API misuse detection method and system based on attention mechanism

Abstract

The invention provides an attention mechanism-based API misuse detection method and system. The method comprises the following steps of 1, obtaining a predetermined quantity code of a detection field; 2, converting the predetermined quantity of codes into training data required by an attention mechanism-based API misuse detection model; 3, training the API misuse detection model based on the attention mechanism by using the training data to obtain model parameters; and step 4, obtaining a to-be-detected code, converting the to-be-detected code into a data format required by the API misuse detection model based on the attention mechanism, enabling the data format converted by the to-be-detected code to pass through the API misuse detection model based on the attention mechanism to obtain a detection result, and generating a detection report for the to-be-detected code. According to the method, a self-attention mechanism is applied to API use protocol learning and API misuse detection, a sample set used by an API is constructed by collecting a large number of Java codes and through static analysis, samples are trained by using a neural network, and a model for automatically detecting API misuse is constructed.

Description

technical field [0001] The present invention relates to the technical field of computer and artificial intelligence, and in particular, to an API misuse detection method and system based on an attention mechanism. Background technique [0002] With the rapid development of computer technology, more and more software frameworks, class libraries, etc. have emerged. Developers only need to call the Application Programming Interface (API) of these frameworks or class libraries to realize various functions. . However, due to reasons such as programmers ignoring the documentation when using the API, or it is difficult to understand the API, misuse of the API occurs, which brings defects or hidden dangers to the software. On the one hand, the use of artificial detection of these APIs is expensive to maintain, and on the other hand, the efficiency is relatively low. [0003] In the current API misuse research algorithm, it can be basically divided into two categories: one is to us...

AI Technical Summary

Problems solved by technology

In addition, in the existing paper research, when the code in the project set is converted into a sequence call, the degree of interdependence between different API elements in a sequence is not exactly the same, that is, in a series of API call sequences, the last element may be It leads to misuse of the first API, such as "getWidgetByName(string.substring(0,10)).setText()" when setting the text or title of the control in Android development. There are multiple APIs in this line of code, resulting in .setText () The misuse should be that the getWidgetByName API cannot get the object or the object does not exist, rather than the string.substring() API element, so elements at different positions in the API call sequence have inconsistent contributions to API misuse

In addition, experienced programmers often pay attention to different misuse defects for different situations when detecting whether the API is misused. For example, when checking object processing, they often pay attention to whether the object is empty. , often pay attention to API parameters and other information that lead to security

Images