Software Accurately Predicts Books' Popularity By Analyzing

The writing process, writing advice, and updates on your work in progress
Post Reply
longknife
Posts: 185
Joined: September 16th, 2012, 2:03 pm
Contact:

Software Accurately Predicts Books' Popularity By Analyzing

Post by longknife » January 11th, 2014, 4:09 pm

Their Sentences

"I'll never agree! That's worse than being a slave in prison," Ella glared as she went out the door.

And this comes from the UK Telegraph:

Scientists find secret to writing a best-selling novel
Computer scientists have developed an algorithm which can predict with 84 per cent accuracy whether a book will be a commercial success - and the secret is to avoid cliches and excessive use of verbs
This particularly hits home:
Less successful work tended to include more verbs and adverbs and relied on words that explicitly describe actions and emotions such as “wanted”, “took” or “promised”, while more successful books favoured verbs that describe thought processes such as “recognised” or “remembered”.
Am I ever going to have to do a lot of revising on current works in progress! Read the articles @ http://www.popsci.com/article/technolog ... -sentences and http://www.telegraph.co.uk/technology/n ... novel.html
Drop by Father Serra's Legacy http://msgdaleday.blogspot.com. Comments always eagerly awaited - but only if you find the item interesting enough to respond to.

User avatar
Quill
Posts: 1059
Joined: March 17th, 2010, 9:20 pm
Location: Arizona
Contact:

Re: Software Accurately Predicts Books' Popularity By Analyzing

Post by Quill » January 12th, 2014, 6:23 pm

I'd take all that with a grain of salt. Cliche, I know.

And let's be clear: both articles point to what was found to be averagely true for Adventure genre, which wouldn't be the same for other genres. As the source article on the study shows (in Figure 1), Figures of Speech (FOS) levels on successful books in some of the other genres are quite different.

And of course the disclaimers of 84% accuracy and other variables that disprove the rule, such as luck and marketing. Also, the test samples were pretty small, and I'm not sure Project Gutenberg is the best cross section. Yes, I know they randomly trolled Amazon for some more modern examples.

Thanks for posting. Real interesting stuff.

User avatar
polymath
Posts: 1821
Joined: December 8th, 2009, 11:22 am
Location: Babel
Contact:

Re: Software Accurately Predicts Books' Popularity By Analyzing

Post by polymath » January 13th, 2014, 12:48 am

The "algorithm" to me seeks out what I label static voice. Static voice is verb and adverb related; however, verbs' strength is their temporal significance. This verb significance topic is covered in The Poetics of Aristotle. Static voice basically uses verbs that describe static actions (1), which describe sensory perception actions (2), where a narrator intervenes with neutral, emotionally flat, fact-based objective mediation (3), and have a nonsignficant temporal tense (4) usually describing an action that has no perfect tense sense, meaning a delineated sequence of time. This action begins now in the immediate past or the now present moment and ends soon therafter. Example below.

(1), (2), (3) and (4): Nero watched Rome burn.

Though the verb is past tense, no time for the action is given. Does Nero just this moment begin to watch or has he been watching for a while? Though past tense, the verb watched is imperfect and easily interpreted as an ongoing action, yet static. How long does he watch? Is watching all he does, not even a thought on his mind of glee or sadness or anger or sorrow or eagerness to exact retribution? No movement or itching or coughing or sipping a glass of wine? Whatever?

The visual at least if not other sensory stimuli of Rome burning are summarized without any expressed emotional commentary. The description gives no sense of the meaning of Nero wathcing Rome burn. Whether Nero is the viewpoint character who idly watches Rome burn or a bystander watches Nero watch Rome burn is not clearly or strongly expressed.

The action of watching is static, Nero makes no movement, makes no active or dynamic sense of the context or texture of Nero watching Rome burn. Like from where? When does he watch? We at least know who, if the Nero who is credited with Rome's demise is the Nero of the sentence subject. Maybe another Nero watches Rome burn, perhaps a more modern Nero. That Rome burns is a what texture, but not what Nero or the narrator feels about Rome burning. Why does Nero watch? Did he start the fire? for example. Then how does he watch, idly, intently, enraptured, bored, fearful?

And the sentence is factual with nothing to interpret or infer that would engage readers' imaginations and critical thought processes. Judicious and timely adverbs at least are called for for their function of expressing commentary, perhaps adjectives as well, and perhaps appositive compound or complex clauses that add meaning to the context and texture.

A simple algorithm would seek out such static verbs that fit the parameters given above. This is not to claim that static verbs are to be excised in every instance. A static sentence serves as a direct and easily understood declaration of a stasis statement. Stasis statements have to be and similar verbs, liketo watch, expressing an ongoing static state of being. They are easy to read and serve when a brief summary is needed and a longer descriptive, dynamic expression would be otherwise burdensome. Static voice has a place, as passive voice does, in storytelling. But dynamic, active voice is a best practice for developing the all-essential illusion of reality within a narrative's persons, its now moments and places and situations--settings, and its dramatic events. These events, characters, and settings are ideally developed using process statements; that is, statements that express dramatic processes in progress.

For example: Fleeing south along the Appian Way, Emperor Nero saw Rome fiercely burn, worried the blame would fall hard on him.
Spread the love of written word.

User avatar
hesperus_lux
Posts: 19
Joined: December 30th, 2013, 10:23 am
Location: Hudson Valley, NY
Contact:

Re: Software Accurately Predicts Books' Popularity By Analyzing

Post by hesperus_lux » February 6th, 2014, 2:20 pm

Less successful work tended to include more verbs and adverbs and relied on words that explicitly describe actions and emotions such as “wanted”, “took” or “promised”, while more successful books favoured verbs that describe thought processes such as “recognised” or “remembered”.
I do tend to avoid adverbs. They weaken writing.
The success of books which favor verbs that describe thought processes over emotions is indicative of our very head driven western society. 'Readers of books' like to be thought of as intelligent and sophisticated, esp those readers who are in the position of evaluating books (like editors). This phenomenon reminds me of the hard, sterile, (and good) writing which followed after Hemingway. Where all things are 'said' by a character and never 'yelled,' 'screamed,' 'mumbled,' etc., even if the character is actually yelling, screaming or mumbling! I have to admit that this style of writing is effective, but I wonder how much longer it will be en vogue.
“Nothing is exactly as it seems, nor is it otherwise.”
- Alan Watts

Post Reply

Who is online

Users browsing this forum: Ahrefs [Bot] and 26 guests