It's taking an array (of strings, since you said it's an array of words; but it would work with numbers or Symbols too) and converting it to an object where both the names and values of the properties are the same. So for example, ["an", "example"] becomes {an: "an", example: "example"}:

const words = ["an", "example"];
const wordDictionary = Object.assign({}, ...words.map((x) => ({ [x]: x })));
console.log(wordDictionary);

It does it by mapping each array element to an object with one property (named by the array element, with the value of the array element) and then assigning all of those objects to a single target object.

For instance, with my ["an", "example"] array, the steps are:

1. Do the mapping, turning:

   ["an", "example"]
   

   into

   [{an: "an"}, {example: "example"}]
   

2. Spread that array out into discrete arguments to Object.assign with a blank target object. So

   Object.assign({}, ...[{an: "an"}, {example: "example"}])
   

   becomes

   Object.assign({}, {an: "an"}, {example: "example"})
   

3. Use Object.assign to assign the properties from those objects to the first one (the blank {} object).

It's worth noting that the code is overcomplicated. It would be much simpler just to use a loop:

const wordDictionary = {};
for (const word of words) {
    wordDictionary[word] = word;
}

Separately, dictionaries like this are better handled by Map instances:

const wordDictionary = new Map();
for (const word of words) {
    wordDictionary.set(word, word);
}

or in this specific case (since the key and value are the same) a Set:

const wordDictionary = new Set(words);