Major Duties:

1. To be actively involved in the research programme as directed by the line manager/project supervisor and focus on developing an enterprise-level solution for risk assessment in complex business networks backed by strong research on the subject-matter.
2. Carry out routine administrative tasks associated with the research project/s to ensure that projects are completed on time.
3. Developing proof-of-concept to justify the research.
4. Carry out appropriate analysis and write up results of own work and lead a new direction as the project progresses.
5. Present regular progress reports on research to members of the research group or external audiences to disseminate and publicise research findings.
6. Use of road mapping/project development tools to share ongoing status updates.
7. Contribute to the production of research reports, publications, and proposals.
8. Any other duties that the programme supervisor may reasonably request.

Essential Criteria:

1. 2.1 Honours Degree (or equivalent) in Applied Mathematics, Computer Science, Electronics, Electrical Engineering, or a closely related discipline.
2. Normally have or be about to obtain a PhD in Computer Science, Applied Mathematics, Electronics, Electrical Engineering, Physics.
3. Relevant experience to include: • Recent, relevant research experience in at least one of the following: risk assessment, AI-modelling, business networks. • Experience developing risk assessment systems. • Demonstrable experience of software development systems (preferably Python/Java/C#).
4. Experience of working effectively as part of a research team in the development and promotion of the research theme.
5. Strong software development skills and proven track record of developing proof of concepts.
6. Sufficient breadth and depth of specialist knowledge in the discipline and of research methods and techniques.
7. Ability to contribute to broader management and administrative processes.
8. Contribute to the School’s outreach programme by links with industry, community groups etc.
9. Practical problem-solving skills, independence of thought and initiative.
10. Ability to assess and organise resources.
11. Ability to communicate complex information in English effectively in oral and written format to technical and non-technical audiences.
12. Ability to build relationships with a wide range of people and roles at different levels of seniority and to influence decision making.
13. Ability to manage self and prioritise workload.
14. A pro-active approach to work and team development.
15. Commitment to continuous professional development.
16. Ability to meet the mobility requirements of the post including the travel to project partners as required by the role.

Desirable Criteria:

1. Strong background in risk assessment systems.
2. Experience of the application of dashboards for multidisciplinary activities.
3. Experience of developing and testing novel algorithms.
4. Mathematical skills for conceptualisation, modelling, optimisation, and analysis of problems.

Major Duties:

1. Undertake research under supervision within the specific research project and as a member of the collaborative research team contribute to develop and evaluate NLP and deep learning models and apply knowledge of relevant research domains along with expert coding skills to platform and framework development projects.
2. Develop/apply scalable algorithms based on state-of-the-art machine learning methodologies and design and evaluate suitable natural language processing models and workflows.
3. Carry out analyses and critical evaluation, in order to interpret, explain and further improve model performance and utility.
4. Engage with the relevant research literature, in order to develop methodologies appropriate to the area of research across a range of platforms and facilities of the wider PwC partnership.
5. Produce high quality research outputs consistent with project aims and commensurate with career stage. This will include collaborating and co-authoring with PI and project team (as appropriate) on outputs.
6. In consultation with the project team, promote research milestones and outputs at national and international conferences and through social media (where applicable).
7. Carry out occasional educational supervision, demonstrating or lecturing duties within the post holder’s area of expertise and under the direct guidance of a member of academic staff.
8. Undertake supplementary duties relevant to the success of the project including administrative duties, presentation of regular progress reports and additional training and development activities as required.

Essential Criteria:

