Semantic API Recommendation via Ontology and Text Preprocessing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing approaches to recommend services using natural language queries are limited by reliance on manual annotation and are not readily applicable to new domains, as they primarily focus on keyword matching and entity relationships, ignoring semantic similarity between queries and services.
Innovation Solution
A framework that generates domain and client-specific application program interface (API) recommendations by preprocessing service descriptions to remove frequent and infrequent words, constructing communities based on semantic similarity, and using ontology components to analyze performance and generate recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If manual annotation and keyword matching are used for service recommendation, then implementation is straightforward, but the system is not readily applicable to new domains and ignores semantic similarity
Solution Approach 1:
The system performs preliminary action by pre-processing service descriptions to extract semantic features and build ontologies before actual query processing. This allows the system to be immediately applicable to new domains without requiring manual annotation for each new domain, as the semantic extraction framework is already in place
Solution Approach 2:
The patent introduces an intermediary layer (semantic feature extraction and ontology components) between the raw service descriptions and the recommendation engine. This intermediary transforms unstructured text into structured semantic representations that can be efficiently queried, enabling domain adaptability without increasing overall system complexity
2Measurement precision
If semantic similarity analysis is implemented, then meaningful recommendations across multiple domains are achieved, but preprocessing complexity increases
Solution Approach 1:
The preprocessing pipeline is segmented into distinct modular components: text cleaning, tokenization, semantic feature extraction, and ontology generation. Each component handles a specific aspect of the processing, making the overall complex task manageable and maintainable while achieving high measurement precision through cumulative refinement
Solution Approach 2:
The system changes parameters by transforming raw text into multiple derived representations (tokens, n-grams, semantic features, ontology nodes). This parameter transformation enables precise semantic similarity measurement while the modular architecture keeps preprocessing complexity controlled
3Loss of information
If frequent and infrequent words are removed from service descriptions, then semantic relationships are better captured, but information loss may occur
Solution Approach 1:
The system selectively extracts only the essential semantic-carrying elements from service descriptions while removing stop words and noise. This extraction process removes redundant text data (articles, prepositions, common verbs) that do not contribute to semantic meaning, thereby reducing information loss while capturing core semantic relationships
Data Source
AI summary
Techniques for generating domain and client-specific application program interface recommendations are provided. In one example, a computer-implemented method comprises modifying, by a device operatively coupled to a processor, a description of a client application program interface by removing text data associated with the description of the client application program interface, resulting in a modified description of the client application program interface. The computer-implemented method can further comprise analyzing, by the device, a performance associated with the client application program interface to generate an ontology based on a semantic similarity between the modified description of the client application program interface and one or more previous descriptions of one or more previous client application program interfaces.