which span across 10 domains
A triple professionally annotated ABSA dataset, specifically for Aspect Category Sentiment Analysis dataset which can be used for Aspect Category Detection (ACD) and Aspect Category Sentiment Classification (ACSC). Aspect category detection (ACD) in sentiment analysis aims to identify the aspect categories mentioned in a sentence. These categories are conceptual, i.e. they do not necessarily explicitly appear in the text, and come from a predefined list of Aspect Categories. Aspect Category Sentiment Classification (ACSC) aims to classify the sentiment polarities of the conceptual aspect categories.
The predefined list of Aspect Categories for this dataset are:
category category_code
0 Account management: Account access account-management.account-access
1 Company brand: Competitor company-brand.competitor
2 Company brand: General satisfaction company-brand.general-satisfaction
3 Company brand: Reviews company-brand.reviews
4 Logistics rides: Speed logistics-rides.speed
5 Online experience: App website online-experience.app-website
6 Purchase booking experience: Ease of use purchase-booking-experience.ease-of-use
7 Staff support: Attitude of staff staff-support.attitude-of-staff
8 Staff support: Email staff-support.email
9 Staff support: Phone staff-support.phone
10 Value: Discounts promotions value.discounts-promotions
11 Value: Price value for money value.price-value-for-money
Annotation
Two annotators have extensive experience in developing manually labelled ABSA datasets for a commercial company, but do not have a formal background/education related to linguistics. The third annotator has a PhD in computational linguistics and is assumed to be an expert tagger.
Annotation Scheme
3.2. Annotation scheme
The FABSA dataset is manually labelled against a hierarchical annotation scheme which consists of parent and child aspect categories (Fig. 2). Each aspect category is associated with a sentiment label (positive, negative and neutral). This creates a total of
(12 × 3) target classification categories.
Following previous work, we adopt a multi-label classification scheme wherein each review is labelled with one or more aspect＋ sentiment label. Accordingly, a single review may contain multiple different aspects and express different (and in some cases contrasting) polarities. Table 3 shows an example of a review which is associated with two different aspect categories and two different and conflicting polarity labels (positive and negative, respectively).