1. 2.1 Honours Degree (or equivalent) in Applied Mathematics, Computer Science, Electronics, Electrical Engineering, or a closely related discipline.
2. Normally have or be about to obtain a PhD in Computer Science, Applied Mathematics, Electronics, Electrical Engineering, Physics.
3. Relevant experience to include: • Relevant recent research experience in at least one of: intelligent systems, artificial intelligence, algorithms development. • Recent experience developing deep learning and/or NLP systems • Working effectively as part of a Research team in the development and promotion of research topics
4. Demonstrable knowledge of Python, and experience with Java, R, C#, or other relevant languages. Knowledge of deep learning or NLP frameworks and tools (e.g. PyTorch, huggingface, spaCy).
5. Demonstrable knowledge of typical algorithms and concepts used in machine learning and deep learning, to include both model architectures and model evaluation.
6. Strong publication record, commensurate with stage of career.
7. Sufficient breadth and depth of specialist knowledge in the discipline and of research methods and techniques.
8. Ability to contribute to broader management and administrative processes.
9. Contribute to the School’s outreach programme by links with industry, community groups etc.
10. Practical problem solving skills, independence of thought and initiative.
11. Ability to communicate complex information in English effectively in oral and written format to technical and non-technical audiences.
12. Ability to build relationships with a wide range of people and roles at different levels of seniority and to influence decision making.
13. Ability to manage self and prioritise workload.
14. A pro-active approach to work and team development.
15. Commitment to continuous professional development.
16. Ability to meet the mobility requirements of the post including the travel to project partners as required by the role.

Desirable Criteria:

1. Strong background in deep learning model development.
2. Experience of the application of AI algorithms and software in multidisciplinary activities.
3. Experience of developing and testing novel algorithms.

Major Duties:

1. Undertake research under supervision within the specific research project and as a member of the collaborative research team contribute to develop 3D environments for visualisation and real time interaction and apply knowledge of relevant research domains along with expert coding skills to platform and framework development projects.
2. Develop/apply highly scalable algorithms based on state-of-the-art machine learning methodologies and design suitable human computer interaction user experimental studies.
3. Carry out analyses, experimental tests, critical evaluation and implementation, and interpretations of experimental data and the literature using methodologies and other techniques appropriate to area of research across a range of platforms and facilities of the wider PwC partnership.
4. Use of roadmapping/project development tools to share ongoing status updates.
5. Produce high quality research outputs consistent with project aims and commensurate with career stage. This will include collaborating and co-authoring with PI and project team (as appropriate) on outputs.
6. In consultation with the project team, promote research milestones and outputs at national and international conferences and through social media (where applicable).
7. Carry out occasional educational supervision, demonstrating or lecturing duties within the post holder’s area of expertise and under the direct guidance of a member of academic staff.
8. Undertake supplementary duties relevant to the success of the project including administrative duties, presentation of regular progress reports and additional training and development activities as required.

Essential Criteria:

1. 2.1 Honours Degree (or equivalent) in Applied Mathematics, Computer Science, Electronics, Electrical Engineering, or a closely related discipline.
2. Normally have or be about to obtain a PhD in Computer Science. Applied Mathematics, Electronics, Electrical Engineering, Physics.
3. Relevant experience to include:
   
   • At least 3 years research experience in at least one of: intelligent systems, artificial intelligence, algorithms development. 
   • 1-3+ years of experience developing AR or VR systems.
   • Working effectively as part of a research team in the development and promotion of the research theme.
   •  Unity debugging experience.
4. Evidence of knowledge of:
   • Scripting languages, Java or Python or C#, and proficient in C++ programming and of how to optimise software for specific headsets and platforms.
   • Unity Game Engine and Editor.
   • Typical algorithms used in AI.
5. Strong publication record, commensurate with stage of career.
6. Ability to contribute to broader management and administrative processes.
7. Contribute to the School’s outreach programme by links with industry, community groups etc.
8. Practical problem solving skills, independence of thought and initiative.
9. Ability to assess and organise resources.
10. Ability to communicate complex information in English effectively in oral and written format to technical and non-technical audiences.
11. Ability to build relationships with a wide range of people and roles at different levels of seniority and to influence decision making.
12. Ability to manage self and prioritise workload.
13. A pro-active approach to work and team development.
14. Commitment to continuous professional development.
15. Ability to meet the mobility requirements of the post including the travel to project partners as required by the role.

Desirable Criteria:

1. Strong background in software application development.
2. Experience of the application of AI algorithms and software in multidisciplinary activities.
3. Experience of developing and testing novel algorithms.

