Ontology-based Why-Question Analysis Using Lexico-Syntactic Patterns

Abstract

This research focuses on developing a method to analyze why-questions. Some previous researches on the why-question analysis usually used the morphological and the syntactical approach without considering the expected answer types. Moreover, they rarely involved domain ontology to capture the semantic or conceptualization of the content. Consequently, some semantic mismatches occurred and then resulting not appropriate answers. The proposed method considers the expected answer types and involves domain ontology. It adapts the simple, the bag-of-words like model, by using semantic entities (i.e., concepts/entities and relations) instead of words to represent a query. The proposed method expands the question by adding the additional semantic entities got by executing the constructed SPARQL query of the why-question over the domain ontology. The major contribution of this research is in developing an ontology-based why-question analysis method by considering the expected answer types. Some experiments have been conducted to evaluate each phase of the proposed method. The results show good performance for all performance measures used (i.e., precision, recall, undergeneration, and overgeneration). Furthermore, comparison against two baseline methods, the keyword-based ones (i.e., the term-based and the phrase-based method), shows that the proposed method obtained better performance results in terms of MRR and P@10 values.

O Ontology-Based Paragraph Extraction and Causality Detection-Based Similarity for Answering Why-Question

“…The output of the paragraph extraction is a list of extracted paragraphs. CA (a set of causality annotations), OSA (a set of original semantic annotations) and ASA (a set of additional semantic annotations) are semantic annotations of a question obtained from question analysis step (Karyawati et al, 2015). Eka Karyawati, Ontology-Based Paragraph … 11…”

Ontology-Based Sentence Extraction for Answering Why-Question

“…The experiments are conducted by generating randomly 10, 20, 30, and 40 questions from the why-question collection in 10 iterations, where the total number of questions available is 5921 why-questions. The question collection is constructed through three steps, firstly, why-questions (i.e., the general domain questions) is collected from the Web; secondly, the why-questions are analyzed to identify general patterns of the why-questions; and thirdly, the why-questions in a specific domain (i.e., Text Retrieval) are generated using the general patterns [2]. The general domain why-questions are collected from the Verberne's data collection [20].…”

“…There are three sets of semantic annotations of a question, which are a set of original semantic annotations (OSA), a set of additional semantic annotations (ASA), and a set of causality annotations (CA) [2]. The OSA is identified from the original question (i.e., inputted question), and the ASA is identified through the query expansion process.…”

“…In our research, the sentence extraction is the second step of answer extraction phase of the three phase why-QA method (i.e., question analysis [2], document retrieval, and answer extraction). The first step of the answer extraction is a paragraph extraction that has a scoring formula similar to the document ranking formula.…”

“…The proposed method is applied in a specific domain (i.e., Text Retrieval domain). The Text Retrieval domain ontology used is this research has been built by [2]. The domain lexicon of the Text Retrieval ontology is used as a basis for identifying the semantic annotations of sentences.…”