Noun phrase (NP) complexity research has shown the effects of both discipline and writing competence on NP complexity in academic writing and has focused more on applied linguistics. Yet few studies examined NPs in the academic writing of computer science (CS), especially in the CS conference abstract writing, in depth. This study fills this gap by investigating the disciplinary preference of NPs through the corpus analysis of 267 published abstracts from a leading CS conference. The authors found that multiple pre-modifiers were the most frequently used device by CS researchers, and attributive adjectives, nouns, and prepositional phrases were fundamental in abstract composition in both CS and applied linguistics. The difference largely lies in the use of devices in later-acquired stages. CS researchers favor more multiple pre-modifiers while their peers in applied linguistics tend to prefer multiple prepositional phrases as post-modifiers. The findings shed light on classroom instruction and future research on NP complexity.

Compiler testing is the most widely used way to assure compiler quality. However, since compilers require a large number of sophisticated test programs as inputs, the existing approaches in compiler testing still have a limited capability in generating both syntactically valid and diverse test programs. In this paper, we propose DeepGen, a deep learning-based approach to support compiler testing through the inference of a generative model for compiler inputs. First, DeepGen trains a Transformer-XL model based on a large corpus of seed programs, and uses the trained model to generate syntactically valid programs. Then, DeepGen adopts a sampling strategy in the inference phase to generate diverse test programs. Finally, DeepGen leverages differential testing on the generated programs to discover compiler bugs. We have evaluated DeepGen over two popular C++ compilers GCC and LLVM, and the results confirm the effectiveness of our approach. DeepGen detects 35.29%, 53.33%, and 187.50% more bugs than three existing approaches, i.e. DeepSmith, DeepFuzz, and Csmith, respectively. In addition, 30.43% bugs detected by DeepGen are not detected by other approaches. Furthermore, DeepGen has successfully detected 38 bugs in the latest development versions of GCC and LLVM; 21 of them have been confirmed/fixed by the developers.