IEEE Technology and Society Magazine - March 2021 - 89

On the one hand, we have that, while the measurement is critical, overmeasuring can be counterproductive [16]. Furthermore, in the case of SIAMES,
and the case of impact measurement, in general,
there are two fundamental aspects to take into
account. On the one hand, the intangibles and, on
the other hand, the measurement of the impact with
the currency as a unit.
One of the methodologies currently used to measure and manage social impact is the SROI, which,
based on traditional cost-benefit analysis, quantifies
social value through indicators associated with the
results achieved. SROI is a process of understanding,
measurement, management, and communication
of the social, environmental, and economic values
generated by an organization [7]. Its purpose is to
examine, qualitatively and quantitatively, the project's process of generating social value. The use
of real measurement reports means that SIAMES
will use, in most cases, reports that use the SROI as
a method to measure impact, because of its strong
implementation.
Although the SROI has positive aspects, the use
of financial proxies can have unintended consequences such as projects that really have a high
impact, but that this refers to intangibles or that
their monetization is low, are relegated to secondary
positions when they should not. In fact, it is already
known, in other areas, that AI tends to increase
human biases instead of correcting them if they are
not taken into account [17].
It seems clear that human participation is capable of obtaining a global vision of the situation
while automated systems lose nuances and aspects
that can be key when assessing the real intention
of a project. Therefore, we must take special care
with these aspects when we propose the elimination of human interventions in the processes we
carry out.
It is essential to have these reflections in a transversal way throughout the development of the
research and try to mitigate their negative effects or,
if this is not possible, clearly expose the deficiencies
that our system presents.

Related work
In the following sections, we try to encompass
multidisciplinary research that includes both social
impact measurement and semantic text mining
and CBR.

March 2021

Social impact
If we look at the research carried out so far that
combines AI and social impact measurement, we
can see that the studies focus either on the ethical
use of AI [18] and its impact [19] or in measuring the
generally negative social impact of different actions
or phenomena not related to the implementation of
social projects, such as in [20]. However, the use of
AI to back social enterprises in measuring positive
social impact is an area still to be explored.
On the whole, the ambiguity of methodologies
currently used to measure social impact generates
problems. There is a trend toward the rationalization
of social projects in both public and private funders
[21] and the requirements on entities undertaking
social projects to formally measure social impact,
thereby providing objective values that can be standardized and verified, have increased. Nevertheless,
although there are countless methods to measure
social impact, the lack of conventions on the indicators to be used makes calculating social value a very
complex and expensive task [5].
The measurement of social impact is a widely
questioned practice, mainly on account of the ambiguity that methodologies for measuring impact suffer from. A frequent object of criticism is the causal
ambiguity of the logic model through which the
activities of the entity are translated into results,
changes, and impacts for each of the stakeholders
[3], [22]. Practitioners with experience in implementing such formal methodologies contend that
impacts are difficult to understand with precision,
far more difficult to calculate [23], [24].
One of the methodologies currently used to measure and manage social impact is the SROI method,
or SROI, which, based on traditional cost-benefit
analysis, quantifies social value through indicators
associated with the results achieved. The mapping
of impact involves defining impact indicators entailing some careful discernment; the choice of indicators is bolstered by a theory of change that holds
assumptions and preferences for how impact can be
achieved. Therefore, identifying indicators is based
on choice and there can be diverse perspectives on
preferences within an organization, with different
stakeholders emphasizing the importance of different activities of different groups of beneficiaries.
Hence, formal methodologies for measuring social
impact that generate objective metrics with an empirical basis are required with the aim of improving the

IEEE Technology and Society Magazine - March 2021

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - March 2021