Major Duties:

1. Undertake research under supervision within the specific research project and, as a member of the collaborative research team, contribute to develop FinTech Compliance Management Tool (CMT) and apply knowledge of relevant domains along with expert coding skills to platform and framework development projects.
2. Develop/apply highly scalable algorithms for FinTech CMT based on state-of-the-art AI methodologies and design suitable user studies.
3. Carry out analyses, experiments, critical evaluation and interpretation of experimental data and the literature using methodologies and techniques appropriate to the area of research across a range of platforms and facilities of the wider PwC partnership.
4. Produce high quality research outputs consistent with project aims and commensurate with career stage. This will include collaborating and co-authoring with PI and project team (as appropriate) on outputs.
5. In consultation with the project team, promote research milestones and outputs at national and international conferences and through social media (where applicable).
6. Carry out occasional educational supervision, demonstrating or lecturing duties within the post holder’s area of expertise and under the direct guidance of a member of academic staff.
7. Undertake supplementary duties relevant to the success of the project including administrative duties, presentation of regular progress reports and additional training and development activities as required.

Essential Criteria:

1. 2.1 Honours Degree (or equivalent) in Computer Science, Applied Mathematics, Electronics, Electrical Engineering, or a closely related discipline.
2. Obtained or be about to obtain a PhD in Computer Science, Applied Mathematics, Electronics, or Electrical Engineering.
3. Relevant experience to include:
   
   • 3+ years experience in deep learning and NLP research and development 
   • working effectively as part of a research team in the development and promotion of a research topic.
4. Demonstrable knowledge of common deep learning models for NLP.
5. Demonstrable knowledge and experience of Python and (Java or C# or C++) programming languages.
6. Strong publication record, commensurate with stage of career.
7. Ability to contribute to broader management and administrative processes.
8. Contribute to the School’s outreach programme by links with industry, community groups etc.
9. Practical problem-solving skills, independence of thought and initiative.
10. Ability to assess and organise resources.
11. Ability to communicate complex information in English effectively in oral and written format to technical and non-technical audiences.
12. Ability to build relationships with a wide range of people and roles at different levels of seniority and to influence decision making.
13. Ability to manage self and prioritise workload.
14. A pro-active approach to work and team development.
15. Commitment to continuous professional development.
16. Ability to meet the mobility requirements of the post including the travel to project partners as required by the role.

Desirable Criteria:

1. Experience in regulatory compliance.
2. Experience in obligation extraction from regulation documents.
3. Strong background in software application development.
4. Experience of the application of AI algorithms and software in multidisciplinary activities.

Major Duties:

1. To be actively involved in the research programme as directed by the line manager/project supervisor and focus on developing an enterprise-level solution for risk assessment in complex business networks backed by strong research on the subject-matter.
2. Carry out research on synchronous and asynchronous automation for business networks and its assistive technologies and report on the findings in discussions with the project supervisor and any associated partners of the project.
3. Carry out routine administrative tasks associated with the research project/s to ensure that projects are completed on time.
4. Developing proof-of-concept wherever applicable to justify the research.
5. Carry out appropriate analysis and write up results of own work and lead a new direction as the project progresses.
6. Present regular progress reports on research to members of the research group or external audiences to disseminate and publicise research findings.
7. Use of road mapping/project development tools to share ongoing status updates.
8. Contribute to the production of research reports, publications, and proposals.
9. Any other duties that the programme supervisor may reasonably request.

Essential Criteria:

1. 2.1 Honours Degree (or equivalent) in Applied Mathematics, Computer Science, Electronics, Electrical Engineering, or a closely related discipline.
2. Normally have or be about to obtain a PhD in Computer Science, Applied Mathematics, Electronics, Electrical Engineering, Physics.
3. Relevant experience to include:
   • At least 3 years research experience in at least one of the following: risk assessment, AI-modelling, business networks. 
   • 1-3+ years of experience developing risk assessment systems.
   • Demonstrable experience of: - software development at the systems (preferably (but not limited to) C/C++/Python/Java/C#).
4. Experience of working effectively as part of a research team in the development and promotion of the research theme.
5. Strong publication record, commensurate with stage of career.
6. Ability to contribute to broader management and administrative processes.
7. Contribute to the School’s outreach programme by links with industry, community groups etc.
8. Practical problem-solving skills, independence of thought and initiative.
9. Ability to assess and organise resources.
10. Ability to communicate complex information in English effectively in oral and written format to technical and non-technical audiences.
11. Ability to build relationships with a wide range of people and roles at different levels of seniority and to influence decision making.
12. Ability to manage self and prioritise workload.
13. A pro-active approach to work and team development.
14. Commitment to continuous professional development.
15. Ability to meet the mobility requirements of the post including the travel to project partners as required by the role.

