MATILDA: Inclusive Data Science Pipelines Design through Computational Creativity

Abstract

This paper argues for developing innovative data science frameworks that render the latest progressions in data engineering and artificial intelligence accessible to non-technical users across diverse fields. Such frameworks would empower these users to leverage advanced data science solutions' capabilities fully. We propose a methodology that merges computational creativity with conversational computing to facilitate an intuitive pathway for non-experts to navigate and derive insights from datasets. We present MATILDA, a platform rooted in creativity-driven data science, and demonstrate its utility in augmenting the data science pipeline's design process through the synergy of human innovation and algorithmic ingenuity.

Keywords

• eol>Data science pipelines
• graph analytics
• knowledge graphs
• computational creativity

1. Introduction

Published in the Proceedings of the Workshops of the EDBT/ICDT 2024 Joint Conference (March 25-28, 2024), Paestum, Italy. * Genoveva Vargas-Solar. † The authors’ list is alphabetical except for the first author. $ [e-mail] (G. Vargas-Solar); [e-mail] (K. Belhajjame); 1Creativity is a process that can combine familiar ideas in new ways, [e-mail] (. J. A. Espinosa-Oviedo); explore the potential within existing conceptual spaces, or [e-mail] (S. Negrete-Yankelevich); form these spaces to allow for previously inconceivable ideas. [e-mail] (J. Zechinelli-Martini) ativity is not considered novelty but the capacity to generate surpris© 2024 Copyright © 2024 for this paper by its authors. Use permitted under Creative Commons ing and valuable ideas that push beyond conventional boundaries CPWrEooUrckReshdoinpgs IhStpN:/c1e6u1r3-w-0s.o7r3g LCicEenUseRAttWribuotironk4s.0hIontpernPatrioonacl e(CeCdBiYn4g.0)s. (CEUR-WS.org) [ 1 ]. proaches contributing to model creative-based processes and friendly design data science pipelines. It discusses how both areas can provide novel ways of designing data science-driven solutions. Section 3 after that describes the challenges of modelling creative-driven data science design processes. It gives the general lines of an approach that can enhance and envision a new way of addressing analytics problems using data and artificial intelligence models. Section 4 introduces a creativity-based data science design platform. It describes the general architecture and functions and shows how it can support the design process of data science pipelines guided by human and computational creativity. Finally, Section 5 concludes the paper and discusses future work.

2. Related work

Addressing the creative design of data science pipelines to make them inclusive for non-experts requires drawing from methods and results from three areas: creative models, friendly data science and provenance. This section gives an overview of the relevant results that provide a scientific background to the project and help to contextualise our objectives.

Creativity-driven systems Artificial intelligence (AI) ofers opportunities to reify and transform how we think about human cognitive capabilities. In this context, computational creativity (CC)2, aims at studying human creativity and building systems that perform in such a way as to be considered creative. Concerning creativity, whether this cognitive capability is individual or collective. CC has moved from the classic individualistic and cognitive model of creativity [ 1 ] to a social and collective creativity model [ 2, 3, 4, 5, 6, 7 ].

Collective-creativity models are concerned with understanding the roles and tasks that diferent agents (both human and non-human) play in a process that requires creativity and how creativity can be measured in this context [ 8, 9 ]. Co-creativity is an excellent approach to establishing eficient, context-aware interactive systems and setting long-term programmes where solutions to problems requiring human and machine collaboration can be studied.

CC has been widely applied in art with systems that promote "artificial" creation. For example, Disco Difusion is an online tool that runs on Google Collab to execute Python programs that, using a learning model, result in "creative" artwork. Language models such as GPT-3 2Computational creativity is the study of building software that exhibits behaviour that would be deemed creative in humans. Such creative software can be used for autonomous creative tasks, such as inventing mathematical theories, writing poems, painting pictures, and composing music (https://computationalcreativity.net are capable of interpreting and generating text. With over 540 billion parameters, the Google Pathways model can explain jokes, follow a chain of reasoning, recognise patterns, perform Q&A sessions on scientific knowledge, and summarise texts. As more parameters are added to language models, the depth of "understanding" they can demonstrate expands. "The painting fool" is a system developed by Simon Colton that draws portraits taking into account emotional information obtained from the subjects being painted through a camera [ 10 ]. NegreteYankelevich and Morales-Zaragoza [ 4 ] propose a model to develop and assess creativity in computational agents embedded in mixed teams. The Apprentice Framework model establishes a series of roles (or levels of responsibility) agents can play within the group over time with the possibility of ascent through the ladder as the system is developed, acquiring thus more responsibility in the creative process. The model also helps identify aspects of the product being produced, at which point the agent is supposed to be creative. By keeping track of both responsibilities and aspects, it is possible to plan and assess the development of the system. Negrete-Yankelevich and Morales-Zaragoza [ 11 ] propose a framework to create animatics. These animated storyboards constitute an essential artefact in producing animations by a team of expert animators that made an award-winning series of one-minute shorts for Mexican TV called Imaginantes (Televisa, "Imaginantes* - YouTube."). The system’s creativity is measured by how well the overall creativity of the team is afected by the system’s performance.

