One of the advantages of bringing a feminist perspective to computational text analysis is that it forces text analysis to expand its scope. A feminist perspective does not merely bring into the field a different approach, among the many that exist, but it transforms the foundations of computational text analysis. In this sense, if you do computational text analysis from a feminist perspective your approach will encompass an ontological, epistemological, methodological, and empirical standpoint. What are the questions that feminism will ask using text analysis? Issues of power, inequality, intersectionality, marginalized groups, oppression, or discrimination. The questions and the topics are not neutral or objective or universal as more mainstream computational text analysis will imply but are all questions that are asked from a particular standpoint.  

A key component of my feminism is negotiation of universals and particulars, from multiple stances. I see the stances as reducible to two categories: experiencer and observer of/interactor with others. To the extent that computational text analysis opens up new ways of seeing oneself and others, I see it as potentially feminism-compatible.
We’ve learned that computational analysis is not neutral–in all phases, from the formulation of research questions, through data selection and preparation, conceptualization, operationalization, and analysis, human researchers bring preconceptions into the process. A feminist text analysis involves self-awareness, transparency, and responsibility for recognizing, naming, and if possible, correcting for, biases.
As such, I think it’s not only possible but imperative. In our algorithmic digital world, when computational analysis is used so pervasively in ways that re-inscribe and reinforce harmful stereotypes, it’s crucial to understand the emergent methods and to incorporate egalitarian concerns and values into research that uses these methods, and into critical responses.
I wonder also if (feminist) text analysis could be a stimulus to alternative viewing, a kaleidoscope-like supplement to close reading, rather than a telescope. DuBois’s chapter openings in The Souls of Black Folk are one of my favorite examples of showing differently:

The moment you define the parameters for a feminist text analysis, they will be outdated. Feminist changes. Text changes. And analysis changes. Identities are not stable; rather they are constantly being iterated and changed by the factors around and outside them. An analysis that seeks to make sense of patterns will capture truth in one snapshot and falsity in another. 

But that doesn’t mean it’s not worth trying. 

I have been thinking through Ralph Ellison’s line: “it was either to live with music or die with noise.” Text analysis aims to pull music from noise; feminist text analysis realizes that whether or not one “finds music” depends on the voice of the musician and the ears of the listener. And that a definition of music will never bring as much fodder for mind and spirit as will an experience of music. 

There need not be a definitive arrival point to signify Feminist Text Analysis. In fact, such a point would bely the iterative nature of feminism, text, and analysis. But projects that acknowledge and thoughtfully codify instability, multiplicity, and other feminist ideals bring us closer to feminist text analysis.