diff --git a/modules/exporter/tsv.go b/modules/exporter/tsv.go
index 9ca2550..d986834 100644
--- a/modules/exporter/tsv.go
+++ b/modules/exporter/tsv.go
@@ -59,7 +59,7 @@ func TsvExport(subjectScoreMap map[string]SubjectStudentCore, output string) {
 	for _, value := range subjectScoreMap {
 		_, subjectExistYet := subjectsDataMap[value.SubjectCode]
 
-		if !subjectExistYet {
+		if !subjectExistYet && value.SubjectCode != "" {
 			subjectInfo := SubjectInfo{
 				SubjectName:   value.SubjectName,
 				SubjectCode:   value.SubjectCode,
diff --git a/modules/scanner/scanner.go b/modules/scanner/scanner.go
index 3a63215..1a82356 100644
--- a/modules/scanner/scanner.go
+++ b/modules/scanner/scanner.go
@@ -101,7 +101,7 @@ func scanFile(input string) map[string]exporter.SubjectStudentCore {
 	}
 
 	// Find the review items. Which is <p> tag with text "Số TC:" and <table> tag
-	allSelection := doc.Find("p:contains(\"Số TC:\"), table")
+	allSelection := doc.Find("p:contains(\"CỘNG HÒA XÃ HỘI CHỦ NGHĨA VIỆT NAM\"), p:contains(\"Số TC:\"), table")
 
 	// Global map to store all scores and subject info
 	var ssScores = make(map[string]exporter.SubjectStudentCore)
@@ -113,12 +113,24 @@ func scanFile(input string) map[string]exporter.SubjectStudentCore {
 		s := allSelection.Eq(i)
 
 		// Check if the current selection is <p> tag
-		if s.Is("p") {
+		// Before each table segments, there are <p> tag with subject data and "header" tag
+		// And the subject data <p> tag usually next to the "header" tag
+		if s.Is("p") && strings.Contains(strings.ToUpper(s.Text()), "CỘNG HÒA XÃ HỘI CHỦ NGHĨA VIỆT NAM") {
+			// Always subject data is in the next <p> tag
+			s1 := allSelection.Eq(i + 1)
+
+			// In some cases, the subject data is not next to "header" tag, or just don't have.
+			//Continue and use the previous subject code
+			if !strings.Contains(s1.Text(), "Số TC:") {
+				continue
+			}
+
 			// Split the string by "Số TC:"
-			parts := strings.Split(s.Text(), "Số TC:")
+			parts := strings.Split(s1.Text(), "Số TC:")
 
 			// The first part contains the course name, split it by ":"
 			courseParts := strings.Split(parts[0], ":")
+
 			subjectName := utils.CleanSubjectName(strings.TrimSpace(courseParts[1]))
 
 			// The second part contains the course credit and code, split it by "Mã học phần:"
diff --git a/utils/scanner.go b/utils/scanner.go
index 80ac1aa..2e7b134 100644
--- a/utils/scanner.go
+++ b/utils/scanner.go
@@ -65,6 +65,11 @@ func CleanSubjectName(value string) string {
 
 	// Remove last part
 	valueParts := strings.Split(value, "-")
+
+	if len(valueParts) == 1 {
+		return value
+	}
+
 	value = strings.TrimSpace(strings.Join(valueParts[:len(valueParts)-1], " "))
 
 	return value