Skip to content

classification

src.generators.committee_stats.classification

Committee classification & aggregation.

Builds the institution name index, classifies each committee member's affiliation to a country / continent / institution, and computes per-area aggregates, yearly time-series, recurring-member rankings and institution timelines.

classify_member(affiliation, prefix_tree, name_index)

Classify a single member's affiliation to a country.

Returns (country, institution_name) or (None, None) on failure.

Source code in src/generators/committee_stats/classification.py 
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
def classify_member(affiliation, prefix_tree, name_index):
    """Classify a single member's affiliation to a country.

    Returns (country, institution_name) or (None, None) on failure.
    """
    aff_lower = affiliation.lower().strip()
    if not aff_lower:
        return None, None

    # Try prefix-tree match first
    matches = prefix_tree.values(prefix=aff_lower)
    if matches:
        uni = matches[0]
        return uni["country"], uni.get("name", affiliation)

    # Fall back to fuzzy matching
    best_match = None
    best_ratio = 0
    for name, uni in name_index.items():
        ratio = fuzz.ratio(name, aff_lower)
        if ratio > best_ratio:
            best_ratio = ratio
            best_match = uni

    if best_ratio > 80 and best_match:
        return best_match["country"], best_match.get("name", affiliation)

    return None, None

classify_committees(all_results: dict) -> dict

Classify all committee members by country, continent, and institution.

Parameters

all_results : dict {conf_year: [{name, affiliation}, ...]}

Returns

dict with keys: by_country, by_continent, by_institution, failed

Source code in src/generators/committee_stats/classification.py 
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
def classify_committees(all_results: dict) -> dict:
    """Classify all committee members by country, continent, and institution.

    Parameters
    ----------
    all_results : dict
        {conf_year: [{name, affiliation}, ...]}

    Returns
    -------
    dict with keys: by_country, by_continent, by_institution, failed
    """
    name_index = _build_university_index()
    prefix_tree = Trie(**name_index)

    by_conf_country: dict = {}
    by_conf_continent: dict = {}
    by_conf_institution: dict = {}
    failed: list = []

    for conf_year, members in all_results.items():
        by_conf_country[conf_year] = defaultdict(int)
        by_conf_continent[conf_year] = defaultdict(int)
        by_conf_institution[conf_year] = defaultdict(int)

        for member in members:
            affiliation = _clean_affiliation(member["affiliation"])
            country, inst_name = classify_member(affiliation, prefix_tree, name_index)
            if country:
                by_conf_country[conf_year][country] += 1
                continent = _country_to_continent(country) or "Unknown"
                by_conf_continent[conf_year][continent] += 1
                by_conf_institution[conf_year][inst_name or member["affiliation"]] += 1
            else:
                failed.append(
                    {
                        "conference": conf_year,
                        "name": member["name"],
                        "affiliation": affiliation,
                    }
                )

    return {
        "by_country": by_conf_country,
        "by_continent": by_conf_continent,
        "by_institution": by_conf_institution,
        "failed": failed,
    }