Multimodal Blog Sentiment Classification Based on MD-HFCE

Table 1

The algorithm for forming a fuzzy exclusive dictionary.

---

Input: Blog text T
Output: Fuzzy sentiment dictionary

---

Process:
(1)  IF t in T [i]:
(2)   t_new [i] = jieba (t)
(3)  END if
(4)  FOR word in t_new [i]:
(5)     = TextRank (t_new [i])
(6)  IF word [i] = word [j] and word [i], word [j] both in t_new [i]:
(7)    k [i] =  + 
(8)  END if
(9)  IF j ≤ 5000:
(10)   FW = k [j] in sort (k [i])
(11)  END IF
(12)  IF word in FW:
(13)   pos = num (word in positive)/NUM
(14)   neg = num (word in negative)/NUM
(15)   other = 1-pos-neg
(16)  END if
(17)  Remove the common and incorrect words from FW to form a final fuzzy dictionary with 593 words, as shown in Table 2
(18)  Print

---