Developing friendly data science solutions.

Friendly data science systems must provide intuitive and interactive access to data processing operations in an agile and visual step-by-step manner [ 12 ]. They should help a user to derive conclusions about the data collection content and identify the potential questions that data can help answer. Through conversational loops and feedback, a friendly exploration and analysis system must calibrate the tasks according to the data’s characteristics and the user’s expertise and expectations. Through metadata collection and user profiling, an exploration and analysis conversation loop should propose actions, insight, and results’ display (and visualisations) that assist the user in completing a given goal.

Discussion. We believe the collective CC can reproduce the collaborative transdisciplinary conditions in which data science solutions are developed. It can drive the proposal of data science solutions combining data, algorithms, and computing resources to model complex systems and contribute to answering research questions to understand and predict them. While computational creativity and conversational techniques have proven efective, they have not been explored with the design policies can impact the quality of life of diferent categories of exploratory data analysis pipelines. Moreover, the of citizens willing to evolve in a given urban area?. Detwo approaches are somewhat opposed because conver- cision makers now call for data scientists’ creativity to sational techniques tend to rely on known territories (i.e. provide studies and mathematical evidence of the kind previously explored data manipulation and analysis ac- of urban changes to be considered in public policies. tions). In contrast, computational creativity allows for exploring unknown territories (data manipulation and Designing a data science pipeline. Datasets and analysis), which may, in some cases, prove more efective. research questions drive the design of DS pipelines.

Our work addresses two challenges. On the one hand, Through a simplified creative scenario using the main adapt and leverage both techniques to design an eficient phases of a DS pipeline: (1) collect or search for datasets and exploratory data analysis pipeline. On the other that can be used for answering a research question, and hand, strike the right balance when creating data analysis then (2) prepare them (explore, clean, engineer) to feed pipelines between ’known’ prior data exploration and one or several Artificial Intelligence (AI) models. (3) analysis actions and ’unknown’ creative actions. Models are trained and tested with dataset fragments. These tasks are calibrated recurrently until specific per3. Creative process for designing formance scores are reached. (4) Results are constantly assessed and eventually considered good enough to be data science driven solutions interpreted by experts, and conclusions are drawn on answering the initial research question to some extent.

Data science pipelines combining machine learning and Sketching a creative process, the elements to consider deep learning are the new query types with specific needs are: What data is needed to answer the research quesregarding how data must be structured and managed. tion and develop a strategy to collect them? How do we The “one all-fits-all” data structure and associated man- transform the initial research question into a quantitaagement functions approach are no longer adapted for tive statement that can be addressed by mathematical or data science queries. Indeed, every query has a specific AI models? Which model families can be pertinent for objective (modelling, prediction), and its design entirely answering the question? How do you design a series of depends on the input dataset and an initial research ques- tasks where data are processed? How do you connect tion (RQ). The data science query is not based on an the results format with the research question statement? explicit knowledge of the data. It includes tasks devoted How do we determine whether results converge? How to mathematically understanding the data; then, the par- do you decide whether results are fair enough for contial results of those tasks determine the design of other sidering an answer? studies devoted to the computation of a model repre- For example, data scientists can film civilians in the senting some hidden knowledge. Given statistical and target urban spaces to collect their behavioural patterns machine learning methods and a target objective, data on how they occupy and evolve along those spaces bescientists rely on libraries that provide methods that they fore and after implementing public spaces. Extract becombine to define a data science pipeline. The results havioural patterns that imply designing a DS pipeline for obtained by this pipeline are never definite, and they are processing videos and detecting civilians, for example, always, to some degree, close to the target. using perceptrons [ 13 ] and behaviour patterns within a

To illustrate the design process of a DS pipeline, con- series of scenes. The patterns can then be classified acsider the following scenario. Consider a trendy decision- cording to properties that detect changes before and after making group willing to adopt a data-driven approach implementing some change. Other possibilities would for designing public policies to enhance citizens’ lives be to run other data collection techniques like questionin urban spaces and reduce energy and economic costs. naires to describe urban civilians’ behaviour through Public policies are intended to modify built environments quantitative variables that can be correlated for detecting to improve them from financial and well-being perspec- changes produced after applying public policies. The postives. Decision-makers know that from the urbanism sibilities are numerous, and they rely on data scientists’ perspective, small changes in the built environment can expertise, on the facilities or not for collecting certain alter how people use the space. For instance, increasing types of data (e.g., video vs questionnaires) and their pedestrian areas in a city downtown close to restaurant knowledge of specific AI models’ families. zones reduces CO2 footprint. Still, it impacts the influx of restaurant customers in the area and lowers real estate Discussion Data Science and Machine Learning Enviprices. Customers can suddenly start preferring restau- ronments provide all the necessary AI models. They are rants close to parking slots. People living in the area can supported by enactment stacks that deal with the storhave problems accessing it and park their cars close to age, fragmentation, indexing and distribution of the data home. The research question is to which extent public required and produced by the tasks composing a pipeline.

