Chukwudi, Festus Uwasomba (2022) Harvesting knowledge from groupware using natural language processing and ontology. Doctoral thesis, Sunway University.
Full text not available from this repository.Abstract
Groupware exist, and they contain expertise knowledge (explicit and tacit) that is primarily for solving problems. Such knowledge should be harvested. A system to acquire on-the-job knowledge of experts from groupware in view of the enrichment of intelligent agents has become a technology that is very much in demand in the field of knowledge technology, especially in this era of textual data explosion, including due to the ever increasing remote work culture. Existing systems are developed based on text corpora from the web, social media, newspapers, textbooks and in particular from other ontologies. However, there is no clear framework with reusable, replaceable, and incremental modules for harvesting procedurals and practices from groupware. As a result, a system that accepts textual data from groupware discussions as input, processes the text and develops an ontology that can be used to enable AI applications remains a challenge. Given this gap, this thesis proposes such a framework to harvest on-the-job knowledge of experts from groupware. In knowledge acquisition, a key is to recognise new concepts in the source vis-à-vis the target ontology. For this, one has to be able to say that the concept is not already in the target ontology, thus not equal, similar or equivalent to an existing concept. Once recognized as a new concept, the concept needs to be inserted/hooked into the target ontology. At present, ontology equality of concepts is still not well-defined and there is no clear consensus on equality, similarity and equivalence of concepts, resulting in difficulties in recognising a new concept as well as for hooking into a target ontology, especially when it is coming from a sentence, which naturally has much less information. Herein, the thesis proposes a facts enrichment approach (FEA) as a novel methodology to rrecognizenew concepts from a sentence (albeit for the moment quite heuristically driven). FEA is also designed to eliminate computational complexities, identification difficulties, and complications of insertion (hooking) of a concept into a target ontology when a concept comes from a sentence. Following the recognition of new concepts, the FEA devises a distinct hooking strategy to adequately position the recognized new concepts into a target ontology. In terms of evaluation, the F1 score of 0.97 means that the technique has quite a high precision and recall rate, which is very much in line with the universal recommendations for the threshold. The results further indicate that the error rate (0.05) and accuracy rate (0.95) are also aligned with the suggestions for the universal thresholds (less than 0.10 and 0.90, respectively). Comparisons with other existing works also hold that the proposed technique can be deemed to be a success. This is based on the comparative analysis of the F1 scores that show that the proposed technique is slightly better than those that are available in the literature. In the future, we plan to fully automate the framework and incorporate the acquisition of tacit knowledge from groupware.
| Item Type: | Thesis (Doctoral) |
|---|---|
| Uncontrolled Keywords: | knowledge technology; ontology; knowledge acquisition; NLP; virtual software development team; |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software |
| Divisions: | Sunway University > School of Engineering and Technology [formerly School of Science and Technology until 2020] > Dept. Computing and Information Systems |
| Depositing User: | Ms Yong Yee Chan |
| Date Deposited: | 27 Sep 2023 01:50 |
| Last Modified: | 27 Sep 2023 01:50 |
| URI: | http://eprints.sunway.edu.my/id/eprint/2382 |
Actions (login required)
 |
View Item |
