Reading StackOverflow Encourages Cheating: Adding Question Text Improves Extractive Code Generation

ACL (NLP4Prog) 2021  ·  Gabriel Orlanski, Alex Gittens ·

Answering a programming question using only its title is difficult as salient contextual information is omitted. Based on this observation, we present a corpus of over 40,000 StackOverflow question texts to be used in conjunction with their corresponding intents from the CoNaLa dataset (Yin et al., 2018). Using both the intent and question body, we use BART to establish a baseline BLEU score of 34.35 for this new task. We find further improvements of $2.8\%$ by combining the mined CoNaLa data with the labeled data to achieve a 35.32 BLEU score. We evaluate prior state-of-the-art CoNaLa models with this additional data and find that our proposed method of using the body and mined data beats the BLEU score of the prior state-of-the-art by $71.96\%$. Finally, we perform ablations to demonstrate that BART is an unsupervised multimodal learner and examine its extractive behavior. The code and data can be found https://github.com/gabeorlanski/stackoverflow-encourages-cheating.

PDF Abstract ACL (NLP4Prog) 2021 PDF ACL (NLP4Prog) 2021 Abstract

Datasets


Introduced in the Paper:

CoNaLa-Ext

Used in the Paper:

CodeSearchNet CoNaLa JuICe

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Benchmark
Code Generation CoNaLa BART W/ Mined BLEU 30.55 # 6
Code Generation CoNaLa BART Base BLEU 26.24 # 8
Code Generation CoNaLa-Ext BART W/ Mined BLEU 35.32 # 1
Code Generation CoNaLa-Ext BART Base BLEU 34.35 # 2

Methods