What are the rules and strategies to combine diferent components that can transform input data into models and predictions that provide quantitative elements to answer initial research questions?

Generative artificial intelligence 3 has started to be consolidated into solutions that give the illusion of creation through interactive and conversational approaches 4. Systems like chat-GPT, in its various versions, mimic conversational and question-answering experiences intended to perform target tasks or produce “new" content based on existing evidence. The principle of this system is synthesising the creation process as an exercise of wrapping together “content" with specific characteristics and considering some constraints to produce artefacts that look, to some extent, novel.

In the case of DS pipelines, the first challenge is to model the creation process behind them. How does someone (a domain expert) state a research question so that a data-driven quantitative study can be run? How are data collected and selected to answer such questions? Which comes first, data or questions? How do we conclude that given datasets representing observations of an object of study are representative enough to produce a model or predict the behaviour of that object? How is the human integrated into the loop and intervene in the design milestones of a DS pipeline?

Challenges and Open Issues. A computational

creativity-based methodology for designing DS pipelines should consider at least the following scientific challenges and associated open problems: the system and the type of feedback to be given by humans.

Intervene the process with an agent by selecting a relevant subprocess where creativity would contribute significantly to the overall solution and assess how it works. Then, try other similar subprocesses and verify again. This bottom-up approach would establish a practice to turn the overall process into a friendly one in a stepwise fashion. • Collecting provenance and data from DS pipelines design tasks: implement processes for data curation, annotation, identification, and quality control in research. • Proposing an ad-hoc computational creativity tool for making DS science pipelines designfriendly for non-data scientists.

4. Towards a Human in the loop creative platform for designing data science pipelines

Figure 1 shows the general architecture of the MATILDA platform that assists people with diferent expertise to follow a creative process for designing DS pipelines given datasets and target research questions. The platform relies on a step-by-step conversational approach based on our previous work [ 12 ] and provides interaction entry points to allow humans feedback, validate and guide the creative process. For each phase of a DS pipeline (data exploration and preparation, fragmentation, training, testing and assessing), the platform suggests possible scenarios that are adopted or not. Therefore the platform • Modelling hybrid (human and nonhuman) relies on a knowledge base representing data science creativity-driven data science pipelines’ design: pipelines, with research questions and data features modpropose a computational creativity model to rep- elled that can be used to propose solutions similar as case resent end-to-end pipeline design. The creativity based reasoning approaches. model can integrate design patterns like the ones presented in [ 14 ] (design, mutant shopping, chorus line, simulation and approximating feedback, entertaining evaluations and no blank canvas).

Depending on the tasks to be designed within a DS pipeline, diferent creativity patterns can best be adapted to address the task. • Define the interaction among humans designing a DS pipeline and artificial system(s) that can take on tasks and propose results. Model the input/output required to feed and expect to/from

1. Data search: given keywords about the topic or

a sample of data to be analysed, the platform relies on queries as answers and exploration techniques to propose related data sets. The platform shows the possible questions associated with data through "queries as answers" techniques. Through an interactive process, a data scientist can converge to a sample of data representative of the type of questions she/he wishes to express (e.g., factual, modelling, prediction, etc.). 2. Designing data exploration and cleaning pipeline: given a dataset, the platform performs a quantitative analysis of the attributes, their dependencies and their values’ distribution. The platform also suggests cleaning and data engineering strategies, allowing data to have specific mathematical properties. The platform gathers information about 3According to the Bing chat-GPT and validated by this paper’s authors: Generative AI refers to a category of artificial intelligence (AI) algorithms that generate new outputs based on the data they have been trained on (www.weforum.org/agenda/2023/02/ generative-ai-explain-algorithms-work/). 4Microsoft Deepspeed https://github.com/microsoft/DeepSpeed/ tree/master/blogs/deepspeed-chat

Our DS creativity platform allows us to study how over

all creativity is afected if computer systems take over diferent roles within the design of data science pipelines.