Desirable Criteria:

1. Strong background in risk assessment systems, synchronous and asynchronous process management.
2. Experience of the application of dashboards for multidisciplinary activities.
3. Experience of developing and testing novel algorithms.
4. Mathematical skills for conceptualisation, modelling, optimisation, and analysis of problems.

Major Duties:

1. Undertake research under supervision within the specific research project and as a member of the collaborative research team contribute to develop and evaluate NLP, Document AI, and deep learning models and apply knowledge of relevant research domains along with expert coding skills to platform and framework development projects.
2. Develop/apply scalable algorithms based on state-of-the-art machine learning methodologies and design and evaluate suitable models and workflows.
3. Carry out analyses and critical evaluation, in order to interpret, explain and further improve model performance and utility.
4. Engage with the relevant research literature, in order to develop methodologies appropriate to the area of research across a range of platforms and facilities of the wider PwC partnership.
5. Produce high quality research outputs consistent with project aims and commensurate with career stage. This will include collaborating and co-authoring with PI and project team (as appropriate) on outputs.
6. In consultation with the project team, promote research milestones and outputs at national and international conferences and through social media (where applicable).
7. Carry out occasional educational supervision, demonstrating or lecturing duties within the post holder’s area of expertise and under the direct guidance of a member of academic staff.
8. Undertake supplementary duties relevant to the success of the project including administrative duties, presentation of regular progress reports and additional training and development activities as required.
9. Carry out occasional educational supervision, demonstrating or lecturing duties within the post holder’s area of expertise and under the direct guidance of a member of academic staff.

Essential Criteria:

1. 2.1 Honours Degree (or equivalent) in Applied Mathematics, Computer Science, Electronics, Electrical Engineering, or a closely related discipline.
2. Obtained or about to obtain a PhD in Computer Science, Applied Mathematics, Electronics, Electrical Engineering or Physics.
3. At least 3 years research experience in at least one of: intelligent systems, artificial intelligence, natural language processing, data science.
4. Demonstrated experience developing deep learning and/or NLP systems.
5. Knowledge of Python, and experience with at least one other programming language (Java, R, C# or similar).
6. Knowledge of deep learning or NLP frameworks and tools (e.g. PyTorch, huggingface, spaCy).
7. Have a working knowledge of typical algorithms and concepts used in machine learning and deep learning, to include both model architectures and model evaluation.
8. Strong publication record, commensurate with stage of career.
9. Sufficient breadth and depth of specialist knowledge in the discipline and of research methods and techniques.
10. Ability to contribute to broader management and administrative processes.
11. Contribute to the School’s outreach programme by links with industry, community groups etc.
12. Practical problem solving skills, independence of thought and initiative.
13. Ability to assess and organise resources.
14. Ability to communicate complex information in English effectively in oral and written format to technical and non-technical audiences.
15. Ability to build relationships with a wide range of people and roles at different levels of seniority and to influence decision making.
16. Ability to manage self and prioritise workload.
17. A pro-active approach to work and team development.
18. Commitment to continuous professional development.
19. Ability to meet the mobility requirements of the post including the travel to project partners as required by the role.

Desirable Criteria:

1. Strong background in deep learning model development.
2. Experience of the application of AI algorithms and software in multidisciplinary activities.
3. Experience of developing and testing novel algorithms.
4. Experience in multi-modal AI.
5. Experience in ML deployment.
6. Experience in data science and statistical methods for the analysis of data.

