An Enhanced SVM Model for Implicit Aspect Identification in Sentiment Analysis
Bibliographic record
Abstract
Opinion Mining or Sentiment Analysis (SA) is a key component of E-commerce applications where a vast number of reviews are generated by customers. SA operates on aspect level where the views are expressed on a specific aspect of a product and have a big influence on the customers’ choices and businesses’ reputation. Aspect Based Sentiment Analysis (ABSA) is the task of categorizing text by aspect and identifying the sentiment attributed to it. Implicit Aspect Identification (IAI) is a subtask of ABSA. This paper aims to empirically investigate how external knowledge (e.g. WordNet) is integrated into SVM model to address some of its intrinsic issues when dealing with classification. To achieve this research goal, we propose an approach to improve Support Vector Machines (SVM) model to deal with IAI. Using WordNet (WN) semantic relations, we suggest an enhancement for the SVM kernel computation. Experiments are conducted on three benchmark datasets of products, laptops, and restaurant reviews. The effects of our approach are examined and analyzed according to three criteria: (i) kernel function used, (ii) different experimental settings, and (iii) SVM behavior towards Overfitting and Underfitting. The research finding of our work is that the integration of external knowledge (e.g. WordNet) is experimentally proved to be significantly helpful to SVM classification for IAI and especially for addressing Overfitting and Underfitting that are considered as two of the main structural SVM issues. The empirical results demonstrate that our approach helps SVM (i) improve its performance for the three considered kernels and under different experimental settings, and (ii) deal better with Overfitting and Underfitting.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".