The platform provides a collaborative environment that integrates an artificial actor within in the creative production process of DS pipelines by data scientists.

their decisions by interacting with the data scien- of datasets and technology for transforming any phetists. This information can be used to keep track nomenon produced in reality into digital data and the of the design process. For now, this is a very variety of algorithms (Mathematical and artificial intelquantitative perspective of the creation process, ligence models), the design of data science solutions reeven if, for future work, we will try to approach mains artisanal. The impact on person-hours and ecocreativity with other perspectives. nomic investment is not anecdotic. The time has come 3. DS pipeline creation: the current platform does to propose methodologies that can formalise the design not rely on existing AI model recommendation of data science solutions and model the “know-how” desystems but on knowledge about the questions veloped by data scientists during the creation process. previously addressed with AI models; it pro- Besides, data science addresses trans-disciplinary chalposes building blocks that can be combined into lenges. It is critical to bridge the gap between technical pipelines. These building blocks could be used to vocabulary, tasks, and the vocabulary of other disciplines answer the questions produced in 1). The build- and users with diferent expertise. This strategy will ing blocks include suggestions on the scores that ensure the usability and acceptability of solutions (i.e. can be used for assessing and calibrating train- pipelines). In summary, data science must become ining phases. The platform is also shared for every clusive and accessible to all. Our work addresses this building block with similar solution contexts in challenge by aiming to adopt computational creativity which they have been used. methods to model the data science design process(es) that combine human and nonhuman creativity.

The platform MATILDA proposed in this paper is based on the original methodologies that we propose. It contributes to creating data science pipelines according to the expectations of knowledge discovery. It is interesting for answering target research questions, the input data’s characteristics and the data scientists’ models. Creativitybased methodologies applied to data science will make it accessible and inclusive to address increasingly complex problems humanity faces.

5. Conclusions and Future Work

The research and development market associated with data science is fuelling the economies of countries in 6. Acknowledgements the world. Almost all sectors in the global economies see data science as a promising alternative to develop The work reported in this paper is performed in the conoriginal solutions to critical societal problems and pro- text of the project FRIENDLY 5 funded by the inter-group mote data-driven decision-making processes that can program of the laboratory LIRIS, Lyon. create know-how and value. Yet, despite the availability

References

[1] M. A. Boden , et al., The creative mind: Myths and mechanisms , Psychology Press, 2004 .

[2] M. L. Maher , Evaluating creativity in humans, computers, and collectively intelligent systems , in: Proceedings of the 1st DESIRE Network Conference on Creativity and Innovation in Design , 2010 , pp. 22 - 28 .

[3] M. L. Maher , Computational and collective creativity: Who's being creative? , in: ICCC, 2012 , pp. 67 - 71 .

[4] S. Negrete-Yankelevich , N. Morales-Zaragoza, The apprentice framework: planning and assessing creativity , Proceedings of the Fifth International Conference on Computational Creativity , 2014 .

[5] R. K. Sawyer , Explaining creativity: The science of human innovation , Oxford university press, 2011 .

[6] A. K. Goel , A. G. de Silva Garza , Special issue on artificial intelligence in design , Journal of Computing and Information Science in Engineering 10 ( 2010 ).

[7] N. Gu , P. Amini Behbahani , A critical review of computational creativity in built environment design , Buildings 11 ( 2021 ) 29 .

[8] A. A. Kantosalo , J. M. Toivanen , H. T. T. Toivonen, Interaction evaluation for human-computer co-creativity: A case study , in: Proceedings of the sixth international conference on computational creativity , Brigham Young University, 2015 .

[9] A. Kantosalo , S. Riihiaho , Experience evaluations for human-computer co-creative processesplanning and conducting an evaluation in practice , Connection Science 31 ( 2019 ) 60 - 81 .

[10] S. Colton , J. W. Charnley , A. Pease , Computational creativity theory: The face and idea descriptive models ., in: ICCC, Mexico City , 2011 , pp. 90 - 95 .

[11] S. Negrete-Yankelevich , N. Morales-Zaragoza, emotion: a system for the development of creative animatics , Proceedings of the Fourth International Conference on Computational Creativity , 2013 .

[12] P. Bethaz , K. Belhajjame , G. Vargas-Solar , T. Cerquitelli, Ds4all: All you need for democratizing data exploration and analysis , in: 2021 IEEE International Conference on Big Data (Big Data) , IEEE, 2021 , pp. 4235 - 4242 .

[13] E. Cruz-Esquivel , Z. J. Guzman-Zavaleta , An examination on autoencoder designs for anomaly detection in video surveillance , IEEE Access 10 ( 2022 ) 6208 - 6217 .

[14] P. Glines , I. Grifith , P. M. Bodily , Software design patterns of computational creativity: A systematic mapping study ., in: ICCC , 2021 , pp. 218 - 221 .