Project Information:

This project aims to secure financial transactions and ensure their accuracy and verification against identifiable criteria of uncertainties. Assessing the current security standards and ensuring that these can tackle the requirements of auditing trials is another direction to explore. The research will advance our understanding of cyber security using quantum methods to determine the aspects of the unpredictability of financial transactions and offer more robust methods to safeguard against quantum vulnerabilities.

Recommended Literature:

[1] Verchyk D, Sepúlveda J. Towards Post-Quantum Enhanced Identity-Based Encryption. In2021 24th Euromicro Conference on Digital System Design (DSD) 2021 Sep 1 (pp. 502-509). IEEE.

[2] Schuld M, Sinayskiy I, Petruccione F. An introduction to quantum machine learning. Contemporary Physics. 2015 Apr 3;56(2):172-85.

[3] Fundira M, Edoun EI, Pradhan A. Adapting to the digital age: Investigating the frameworks for financial services in modern communities. Business Strategy & Development. 2024 Mar 1.

[4] Kong I, Janssen M, Bharosa N. Realizing quantum-safe information sharing: Implementation and adoption challenges and policy recommendations for quantum-safe transitions. Government Information Quarterly. 2024 Mar 1;41(1):101884.

[5] Selvaganesh R, Sriram KA, Venkatesh K, Teja KS. Secure data storage based on efficient auditing scheme. In Artificial Intelligence, Blockchain, Computing and Security Volume 1 2024 (pp. 964-968). CRC Press.

Project Information:

• Multimodal Transformers for document understanding: We will rigorously assess existing multimodal transformer architectures on document understanding tasks like information retrieval, question answering, and summarisation. This analysis will identify limitations and opportunities for improvement. We will then propose extensions to incorporate modality-aware attention mechanisms, temporal reasoning for sequential elements, and knowledge-guided fusion strategies to enhance the transformer's ability to integrate diverse information.

• New Deep Learning Architectures for document understanding: Based on the insights from study 1, we will explore deep learning paradigms beyond transformers to synergistically integrate their strengths with other architectures like convolutional neural networks for visual features or graph neural networks for relational content. We will design novel architectures that progressively build a unified representation of the document, extracting local features from individual modalities and then fusing them to capture overarching semantic relationships. We will develop mechanisms that explicitly model the interdependencies between modalities, guiding the model's focus on relevant information across different channels.

• Rigorous Evaluation and Applications: The proposed methods will be rigorously evaluated on benchmark datasets encompassing diverse document types and tasks. Metrics will include retrieval accuracy, question answering F1 scores, and summarisation quality. We will explore the practicality of the developed models in real-world scenarios, such as automatic document indexing, knowledge base construction from multimodal documents, and intelligent document search engines.

Project Information:

The main specific aim of this project is to deliver a transparent, fair, understandable and accountable digital worker driven by deep learning.

DNNs are increasingly used in place of traditionally engineered software in many areas. This project will explore the most promising approaches for the development of a digital worker to perform professional business support services that have a high degree of reliability. However, DNNs are complex non-linear functions with algorithmically generated (and not engineered) coefficients, and therefore are effectively “black boxes”. Hence, in this project, we are not simply developing DNNs for digital worker tasks but also carrying out quantitative evaluation against stringent metrics to evaluate the performance of specific tasks (to be identified as part of the project.)

The project will also seek to identify a rationale for why a digital worker DNN performs in a particular manner so that stakeholders may understand and appropriately trust the work done by a digital worker.

Once the broad stroke of the capabilities of a DNN based DW have been proved (against statistical benchmarks) the ethics of deploying such a technology will be explored so as to provide evidence for or against its use, or when or when not to use it, for example in providing a backup service when the equivalent human service may be unavailable.

The key technological objective is:

• To develop a DNN for performing a defined digital worker task using benchmark case studies.

The key ethical objectives are:

• To generate evidence for explaining the digital worker DNN and any result from the worker.
• To examine the effects of deployment of DW are full or partial replacements for human workers.

