When the Holy Qur’ān was revealed to Prophet Muhammad PBUH, Quraish and the surrounding areas were people who spoke Arabic fluently. However, when the Arabs first heard the Holy Qur’ān, they were astonished by its eloquence. They had never heard such a wonderful speech in their lives. Their sentiments convinced them that such a beautiful and impressive speech could only be a divine diction, not a human creation.
Nowadays, most non-Arabic speakers find it difficult to understand the metaphoric expressions in the Holy Qur’ān. Instead, they do a long process of comparing various exegeses such as Tafsir ibn Kathir, Tafsir al-Qurtubi, Tafsir al-Tabari, etc. which in turn is a cumbersome process for non-specialists.
What it does
Basically, the app displays the Arabic Āyah of the Qur’ān, its pronunciation in English, and the translated text of the meaning of the Qur’ān. So that non-Arab users can read the Qur’ān easily and comprehend its meaning. During this process, the metaphor detector updates the level of the status bar below based on the context. Thus, the user can view the various interpretations of this Āyah, and not just getting the literal meaning of it.
How we built it
The app is the practical application of a set of criteria for the identification of metaphors in the Holy Qur’ān. A two-step methodology was employed. First, candidate metaphors were identified by checking the Holy Qur’ān exegesis. Second, each of the suggested criteria was applied to each candidate metaphor. A candidate metaphor passing a criterion was assigned a mark along the continuum of metaphoricity. Then a total number of marks was calculated to arrive at the degree of metaphoricity of each candidate within every criterion.
We designed a simple and straightforward interface to lend a helping hand to the user while using the app.
We linked the user interface to Quran and Tafseer APIs. For every verse, we calculate the metaphors with the Metaphor Detector model and present the score to the user.
Metaphor detector is built on a set of rules for the computational identification of metaphor in the Holy Qur’ān. It is recommended for non-Arab users to be able to comprehend the intricacies of the meanings of the Holy Qur’ān.
Challenges I ran into
The app is based on a theoretical framework for identifying metaphors in the Holy Qur’ān. The proposed framework consists of a set of criteria and linguistic markers namely, semantic criterion including basic/non-basic meaning, culture-bound, and collocational criteria; others were grammatical criterion; morphological criterion and the last was related to the frequency of occurrence. The process of arriving at and proving the aforementioned criteria along with using them as input rules for computer software was a very challenging task.
Accomplishments that I'm proud of
This work represents a novel direction for computational identification of metaphor. Computer software for processing an entire corpus (selected Sūrahs from the Holy Qur’ān) that could yield a list of potential metaphors would thus seem to be a welcome addition to the set of tools currently available to non-Arab users towards a clear understanding of the meanings of the Holy Qur’ān.
What I learned
The process of developing metaphor identification criteria was a long journey of hard work. Computer software rules based on the six constructed criteria for the computational identification of metaphor in the Holy Qur’ān were developed. The total number of candidate metaphors and the criteria against which the degree of metaphoricity was calculated were provided. From this, a detailed description of each criterion and how it was applied was studied. Then, accordingly, a software rule as to how this criterion constituted a suggested formula for the software input was provided. Computation of all marks assigned to each individual candidate was computed along with a score from 1-17. According to this computation, the degree of metaphoricity of a candidate metaphor was determined. The degree of metaphoricity of candidates was calculated as follows: the lower the total score of each candidate in criteria, the lower the degree of metaphoricity of the candidate metaphor in question, and vice versa; the higher the total score of a candidate, the higher the degree of its metaphoricity.
I learned that “Metaphoricity is gradable”. Some metaphors are more metaphorical or primary while others are secondary. The underlying assumption of this indicator of ‘degree of metaphoricity’ is that some metaphors have a potential of denoting metaphorical meaning stronger than others and are therefore considered of a degree of metaphoricity higher than the others.