Automatically Classifying Non-functional Requirements with Feature Extraction and Supervised Machine Learning Techniques
Bibliographic record
Abstract
Abstract. Context and Motivation: Non-functional requirements (NFRs) of a system need to be classified into different types such as usability, performance, etc. This would enable stakeholders to ensure the completeness of their work by extracting specific NFRs related to their expertise. Question/Problem: Because of the size and complexity of requirement specification documents, the manual classification of NFRs is time-consuming, labour-intensive, and error-prone. We thus need an automated solution that can provide a highly accurate and efficient categorization of NFRs. Principal ideas/results: In this investigation, using natural language processing and supervised machine learning (SML) techniques, we investigate with feature extraction techniques including Part Of Speech-tagging based, Bag of Words (BoW) ,and Term Frequency-Inverse Document Frequency (TF-IDF) combined with SML algorithms including Support Vector Machine (SVM), Stochastic Gradient Descent (SGD) SVM, Linear Regression (LR), Decision Tree (DT), Bagging DT, Extra Tree, Random Forest (RF), Gaussian Naïve Bayes (GNB), Multinomial Naïve Bayes (MNB), and Bernoulli Naïve Bayes (BNB). Contribution: The proposed strategy consists of three different combinations of the above-mentioned techniques. SVM with TF-IDF, LR with POS and BoW, and MNB with BoW all achieved recall values higher than 0.90, precision values above 0.87, and execution times less than 0.1s. In addition, we validated these classifiers using a case-study dataset where they promise results of recall values over 0.90 and precision values over 0.92.
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 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".