Project Information:

VR and AR demonstrate the potential to improve learning practices under certain conditions by introducing features such as haptics, enhanced visual information, and intelligent tutoring, that combine for a unique and dynamic pedagogical tool. With the advent of VR and AR, virtual systems could offer ubiquitous, effective, and affordable solutions to both learn and manage corporate systems and compliance processes. Virtual systems could one day be ubiquitous, and it will be important to observe the extent to which participants are learning for personalisation and optimisation within business process applications.

This project is required to build the necessary hardware and software to enable the collaborative and virtual presences required by systems such as those envisaged above, using a multi sensorial VR approach. The project also requires the development and examination of example scenarios. This requires the development of metrics against which to measure the effectiveness of such systems in the context of the example scenarios using techniques that provide quantitative assessment that are statistically rigorous.

There is a developing corpus of literature dedicated to EEG analysis techniques for human-machine systems, yet applications specifically for decision-making contexts are currently limited. Data is typically acquired directly from the human component of the system via the EEG headset, which is analysed. For example, the P300 signal is frequently used in medical science and neuroscience applications, but its potential for assisting in analysing learning and memory in a training, learning and decision-making context is promising. The relationship between the P300 and working memory suggests that the signal could be a valuable metric for detecting retention in virtual training systems, but there is little research in these applied environments evaluating the suitability for this purpose.

We propose using appropriate qualitative and quantitative analysis, including the analysis of appropriate user memory signals, to prove the VR, AR and AI can be combined to enhance the decision-making process within a business process.

Further Reading
Investigating the P300 Response as a Marker of Working Memory in Virtual Training Environments
Simpson, T. G. & Rafferty, K., 06 Apr 2021, (Early online date) In: IEEE Transactions on Human Machine Systems. 13
Evaluating the Effect of Reinforcement Haptics on Motor Learning and Cognitive Workload in Training
Simpson, T. G. & Rafferty, K., 31 Aug 2020, International Conference on Augmented Reality, Virtual Reality and Computer Graphics. Springer, p. 203-211 (Lecture Notes in Computer Science)(Lecture Notes in Computer Science).

Project Information:

In this project we will investigate a deep learning centred approach to deontic information extraction. We will design a deep learning architecture, drawing on the state of the art in deep learning and deontic information extraction, which can learn a deontic information model. The deontic information model can then extract deontic information from regulatory documents.

Objectives:

• Investigate deontic information modelling using deep learning
• Investigate deontic information extraction
• Conduct a case study using regulatory documents

Further Reading

1. Zhengrong Guo, Gongde Guo, Hui Wang (2022. Question-answer pair generation method based on key-phrase extraction and answer filtering. Under review.
2. Devlin, J., Chang, M. W., Lee, K., & Toutanova, K. (2018). Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805.
3. Elwany, E., Moore, D., & Oberoi, G. (2019). Bert goes to law school: Quantifying the competitive advantage of access to large legal corpora in contract understanding. arXiv preprint arXiv:1911.00473.
4. Hegel, A., Shah, M., Peaslee, G., Roof, B., & Elwany, E. (2021). The Law of Large Documents: Understanding the Structure of Legal Contracts Using Visual Cues. arXiv preprint arXiv:2107.08128.
5. Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., ... & Stoyanov, V. (2019). Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692.
6. Ormerod, M., Martínez-del-Rincón, J., Robertson, N., McGuinness, B., & Devereux, B. (2019, August). Analysing representations of memory impairment in a clinical notes classification model. In Proceedings of the 18th BioNLP Workshop and Shared Task (pp. 48-57).
7. Rogers, A., Kovaleva, O., & Rumshisky, A. (2020). A primer in bertology: What we know about how bert works. Transactions of the Association for Computational Linguistics, 8, 842-866.
8. Rosa, G. M., Rodrigues, R. C., de Alencar Lotufo, R., & Nogueira, R. (2021, June). To tune or not to tune? zero-shot models for legal case entailment. In Proceedings of the Eighteenth International Conference on Artificial Intelligence and Law (pp. 295-300).
9. Yoshioka, M., Aoki, Y., & Suzuki, Y. (2021, June). BERT-based ensemble methods with data augmentation for legal textual entailment in COLIEE statute law task. In Proceedings of the Eighteenth International Conference on Artificial Intelligence and Law (pp. 278-284).

