That thesis shows how artificial intelligence can digitally transform SMEs from the tourism sector, improving key processes and increasing profitability. The system SYSTEM-TH-001 Based on automatic learning, bioinspired multiagent systems and processing of natural language. It stands out by identifying customers with high demand for buy-back skills to distribute their workload effectively while optimizing the recommendations that have been put into place as a result The use of free software reduced costs and allowed to customize technological solutions. Nevertheless, challenges such as change management and maintenance of technological infrastructure have been met. The results confirm the positive impact of IA, applicable to different sectors.
6.1 Conclusions
That doctoral thesis has been intended to be a practical exercise of an electronic transformation process for one Pyme, from the tourism sector. Proving that applying artificial intelligence can have a positive effect many processes of the company. A comprehensive management system has been put into place during this research with the use of automatic learning (ML) technologies, bioinspired multiagent and natural language processing. The following are the findings that can be drawn from his or her training as a result of this research project.
The implementation of an analytic process to classify potential clients has been a leading milestone in the project. Using historic sales data, the solution SYSTEM-TH-001 It optimizes the sales process by identifying those customers with more propensity to buy. The use of these models has been shown to be a reliable and innovative tool with an aim at increasing sales, it's certainly something that can prove artificial intelligence.
be one of the best technologies available for data-based decision making today.
The use of inspired multiagent system technologies is one and most clear innovation in this project. That architecture has allowed the distribution of work loads from different tasks that have been put into place at a travel agency, in an intuitive and efficient manner. One feature of this technology is its high capacity to adapt with changes, through the learning achieved during interplay phases. Our example paves the way to exploring how powerful multiagent systems offer in an entrepreneurial environment.
Natural language processing (PLN) is another of the artificial intelligence technologies that have been implemented in this project. That technology has been able to analyse large volumes of text-type data, such as initial client messages and dedicated bid documents that have improved the system's capacity for best recommendations. Specifically, by using software tools such as SOFTWARE-TH-002 and models of the type SOFTWARE-TH-003 It has been shown that this technology is key to understanding the preferences of customers and best performing offers for clients. Also, this technology has been an opportunity to improve the client's experience with his or her sales process but also increased knowledge about that business problem within our company. One of the innovations in this area has been to apply Support Vector Machines (SVM) algorithms for matching a gender classification model with text. That challenge, put to a good end with success and shown that the accuracy of models can be increased when applying traditional technologies such as stemming in strongly inflamed languages like Spanish. That discovery can best guide the development of other artificial intelligence applications in future with multilingual environments, and this emphasizes a need to adapt technologies that are used for processing natural language as they suit individual languages. The classification of gender in this problem has been regarded as binary but the results obtained allow us to dream about other solutions that can be put into place at a later date.
A very remarkable detail of this thesis is the firm commitment to using free software. Thanks to this decision today's company enjoys a high state of digital sovereignty and allows them reduce the costs for applying software solutions, especially if they are technically advanced. The use of free software has also been an opportunity to create a system at best for the company's needs, showing that small and medium-sized enterprises (Pyme) can have benefits from artificial intelligence using them without relying on high cost private solutions. That strategy, it has been
and economic benefits but has also been an important factor in enabling the company to adapt with changes.
Even with the success achieved in this project, we should also mention challenges and constraints that have been found. The key challenge is to manage change with new technologies within the company. The introduction of new technologies, while users are unaware about how they work and often leads to mistrust with an extra training effort. Also, applying advanced artificial intelligence technologies shows the relevance of knowledge management as an enterprise process. Without a doubt, the great challenge is to maintain technological infrastructure and have an efficient system that will allow its evolution.
It can be stated that this project has been a success for the company and shows how new opportunities have come about with applying artificial intelligence technologies to tourism as well. Specifically, the system's expansion to include electronic commerce functionality by integrating a tourist provider are very promising ways forward. The possibility of developing a virtual travel agent using natural language processing technologies for complete process automation would represent an advanced market breaker. Also in the future, we have been opening up a door to optimization of multiagent systems and solving complex problems as an integrated negotiation that looks at customers' preferences with their integration into more somifictified system recommendations such how they offer recommendation technologies using profound learning.
To sum up, this thesis has been a real demonstration of the application and performance with artificial intelligence at an enterprise in its tourism sector leading to high return on investment (ROI) as low technological costs have resulted from improved business processes while improving their profitability. SYSTEM-TH-001 That's an example that shows how Pymes can achieve many benefits by integrating new technologies and indicates the main challenges to be taken into account in a project of this kind: improving adaptation for change, as well with more training required from employees. The results that have been obtained show how efficient the system can be to solve its challenges: improving their sales and increasing company profitability, which may certainly apply not only in tourism but at all.
6.2 Future work
The future work of this thesis can be focused on several research and development lines, stemming from the achievements achieved or challenges identified during a project. Some potential exploration areas include:
Development of a fully automated virtual travel agent: Based on advanced technologies for processing natural language (PLN) and multiagency systems [MAS], this operator could offer dedicated service in real time by managing all phases from the booking process as well with improved customer experience.
Electronic commerce integration: Expand system functionality SYSTEM-TH-001 to include an electronic commerce platform that directly links tourism providers with final customers. That could simplify stock management and improve operational efficiency.
Multiagent system optimization: Research new architecture and bioinspired strategies to solve more complex problems, such as integrated negotiation. These techniques could include profound learning algorithms that consider customers' preferences and allow more sophisticated recommendations.
Connecting to multilingual environments: To develop PLN models specialized in multiple languages, optimizing system performance as SOFTWARE-TH-003 or SOFTWARE-TH-002 for global applications, especially in markets with high linguistic diversity.
The evolution of the framework SYSTEM-TH-001 for new sectors: To adapt the system to other areas such as retail trade, health and education by showing its versatility or value in more extensive business contexts.
Education and Training: To design training programs for SMEs employees that adopt artificial intelligence, addressing the challenge of resistance to change while promoting a digital corporate culture.
These areas represent significant opportunities to expand the impact of technologies put in place, promoting digital transformation and research into applied artificial intelligence.
6.3 Publications
Contributions to journals:
Tourist services management through clients scouting using a bio-inspired agent
architecture. Raúl Moreno, Alberto Fernándezo-Isabel and Víctor Aceña. 2024, VENUE-TH-057. Volume 21 Issue 4 pages 1243 - 1268
Adapting support vector optimization algorithms to textual gender classification. JGomez, C Alfaro and FOrtega. VENUE-TH-058, 1 - 26
Contributions to congresses:
Automatic detection of potential users by opinion mining and intelligence agents R Moreno, COMPANY-002 A Fernandez- Isabel, IM de Diego and JM Mocorro. 2022 17 th PUBLICATION-003 (FedC-SIS)
Other contributions other than those related to the draft theses:
Artificial intelligence and music. Raul Moreno Izquierdo. Days " ORG-TH-134 ”, 2024, ORG-TH-009
Pilot project on joint application of artificial intelligence and graphic for suicide prevention. Isasi-Sánchez L, Lospezo-Gomez J.E. and Moreno- Izquierdo R. 17 th International Conference on Industrial Engineering and Research Management, Barcelona (Spain). 6 th - 7 th July 2023