In a broader sense, failure to identify risk can result in inputs remaining undiagnosed for a longer period, leading to significant and more pervasive deficiencies over time. Thus, it is desired to develop methods to comprehend these entanglements of data from multiple sources in synchronous or asynchronous business networks and to provide solutions to flag the potential risks. In addition, reducing false positives without compromising the accuracy of risk mapping is an additional challenge that must be addressed as part of this project.

Project Information:

Business networks with multiple data points are challenging to map and profile for risks. This project is anticipated to investigate executable business processes over a complex network containing various synchronous and asynchronous data sources. The project will expand the understanding of the risk landscape and generate valid methods that can replace the existing clause approaches with more comprehensive and novel AI-assisted methods that can dynamically adjust the criteria without affecting the system's accuracy. The project will explore business processes and networks from multiple directions – link analysis, relationships between data and suspicious workflow, security risks, to name a few. Such provisioning will ensure that business decisions are taken based on true data by considering potential risks associated with the processes.

The project aims to develop methods of defining checks and clauses with the assistance of computational game theory to reduce the false positives in the risk flagging system and introduce potential reasoning capabilities within the mappings for synchronous and asynchronous business networks that support coordinated decision-making.

Further Reading:

1. Yang W. Research on Risk Intelligent Assessment Method of IT Operation and Maintenance Based on Cloud Computing. In2022 IEEE 2nd International Conference on Power, Electronics and Computer Applications (ICPECA) 2022 Jan 21 (pp. 900-903). IEEE.
   
2. Stergiopoulos G, Dedousis P, Gritzalis D. Automatic analysis of attack graphs for risk mitigation and prioritization on large-scale and complex networks in Industry 4.0. International Journal of Information Security. 2022 Feb;21(1):37-59.
3. Vassilev V, Donchev D, Tonchev D. Risk assessment in transactions under threat as Partially Observable Markov Decision Process. InOptimization in Artificial Intelligence and Data Sciences 2022 (pp. 199-212). Springer, Cham.
   
4. Shareef AS, Rubasundram GA. Predicting Financial Statement Fraud using Artificial Neural Networks. In Artificial Intelligence and Big Data for Financial Risk Management 2022 Aug 31 (pp. 17-35).
   
5. Cheng D, Niu Z, Li J, Jiang C. Regulating systemic crises: Stemming the contagion risk in networked loans through deep graph learning. IEEE Transactions on Knowledge and Data Engineering. 2022 Mar 25.

Project Information:

In this project, the goal is to build on recent progress in deep learning and natural language processing to develop methods and systems for processing information in commercial document data and using the resultant representations to generate useful, task-relevant knowledge, for example, through text summarization, text generation, and question answering. The project will build on modern foundational models and architectures in machine learning, and in particular transformer models, such as BERT (Devlin et al 2019) and RoBERTa (Liu et al 2019). Particular problems that are to be tackled within the business services domain include defining Service-Level Agreements (SLAs), a stage in the finalisation of a contract between a service provider and a client. They are defined at different levels and used by organisations internally as well as in the supplier/customer relationship. All SLA use terminology and commonality of vocabulary that ensures the same quality of service across different units in an organisation as well as across multiple locations and subcontract work. Because of their ubiquity and importance in business services, it is costly, in terms of staff time, to ensure verification and compliance. 

A key challenge in this project is to develop models that process textual data in a way which is transparent, understandable and accountable. The large-scale transformer models that are now ubiquitous in Natural Language Processing research are essentially “black boxes”, where the representations of linguistic meaning and the basis for output decisions by the model are not readily explainable to an end user. Currently, therefore, there is intensive research on developing a wide range of statistical methods and other analysis techniques to better quantify and explain such models (Rogers et al 2020). 

In addition to the development of deep learning architectures and model explainability methods, an additional component of this project is to develop an assessment metric for the implemented intelligent systems and benchmark them against current practices. In particular, this will involve evaluating error rates and making recommendations for the depth of deployment of intelligent agents in practice.

References
Devlin, J., Chang, M. W., Lee, K., & Toutanova, K. (2018). Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805.
Elwany, E., Moore, D., & Oberoi, G. (2019). Bert goes to law school: Quantifying the competitive advantage of access to large legal corpora in contract understanding. arXiv preprint arXiv:1911.00473.
Hegel, A., Shah, M., Peaslee, G., Roof, B., & Elwany, E. (2021). The Law of Large Documents: Understanding the Structure of Legal Contracts Using Visual Cues. arXiv preprint arXiv:2107.08128.
Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., ... & Stoyanov, V. (2019). Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692.
Ormerod, M., Martínez-del-Rincón, J., Robertson, N., McGuinness, B., & Devereux, B. (2019, August). Analysing representations of memory impairment in a clinical notes classification model. In Proceedings of the 18th BioNLP Workshop and Shared Task (pp. 48-57).
Rogers, A., Kovaleva, O., & Rumshisky, A. (2020). A primer in bertology: What we know about how bert works. Transactions of the Association for Computational Linguistics, 8, 842-866.
Rosa, G. M., Rodrigues, R. C., de Alencar Lotufo, R., & Nogueira, R. (2021, June). To tune or not to tune? zero-shot models for legal case entailment. In Proceedings of the Eighteenth International Conference on Artificial Intelligence and Law (pp. 295-300).
Yoshioka, M., Aoki, Y., & Suzuki, Y. (2021, June). BERT-based ensemble methods with data augmentation for legal textual entailment in COLIEE statute law task. In Proceedings of the Eighteenth International Conference on Artificial Intelligence and Law (pp. 278-284).

Project Information:

In this PhD project, we will build on existing work on question-answer generation and natural language processing in the School to go a big step further – generating questions and their answers (e.g., in the form of question-answer pairs) from text automatically, and performing automatic testing based on user submitted documents. We will also evaluate through a case study in an industrial context. Generating high-quality question-answer pairs is a key part of this project and provides the basis for automatic testing. In general question-answer pair generation will improve the performance of question answering systems, help acquire knowledge from text, and promote machine reading comprehension.

Objectives:

• Investigate automatic question-answer pair generation based on deep learning and knowledge graph
• Investigate automatic testing based on question-answer pairs and user submitted documents
• Conduct a case study in an industrial context.

Further Reading:

1. Zhengrong Guo, Gongde Guo, Hui Wang (2022. Question-answer pair generation method based on key-phrase extraction and answer filtering. Under review.
   
2. Devlin, J., Chang, M. W., Lee, K., & Toutanova, K. (2018). Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805.
   
3. Elwany, E., Moore, D., & Oberoi, G. (2019). Bert goes to law school: Quantifying the competitive advantage of access to large legal corpora in contract understanding. arXiv preprint arXiv:1911.00473.
   
4. Hegel, A., Shah, M., Peaslee, G., Roof, B., & Elwany, E. (2021). The Law of Large Documents: Understanding the Structure of Legal Contracts Using Visual Cues. arXiv preprint arXiv:2107.08128.
   
5. Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., ... & Stoyanov, V. (2019). Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692.
   
6. Ormerod, M., Martínez-del-Rincón, J., Robertson, N., McGuinness, B., & Devereux, B. (2019, August). Analysing representations of memory impairment in a clinical notes classification model. In Proceedings of the 18th BioNLP Workshop and Shared Task (pp. 48-57).
   
7. Rogers, A., Kovaleva, O., & Rumshisky, A. (2020). A primer in bertology: What we know about how bert works. Transactions of the Association for Computational Linguistics, 8, 842-866.
   
8. Rosa, G. M., Rodrigues, R. C., de Alencar Lotufo, R., & Nogueira, R. (2021, June). To tune or not to tune? zero-shot models for legal case entailment. In Proceedings of the Eighteenth International Conference on Artificial Intelligence and Law (pp. 295-300).
   
9. Yoshioka, M., Aoki, Y., & Suzuki, Y. (2021, June). BERT-based ensemble methods with data augmentation for legal textual entailment in COLIEE statute law task. In Proceedings of the Eighteenth International Conference on Artificial Intelligence and Law (pp. 278-284).